Englische Übersetzung für "Ati und nVidia im Optimierungswettstreit"

[GloomY]

Einige englischsprachige Webseiten (nV News (http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=15483) und Beyond3D (http://www.beyond3d.com/forum/viewtopic.php?t=7072)) wären sicher über eine englische Version des Artikels erfreut. Wer hätte erstens Lust und hält sich zweitens für fähig, einen Teil davon zu übersetzen? Ich würde auch mithelfen, habe aber keine Lust, alle neun Seiten alleine zu machen. Wenn sich ausser mir noch mindestens zwei oder drei Andere melden würden, dann könnten wir das in Angriff nehmen.

Leo hat zwar gemeint, dass es eher wenig zusätzliche Besucher für 3DCenter bringen würde, aber immerhin könnte man das Ansehen vielleicht etwas verbessern, während der Artikel sonst auf Grund seiner Einsprachigkeit eher weniger Beachtung bekommen würde.

Als Übersetzungshilfe kann ich Dict.leo.org (http://dict.leo.org/?lang=de) empfehlen, wo man sehr leicht und schnell Wörter nachschlagen kann, wenn sie einem entfallen sind.
Und denkt daran: Je mehr Leute sich melden, desto weniger Arbeit wird es für jeden. Also los ;)

[Endorphine]

Ich melde einfach mal Interesse an. Unverbindlich. Ein paar sollten sich aber schon noch finden.

[mapel110]

ich auch.

[GloomY]

Dank an euch beide =)

Am Besten aber mal bis heute abend warten, wie viele Leute sich bis dahin gemeldet haben. Dann können wir den Artikel aufteilen und jeder kann loslegen.

[Endorphine]

Hab grad mal ein paar Zeilen probiert. Der Start ist wie immer holprig, aber dann kommt man schon rein =) Bei mir hapert es ab und zu an der Grammatik, aber da können wir uns ja gegenseitig am Ende korrigieren.

Edit: Hmm, manche Sätze von Leo sind schon im Deutschen recht schwierig. Das könnte sehr zeitintensiv werden, da man auch nicht richtig ins englische denken reinkommt dadurch. Ich möchte mich mal noch nicht festlegen, ob ich mitmache.

Edit #2: Warum nicht einfach die englischen Seiten einen Link zur maschinellen Übersetzung posten lassen? Das ist eigentlich recht verbreitet und stellt auch keinen grossen Nachteil dar, da die meisten sowieso nur die Benchmarkdiagramme und Bilder durchsehen und vielleicht noch das Fazit lesen.

[mapel110]

Original geschrieben von Endorphine
Hab grad mal ein paar Zeilen probiert. Der Start ist wie immer holprig, aber dann kommt man schon rein =) Bei mir hapert es ab und zu an der Grammatik, aber da können wir uns ja gegenseitig am Ende korrigieren.

Edit: Hmm, manche Sätze von Leo sind schon im Deutschen recht schwierig. Das könnte sehr zeitintensiv werden, da man auch nicht richtig ins englische denken reinkommt dadurch. Ich möchte mich mal noch nicht festlegen, ob ich mitmache.

ich denke, solange der sinn erhalten bleibt, können wir uns auch wengier förmlich ausdrücken.
solange keine slang wörter von uns reingemischt werden ;)

[Aqualon]

Original geschrieben von mapel110
ich denke, solange der sinn erhalten bleibt, können wir uns auch wengier förmlich ausdrücken.
solange keine slang wörter von uns reingemischt werden ;)

Ich wollte ja auch erst mithelfen, aber da ich bei manchen Sätzen schon in Deutsch leichte Verständnisprobleme habe, lass ich das lieber bleiben. Technisch korrekt soll der Artikel ja trotzdem sein, selbst wenn er nicht so "geschnörkelt" übersetzt wird.

Aqua

[Endorphine]

Ich hab einfach mal bei Seite 5 angefangen, um den Arbeitsaufwand abschätzen zu können. Versucht es auch mal, es ist durch Leos komplexes Deutsch nicht so einfach wie man denkt.

Ich traue mich mal, das zu posten, was bei rausgekommen ist (nicht hauen ;))

Deutsch:ATi-Optimierungen

Es wurde auch bei diesen Benchmarks extra ein 16x anisotroper Filter gewählt, um eher Grafikchip-limitierte Szenarien zu erreichen, da sich so Treiber-Optimierungen am ehesten erkennen lassen:

Die Situation präsentiert sich hier ungleich aufgeräumter als bei den nVidia-Benchmarks. Dennoch läßt sich erkennen, daß auch ATi bei den beiden 3DMarks ein paar in ihren Auswirkungen zwar geringfügige, aber nichts desto trotz vorhandene Optimierungen mitlaufen läßt. Denn obwohl die Leistungs-Unterschiede bei den beiden 3DMarks zwischen den Messungen mit originalen Treiber und den Messungen mit modifizierten Treiber recht gering sind, [...]


Englisch. Bitte nicht in der Luft zerreissen, das ist nur ein kleiner Anfang. Ich brauche bei sowas sowieso immer etwas Zeit, um da wieder rein zu kommen. Nur, um mal den Schwierigkeitsgrad der Aufgabe aufzuzeigen:
ATI optimizations

We chose 16° anisotropic filter for these benchmarks to move the bottleneck more towards the graphics card so that driver optimizations are easier to recognize.

The situation we got here is more unequally distributed than in the nvidia benchmarks. Thus we can discover that there are minor but existant effects. Performance differences in both 3dmarks are quite small[...]

[Aqualon]

Original geschrieben von Endorphine
[...]more unequally distributed[...]


Damit sagst du das Gegenteil von dem, was eigentlich gesagt werden soll. Ins deutsche zurückübersetzt heisst das nämlich sowas wie "noch ungleicher verteilt".

Und der eigentliche Sinn des deutschen Satzes war doch, dass bei ATI nicht so viele gravierende Unterschiede zwischen Standard und Anti-Cheat auftreten wie bei Nvidia, oder?

Aber ansonsten finde ich den Teil ordentlich übersetzt. Zwar nicht ganz so "ausschweifend" wie das Original, aber der Sinn bleibt erhalten und es ist leicht lesbar.

Aqua

[Endorphine]

Ja, solche Details müsste man dann korrigieren und evtl. ausdiskutieren. Und das ist wirklich eine Menge Arbeit. Deshalb - lohnt sich der Aufwand wirklich? Ich glaube, es wäre sinnvoller wenn man einfach etwas Werbung für den Artikel macht und ihn mit maschinellen Übersetzungen verlinkt. Das hat sich bis jetzt eigentlich ganz gut bewährt in der Praxis. Und da Englisch als germanische Sprache dem Deutschen sowieso sehr nahe steht ist das Ergebnis IMHO auch tragbar und wir könnten uns die Arbeit sparen. Es ist ja nicht so wie japanisch -> Englisch. Wie seht ihr das?

Edit: Es sei denn, die Übersetzerbots verheddern sich bei Leos ausschweifenden Schachtelsätzen ;)

[Aqualon]

Hab mich mal an Seite 8 versucht:

Deutsches Original

Da jene generelle Optimierung des anisotropen Filters seitens ATi aber eben auch für alle anderen Anwendungen zutrifft, haben wir die schon vorbenutzten Benchmarks außerhalb von Unreal Tournament 2003 ebenfalls noch einmal unter 16x anisotropen Filter durchlaufen lassen. Der anisotrope Filter wurde dabei einmal regulär über das Control Panel erzeugt, was einen bilinearen/trilinearen anisotropen Filter ergibt, und einmal über das rTool 0.9.9.6d, was einen reinen trilinearen anisotropen Filter ergibt:



Manuelle Englische Übersetzung

Given that ATI´s general optimisations of the anisotropic filter also apply for all other apllications, we ran the former used benchmarks except Unreal Tournament 2003 once again with the 16° anisotropic filter. The anisotropic filter was once regularly adjusted with the Control Panel, which leads to a bilinear/trilinear anisotropic filter and once with rTool 0.9.9.6d, which results in a pure trilinear anisotropic filter:



Übersetzung von Altavista

Since that general optimization of the anisotropic filter applies on the part of ATi however evenly also to all other applications, we left continuous bench mark already before-used outside the filter anisotropic of unreal Tournament 2003 likewise again under 16x. The anisotropic filter was once regularly produced thereby over the control panel, which results in a bilinearen/trilinearen anisotropic filter, and once over rTool the 0.9.9.6d, which results in a pure trilinear anisotropic filter:


Die maschinelle Übersetzung klingt also gar nicht mal so übel.

Aqua

[Aqualon]

Original geschrieben von Endorphine
Ja, solche Details müsste man dann korrigieren und evtl. ausdiskutieren. Und das ist wirklich eine Menge Arbeit. Deshalb - lohnt sich der Aufwand wirklich?


Nachdem die Übersetzung von Altavista gar nicht mal so übel geraten ist, würde es vielleicht sogar ausreichen nen Link zu ner Übersetzungsseite anzugeben + nen Link zu nem speziellen englischen Thread hier im Forum, wo man Verständnisfragen stellen könnte.

Aqua

[Endorphine]

Original geschrieben von Aqualon
Nachdem die Übersetzung von Altavista gar nicht mal so übel geraten ist, würde es vielleicht sogar ausreichen nen Link zu ner Übersetzungsseite anzugeben + nen Link zu nem speziellen englischen Thread hier im Forum, wo man Verständnisfragen stellen könnte.

Aqua Das wäre wohl das vernünftigste.

Hier der entsprechende Google-Link: http://translate.google.com/translate?u=http%3A%2F%2Fwww.3dcenter.org%2Fartikel%2Fati_nvidia_treiberoptimier ungen%2F&langpair=de%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools

Stellenweise holprig, aber noch gut lesbar IMHO.

[Aqualon]

Original geschrieben von Endorphine
Stellenweise holprig, aber noch gut lesbar IMHO.

Aber mit netten Stilblüten, aus Botmatch wurde Offering match *gg*

Google und Altavista benutzen anscheinend auch die selbe Übersetzungsengine. Der Teil von Seite 8, den ich übersetzt hab, ist haargenau gleich bei den beiden.

Aqua

[GloomY]

Ich habe die Erfahrung gemacht, dass die automatische Übersetzung recht schlecht ist. Z.B. hat AMDZone den ersten Absatz des Athlon64 OC Artikel maschinell übersetzt. Das Original lautete:Dadurch, daß bei den K8-Prozessoren Athlon 64 und Opteron der FrontSideBus generell wegfällt, weil das Speicherinterface nicht mehr wie beim Athlon XP oder Pentium 4 im Mainboard-Chipsatz sitzt, sondern im Prozessor selber, werden natürlich auch alle bisherigen Erfahrungen und Weisheiten bezüglich des Overclockings bei diesen neuen Prozessor über den Haufen geworfen oder aber zumindestens in Frage gestellt. Denn wie übertaktet man beim Athlon 64, wenn dieser gar keinen FrontSideBus mehr besitzt, sondern ausschließlich nur noch einen Speichertakt?Daraus wurde dann:Because with the K8-Prozessoren Athlon 64 and Opteron of the FrontSideBus is generally omitted, because the memory interface no more does not sit as with the Athlon XP or Pentium 4 in the Main board chip set, but in the processor, naturally also all past experiences and pointingnesses become concerning the Overclockings with this new processor over the heaps thrown or however zumindestens in question posed. Because how one over-clocks 64, if this possesses no more FrontSideBus, with the Athlon but excluding only a storing act?Ich persönlich finde das Ergebnis ziemlich schlecht. Daher sehe ich die Notwendigkeit einer manuellen Überstetzung. Von mir aus kann man ja auch eine maschinelle Übersetzung nehmen und dann deren Ungereimtheiten ausbessern, hauptsache das Endergebnis ist einigermaßen lesbar.

Ich möchte auch noch mal betonen: Wir wollen keinen perfekten Zeitungsartikel machen. Solange der Sinn rüberkommt, wird man uns Fehler durchaus verzeihen. Bei einer automatischen Übersetzung bin ich mir aber nicht sicher, ob hier nicht an der ein oder anderen Stelle ziemlicher Müll herauskommt, also der Sinn verdreht wird.

[Endorphine]

Es stellt sich eben die Frage von Aufwand vs. Nutzen. Lohnt es sich, dass wir alle hier viel Zeit und Mühe in eine Übersetzung stecken, wenn dann doch kaum jemand den Text wirklich liest und eine maschinelle Übersetzung 95 % des Inhalts ebenso rüberbringt?

Links zu den Übersetzerbots sind ja recht weit verbreitet, IMHO nicht unbedingt zu Unrecht.

[mapel110]

Original geschrieben von Endorphine
Es stellt sich eben die Frage von Aufwand vs. Nutzen. Lohnt es sich, dass wir alle hier viel Zeit und Mühe in eine Übersetzung stecken, wenn dann doch kaum jemand den Text wirklich liest und eine maschinelle Übersetzung 95 % des Inhalts ebenso rüberbringt?

Links zu den Übersetzerbots sind ja recht weit verbreitet, IMHO nicht unbedingt zu Unrecht.

sehe es doch als übung für dich. lernst unter umständen neue vokaln =)

[Aqualon]

Original geschrieben von GloomY
Ich persönlich finde das Ergebnis ziemlich schlecht. Daher sehe ich die Notwendigkeit einer manuellen Überstetzung. Von mir aus kann man ja auch eine maschinelle Übersetzung nehmen und dann deren Ungereimtheiten ausbessern, hauptsache das Endergebnis ist einigermaßen lesbar.

Bei dem Beispiel ist es wirklich sehr schwer verständlich um was es überhaupt geht, da wär ne Nachbearbeitung dringend notwendig.

Die Google-Übersetzung als Vorlange zu nehmen und die dann zu überarbeiten dürfte wohl die beste Lösung sein.

Aqua

[mapel110]

Original geschrieben von Aqualon
Bei dem Beispiel ist es wirklich sehr schwer verständlich um was es überhaupt geht, da wär ne Nachbearbeitung dringend notwendig.

