What about P benchmarks ???

[Umfriend]

Wasn't Dr. Mordid the one who had "G400 + (classified) in his signature? What has happened to that? Was the RT-2000 classified?

[Dr Mordrid]

Nope. (classified) just moved to another system.

Dr. Mordrid

[Dr Mordrid]

I use Premiere with the Matrox RT products because it's required. The realtime feature controls are implemented by a Premiere plugin and this is the only editor they support in Windows.

Premiere works quite nicely with the RT cards as their extra features tend to smooth over some major weaknesses of Premiere itself, one of which is that IT'S SLOW AS MUD.

In non-realtime editing I use Uleads MediaStudio Pro 6.5 and most of their other editing software along with Boris RED, SpiceMaster & other plugins. MSPro 6.5 is quite fast to work with and produces very nice results.

Dollar for Dollar MSPro 6.5 a better buy than Premiere, is easier to use, is MUCH faster and will work on a wider variety of systems. It also will work with both DV type 1 and DV type 2 files. Premiere does not support type 1 properly and is nothing short of a total memory HOG.

I haven't had the need for or interest in Avid Xpress DV. I have enough learning curves to deal with right now due to having several betas running at once.

Dr. Mordrid

[Greebe]

I take it you're putting on the hip waders before entering the Lab aye Doc?

[SpiralDragon]

Thanx Doc.

i have been working on Premier since 1996.... also worked on Edit ... Media 100 and the various Avid offerings... to be honest the smoothest of them all so far is AVID IMHO..... but my problem lies in the lack of hardware offerings for the DV version and i am not about to go spending $15,000 for the regular express version (the DV version is $500 for students) ..... but to be honest having now worked 10 films on AVID doing even a simple job in premier makes me feel bad .. then again with a product from M like the RT2x00 is a compeling reason to move back to premier.... i am curious cause i am about to spend a large summ of money on a new system and am at a loss as to what dirction i should go

[Dr Mordrid]

Yean Greebe. Hip deem in Premiere generated mud at times, but it's fun when something breaks the monotony

Dr. Mordrid

[Edward]

Originally posted by Dr Mordrid
Benchmarks don't mean much when NVIDIA cheats their way through them

If you run 3DMark2001SE on an NVIDIA card and omit the title screens the score drops. The difference is huge; as in 600-1000 points using the Ti4600.

Wonder why? Read this thread;

http://www.nvnews.net/forum/showthread.php?threadid=15459

Maybe the NVIDIA drivers are doing a sneaky? Seems possible as MadOnion is "researching" the possiblity that NVIDIA's drivers are the root of this "issue"

Dr. Mordrid

Thats defeniltly not true, at least not so much, I lost only 350 points with a GF4Ti4600 and 500 points with a R8500128 mb, there are other ways to cheat, look at the smoke en lightflares in some drivers and mipmap borders

[Dr Mordrid]

So because you lost less than with a Radeon there isn't a problem? Even when other boards don't share this "feature"?

Hmm......

Dr. Mordrid

[Edward]

Originally posted by Dr Mordrid
So because you lost less than with a Radeon there isn't a problem? Even when other boards don't share this "feature"?

Hmm......

Dr. Mordrid

Who said it was not a problem I'm only pointing out that it's not a Geforce(only) fault, and probely a 3DMarkfault.

Btw, madonion has released a patch to solve this issue and guess what, the scores have not dropped

But every company is focussing on benchmarkprogs, you can call it optimizing or cheating.

ATI uses a certain string to optimize all Q3-engine games, I disable them all the way up until now.

I rather have some lower fps and top noth IQ instead of banded fog and banded transparant textures and few more fps.

I'll be surpriced if Matrox don't walk that way

[Mendini]

bug fix

here's what I found on http://www.rivastation.com

6/13/02 3D Mark 2001 Bug

Die Website The Inquirer berichtet von einem Bug in 3D Mark 2001, der NVIDIA basierten Grafikkarten zu leicht besseren Ergebnissen verhilft:

Virtually all graphics card testers rely on this suite of tests to compare the performance of competing cards. And the graphics card makers are well aware of this.

But Nvidia's high end graphics chipsets produce better results in the 3D Mark 2001 than they ought to, although the firm claims that this is a bug in the driver which it is fixing today.

Documents THE INQUIRER has seen show that running the tests without the 3D Mark test screens on all nVidia's high-end graphics cards, produces lower scores in the tests than with the 3D Mark splashes turned on.

Because it's a problem in the benchmarking software it seems that Nvidia is now pulling out all the stops to fix the bug, which produces different results than in real world applications.


So it seems like Nvidia is woking on a fix atm

[Edward]

Why border what nV does, I want Parhelia benchmarks for god sake , come on, this is where we're all waiting for, Unreal 2002 perf test, Morrowind, Aquanox, with aniso and fa, there we need it for.

[notagain]

no official bench will be publish within 5 days
and there for anything that could be say would be 'craps'

so let's do some bet then: I 'm viewing in my crystal ball

aquanox 1024*768 with 16*FAA and all the possible rest at max @60 fps

[Kenshin]

lol all compagny optimise their driver for bench especially for 3dmark
Ati does the same thnik with special driver who deterior quality
u are funny
(sorry for my bad english i'm french)

[Marshmallowman]

I wish there was a benchmark where you set a frame rate(eg 85fps), and then you count the missed frames...hence find the frame rate the card can support.

The way I see it on my geforce is the fps goes through the roof when nothing is happening(whats the point), and then when the shit hits the fan it stalls...which is BS because the non intensive stuff boosts your reported FPS so much it hoodwinks users into thinking its faster, but in REAL terms its NOT.

[j3martin]

There is some info on www.hardocp.com. Apparently a German website posted their Parhelia numbers before they were supposed to. They've apparently taken the numbers down but HardOCP reports on what was seen.