Die Google-Übersetzung als Vorlange zu nehmen und die dann zu überarbeiten dürfte wohl die beste Lösung sein.

Aqua

das mach ich sowieso. ich nehm google oder babblefish als grobe übersetzung zur hilfe, schreibe den text an sich aber neu.

[Kakarot]

ATi-Optimierungen

Es wurde auch bei diesen Benchmarks extra ein 16x anisotroper Filter gewählt, um eher Grafikchip-limitierte Szenarien zu erreichen, da sich so Treiber-Optimierungen am ehesten erkennen lassen:

Die Situation präsentiert sich hier ungleich aufgeräumter als bei den nVidia-Benchmarks. Dennoch läßt sich erkennen, daß auch ATi bei den beiden 3DMarks ein paar in ihren Auswirkungen zwar geringfügige, aber nichts desto trotz vorhandene Optimierungen mitlaufen läßt. Denn obwohl die Leistungs-Unterschiede bei den beiden 3DMarks zwischen den Messungen mit originalen Treiber und den Messungen mit modifizierten Treiber recht gering sind, können diese Unterschiede jedoch bei der Konstanz dieser beiden Benchmarks, welche kaum Schwankungen von ± 10 Punkten produzieren, unmöglich Meßtoleranzen sein.

We choose a 16° anisotropic filter for these benchmarks in order to create a scenario which is limited by the graphics card for the reason to recognise driver optimisations more easily.

This Situation looks more tidy in addition to benchmarks by nVidia.
By running both 3DMark benchmarks we have recognised minor optimisations used by ATi which can be identified clearly even if the effects are insignificant. It's impossible that the variation of +/- 10 points are measuring tolerances, because of the permanence both 3DMark benchmarks have, even if there is a slight gap between measurements based on the original drivers on the one hand and the modified drivers on the other hand.

just my 2 cents

[GloomY]

@Aqualon: Mach' doch mit. Wie ich sehe bist du auch nicht ganz unwissend, was die englische Sprache angeht. Und Fehler finden wir schon zusammen, wenn wir uns gegenseitig helfen.

Eine vollständig automatische Übersetzung wurde ja schon in dem Thread bei nV News als nicht gut bezeichnet:blah it's this stupid translator it puts words in wrong placesSo wie's aussieht, ist das alleine einfach nicht akzeptabel für englischsprachige Leser.

@Endorphine: Wenn dir der Aufwand zu groß ist, dann lass es sein. Das nimmt dir keiner übel, kein Problem. Auch wenn ich es natürlich schade finden würde, weil wir jede helfende Hand gebrauchen können. Jeder muss aber wissen, was er mit seiner Zeit machen will...

btw: Zeckensack war schon eine Woche lang nicht mehr im Forum. :( Den hätte ich sonst noch gerne angeschnorrt :D

[Aqualon]

Original geschrieben von GloomY
@Aqualon: Mach' doch mit. Wie ich sehe bist du auch nicht ganz unwissend, was die englische Sprache angeht. Und Fehler finden wir schon zusammen, wenn wir uns gegenseitig helfen.

Ich nehm mir dann mal Seite 8 an, mehr dürfte ich wohl zeitlich vorerst nicht schaffen.

Wie schauts eigentlich mit ner Deadline aus? Allzulang sollten wir uns wohl nicht Zeit lassen.

Aqua

[Endorphine]

Original geschrieben von GloomY
btw: Zeckensack war schon eine Woche lang nicht mehr im Forum. :( Den hätte ich sonst noch gerne angeschnorrt :D Schnorr' doch nggalai an =)

[Aqualon]

Übersetzung von Seite 8:

Given that ATi´s general optimizations of the anisotropic filter also apply for all other applications, we ran the former used benchmarks except Unreal Tournament 2003 once again with the 16x anisotropic filter. The anisotropic filter was once regularly adjusted over the control panel, which leads to a bilinear/trilinear anisotropic filter and once over rTool 0.9.9.6d, which results in a pure trilinear anisotropic filter:


Diagramm ATi-Treiber 03.4


The other benchmarks however partly produced very irritating results: The fact that there´s no impact at all on some benchmarks suggests a possible CPU limitation and is therefore not further notable. But the definitely higher results produced under the two 3DMarks with the assigned high-order filtering are at the moment absolutely not explainable. Perhaps we return again later to this phenomenon, but for the instant we don´t want to keep ourself busy with the two 3DMarks, since both benchmarks are considered by Ati & nVidia to be "in need" for application-specific optimizations and so became uninteresting as comparison instruments between graphics cards anyway.

Apart from these two exceptions at least in AquaNox 2 and Max Payne a few lower results can be observed as they were expected to be seen with the high order filtering produced by rTool. The difference between the bilinear/trilinear anisotropic filter by control panel and the pure trilinear anisotropic filter by rTool is however clearly smaller than the difference measured with the Flybys in Unreal Tournament 2003.

Finally there´s a need to check if these observed performance differences also apply to the newer drivers 03.5 and 03.6. For that purpose we ran some selected Unreal Tournament 2003 Flyby-Benchmarks once again to compare these drivers with the 03.4 driver. But there shouldn´t be any real differences, because ATi´s filter-optimization is - as shown before - identical within the driver versions 03.2 to 03.6:


Diagramm ATi-Treiber 03.4, 03.5 & 03.6 @ UT2003


Thus it can be verified that the ascertained performance differences of the driver 03.4 with the anisotropic filter are transferable 1:1 to the newer drivers 03.5 and 03.6. An additional control of real screenshots with the driver version 03.6 brought no new realizations: Also with this driver version no further differences except the bilinear/trilinear anisotropic filter showed up between the filter produced by Ati control panel and that one produced by rTool (therefore we also did without these screenshots).

---

Hoffe mal fürn Anfang ist das nicht allzu übel ;)

Kritiken sind sehr willkommen, ist ja meine erste Übersetzung dieser Art.

Aqua

Edit: optimisation durch optimization ersetzt.

[Endorphine]

Ich mach grad Seite 4. Seite 5 kann Kakarot ja weitermachen.

[Aqualon]

Ich übernehm dann auch Seite 7, wenn ich schon mal drin bin.

Aqua

[Aqualon]

Übersetzung von Seite 7:

Fortunately the ATi driver leaves the user a small back door, if a genuine tilinear filter is prefered. Because when we controlled the anisotropic filter not over the ATi control panel, but over the TweakTools aTuner 1.4.32.4465 (comes with a experimental but in our benchmarks nevertheless efficient Radeon-Support) or the rTool 0.9.9.6d (the previous version 0.9.9.6c wasn´t able to do that!), we had no problems to unite the trilinear and the anisotropic filter (compare with the original ATi-Screenshot on the right):


Screenshots ATi 03.4 + rTool 8x AF & ATi 03.4 8x AF


At this place, there was of course the question, wether there exists other differences between those two pictures aside of the bilinear or trilinear filter. After detailed investigation of appropriate "normal" screenshots (i.e. without colored MipMaps) we could however detect only one and in addition absolutely minimum difference (MouseOver effect by Javascript, alternatively a click opens both screenshots):


MouseOver Screenshots ATi 03.4 8x AF & ATi 03.4 + rTool 8x AF


If one looks at the edge to the abyss quite at the right edge of the abyss, then minimum differences can be seen in favor of the rTool. Since all other screenshots recorded by us including those with colored MipMaps however showed no further differences, one can roughly say that there is probably no difference in image quality between the bilinear/trilinear anisotropic filtering by the control panel and the pure trilinear anisotropic filtering by rTool, with regard to static elements. Excluded from that is the effect of Mip-Banding, which can only be seen in motion.

By the way there was finally a possibility at our disposal to control, how much computing power ATi saves with the bilinear/trilinear anisotropic filtering by control panel compared to the pure trilinear filtering by rTool. For this purpose we simply ran some benchmarks with the 16x anisotropic filter. The anisotropic filter was once regularly produced by the control panel, which results in a bilinear/trilinear anisotropic filter and once over rTool 0.9.9.6d, which results in a trilinear anisotropic filter.

We could observe by the way, that during these benchmarks the settings of the control panel and aTuner resp. rTool interfered with one another. In each case just the settings of the program run first were accepted and only a restart of windows solved this problem. First we ran some Unreal Tournament 2003 benchmarks:


Diagramm ATi-Treiber 03.4 @ UT2003


Thus the performance advantages, which ATi gains by this general optimization of the anisotropic filter, can at least be exactly classified for Unreal Tournament 2003. In the Botmatches and also in the Timedemos (replays of real gaming) changes can hardly be seen, which results from the general CPU limitation of the game. The fact that the general optimization of the anisotropic filter works nevertheless is noticeable in the Flyby-Benchmarks which - free from CPU limitations - shows about 20 per cent difference between bilinear/trilinear anisotropic filter (control panel) and pure trilinear anisotropic filter (rtool).

---

Das wär dann Seite 7 und für heute hab ich erstmal genug :buck:

Aqua

Edit: Die richtige Seite sollte man doch drüberschreiben :weg:
Edit 2: optimisation durch optimization ersetzt.

[Endorphine]

page 4 - ready 4 c&p =)

<p style="font-family:Copperplate Gothic Light,Georgia,Narkisim;font-size:15pt;color:#7700FF">
<b>&nbsp;&nbsp;&nbsp;nVidia-optimizations&nbsp;(contin.)</b></p>

<p align=justify>The most important fact here is that nVidia uses pseudo-trilinear filtering with a high portion of bilinear filtering. This alone would be no special behavior, but the comparison to other Direct3D applications indicates clearly that nVidia uses strict trilinear filtering in the normal case. The texture filtering behavior of UT2003 with current nVidia drivers is an exception to this rule and this means that there must be some sort of application-specific optimization. The following figures support this argumentation <i>(from top to bottom: Direct3D AF-Tester, Devastation Demo, Serious Sam: The Second Encounter, 3DMark03)</i>:</p>

<p align=center><img src="/images/ati_nvidia_treiberoptimierungen/screenshot13.jpg" width=440 height=329 border=0 alt="Direct3D AF-Tester"></p>
<br />
<p align=center><img src="/images/ati_nvidia_treiberoptimierungen/screenshot14.jpg" width=440 height=199 border=0 alt="Devastation Demo"></p>
<br />
<p align=center><img src="/images/ati_nvidia_treiberoptimierungen/screenshot15.jpg" width=440 height=310 border=0 alt="Serious Sam: The Second Encounter"></p>
<br />
<p align=center><img src="/images/ati_nvidia_treiberoptimierungen/screenshot16.jpg" width=440 height=526 border=0 alt="3DMark03"></p>

<p align=justify>Especially the second screenshot (taken from Devastation Demo) is of great interest. Devastation Demo uses the same Engine like Unreal Tournament 2003 and most notably the same method of mipmap-colorization but shows different texture filtering: smooth transitions from one mipmap to another - proper trilinear filtering. This fact definitely shows that nvidia's driver 44.65 recognizes the execution of UT2003 and alters the filtering technique from strict trilinear filtering to application-specific pseudo-trilinear filtering.</p>

<p align=justify>Finally it remains to be resolved to which drivers our conclusions apply to. We also took screenshots with driver versions 44.03, 44.67, 44.68, 44.71 and 44.90. They are virtually identical to the ones shot with version 44.65, so there's no use of publishing them here. We also verified the test for the reference filter quality with driver versions 44.03 and 44.90 which also delivered the same output. There is no visible rendered difference between these driver versions.</p>

<p align=justify>The only question left is whether the investigation of benchmarked results from the three benchmarks - which receive application-specific optimization - result in measurable performance differences between individual driver versions. This would point out possible changes in already discovered application-specific optimizations. Therefor these three benchmarks with application-specific optimizations, again with driver versions 44.03 and 44.90, respectivly with and without anti-cheat script:</p>

<!-- Tabellen -->

<p align=justify>Thus it can be stated that all said before applies to all drivers from version 44.03 (current official driver) to 44.90 (latest unofficial version).</p>

[Kakarot]

Page 5
ATi-optimisations


ATi-optimisations

We choose a 16° anisotropic filter for these benchmarks in order to create a scenario which is limited by the graphics card for the reason to recognise driver optimisations more easily:
[ img ]...[ /img ]

This situation looks more tidy in addition to benchmarks by nVidia.
By running both 3DMark benchmarks we have recognised minor optimisations used by ATi which can be identified clearly even if the effects are insignificant. It's impossible that the variation of +/- 10 points are measuring tolerances, because of the permanence both 3DMark benchmarks have, even if there is a slight gap between measurements based on the original drivers on the one hand and the modified drivers on the other hand.

In these two cases it's the same as we mentiond at the benchmarks by nVidia: To use application-specific optimisations along with theoretical benchmarks is absurd and should be called unauthorized optimisation. It's not in the sense of theoretical benchmarks to benchmark higher because of this specific-optimisations. In this case there is no need to have a screenshot with lower image quality to underline our statement, because even the fact that application-specific optimizations are used is enough to prove our point of view.

However it's quite interesting in which way the newer version of the Ati drivers act regarding to the "optimisations" we mentioned before. In addition to that we benched both 3DMarks with newer driver versions called 03.5 and 03.6. In case of the 03.5 driver we made an extra test in terms of the modified driver version. The 03.6 driver refused to work with the "anticheat" script, so we have no choice and except the standard benchmark score:
[ img ]...[ /img ]

These results dilute our statment mentioned before.
As Ati announced further, there are no application-specific optimisations for 3DMark03 within the 03.5 driver version.
However current ATi driver versions still benefit from the application-specific optimisations used for 3DMark2001 - including the driver version 03.5 and probably the newest driver version 03.6.

[Aqualon]

@Kakarot:

Könntest du vielleicht die Seite 5 in einem Posting zusammenfassen und es auch dazuschreiben, welche Seite es ist?

Dann wärs auch leichter das Korrekturzulesen und später zusammenzufassen.

Aqua

[Kakarot]

geht klar, kommt sofort, bin noch etwas am korrigieren. :)

[GloomY]

Super, dass ihr so fleissig seid =)

Ich mach' mich dann mal an Seite 6 ran.
Original geschrieben von Endorphine
Schnorr' doch nggalai an =) Jep, den hab' ich im Thread im Mod-Forum auch schon vorgeschlagen, schreibe ihm aber sicherheitshalber noch eine PM.

edit: Seite 6

ATI-Optimizations (continued)

Actually this article could almost end here. Almost - because the ATI driver finally exposes a surprise which - at first - does not show any differences in benchmarks. Because actually we only wanted to take some ATI screenshots in comparison with the nVidia shots shown on the pages before in order to show how to filter "correctly". On the "real" screenshots there was nothing to expose, but the screenshots with colored mipmaps show surprising results when activating anisotropic filtering (click opens lossless compressed screenshots in 1024x768 resolution, on the right side the same nVidia image with anti-cheat script):

[image]

Firstly, a clear bilinear filter shows up here - although ATI's quality filter mode should definitely filter trilinear. However, a weak blue colouring before the clear blue colouring is still noticeable, which indicates still a partial trilinear filtering. In order to get the cause for this behavior, we regarded in comparison to the "normal" image an additional purely bilinear filtered image, where definitely all textures are thus only filtered bilinear (created with rTool by the way).

[image]

The difference can be recognized precisely and permits apparently only the conclusion that ATi indeed filters some textures trilinear, others in contrary only bilinear. In order to get to know this more exactly, we endeavored the D3D AF-Tester, which is capable of indicating the mipmaps for different texture stages. Thereby a clear difference between the texture stage 0 and all other texture stages showed up under trilinear anisotropic filter:

[image]

The driver version 03.6 used for this test already suggests that this behavior is not limited to a single driver. In Fact all ATi drivers starting from the version 03.2 up to the most current driver version 03.6 behave identical in the same manner like the screenshots show here. In order to allow comparability the nVidia images of the driver version 44.03 (results are equal through driver version 44.90) are shown below, where all texture stages are perfectly equal trilinear filtered in the quality filter mode (with the exception of Unreal Tournament 2003, as already mentioned before)

[image]

Thus it can be stated that the bilinear optimization under the anisotropic filter seen on Unreal Tournament 2003 screenshots is not an application-specific optimization, but a technique generally used by ATI for the anisotropic mode for all applications. More precisely, with ATi's anisotropic filter only the basis texture (texture stage 0) is filtered trilinear, whereas all other textures (texture stages 2-7) are only bilinear filtered. This behavior of the ATi driver thereby applies to all applications if these are using the anisotropic mode.

Similar to nVidia this general optimization of the anisotropic filter is based on the fact that also current ATi graphic chips only have bilinear filtering texture units, thus needing twice the rendering power for trilinear filtering und hence saving enormous computation efforts with bilinear filtering. Beside that, optimizations take advantage of the usually not needed trilinear filter for lightmaps or small bumpmaps, because due to their kind and/or their size hardly any mip banding effects can occur, which would make a trilinear filter necessary.

Nevertheless, ATi however ignores in this context that there are also different textures (for example detail textures) beside the basis texture, which can show mip banding and therefore trilinear filtering should be used. As a matter of course the same ATi hardware can filter all textures correctly trilinear, which is obvious when deactivating the anisotropic filter (on the left 8x anisotropic filter, right without anisotropic filter):

[image]

As already mentioned, an application-specific optimization is definitely not present here, since not only Unreal Tournament 2003 uses this mixture of trilinear filtering for the basis texture and bilinear filtering for the rest of the textures but the ATi driver generally dictates this for all applications. Nevertheless, the question must be asked, whether this general optimization for the anisotropic filter may really be called a "trilinear filter" and whether the whole tweaking possibly tends toward illegal optimization. Because this general optimization method does not appear to be alright at least for us.

[mapel110]

bin an seite 1 dran :)

[mapel110]

*posting-schind*
hab ich mir verdient :)


SEITE 1

ATI & nVidia in "optimizations" contest

In course of our upcoming graphics cards roundup of the current highend chips of nVidia and ATI,
we stood and stand in front of the unusual question for the first time, which benchmarks we can use.
Due to the last information about driver-"optimizations" we unfortunately can not be sure
in which benchmarks the chip manufactures use permitted or unauthorized optimizations.

The main course of this uncertainty is an article from digit-life,
which proofs that ATI and nVidia use unauthorized driver optimizations in the 3dmark2001.
But also some benchmakr results from our 3dcenter-forum under the employment of the anticheat script provided by digit life in the before mentioned article,
which can remove application-specific optimizations from the drivers, turned us precariously,
because they proof that there is a clear performance difference between the original and modified anticheat drivers in 3dmark03, unreal tournament 2003 and codecreatures benchmark.

In order to achieve a certain safety in our benchmarks for our next graphics chips roundup, we have rebuilt the test scenario of digit life for a larger amount of benchmarks,
for the original (nvidia 44.65 and ATI 03.4, like in the article of digit-life) and for the anticheatscript modified drivers.

The fact that this AntiCheat Scripte possibly cannot go around all optimizations is perfectly conscious to us thereby -
however this scripts is momentarily also the only one and best to accomplish this task.
also a "result" of this script has not to be a unauthorized optimization - it proofs only that there is a optimization present.
whether these optimization are permitted or without permission can only be proofed by screenshots.

This article assumes the fact that the AntiCheat Scripte of digit Life perfectly fulfills its function of the deactivation of application-specific optimizations and has no unwanted side effects.
When there is a proof that this script cannot do so, some statements of this article would become obsolete.

Unfortunately, these anticheat scripts works only under Direct3D, so we used Serious Sam: The Second Encounter and also Counter Strike in the Direct3D-API.
A test of the opengl versions of these games and of pure opengl games is not possible with these scripts.

In order to understand this article it's necessary to know what bilinear and trilinear filtering is.
In addition the reading of our basic article "diagram filter: Bilinear to anisotropic in the detail" is recommended.

NVIDIA OPTIMIZATIONS

In the following benchmarks we used 8x anisotropic filtering to ensure grafikchip limitated scenaries
so driver optimizations can be recognized easier.

<grafik>

It shows clearly, that both 3dmarks and also Unreal Tournament 2003 achieve remarkable lower scores.
The nvidia driver has definitly application specific optimizations for these games. For all the other games it seems not to be the case
- or however the anticheat scripts can't recognize these application-specific optimizations for the other games, what cannot be really excluded unfortunately.

The realization that a application-specific optimization is present in 3DMark2001, 3DMark03 and Unreal Tournament 2003 proofs not that these optimizations are unauthorized - nearly anyhow.
Application specific optimizations should not affect theoretical benchmarks like the 3dmarks because the majority thinks this is unauthorized optimizations.
The hardware should only do what the developer of the theoretical benchmark invented.

Any simplifications, even if they do not change the image quality, contradict this principle and actually occur the sense of a theoretical benchmark.
The futuremark developer have exactly the same opinion. Any application specific optimization is an "unauthorized optimization",
which opinion we would like ourselves to attach regarding theoretical benchmark programs.
http://www.futuremark.com/companyinfo/3dmark03_audit_report.pdf

To mark of cource our benchmarks use the 3dmark03 built 330, which was published by futuremark to disable the optimization of nvidia and ATI.
Unfortunately futuremark hasn't discovered all "optimizations" of nvidia - and the "optimizations" in the 3dmark2001 obviously not at all.

/edit
ich hab die links noch nicht mit drin. und ich hab noch keine gute übersetzung für "unerlaubte optimierung" gefunden. ich bitte um vorschläge. derzeit stehts als "bad optimization" drin :)

/edit2
bad durch unauthorized ersetzt

[Aqualon]

Original geschrieben von mapel110
ich hab noch keine gute übersetzung für "unerlaubte optimierung" gefunden. ich bitte um vorschläge. derzeit stehts als "bad optimization" drin :)

Wie wärs mit unauthorized optimisations? Es kann ja nicht im Sinn des Hersteller eines Programm sein, dass die Grafik optimierungsbedingt schlechter aussieht, als in anderen Spielen, die nicht mit ner Optimierung bedacht wurden.

Aqua

[GloomY]

Original geschrieben von Aqualon
Wie wärs mit unauthorized optimisations? Es kann ja nicht im Sinn des Hersteller eines Programm sein, dass die Grafik optimierungsbedingt schlechter aussieht, als in anderen Spielen, die nicht mit ner Optimierung bedacht wurden.

Aqua Ich hab's mit "illegal optimizations" übersetzt. Allerdings gefällt mir "unauthorized" irgendwie doch besser. "Illegal" klingt so juristisch. Ich werd's bei meinem Teil ändern.

btw: "optimizations" oder "optimisations"? Nehmen wir britisches oder amerikanisches Englisch?

[Endorphine]

Ich bin für durchgängig amerikanisches Englisch. Die Zielgruppe ist einfach um Grössenordnungen bedeutender.

[mapel110]

Original geschrieben von Endorphine
Ich bin für durchgängig amerikanisches Englisch. Die Zielgruppe ist einfach um Grössenordnungen bedeutender.

hab ich auch so gemacht. allerdings eher unbewusst. wusste garnicht, dass es bei dem wort zwei schreibweisen gibt. mal wieder was gaylernt :)

/edit
habs editiert.
soweit zufieden mit meinem teil ?

[Aqualon]

Original geschrieben von Endorphine
Ich bin für durchgängig amerikanisches Englisch. Die Zielgruppe ist einfach um Grössenordnungen bedeutender.

Hab optimisation durch optimization ersetzt. Allerdings hab ich nicht wirklich drauf geachtet, ob ich bei Wörtern die amerikanische oder englische Schreibweise verwendet habe.

Aqua

[Kakarot]

Original geschrieben von GloomY
btw: "optimizations" oder "optimisations"? Nehmen wir britisches oder amerikanisches Englisch?
Seite 5 ist britisches Englisch, mir ist es eigentlich egal, obwohl mir britisches Englisch besser gefällt.

Unerlaubte Optimierung = interdicted optimisation (verbotene Optimierung)

edit: unauthorized optimisation, gefällt mir besser. Hab's dahingehend geändert.

[mapel110]

was fehlt jetzt noch ?
seite 1, 4, 5, 6, 7 und 8 haben wir ja.
noch irgendwas in arbeit ?

/edit
ich mach mich mal an seite 2. das ist nicht soviel :)

[Aqualon]

Hab heute leider keine Zeit, könnte mich dann wahrscheinlich erst wieder morgen Abend an ne Übersetzung machen.

Aqua

[mapel110]

seite2

nVidia-Optimizations (cont.)

Because of that we can go on without image quality comparision in 3Dmark, since alone the power measurement is sufficient here as proof for an unauthorized optimization. Expressly this includes not Unreal Tournament 2003. Here are application specifis optimizations on the one hand legal and on the other hand very welcome, but only if they achieve higher performance without dropping down the image quality. In this case it would be an unauthorized optimization.

Therefor we employed us deeper with Unreal Tournament 2003 to find out whats the reason for the higher performance of the original driver compared to the unoptimized (modified) drivers. At first we had to find out if this performance difference appears only in special flybys, botmachtes or timedemos, or if they are common for the whole game. To do so, we testet all the flybys, botmachtes and timedemos available, at first without anisotropic filtering.

<grafik>

and the same with 8x anisotropic filtering

<grafik>

From this it can be concluded now that nVidia has - despite of one or another outliers - probably a general optimization for Unreal Tournament 2003 and no optimization only for certain Flybys or only for the anisotropic filter. So it is not possible to make these optimizations dont working e.g. with a costum timedemo. We testet the timedemos "overkill" and "cheat" which are not known by nvidia, nevertheless they benefit from the optimizations, so we can conclude that this concerns the whole game.

[mapel110]

seite 3 und 9 auch in arbeit.
stay tuned :)

seite 9

In the course of the discussion of a first version of this article within the internal range of the 3DCenter of forum: -) concerned Demirug (programmer of the D3D AF tester) again more exactly the application-specific optimization of unreal Tournament 2003. Nvidia uses the filename "ut2003.exe" to apply the special optimizations (can easily be proofed by renaming other applications, results in higher performance).To proof the image quality the "D3D AF-tester" should be renamed to ut2003.exe to show the nvidia optimizations.

But there was a "technical" problem which showed quiet unexpected another "optimization" of nVidia. Renaming the actual version 1.1 of the "D3D AF Tester" (which was used in the first version of this article) causes nothing - the results stay the same. Only as Demirug presented a new version of the D3D of AF tester, this effect disappeared and it could be determined the image quality of unreal Tournament 2003 after renaming in " UT2003.exe " .

Completely obviously nvidia has built in a small security which causes only renaming the "D3D AF tester" to "UT2003.exe" not to show this effect. The only reason for doing this is nvidia dont want the hardware reviewers to proof the bad filter quality under Unreal Tournament 2003 with the "D3D AF tester"(which is btw a tool recommend by nVidia).

This however only besides. Technically seen demirug came up with a brandnew version of the "D3D AF tester" which let him get the following results in the three nVidia filtermodi under Unreal Tournament 2003:

Quality filter:
The trilinear filtering on the stage 0 used only for approx. half of all pixels.
The trilinear filtering on stage 1-7 is only a completely narrow volume between the MipMaps.
Textures on the stage 0 get the maximum level at anisotropic filtering.
Textures on stage 1-7 get maximally a level at anisotropic filtering of 2.

Performance filter
Trilinear filtering is missing any longer, one gets only a bilineare filtering.
Textures on the stage 0 get the maximum level at anisotropic filtering.
Textures on stage 1-7 get maximally a level at anisotropic filtering of 2.

High Performance filter
Trilinear filtering is missing any longer, one gets only a bilineare filtering.
Textures on the stage 0 get the maximum level at anisotropic filtering -
however only with the filtering, which is used anyway during the attitude (in principle angle dependence with this filter mode).
Textures on stage 1-7 get no anisotropic filtering.

This observation applies only to unreal Tournament 2003. In all other applications (after present level of knowledge!) the nVidia driver behaves " normally " and implements for all texture stages the computation of the trilinear anisotropic filter in such a way, as this was requested by control panel or Tweaktool.
/////////////////////////////////////////
Fazit seite 9
RESULT

The following can now be summarized:

NVidia uses in 3DMark2001 and 3DMark03 application-specific optimizations in the drivers 44,03 to 44,90, which under first benchmark approx. 16 per cent and under second benchmark approx. 62 per cent (!) performance advantage under 8x anisotropic filter bring in. Whether these application-specific optimizations bring disadvantages in the image quality with itself was not tested, since application-specific optimizations do not have anyway absolutely to be in theoretical benchmark.

NVidia uses further under unreal Tournament 2003 a application-specific optimization in the drivers 44,03 to 44,90, which brings a performance advantage in the flybys about approx. 57 per cent (!) under 8x anisotropic filter. How much these 57 per cent however without and how much with image quality loss were reached, could with our possibilities not be determined and must remain open. To be proven knew however is that nVidia at least uses a pseudotrilinear filter particularly under Unreal Tournament 2003, as well as partital MipMaps shifts to the rear in the anisotropic filter and partly the full anisotropic filter does not work. Further it could be proven that this is not the normal condition of the nVidia drivers, since these filter cleanly trilinear anisotropic in all other applications. Thus it is considered as proven that the application-specific optimization of Unreal Tournament 2003 brings a definitely worse image quality with itself, even if the difference is quite small.

ATi used a application-specific optimization, which brings in approx. 2 per cent performance advantage for this bench mark under 16x anisotropic filter under 3DMark03 up to the driver 03,4. In the drivers 03,5 and 03,6 this optimization is no longer ascertainable, so we can see this case as done.

ATi uses a application-specific optimization in the drivers 03,4 to 03,6, which brings in approx. 4 per cent performance advantage under 16x anisotropic filter under 3DMark2001. Whether disadvantages bring this application-specific optimization in the image quality with itself was not tested, since application-specific optimizations do not have anyway absolutely to be in theoretical bench mark.

ATi used further a general optimization under the anisotropic filter in the drivers 03,2 to 03,6, which brings in approx. 20 per cent performance advantage under 16x anisotropic filter in the Flyby measurements of unreal Tournament 2003. In other benchmarks however as well do not fasten as at all. ATi filters the basis texture when anisotropic filtering only trilinear, further textures however only bilinear. Disadvantages in the image quality outside of this bilinear/trilinearen combining filter could not be proven. Nevertheless the question must open-remain whether this general optimization of the anisotropic filter likewise does not represent an unauthorized optimization, since the ATi own requirement of the "trilinear filtering" is not finally reached. Also offers here nVidia - except for the determined exception in unreal Tournament 2003 mind you - a trilinear filter in principle for all texture stages.

Generally it can be recommend not to use 3dmark for benchmarking because of the determined application-specific optimizations on the part of ATi and nVidia.And completely generally keep the eyes open far for the existing image quality - particularly with new drivers - in the future rather examine twice, than assuming blindly everything will already fit somehow.

So we are in an very tricky situation in Unreal Tournament 2003. Surely nVidia uses an unauthorized optimization - this attacks however the static image quality only marginally, but replaced " only " the trilinear filter by a pseudotrilinear. That is further an unauthorized application-specific optimization, but on the other hand this is also - roughly compared - the normal condition at ATi under the anisotropic filter. ATi also replaces the trilinear filter in this case , at ATi simply by a bilinearen/trilinearen combining filter.

The situation in Unreal Tournament 2003 is nearly the same for nVidia and ATI under anisotropic filter since both use no complete trilinear filter. Theoretically we could go on without remarking that this specific optimization is an unauthorized one at nvidia and a partical unauthorized at ATI. Both chip manufacturers have optimized down the image quality so far that the image quality is nearly again alike ironically. nVidia has " created " an unauthorized optimization which looks not worse than the ATi normal condition; -).


However as a result of this comparison naturally again some other problems arose: Because without anisotropic filter ATi filters normally (purely trilinear) under unreal Tournament 2003,
while nVidia uses the pseudotrilinear filter. But other benchmarks show an disadvantage for nvidia in image quality. Because while nVidia filters outside from unreal Tournament 2003 everywhere clearly trilinear at anisotropic filtering, there are the observed bilinearen/trilinearen mixes at ATi in all applications.If one regards however this latter comparison as generally fair, then also the application-specific optimization is not unauthorize under unreal Tournament 2003 any more, since it does in the long run only the same.

The whole thing reminds a gordian knot, which you can smash but not solve. Surely it would be the simplest to set the bilineare/trilineare anisotropic filtering of the ATi driver as unauthorized optimization and to use only the pure trilinear filter offered by tools like "rtool". then we would have weapons equally (with exception Unreal Tournament 2003). But we have also to consider that most of the users take the control panel for switching the anisotropic filter and not any externat tools like rtool or atuner.

If one makes the benchmarks only with such external tools he works past the control panel and so past to most of the users. Because of that we cannot say how we will benchmark in our upcoming high end graphics cards roudup. Maybe this desicion will come up in the discussion to this article.

Surely it would be better both graphics chip developer leave the quality modes in their drivers untouched and any optimizations whether application specific or general would be build into specially modes. Probably nobody has something against optimizations, which have hardly a visible quality loss and bring in a performance advantage between 20 and 57 per cent. If we should find such extra modes in future drivers, even only for individual games usable, we would bench this probably additionally to the pure quality mode. Only the quality mode should not use such optimizations, if both chip developers want to assume to be outriders of image quality.



///////////
to be continued. kann man schon mal gegenlesen :)
mach nachher weiter.

/edit
mal die zeilenumbrüche entfernt :)
FERTIG =)

[Leonidas]

Bitte PM, wenn fertig.

[GloomY]

So, wie sieht's aus, Mapel? Wieviel hast du schon von Seite 3?

Wir sollten langsam mal fertig werden. Obwohl ich momentan nicht wirklich viel Zeit habe (Klausur), würde ich das hier gerne zu Ende bringen und notfalls auch noch Seite drei machen.

edit: Seite 3

Now the question arises, from where the performance comes - that is, whether image quality was sacrificed somewhere. Therefore, we recorded a small demo (the former for performance measurement used "Cheat"-Demo), with which we could create everytime reproduceable screenshot positions and thus were able to produce the following screenshots, first again without anisotropic filter (click opens lossless compressed screenshot in 1024x768 resolution):

[image]

... and now with 8x anisotropic filter (click opens lossless compressed screenshot in 1024x768 resolution):

[image]

If we are honest, there are not many differences visible, even when looking precisely. By the way, this applies to the rest of the other screenshots we created for this purpose in Unreal Tournament 2003. NVidia's driver optimization for Unreal Tournament 2003 seems to cost only very marignal image quality. However, at least in the picture with 8x anisotropic filter some differences are identifiable (MouseOver effect with Javascript, alternative: Click opens both screenshots):

[image]

In the difference comparison it can be seen here clearly that the rear dunes on the picture of the original 44.65 nVidia driver appear to be covered with a veil of blur. Only the modified driver with the anticheat script removes this veil of blur and offers an appropriate image quality matching the 8x anisotropic filter setting. Here, nVidia thus clearly sacrificed a part of the image quality in order to achieve higher performance, even if this effect apparently occurs only under 8x anisotropic filter and then applies to some parts of the picture only.

In addition here is another difference comparison from the same demo, only at another screenshot position (Mouseover effect with Javascript, alternative: Click opens both screenshots):

[image]

Here one must look very exact in order to be able to identify the differences. Nevertheless, when regarding exactly differences between the two screenshots show up at the edges to the abyss in the center and at the right border of the image. However, we do not evaluate them as fairly tragic due to only quite small differences. Thus, we must state at this time that nVidia lowered the image quality under Unreal Tournament with an application-specific optimization in driver 44.65. This however - up to this time of the article - results in a only minimum worse overall image quality.

However, there are naturally also image quality characteristics, which cannot be be caught by screenshots, because they only are visible in motion. To this category belongeds primarily the point of the bilinear or trilinear filtering, that is the question how well the graduations between the individual mipmaps are covered. Actually, nothing should go wrong here, because the nVidia driver promises a clean trilinear filtering with the filter setting "Quality", with which the individual mipmaps stages should gently pass into each other. Anyway, we looked at the above screenshots again with colored mipmaps, first without anisotropic filter. The shots were produced with the command "firstcoloredmip 1" (click opens for lossless compressed screenshot in 1024x768er resolution):

[image]

and then with 8x anisotropic filter (click opens lossless compressed screenshot in 1024x768 resolution):

[image]

Indeed this is now a surprising result. Because nVidia obviously doesn't uses the normal trilinear filter here - particularly only in Unreal Tournament 2003 - , but apparently a bilinear filter with a small portion of trilinear filtering, ergo a pseudotrilinear filter. Thus, it becomes finally clear, from where the partial high performance differences between original and modified driver come, because the bilinear filter costs the GeForceFX chips only half the rendering power.

This is primarily because all GeForceFX chips have only bilinear texture units. For trilinear filtering either both texture units or two Rendering passes are used then. In short: With Multitexturing and/or with anisotropic filtering the filling rate and therefore the computing power of the graphic chip is cut in half with trilinear filtering. By dramatically reducing the fraction of trilinear filtering in Unreal Tournament 2003 by nVidia, a substantial amount of fillrate is saved.

Roughly estimated (not checked) this would look the following way: When regarding full trilinear filtering once as 100 per cent performance, for bilinear filtering only 50 per cent performance would naturally be necessary. For nVidia's "optimized" method for Unreal Tournament 2003 this would perhaps result in only approx. 60-65 per cent needed performance. The mip-banding (transition between two mipmaps) is nevertheless still quite well covered by the narrow trilinear range.

Additionally the colored screenshots shows under anisotropic filtering that with the "optimization" the first mip-stages already begin noticeably nearer in individual cases (does not always apply according to other screenshots taken by us). This is not the way correct filtering with mipmapping should work. By the way, this is the explanation for the image quality difference in these places, which can be seen on the "real" screenshots.

uff, fertisch :jump1: :D

[mapel110]

ich hab leider keine zeit mehr, um alles nochmal durchzulesen. ich könnte das erst am montag wieder machen.

wäre also nit schlecht, wenn jemand also nochmal alle seiten auf fehler durchsucht.
jedenfalls ist die übersetzungsarbeit damit erstmal beendet. :)

[GloomY]

Jep, jetzt muss sich nur jemand finden, der sich das Ganze nochmal durchliesst. Ich hab' nachher noch 'ne Klausur und bekommen daher ausser zig numerischen Verfahren momentan nichts in meinen Schädel hinein ;)

Ich werde nach der Klausur mich aber mal dransetzen. Vielleicht kann da der eine oder andere schon mal ein bisschen Vorarbeit leisten...

[GloomY]

Ok, ich hab's leider gestern doch nicht mehr geaschafft und fahre jetzt für ein paar Tage weg. Die Übersetzungen sind ja alle gemacht, jetzt bräuchte man nur noch jemanden, der sich diese nocheinmal durchliesst. Meine Teile sind imho in Ordnung, ich hab' mir recht viel Mühe gegeben, also sag' ich jetzt mal als Motivationshilfe, dass man sich den Aufwand bei mir sparen kann ;)

Es wäre echt gut, wenn das demnächste fertig würde, weil es wirklich nicht mehr viel Arbeit ist.

[Aqualon]

Ich bin leider momentan auch anderweitig beansprucht und werde bis Ende August wohl nicht mehr allzuviel Zeit für Forenbesuche aufwenden können.

Aqua

[Leonidas]

Ich danke erstmal allen wie verrückt. Hier ist eine *unkontrollierte* Vorversion, die noch nicht verlinkt und damit noch nicht offiziell ist:
http://www.3dcenter.org/artikel/ati_nvidia_treiberoptimierungen/index_e.php

[aths]

:-)

Wettlauf würde ich nicht mit Wettlauf (race) sondern Wettbewerb (contest) übersetzen.

[mapel110]

nö, ich finde das passt.

race macht das ganze etwas "lockerer" und imo auch interessanter :kicher:

/edit
ich les jetzt mal

[Popeljoe]

Hab mir Seite 1 mal durchgesehen und ein paar Anmerkungen:
Seite 1 Zeile 2: ...which to use. oder ...which benchmarks we should use.
Seite 1 Zeile 8: ...turned us precausely. They proof...
Seite 1 Zeile 29: ...These applications does not seem to work in the rest of the tested games
Seite 1 Zeile 33: ...does not proof, that these applications are unauthorized
Seite 1 Zeile 34: ...because most people think,that these apps are only optimizations for treating and thus unusefull for real tests.
Seite 1 Zeile 40: We would like to mark, that we used 3Mark03 version 330...
Irgendwie gefäält mir dieses "unauthorized" nicht so richtig, im Moment finde ich aber nix besseres.
Mache jetzt weiter...
Hoffe, dass es gewollt ist...
Popeljoe

Seite 2 Zeile 1: ...,since only the ...
Seite 2 Zeile 2ff: In fact this does not include Unreal Tournament 2003: in a real game these application specific optimisations are legal and welcome, if they burst the performance without loss of image quality!
Such Optimisations are unallowed if the result is a loss of image quality!
To find out the reasons for the higher performance of the original driver compared to the unoptimized (modified) drivers in Unreal Tournament 2003, we had to dig a little deeper.

Seite 2 unten: Out of these tests it can be concluded, that nVidia has - despite of one or another outliers - probably built a general optimization for Unreal Tournament 2003 and no optimization only for certain Flybys or only for the anisotropic filter.
Those, who think, they can cheat out these optimizations by using so called "custom timedemos", are wrong.
Despite the fact, that the used timedemos "Overkill" and "Cheat" are unknown by nVidia, they worked with the optimisations. So we conclude, that the optimisations work in the whole game.

[mapel110]

Original geschrieben von Popeljoe
Seite 1 Zeile 2: ...which to use. oder ...which benchmarks we should use.
Seite 1 Zeile 8: ...turned us precausely. They proof...


ack. das kann geändert werden. der rest imo not.
trotzdem danke soweit :)

/edit
wir haben uns im übrigen auf die amerikanische schreibweise geeinigt. also mit Z. optimizations ;)

[Popeljoe]

Ami Schreibweise, alles klar!
Wolte euch auch nicht dazwischenhauen, aber einiges war etwas schlecht zu lesen.
Z.B. Bandwurmsätze und einige Sachen sind zu "deutsch" geschrieben, sprich zu direkt übersetzt.
Falls gewünscht: bin grad bei Seite 3 und hab auch bis 16.00 noch Zeit.
Popeljoe

[Popeljoe]

Seite 3:
Seite 3 Abs.2: To be honest,…
Seite 3Abs.3: In comparison the difference can be seen clearly: the rear dunes on the picture of the original 44.65 nVidia driver appear to be covered with a veil of blur.
Only the modified driver with the anticheat script removes this veil of blur and offers an appropriate image quality matching the 8x anisotropic filter setting. In this case, nVidia has clearly sacrificed a part of the image quality in order to achieve higher performance, even if this effect apparently occurs only under 8x anisotropic filter and then applies to some parts of the picture only.
Seite 3 Abs.4: In this case you will have to take a very close look, to be able to identify the differences. Nevertheless, when regarding exactly differences between the two screenshots show up at the edges to the abyss in the center and at the right border of the image. However, we do not evaluate them as fairly tragic due to only quite small differences. Thus, we must state at this time that nVidia lowered the image quality under Unreal Tournament with an application-specific optimization in driver 44.65. This results - up to this point of the article - in only a minimum worse overall image quality.
However, there are also image quality characteristics (naturally), which cannot be be caught by screenshots, because they are only visible in motion. This depends primarily on the bilinear or trilinear filtering, so the question is, how well the graduations between the individual mipmaps are covered. Actually, nothing should go wrong here, because the nVidia driver promises a clean trilinear filtering with the filter setting "Quality", with which the individual mipmaps stages should gently pass into each other. Anyway, again we took a look at the screenshots above, this time with colored mipmaps, first without anisotropic filter. The shots were produced with the command firstcoloredmip 1
Seite 3 Abs.4:
Now this is, indeed, a surprising result. Because nVidia obviously don't use the normal trilinear filter here - particularly only in Unreal Tournament 2003 - , but apparently a bilinear filter with a small portion of trilinear filtering, ergo a pseudotrilinear filter. Thus, it becomes finally clear, where the partial high performance differences between original and modified driver comes from. Using the bilinear filter costs the GeForceFX chips only half of the rendering power.
This is primarily because all GeForceFX chips have only bilinear texture units. For trilinear filtering either both texture units or two Rendering passes are used. In short: With Multitexturing and/or with anisotropic filtering the filling rate and therefore the computing power of the graphic chip is cut in half with trilinear filtering. By dramatically reducing the fraction of trilinear filtering in Unreal Tournament 2003 by nVidia, a substantial amount of fillrate is spared.
We only estimated (not mathematically checked!) the following: When regarding full trilinear filtering once as 100 per cent performance, for bilinear filtering only 50 per cent performance would naturally be necessary. For nVidia's "optimized" method for Unreal Tournament 2003 this would perhaps result in only approx. 60-65 percent need of performance. The mip-banding (transition between two mipmaps) is nevertheless still quite well covered by the narrow trilinear range.

In addition the colored screenshots show, under anisotropic filtering, that with the "optimization",in some cases the first mip-stages already start noticeably closer (earlier) (does not always apply according to other screenshots taken by us). This is not the correct way the filtering with mipmapping should work. By the way, this is the explanation for the image quality difference in these places, which can be seen on the "real" screenshots.

S.4 Anfang:The most important fact here is that nVidia uses pseudo-trilinear filtering with a high portion of bilinear filtering. This wouldn’t cause much trouble, if it would be the standard mode for Gforce FX Chips. It gets problematic in comparison with other Direct3D applications, because nVidia uses strict trilinear filtering in the normal mode. The texture filtering behavior of UT2003 with current nVidia drivers is an exception to this rule and this means that there must be some sort of application-specific optimization. The following figures support this argumentation (from top to bottom: Direct3D AF-Tester, Devastation Demo, Serious Sam: The Second Encounter, 3DMark03):


Mehr braucht nicht.:(

Popeljoe

[GloomY]

@Popeljoe: Ich finde es ja gut, dass du auch ein Auge aufmachst, aber hebe deine Änderungen bitte hervor, wenn du ganze Sätze quotest. Das macht es einfacher, die Änderungen nachzuvollziehen.

btw: Es heisst imho "half the rendering power" und nicht "half of the rendering power".

Ansonsten finde ich es nicht umbedingt sinnvoll, so viele einzelne Sätze zu machen. Sicher ist es im Englischen nicht üblich, lange verschachtelte Sätze wie im Duetschen zu machen, aber imho sind diese hier nicht so extrem lang und daher noch akzeptabel.
Und wenn du aus Haupt- und Nebensatz eingeleitet mit Konjunktion zwei Hauptsätze machst und die Konjunktion einfach weglässt, sind die einzelnen Sätze zwar kürzer, aber die Beziehung, die zwischen diesen beiden bestand, ist nun nicht mehr umbedingt ersichtlich, so wie es im deutschen Original der Fall war. (hier: "Thus, it becomes finally clear, where the partial high performance differences between original and modified driver comes from. Using the bilinear filter costs the GeForceFX chips only half of the rendering power.")
Imho sollte man dort nicht umbedingt etwas ändern.

btw: Ich schaue die anderen Sachen natürlich auch noch durch...

[Popeljoe]

Ich hab mir ganz einfach den Text gegriffen und die Sachen verbessert, die imo unverständlich waren!
Ich wollte hier auch keinen angepissen oder mich profilieren!
Einzelsätze habe ich nur anfangs gezielt und unter Angabe der Textstelle rausgegriffen.
Die folgenden Seiten sind eigentlich vollständig durchgesehen.

Und "half the rendering power" ist asselamerikanisch, wenn du dies in einer sozusagen "Veröffentlichung" haben willst: Bitte!
Außerdem schreibst du ja selber "Using the bilinear filter costs the GeForceFX chips only half of the rendering power ".
Popeljoe

[Popeljoe]

Achso: unerlaubt übersetzt mit unauthorized oder unallowed?
Ich hab dab unallowed geschrieben, weil NV sich ja quasi selbst dazu authorisiert hat!:D

Ansonsten isses heiss!
Popeljoe

[Leonidas]

Original geschrieben von Popeljoe
Hab mir Seite 1 mal durchgesehen und ein paar Anmerkungen:
Seite 1 Zeile 2: ...which to use. oder ...which benchmarks we should use.
Seite 1 Zeile 8: ...turned us precausely. They proof...
Seite 1 Zeile 29: ...These applications does not seem to work in the rest of the tested games
Seite 1 Zeile 33: ...does not proof, that these applications are unauthorized
Seite 1 Zeile 34: ...because most people think,that these apps are only optimizations for treating and thus unusefull for real tests.
Seite 1 Zeile 40: We would like to mark, that we used 3Mark03 version 330...




Ganz ehrlich: Es gibt in diesem Dokument keine Zeilen im gewöhnlichen Sinne. Ich weiss nicht, wo ich diese Textstellen suchen soll.

[Leonidas]

Ich denke, unpermitted ist das beste Wort.

[Leonidas]

Neue Version, hab unterschiedliche schreibweisen einiger wörter geglättet:
http://www.3dcenter.org/artikel/ati_nvidia_treiberoptimierungen/index_e.php



Ich warte nur noch auf den offiziellen 45.23. Wenn er genauso ist wie der inoffizielle 45.23, kommt noch ein deutscher Nachtrag und dann werden beide Artikel-Versionen ineinander verlinkt.

[Kakarot]

Seite 5, erster Satz - ist noch britisches Englisch... s durch z ersetzen.

[Popeljoe]

Original geschrieben von Leonidas
Ganz ehrlich: Es gibt in diesem Dokument keine Zeilen im gewöhnlichen Sinne. Ich weiss nicht, wo ich diese Textstellen suchen soll.
Ganz ehrlich: nimm den englischen Text, auf den du in deinem Link verwiesen hast und zähl die Zeilen von oben nach unten!
Das macht man gewöhnlich so!:D
Popeljoe
p.s.: unpermitted triffts, hab die ganze Zeit gegrübelt!

[Quasar]

Original geschrieben von Leonidas
Ich denke, unpermitted ist das beste Wort.

Diese Wort existiert im Englischen nicht. "illegal" oder "unfair" oder etwas ähnliches wäre IMO besser.

"precausely" im zweiten Absatz gibt's auch nicht.

"limitated" bei nVidia Optimizations ist auch ein vertipper...

Und ganz allgemein: Keine Auslass-Apostrophierung in schriftlicher Sprache... "can't" "don't" sind absolute no-nos, auch wenn's vielleicht einige Cowboy-Seiten so machen.

Und passt auf Numerus- und Genus-Konsistenz zumindest innerhalb eines Satzes auf. Und nach "any" steht meist der Singular.

Any application specific optimization is an "unpermitted optimization", which opinion we would like ourselves to attach regarding theoretical benchmark programs.
"Any application specific optimiziation is an illegal optimization" - a thesis we strongly support regarding synthetic benchmarks.
Falls ich den englischen Satz richtig gedeutet habe und man sich einer Meinung anschliessen will.

"cource" ist auch falsch (erste Seite letzter Absatz)

[zeckensack]

Original geschrieben von Quasar
Diese Wort existiert im Englischen nicht. "illegal" oder "unfair" oder etwas ähnliches wäre IMO besser.

"precausely" im zweiten Absatz gibt's auch nicht. "illegitimate" (üpp (http://www.infoplease.lycos.com/ipd/A0485960.html), #2)
oder ganz banal
"not allowed"

[Popeljoe]

Original geschrieben von Quasar
Diese Wort existiert im Englischen nicht. "illegal" oder "unfair" oder etwas ähnliches wäre IMO besser.

"precausely" im zweiten Absatz gibt's auch nicht.
Oh manno Quasar:
fang du jetzt nicht auch noch an!
Ich hab schon diverse PNs mit Gloomy und Mapel gewechselt um ja alle persönlichen Fußfallen zu umgehen...
Klappt aber nicht: s.o.
Popeljoe

[Quasar]

Noch ein heisser Tip: Die englische Satzstruktur ist zu verschieden von der deutschen, insbesondere bei längeren und verschachtelten Sätzen, um einfach den deutschen Satzbau und Satzteilfolgen 1:1 übersetzt und in der Stellung der Worte angeglichen, übernommen zu werden.

In dem Sinne: Mut zum UMFORMULIEREN!

edit:
Seite zwei, unterm Diagramm: "outliers" Das Wort gibt's zwar, aber kein normaler Mensch wird es kennen...

[zeckensack]

Seite 1, Absatz 1
"
During the course of preparing our upcoming graphics cards roundup of the current highend chips of nVidia and ATi, we wondered, and still wonder which benchmarks to use. This situation is unprecedented. Due to the last information about driver "optimizations" we unfortunately cannot* be sure for which benchmarks the chip manufactures have optimized, either legitimately or not.
"

*"can not" wird zusammengezogen.

"
The main cause of this uncertainty is an article by Digit-Life which proves that ATi and nVidia use unpermitted* driver optimizations for 3DMark2001. Additionally some benchmark results from the 3DCenter forums generated with the AntiCheat scripts provided by Digit-Life in the aforementioned article to remove application specific optimizations from the drivers, raised our suspicion. They prove that there is a clear performance difference between the original and modified AntiCheat drivers in 3DMark03, Unreal Tournament 2003 and Codecreatures benchmark.
"

*:)

[zeckensack]

Original geschrieben von zeckensack
Seite 1, Absatz 2
"
The main cause of this uncertainty is an article by Digit-Life which proves that ATi and nVidia use unpermitted* driver optimizations for 3DMark2001. Additionally some benchmark results from the 3DCenter forums generated with the AntiCheat scripts provided by Digit-Life in the aforementioned article to remove application specific optimizations from the drivers, raised our suspicion. They prove that there is a clear performance difference between the original and modified AntiCheat drivers in 3DMark03, Unreal Tournament 2003 and Codecreatures benchmark.
"

*:) Würde ich gerne mit der Kettensäge weiter umstrukturieren :)

"
The main cause of this uncertainty is an article by Digit-Life which proves that ATi and nVidia use special driver optimizations for 3DMark2001. Some benchmark results from 3DCenter forums raised even more suspicion. These were generated with the AntiCheat scripts Digit-Life have provided along the aforementioned article. The scripts remove application specific optimizations from drivers and we've seen a clear performance disparity between the original and modified drivers in 3DMark03, Unreal Tournament 2003 and Codecreatures benchmarks.
"

[Quasar]

In order to achieve a certain safety in our benchmarks for our next graphics chips roundup, we have rebuilt the test scenario of Digit-Life in order to provide you with a larger amount of benchmarks, for the original (nVidia 44.65 and ATi 03.4, as in the article by Digit-Life) and for the AntiCheat script modified drivers.

The fact that the AntiCheat scripts possibly cannot circumvent all optimizations is absolutely present to us -at the moment however this script is unique and solely able to aid us in accomplishing[7b] this task. Also a [b]positive result derived from the use of this script does not necessarily indicate an unpermitted* optimization - it only proofs that there is a optimization present - whether permitted or unpermitted can only be proofed by screenshots.

[Quasar]

This article assumes for fact that the AntiCheat scripts of Digit-Life perfectly fulfill their function of deactivating application-specific optimizations and has no unwanted side effects. If there is proof to be found that this script cannot work without generating these unwanted side effects, some statements of this article would become obsolete.

Unfortunately, these AntiCheat scripts work only in Direct3D, so we used Serious Sam: The Second Encounter and also Counter-Strike with their Direct3D API. A test of the OpenGL versions of these games and of pure OpenGL games is not possible with these scripts.

[Quasar]

Zum Titel:
"ATi and nVidia: Battle of wits for most optimized drivers" wäre mein Vorschlag.

edit:
So, jetzt muss ich aber schlafen!

[zeckensack]

Seite 1, Absätze 3&4
"
To get some confidence in our own benchmarks for our next graphics chip roundup, we'll largely follow Digit-Life's testing methology for our benchmarks, using both the original drivers (nVidia 44.65 and ATi 03.4, just like in Digit-Life's article) and the AntiCheat script modified versions.

We're perfectly aware that the AntiCheat scripts may not be sufficient to defeat all optimizations - however at this time they're the best tool for the job, by virtue of being the only one available. These scripts may also have the unwanted side effect of removing perfectly legal optimizations - in this case we can only detect that there is an optimization. Whether it's valid or not can only be judged by inspecting screenshots.
"

[zeckensack]

Seite 1, Absätze 5&6 frei nach Quasar:
"
This article relies on Digit-Life's AntiCheat scripts completely deactivating all application-specific optimizations without unwanted side effects. Should this assumption be proven to be false, some statements of this article will become obsolete.

Unfortunately, these AntiCheat scripts work only for Direct3D, so while testing Serious Sam: The Second Encounter and Counter-Strike we used their Direct3D renderers. The scripts will not affect these games' OpenGL renderers, nor will they affect pure OpenGL games.
"

[zeckensack]

S1, A7&8

"
Bislang fehlender siebter Absatz (nicht gewünscht wg Link?):
Parts of this article will require a deeper understand of bilinear and trilinear texture filters. We recommend our own Filters: from bilinear to anisotropic (http://hier.der.link/) as further reading.

nVidia optimzations

In the following benchmarks we used 8x anisotropic filtering to ensure graphics chip limited scenarios, so driver optimizations become more apparent:

Diagramm
"

[zeckensack]

S1, A9&10
"
There clearly is a remarkable impact on 3DMarks and Unreal Tournament 2003 scores. The nVidia driver definitely contains application specific optimizations for these games*. For all other** games it does not seem to be the case - or rather the AntiCheat script can't recognize any application-specific optimizations for the other games, we unfortunately cannot entirely dismiss this possibility.

*3DMark=Spiel? So steht's jedenfalls auch in der deutschen Fassung.
**entweder "all other" oder "the other". IMO keinesfalls "all the other", das ist zu deutsch.


The observation that application specific optimization is present for 3DMark2001, 3DMark03 and Unreal Tournament 2003 does not yet prove that these optimizations are unpermitted* - but there's a catch. Application specific optimizations should not affect theoretical benchmarks like the 3DMarks because the majority thinks these are unpermitted optimizations. The hardware should only do what the developer of the theoretical benchmark intended.


* :naughty:
"

[zeckensack]

S1, A11&12

"
Any simplification, even if it does not change image quality, contradict this principle and actually circumvent the purpose of a theoretical benchmark. The 3DMark developers Futuremark share the same opinion ("3DMark03 Audit Report" (http://ich.bin.ein.link/), PDF, 1 MB). Regarding theoretical benchmarks, any application specific optimization is an "unpermitted optimization", an oppinion we strongly agree with.

We should add that all our 3DMark03 testing has been performed with build 330, which was published by Futuremark to disable the optimization of nVidia and ATi. Unfortunately Futuremark haven't discovered all of nVidia's "optimizations" - and apparently they didn't discover the "optimizations" for 3DMark2001 at all.
"

[Leonidas]

Original geschrieben von Quasar
Diese Wort existiert im Englischen nicht. "illegal" oder "unfair" oder etwas ähnliches wäre IMO besser.




Mmh. Babylon hat es mir soforr übersetzt. Und illegal oder unfair sind weit von der Bedeutung des deutschen Originals entfernt.

[Leonidas]

Original geschrieben von zeckensack
"illegitimate" (üpp (http://www.infoplease.lycos.com/ipd/A0485960.html), #2)
oder ganz banal
"not allowed"


Ich nehm jetzt unallowed

[Leonidas]

Original geschrieben von Quasar
edit:
Seite zwei, unterm Diagramm: "outliers" Das Wort gibt's zwar, aber kein normaler Mensch wird es kennen...



Sorry, Quasar: Aber alle Deine Anmerkungen sind nutzlos, wenn Du keine Alternative vorgibst.

[Leonidas]

Original geschrieben von Quasar
In order to achieve a certain safety in our benchmarks for our next graphics chips roundup, we have rebuilt the test scenario of Digit-Life in order to provide you with a larger amount of benchmarks, for the original (nVidia 44.65 and ATi 03.4, as in the article by Digit-Life) and for the AntiCheat script modified drivers.

The fact that the AntiCheat scripts possibly cannot circumvent all optimizations is absolutely present to us -at the moment however this script is unique and solely able to aid us in accomplishing[7b] this task. Also a [b]positive result derived from the use of this script does not necessarily indicate an unpermitted* optimization - it only proofs that there is a optimization present - whether permitted or unpermitted can only be proofed by screenshots.



Wenn man ganze Absätze übersetzt, ist es nicht notwendig, auf die Änderungen hinzuweisen, ich übernehme ganze Absätze auch komplett. Thx.

[Leonidas]

Original geschrieben von zeckensack

"
Bislang fehlender siebter Absatz (nicht gewünscht wg Link?):

"


Jap.

[Leonidas]

Bis hier hin: Alles eingearbeitet. Thx @ all.

http://www.3dcenter.org/artikel/ati_nvidia_treiberoptimierungen/index_e.php

[Exxtreme]

Original geschrieben von Leonidas
Bis hier hin: Alles eingearbeitet. Thx @ all.

http://www.3dcenter.org/artikel/ati_nvidia_treiberoptimierungen/index_e.php
Ist der jetzt final?

[Leonidas]

Original geschrieben von Exxtreme
Ist der jetzt final?


Nein. Wer weiter ändern will: Ich baue alles ein, was noch an Änderungen kommt. Artikel geht wahrscheinlich erst Online, wenn 45.23 draussen ist, weil ich in einem Vorwort diesen 45.23 erwähnen will (muss).

[Aqualon]

Seite 7, 1. Satz:

[...]if a genuine trilinear filter[...]

Aqua

[Exxtreme]

Letze Seite, Zeile 2:

"...the application-specific optimization of Znreal Tournament 2003...

[GloomY]

Seite 5, 3. Absatz:
It's not in the sense of theoretical benchmarks to benchmark higher because of this specific optimizationswürde ich durch das ersetzen:It's not the sense of theoretical benchmarks to score higher because of these specific optimizationsEin Absatz weiter:However it's quite interesting in which way the newer version of the ATi drivers act regarding to the "optimizations" we mentioned beforeersetzen durch:However, it's quite interesting in which way the newer versions of the ATi drivers behave concerning the two discovered "optimizations".(Im deutschen Original steht was von zwei gefundenen Optimierungen)

gleicher Absatz:so we have no choice and except the standard benchmark score:würde ich zu folgendem umformen:so we have no other choice than accepting the standard benchmark score.except != accept ;)

Btw: Wir sollten uns grundsätzlich darauf einigen, ob wir "nvidia", "nVidia" oder sonst irgend eine andere Variante benutzten und das einheitlich von der ersten bis zur letzten Seite verwenden. Gleiches gilt für "ATI" bzw. "ATi".

edit: Seite 7, zweiter Absatz: wether -> whether
Ausserdem würde ich "whether there exists other..." in "whether there are other" (there are = es gibt) abändern.

gleicher Absatz: we could however detect only one and in addition absolutely minimum differenceklingt so, als ob es ausser dem einen gefundenen Unterschied doch noch eine anderen gibt.
Mein Vorschlag:
we could however detect only one single minor difference.

[mapel110]

freut mich zu sehen, dass ich nicht der einzige bin, der fehler gemacht hat :D

@z-bag

"all the other" mein ich aber auch aus den engl. sprachgebrauch zu kennen. :kratz:

ich hatte den ganzen artikel mehr oder weniger überflogen. ich hätt ihn so online gestellt. jetzt fehlt nurnoch ngglai, der das ganze nochmal übersetzt :D

[Leonidas]

Original geschrieben von GloomY
Btw: Wir sollten uns grundsätzlich darauf einigen, ob wir "nvidia", "nVidia" oder sonst irgend eine andere Variante benutzten und das einheitlich von der ersten bis zur letzten Seite verwenden. Gleiches gilt für "ATI" bzw. "ATi".



Wie schreibe ich in 3DC alles? nVidia und ATi.

[Leonidas]

Alles eingearbeitet. Es gibt im deutschen Original 3 neue Absätze:



Die Aussagen dieses Artikels gelten für folgende Treiberversionen, sofern nicht in Ausnahmen etwas anderes im Text steht: nVidia 44.03, 44.65, 44.67, 44.68, 44.71, 44.90, 45.23 und 45.24 sowie ATi 03.4, 03.5 und 03.6. Original wurde dieser Artikel mit den Treiber-Versionen nVidia 44.65 und ATi 03.6 begonnen und dann mit den Treiber-Versionen nVidia 44.03, 44.67, 44.68, 44.71 und 44.90 und ATi 03.5 und 03.6 gegengecheckt. Mit Update vom 13. August 2003 sind dann zusätzlich noch die Treiber-Versionen nVidia 45.23 und 45.24 gegengecheckt worden, welche aber keine Veränderungen zu den anderen getesteten Treibern aufweisen.




Nachtrag vom 14. August 2003:

Die soeben von nVidia veröffentlichen neuen offiziellen Treiber 45.23 und 45.24 enthalten - nach einem kurzen Gegentest - keine Änderungen, welche diesen Artikel betreffen. Die festgestellte Applikations-spezifische Optimierung von Unreal Tournament 2003 ist jedenfalls zweifelsfrei weiterhin in diesen Treibern enthalten. Der einzige Unterschied wäre, daß es nun - wieder einmal - nicht mehr möglich ist, durch Umbenennen des D3D AF-Testers in "UT2003.exe" die Filterqualität von Unreal Tournament 2003 mit dem D3D AF-Tester darzustellen.

Aus der Diskussion zum Artikel hat sich im übrigen eine Erklärung dafür ergeben (Thx @ BlackBirdSR), weshalb manche ATi-Benchmarks mit dem vollen trilinearen Filter per rTool schneller sind als mit bilinearem/trilinearem Filter per Control Panel: Letzteres schreibt seinen Filter der Anwendung vor, egal was diese anfordert, während rTool seinen vollen trilinearen Filter nur dann liefert, wenn nicht direkt etwas anderes angefordert wurde. Konkret bedeutet dies also, daß das rTool bei einer Anwendung, welche explizit nur einen bilinearen Filter verlangt, auch nur einen bilinearen Filter liefert - eigentlich sollte es ja auch genauso sein.

Wird der Filter allerdings vom ATi Control Panel gesteuert, so liefert dieses auch auf eine klare Anforderung eines trilinearen oder bilinearen Filters immer nur den aufgezeigten bilinearen/trilinearen Mix. Normalerweise sind somit die Benchmarks mit rTool langsamer, weil mit diesem normalerweise voll trilinear Filter gefiltert wird - bei Anwendungen, welche oftmals aber einen bilinearen Filter anfordern, kann sich dieses Verhältnis aber wie beobachtet auch drehen. Solche Anwendungen sind beispielsweise die beiden 3DMarks und Counter-Strike - und genau dort waren die rTool-Benchmarks schließlich schneller als jene per Control Panel.

[Xmas]

Original geschrieben von Quasar
Also a positive result derived from the use of this script does not necessarily indicate an unpermitted* optimization - it only proofs that there is a optimization present - whether permitted or unpermitted can only be proofed by screenshots.
Das Verb ist aber immer noch to prove, nicht to proof ;)



Original geschrieben von Quasar
Und ganz allgemein: Keine Auslass-Apostrophierung in schriftlicher Sprache... "can't" "don't" sind absolute no-nos, auch wenn's vielleicht einige Cowboy-Seiten so machen.
Da muss ich jetzt aber widersprechen. Im absolut formalen Englisch, also Juristisches/Geschäftsverkehr/technische Dokumentation, würde ich zustimmen, aber im "Gebrauchsenglisch" sind diese Varianten absolut legitim.

[Quasar]

Original geschrieben von Leonidas
Mmh. Babylon hat es mir soforr übersetzt. Und illegal oder unfair sind weit von der Bedeutung des deutschen Originals entfernt.

www.dictionary.com, dict.leo.org kennen es nicht, mein Webster's hingegen schon (S.2505, Spalte 3,2. Hälfte). Komisch, aber gut.

Gibt's dann wohl irgendwie doch :)

[zeckensack]

The observations in this article apply to the following driver sets, unless explicitly stated otherwise: nVidia 44.03, 44.65, 44.67, 44.68, 44.71, 44.90, 45.23 and 45.24, and ATi 03.4, 03.5 und 03.6. Initially this article was started using the 44.65 nVidia drivers and 03.6 ATi drivers, the 44.03, 44.67, 44.68, 44.71 and 44.90 nVidia drivers and 03.5 und 03.6* were checked later. August 13 2003 we've also checked nVidia versions 45.23 and 45.24, but these perform just like the other drivers we tested earlier.

*doppelt genannt. "Zuerst 03.6, dann 03.6"
Btw, warum 03.x?

[Leonidas]

Original geschrieben von zeckensack

*doppelt genannt. "Zuerst 03.6, dann 03.6"
Btw, warum 03.x?


Mit ATi 03.4 begonnen. Gefixt.

0 ... 3DC-Spleen -> ATi hatte seine Treiber-Versionen ursprünglich mit einer 0 davor angegeben.

[Quasar]

Original geschrieben von Xmas
Das Verb ist aber immer noch to prove, nicht to proof ;)




Da muss ich jetzt aber widersprechen. Im absolut formalen Englisch, also Juristisches/Geschäftsverkehr/technische Dokumentation, würde ich zustimmen, aber im "Gebrauchsenglisch" sind diese Varianten absolut legitim.

Dem ersten stimme ich zu, es war halt schon spät....

Zum zweiten: Frag' mal die muttersprachigen Dozenten an der Hamburger Uni, die sagen dir etwas anderes...

[Quasar]

Original geschrieben von Leonidas
Nachtrag vom 14. August 2003:

Die soeben von nVidia veröffentlichen neuen offiziellen Treiber 45.23 und 45.24 enthalten - nach einem kurzen Gegentest - keine Änderungen, welche diesen Artikel betreffen. Die festgestellte Applikations-spezifische Optimierung von Unreal Tournament 2003 ist jedenfalls zweifelsfrei weiterhin in diesen Treibern enthalten. Der einzige Unterschied wäre, daß es nun - wieder einmal - nicht mehr möglich ist, durch Umbenennen des D3D AF-Testers in "UT2003.exe" die Filterqualität von Unreal Tournament 2003 mit dem D3D AF-Tester darzustellen.

Supplement as of august 14th 2003:
The most recently released driver sets 45.23 and 45.24 from nVidia do not contain - verified by a quick cross-check - any changes in regard to this article, as the demonstrated application-specific optimizations for Unreal Tournament 2003 are still present. The only difference is that it is impossible now to prove the existence of said optimizations by changing the executables name of our D3D AF-Testing Programm to 'UT2003.exe".

[zeckensack]

Added on August 14 2003:

Judging by a quick test, the 45.23 and 45.25 driver sets just released by nVidia don't change anything that would affect this article. The specific optimizations for Unreal Tournament 2003 we noted are without any doubt still in these drivers. The sole difference is that it's impossible - again - to reproduce Unreal Tournament 2003's filtering quality by renaming D3D AF-Tester to "UT2003.exe".

While discussing the article (thx @ BlackBirdSR) we discovered a reasonable explanation why some benchmarks score higher on ATi cards with full trilinear filtering (set with rTool) than with the regular bilinear/trilinear filtering achieved through the control panel settings: the latter forces its filtering settings on applications, regardless of what they try to select, while rTool will only deliver full trilinear if applications don't try to select something else. In a nutshell, rTool will still allow bilinear filtering if an application explicitly requests it - the way it should be.

[Quasar]

Original geschrieben von Leonidas
Aus der Diskussion zum Artikel hat sich im übrigen eine Erklärung dafür ergeben (Thx @ BlackBirdSR), weshalb manche ATi-Benchmarks mit dem vollen trilinearen Filter per rTool schneller sind als mit bilinearem/trilinearem Filter per Control Panel: Letzteres schreibt seinen Filter der Anwendung vor, egal was diese anfordert, während rTool seinen vollen trilinearen Filter nur dann liefert, wenn nicht direkt etwas anderes angefordert wurde. Konkret bedeutet dies also, daß das rTool bei einer Anwendung, welche explizit nur einen bilinearen Filter verlangt, auch nur einen bilinearen Filter liefert - eigentlich sollte es ja auch genauso sein.

Arising from the discussion already taking place in our forums there is a likely explanation (Thx@BlackBirdSR) as to why some benchmarks do produce higher scores with full trilinear filtering forced by rTool compared to the bilinear/trilinear-mix activated in the control panel: The control panel does force its filter setting over the applications' request whereas rTool makes the driver use full trilinear anisotropy only when the application does not request something different.
This however means in fact, that rTool only uses bilinear anisotropy if the application itself requests a bilinear filter - just the way it was meant to be [played?].

[zeckensack]

If however the filtering is controlled by the ATi control panel it will ignore specific application requests for bilinear and trilinear filters and always deliver the bilinear/trilinear filtering mix we've shown. Usually this leads to lower benchmark scores when using rTool because this way there's full trilinear filtering most of the time - but for applications that do request other filters this situation can reverse, as we've seen. Among these applications are both of the 3DMarks and Counter Strike - these were exactly the ones where the rTool benchmarks finally outscored those controlled through the panel.

[Quasar]

Original geschrieben von Leonidas
Wird der Filter allerdings vom ATi Control Panel gesteuert, so liefert dieses auch auf eine klare Anforderung eines trilinearen oder bilinearen Filters immer nur den aufgezeigten bilinearen/trilinearen Mix. Normalerweise sind somit die Benchmarks mit rTool langsamer, weil mit diesem normalerweise voll trilinear Filter gefiltert wird - bei Anwendungen, welche oftmals aber einen bilinearen Filter anfordern, kann sich dieses Verhältnis aber wie beobachtet auch drehen. Solche Anwendungen sind beispielsweise die beiden 3DMarks und Counter-Strike - und genau dort waren die rTool-Benchmarks schließlich schneller als jene per Control Panel.

If, however, the texture filter is controlled by ATi's Control Panel, it is overridden by it meaning that even if the application request a trilinear texture filter it only receives the pointed out bilinear/trilinear mixed-filter. Normally, that means benchmarks or game tests exhibit slower performance with forced trilinear filtering via rTool but if there is an application that, by default, uses a bilinear filter such as 3DMark2001 or Counter-Strike things change to the opposite as the benchmarks of the above mentioned applications did show.

[Leonidas]

Ich danke allen wie verrückt. Weitere Änderungen sind immer willkommen, aber ich mach die Sache jetzt trotzdem erstmal offiziell.

[Quasar]

Du kannst ja einen kleinen Disclaimer davorsetzen...

This translation has been made possible by our community who wanted me to provide this articles' information to more than just the german-speaking audience. ;)

[zeckensack]

In closing, we'd like to apologize for our language skills. They suck.

*kicher*

[Leonidas]

Passt.

[Quasar]

Da du ja weiter feilen wolltest:

Seite 1:
"contradict this principle and actually circumvent the purpose of a theoretical benchmark."
- contraditc_s_

"any application specific optimization is an "unallowed optimization", an oppinion we strongly agree with."
- opinion



Seite 2:
" since only the power measurement is sufficient here as proof for an unallowed optimization."
- performance check /performance discrepancy

" optimizations are legal and welcome, if they burst the performance without loss of image quality!"
- boost the performance

" Such Optimisations are unallowed"
- optimizations

" performance difference appears only in special flybys, botmatches or timedemos, or if they are common for the whole game."
- only appears [...] or if they do in the whole game.

"despite of one or another outliers"
- outlier

"Those, who think, they can cheat out [...], are wrong"
- Those who think (ohne Kommata)they can counter-cheat [...] are wrong (ohne Komma)



Seite 3:
" from where the performance comes - "
- where the extra-performance comes from (oder)
- where the extra-performance is generated

" that is, whether image quality was sacrificed somewhere."
- streiche "whether" (da müsste ein "or" folgen), setze "if"

"(the former for performance measurement used "Cheat"-Demo)"
- the Cheat-Demo used for performance measurement on the previous page

"with which we could create everytime reproduceable screenshot positions"
- which enabled us to exactly recreate a single screenshot position

"To be honest, there are not many differences visible, even when looking precisely."
- Honestly, even when taking a very scrutinizing looks, there is just not much visible difference

"By the way, this applies to the rest"
- This applies, by the way, to the rest

"seems to cost only very marignal image quality"
- marginal

" 8x anisotropic filter some differences are identifiable"
- can be identified

"In comparison the difference can be seen clearly:"
- In direct comparison the differences are clearly visible:"

"In addition here is another difference comparison from the same demo"
- In addition, here is...

"In this case you will have to take a very close look, to be able to identify the differences."
-In this case you will have to take a very close look to be able to identify the differences.

" Nevertheless, when regarding exactly differences between the two screenshots show up at the edges to the abyss in the center and at the right border of the image."
- Nevertheless differences show up at the edges to the abyss in the center and at the right border of the image when looked upon very carefully.

" However, we do not evaluate them as fairly tragic due to only quite small differences"
- We do not, however, consider these minor differences as very serious degradations of the image quality.

" This results - up to this point of the article - in only a minimum worse overall image quality."
- [...] in only a slightly worse overall image quality.

"However, there are also image quality characteristics (naturally), which cannot be be caught by screenshots,"
- ein "be" weg

"When regarding full trilinear filtering once as 100 per cent performance, for bilinear filtering only 50 per cent performance would naturally be necessary."
- per cent kann man m.E. besser so schreiben: %, oder aber, wie später: percent

" The mip-banding (transition between two MipMaps) is nevertheless still quite well covered by the narrow trilinear range."
- still covered quite well



Seite 4:
" The following figures support this argumentation"
- this line of argument

"Smooth transitions from one mipmap to another"
- into another

"The only question left is whether the investigation of benchmarked results from the three benchmarks"
- The only question that's left is, wheter or not the obtained results from the three benchmarks used - which receive [...]

"Therefor these three benchmarks with application-specific optimizations, again with driver versions 44.03 and 44.90, respectivly with and without AntiCheat script:"
- Therefore [...] respectively

"Thus it can be stated that all said before applies to all drivers from version 44.03 (current official driver) to 44.90 (latest unofficial version)."
- Judging from these results, it's safe to say, that [...]

to be continued...

[Leonidas]

Thx. Eingearbeitet.

[Quasar]

Seite 5:
" It's impossible that the variation of +/- 10 points are measuring tolerances, because of the permanence both 3DMark benchmarks have"
- +/- 10 points are to be accounted for measuring tolerances [...] consistency that both 3DMarks show.

" In addition to that we benched both 3DMarks with newer driver"
- Gibt es "benched" echt? Ich dachte, das wäre eine deutsche Erfindung? vielleicht "benchmarked" oder "tested"

"These results dilute our statment mentioned before. As ATi announced further, there are no application specific optimizations for 3DMark03 within the 03.5 driver version."
- [...]statement mentioned before, since ATi already had announced, that there are no[...]



Seite 6:
"Firstly, a clear bilinear filter shows up here - although ATi's quality"
- In the first place, a clear...

" In order to get the cause for this behavior, we regarded in comparison to the "normal" image an additional purely bilinear filtered image, where definitely all textures are thus only filtered bilinear."
- In order to find the cause [...], we compared an additional, purely bilinear filtered screenshot to the normal image. (Den Rest mit "thus.. " würde ich weglassen, "purely" sagt schon alles aus).

"The difference can be recognized precisely and permits apparently only the conclusion that ATi indeed filters some textures trilinear, others in contrary only bilinear."
- [...] and leads to the apparent conclusion [...] while others (in contrary) are filtered only bilinear.

"up to the most current driver version 03.6 behave identical in the same manner like the screenshots show here."
- entweder "identical" oder "in the same manner" weglassen, da doppelt

" are shown below, where all texture stages are perfectly equal trilinear filtered in the quality filter mode"
- [...]are filtered equally and solidly trilinear in the drivers' quality mode.

" This behavior of the ATi driver thereby applies to all applications if these are using the anisotropic mode."
- "if these are" kann wech

"Similar to nVidia this general optimization of the anisotropic filter is based on the fact that also current ATi graphic chips only have bilinear filtering texture units,"
- that current ATi graphic chips also only have...

" saving enormous computation efforts with bilinear filtering."
- saving [...] ressources. "efforts" macht man, man spart sie nicht.

"Nevertheless, ATi however ignores in this context that there are also different textures (for example detail textures) beside the basis texture, which can show mip-banding and therefore trilinear filtering should be used."
- Nevertheless und however sind zusammen doppelt gemoppelt: however raus und "base textures", oder? (Paar zeilen später nochmal)



Seite 7:
"f a genuine trilinear filter is prefered"
- preferred

" Because when we controlled the anisotropic filter not over the ATi control panel, but over the TweakTools"
- when we enabled [...] not via ATis control panel but used the Tweaktools [...]

"Thus the performance advantages, which ATi gains by this general optimization of the anisotropic filter, can at least be exactly classified for Unreal Tournament 2003."
- [...]advantages gained by ATi's optimization of anisotropic filtering can at least be quantified for Unreal[...]
später das "per cent" nochmal




Seite 8:
"Given that ATi´s general optimizations of the anisotropic filter also apply for all other applications, we ran the former used benchmarks except Unreal Tournament 2003 once again with the 16x anisotropic filter."
- [...] applies to all other[...] the previously used[...]

"The other benchmarks however partly produced very irritating results:"
- "irritating" bedeutet ärgerlich, erzürnend... besser wäre hier "unnerving", wenn man wirklich beim deutsch irritierend bleiben wollte oder aber "confusing", was die Sache m.E. am besten trifft.

" CPU limitation and is therefore not further notable"
-das deutsche "nicht weiter tragisch" wäre hier besser mit "and therefore negligible" getroffen.

"An additional control of real screenshots with the driver version 03.6 brought no new realizations:"
- [...]no new revelations. Oder: "03.6 revealed nothing new."



Seite 9:
" (can easily be proofed by renaming other applications, results in higher performance). To proof"
- wie Xmas schon sagte: "proved" und "prove"

" "technical" problem which showed quiet unexpected another "optimization" of nVidia."
- quite unexpected

" Renaming the actual version 1.1 of the D3D AF-Tester (which was used in the first version of this article) causes nothing - the results stay the same."
- caused nothing [...] results stayed the [...]

"Only as Demirug presented a new version of the D3D AF-Tester, this effect disappeared and it could be determined the image quality of unreal Tournament 2003 after renaming in "ut2003.exe".
- Only after [..] it could be used to determine (besser: evaluate)[...]Unreal Tournament 2003 after renaming the executable to "ut2003.exe"

"Completely obviously nvidia has built in a small security which causes only renaming the D3D AF Tester to "ut2003.exe" not to show this effect."
-It is very obvious that nVidia has built in (besser: integrated) a small security token which prohibits the D3D-AF Test program to show the UT2003-Optimization just by renaming the tool.

" reviewers to proof the bad filter quality under Unreal Tournament 2003 with the D3D AF-Tester (which is btw a tool recommend by nVidia)."
- "prove"[...] recommended. Aber war das nicht das Tool von XMas, was ich in den Reviewers Guides von nV sah?

"This however only besides."
- On an interesting sidenote, Demirug came up (technically seen am besten rauslassen m.E.)

"Performance filter
Trilinear filtering is missing any longer, one"
- is missing still (weiter unten nochmal)

"Vidia uses in 3DMark2001 and 3DMark03 application specific optimizations in the drivers 44.03 to 44.90, which under first benchmark approx. 16 per cent and under second benchmark approx. 62 per cent (!)"
-"per cent" nochmal, und später wieder...

"How much these 57 per cent however without and how much with image quality loss were reached, could with our possibilities not be determined and must remain open."
- To exactly determine how much of these 57% were reached only with loss of image quality and how much were reached with identical screen output is beyond our possibilites and must remain an open question.

" To be proven knew however is that nVidia at least uses a pseudo trilinear filter[..]"
- It was clearly shown, however, that nVidia [...]

"as well as partly MipMaps shifts to the rear in the anisotropic filter and partly the full anisotropic filter does not work."
- [...]partial MipMap shifts [...] and in some cases the full[...]did not work.

"Thus it is considered as proven that the application specific"
- [...] considered proven [...].

" In the drivers 03.5 and 03.6 this optimization is no longer ascertainable, so we can see this case as done."
-[...]is no longer verifiable[...] so we can rest this case.

"Whether disadvantages bring this application-specific optimization in the image quality with itself was not tested, since application-specific optimizations do not have anyway absolutely to be in theoretical benchmarks."
Wheter this application-specific optimization uses degradations in terms of image quality was not tested, since we consider any optimization for a synthetic benchmark an attempt to deceive the public.

" which brings in approx. 20 per cent performance advantage under 16x anisotropic filter in the flyby measurements of Unreal Tournament 2003. In other benchmarks however as well do not fasten as at all."
- which yields an approximately 20% performance advantage under 16x anisotropic filter in the FlyBy-Benchmarks of Unreal Tournament 2003 while producing absolutely no effect in some of the other benchmarks.

"ATi filters the basis texture when anisotropic filtering only trilinear, further textures however only bilinear."
[...]base texture[...] any further textures (stages)[...]

"Disadvantages in the image quality outside of this bilinear/trilinearen combining filter could not be proven."
-[..] quality apart from this bilinear/trilinear filter-mixture[...]

" Nevertheless the question must open-remain whether this general optimization of the anisotropic filter likewise does not represent an unallowed optimization, since the ATi own requirement of the "trilinear filtering" is not finally reached."
- Nevertheless, the question as to whether this general optimization of the anisotropic filter does not represent an unallowed optimization in itself has to remain unanswered, since ATi's postulated "trilinear filtering" is not achieved.

" Also offers here nVidia - except for the determined exception in Unreal Tournament 2003 mind you - a trilinear filter in principle for all texture stages."
Also, nVidia offers here - except for the shown execption in Unreal Tournament 2003 - a normal trilinear anisotropy for all texture stages.

to be continued...

[Quasar]

" And completely generally keep the eyes open far for the existing image quality - particularly with new drivers - in the future rather examine twice, than assuming blindly everything will already fit somehow."
- And, for the future in general, keep a keen eye on for the delivered image quality - especially with new drivers - and rather double-check on your screen output instead of leaving it all up to the GPU vendors.

" this attacks however the static image quality only marginally, but replaced "only" the trilinear filter by a pseudo trilinear."
-this, however, degrades the static [...]

" That is further an unallowed application-specific optimization, but on the other hand this is also - roughly compared - the normal condition at ATi"
- Furthermore, this is an unallowed [...]

" the normal condition at ATi under the anisotropic filter. ATi also replaces the trilinear filter in this case, at ATi simply by a bilinear/trilinear combining filter."
- the normal condition with ATi und anisotropic filtering, as they are replacing the trilinear filter with a bilinear/trilinear-combo by default.

"Both chip manufacturers have optimized down the image quality so far that the image quality is nearly again alike ironically."
-Ironically, both chip manufacturers have de-optmized the image quality so far, that they find themselves facing each other at knee-level again.

"However as a result of this comparison naturally again some other problems arose:"
-As a result of this comparison, however, some other problem arose:

"The whole thing reminds a gordian knot, which you can smash but not solve."
-[...]cut but not untie.

"Surely it would be better both graphics chip developer leave the quality modes in their drivers untouched and any optimizations whether application specific or general would be build into specially modes."
-Surely, it would be desirable if both graphics chip developerts left the quality modes in their drivers untouched[...]

"Probably nobody has something against optimizations, which have hardly a visible quality loss and bring in a performance advantage between 20 and 57 per cent."
- Nobody would likely be argueing against optimizations[...] and yield a performance[...] percent (oder:%)

"Only the quality mode should not use such optimizations, if both chip developers want to assume to be outriders of image quality."
- [...] quality modes[...]want to assume the role of outriders of image[...]


ich glaube, ich habe fertig das gröbste...

[Leonidas]

Original geschrieben von Quasar
- "prove"[...] recommended. Aber war das nicht das Tool von XMas, was ich in den Reviewers Guides von nV sah?



Kann ich nicht beantworten.

[Leonidas]

Thx. Eingearbeitet.

[Exxtreme]

Also die Leute im R3D-Forum nehmen diesen Artikel positiv auf. :)

http://www.rage3d.com/board/showthread.php?s=&threadid=33704460

[Exxtreme]

Excellent article on NVidia and ATi Optimisations (http://www.beyond3d.com/forum/viewtopic.php?t=7456&start=0&postdays=0&postorder=asc&highlight=) @ B3D. :up:

Leo, die Leute im B3D- und R3D-Forum meinen, daß du den Artikel ergänzen solltest.
Und zwar erlaubt der ATi-Treiber die volle trilineare Filterung wenn man im Controlpanel "Standardeinstellung" einstellt und den AF im Spiel selbst regelt.

[Leonidas]

Original geschrieben von Exxtreme
Excellent article on NVidia and ATi Optimisations (http://www.beyond3d.com/forum/viewtopic.php?t=7456&start=0&postdays=0&postorder=asc&highlight=) @ B3D. :up:

Leo, die Leute im B3D- und R3D-Forum meinen, daß du den Artikel ergänzen solltest.
Und zwar erlaubt der ATi-Treiber die volle trilineare Filterung wenn man im Controlpanel "Standardeinstellung" einstellt und den AF im Spiel selbst regelt.


Sollte ich sicher ergänzen. Text würde ungefähr so aussehen:


Es gibt zur ATi-Optimierung unter dem ansisotropen Filter den berechtigten Einwand, daß man hier diese Optimierung im von uns gewählten Testspiel Unreal Tournament 2003 auch mit Boardmitteln umgehen kann. Dazu schaltet man den anisotropen Filter im ATi Control Panel aus und diesen dann wieder in den ini-Dateien des Spiels an (UT2003.ini, LevelOfAnisotropy). Allerdings hat dies zweierlei Nachteile: Erstens hat es sich eingebürgert, den anisotropen Filter über Control Panel oder Tweaktools einzustellen, nicht aber im Spiel. Dies hat auch den Vorteil, daß man alle Spiele generell mit einer AF-Einstellung spielen kann. Für einen Spieler, welcher beispielsweise generell alles mit 8xAF spielt, wäre es umständlich, extra für Unreal Tournament 2003 AF im Control Panel wieder zu deaktivieren und für andere Spiele wieder zu aktivieren.

Der zweite Nachteil ist dann noch viel schwerwiegender: ATi optimiert schließlich den anisotropen Filter für alle Direct3D-Games. Selbst daß es für unser Testspiel Unreal Tournament 2003 eine Lösung mit Boardmitteln gibt, kann dies nicht darüber hinwegtäuschen, daß diese Lösung mit den wenigsten Direct3D-Spielen funktioniert, da die wenigsten eigene Einstellungen für den anisotropen Filter mitbringen. Insofern kann die Applikation-Einstellung des ATi-Treibers auf keinen Fall als generelle Lösung für die ATi-Optimierung des anisotropen Filter angesehen werden.


Kann das jemand übersetzen? Danke.

[bluey]

meiner meinung nach sind Optimierung bestimmter Engines gar nicht mal sooo schlecht. Dann hält zumindest meine Grafikkarte länger ;D . Nur sollten diese Manuel, aus sportlicher Fairness abschaltbar sein.

[Leonidas]

Original geschrieben von DocterD
meiner meinung nach sind Optimierung bestimmter Engines gar nicht mal sooo schlecht. Dann hält zumindest meine Grafikkarte länger ;D . Nur sollten diese Manuel, aus sportlicher Fairness abschaltbar sein sein.


Kein Problem mit dieser Meinung. Optimierungen sind immer gut für ältere Karten. Nur eben sollte man sie auf Wunsch - wie Du schon sagtest - auch abschalten können.

[aths]

Original geschrieben von Leonidas
Kein Problem mit dieser Meinung. Optimierungen sind immer gut für ältere Karten. Nur eben sollte man sie auf Wunsch - wie Du schon sagtest - auch abschalten können. Ihr werft hier die Begriffe "Optimierung" und "Cheat" durcheinander :)

Wer Füllrate sparen will, kann in UT den trilinearen Filter ganz legal abschalten und damit die Leistung auf Kosten der Qualität optimieren.

[zeckensack]

Original geschrieben von Leonidas
Es gibt zur ATi-Optimierung unter dem ansisotropen Filter den berechtigten Einwand, daß man hier diese Optimierung im von uns gewählten Testspiel Unreal Tournament 2003 auch mit Bordmitteln umgehen kann. Dazu schaltet man den anisotropen Filter im ATi Control Panel aus und diesen dann wieder in den ini-Dateien des Spiels an (UT2003.ini, LevelOfAnisotropy). Allerdings hat dies zweierlei Nachteile: Erstens hat es sich eingebürgert, den anisotropen Filter über Control Panel oder Tweaktools einzustellen, nicht aber im Spiel. Dies hat auch den Vorteil, daß man alle Spiele generell mit einer AF-Einstellung spielen kann. Für einen Spieler, welcher beispielsweise generell alles mit 8xAF spielt, wäre es umständlich, extra für Unreal Tournament 2003 AF im Control Panel wieder zu deaktivieren und für andere Spiele wieder zu aktivieren.
Regarding ATI's anisotropic filtering optimizations, we should add that in Unreal Tournament 2003, the game we selected four our tests, that these can be circumvented without resorting to third party tools. To do so, anistropic filtering must be disabled in ATI's control panel and set it in the game's ini file (UT2003.ini, LevelOfAnisotropy). This approach has its caveats: first, it's common practice to control anisotropy through the control panel or tweak tools, not in games. This also allows playing games with one AF level. Gamers wishing to play everything at 8xAF, it would be an inconvenience to turn it off in the control panel before playing Unreal Tournament 2003 and to reenable it for other games.
Der zweite Nachteil ist dann noch viel schwerwiegender: ATi optimiert schließlich den anisotropen Filter für alle Direct3D-Games. Selbst daß es für unser Testspiel Unreal Tournament 2003 eine Lösung mit Boardmitteln gibt, kann dies nicht darüber hinwegtäuschen, daß diese Lösung mit den wenigsten Direct3D-Spielen funktioniert, da die wenigsten eigene Einstellungen für den anisotropen Filter mitbringen. Insofern kann die Applikation-Einstellung des ATi-Treibers auf keinen Fall als generelle Lösung für die ATi-Optimierung des anisotropen Filter angesehen werden.Second, and more significant: ATI optimizes anisotropic filtering for all Direct3D based games. Even though there's a way around that for Unreal Tournament 2003, this won't improve the situation for the vast majority of Direct3D based games, where there simply aren't any anisotropy controls. This workaround for one game can't be regarded as a general solution for ATI anisotropy optimizations.



Nicht ganz exakt, aber vom Sinn her sollte das IMO passen :)

[Leonidas]

Super-Thx.

[zeckensack]

Leo, noch ein paar kleine Korrekturen :sulkoff:

Original geschrieben von zeckensack
Regarding ATI's anisotropic filtering optimizations, we should add that in Unreal Tournament 2003, the game we selected four our tests, <"that" streichen> these can be circumvented without resorting to third party tools. To do so, anistropic filtering must be disabled in ATI's control panel and set <"it" streichen> in the game's ini file (UT2003.ini, LevelOfAnisotropy). This approach has its caveats: first, it's common practice to control anisotropy through the control panel or tweak tools, not in games. This also allows playing games with one AF level. For gamers wishing to play everything at 8xAF it would be an inconvenience to turn it off in the control panel before playing Unreal Tournament 2003 and to reenable it for other games.

Second, and more significant: ATI optimizes anisotropic filtering for all Direct3D based games. Even though there's a way around that for Unreal Tournament 2003, this won't improve the situation for the vast majority of Direct3D based games, where there simply aren't any anisotropy controls. This workaround for one game can't be regarded as a general solution for ATI anisotropy optimizations.

[Leonidas]

Done.

[Quasar]

Noch einige Kleinigkeiten:

Regarding ATI's anisotropic filtering optimizations, we should add that in Unreal Tournament 2003, the game we selected for our tests, these can be circumvented without resorting to third party tools. To do so, anisotropic filtering must be disabled in ATI's control panel and set in the game's .ini file (UT2003.ini, LevelOfAnisotropy). This approach has its caveats: first, it's common practice to control anisotropy through the control panel or tweak tools, not in games. This also allows playing games with one AF level. For gamers wishing to play everything at 8xAF it would be an inconvenience to turn it off in the control panel before playing Unreal Tournament 2003 and to reenable it for other games.

Second, and more significant: ATI optimizes anisotropic filtering for all Direct3D based games. Even though there's a way around that for Unreal Tournament 2003, this won't improve the situation for the vast majority of Direct3D based games, where there simply aren't any anisotropy controls. This workaround for one game can't be regarded as a general solution for ATI anisotropy optimizations.

[zeckensack]

Oh Gott :o

Danke :wink: