32-Bit worse than 16-Bit, broken display output...

[Colourless]

32-Bit worse than 16-Bit... you are probably thinking I'm on crack, but let me tell you it's true, and it's just a symptom of what turns out is a very bad hardware bug in the Voodoo 5 6000.

So, do you want to know what exactly this problem is?

If so, you might want to read this 1800 word essay that I wrote about this rather interesting and worrying problem I found with Voodoo 5 6000s. (hense my topic from yesterday about RGB signal outputs).

Head over to my website and read it right now (or later if you must )
http://www.users.on.net/triforce/illuminate/v56kprobs.html

You might be amazed.

-Colourless

 

[Fuz]

I am not sure if you have already told us or not, but how on earth did you get your hands on one of those?

Also, what the hell you doing with a 14" monitor? That is the most odd pairing ever! Voodoo 5 6000 with a 14" monitor!

 

[BRiT]

I am not sure if you have already told us or not, but how on earth did you get your hands on one of those?
!

From what I recall, he was given one from 3dfx/Quantum3D for his idea/creation of what became performance-trilinear filtering that is possible when using 4 or more VSA chips.

--|BRiT|

 

[Ailuros]

Colourless,

I'm fairly sure you tried other games too outside of q3a and Deus Ex. The latter is really a bad case example investigating such an issue since it has a known gamma problem.

 

[darkblu]

thanks for sharing your sophisticated investigation, colourless, good job!

 

[Althornin]

nice job, very interesting.
maybe a reason it was delayed, etc.

And yes, Colourless got the V56k for coming up with the idea of performance trilinear.

 

[Colourless]

Colourless,

I'm fairly sure you tried other games too outside of q3a and Deus Ex. The latter is really a bad case example investigating such an issue since it has a known gamma problem.

Why do you think I wrote my own program to test for the problem? It wasn't exactly possible to know what was going on otherwise. I noticed a problem in the games, and then created my own program to examine exactly what was going on. DeusEx's big problem is just that is it so dark. Quake 3's problem is it plays with the Gamma Table to create Overbright bits. Both of these games show the problem more clearly than others (i.e. other games wouldn't need such high gamma correction). If there is no gamma correction, the problem isn't noticeable.

And yes, the card got given to me as a gift because of the performance trilinear stuff.

 

[trollop]

o_o

There are some Voodoo 5 5500 drivers out there that have Analog SLI as an option, and it causes that exact same thing to happen when you turn it on. Colors are more vivid and lifelike but they also often bleed and cause all sorts of horrid banding and the like with FSAA on. I believe I have them, if anyone is interested.

 

[Colourless]

Most interesting. On the x3dfx board i speculated that it might be occuring due to Analog SLI. I can't exactly test it though since I don't have a 5500 and digital SLI doesn't work on 6000s.

 

[Nagorak]

How does the V5 6K stack up against the Nvidia cards of the day? I'm curious to know how it would have done had it been released.

 

[Typedef Enum]

http://www.ve3d.com/hw/previews/v5_6000/5.html

2xFSAA
10x7: 69.9

4xFSAA
10x7: 35.7

Recall they ended up doing the 'world exlusive' on Quantum drivers, and not tuned for 3D gaming.

If you were to compare it to the competition...Recall, this preview happened in December 2000, but it was expected that the V5-6000 would 'probably' be released the Summer of 2000 (at least, according to their PR people). It would have gone to battle against the GTS Ultra.

Again, totally unscientific (different configurations, etc)...but nonetheless...

http://www.nvnews.net/previews/geforce2_ultra/fsaa.shtml

1.5x1.5 FSAA
10x7: 52.8

2x2 FSAA
10x7: 32.6

In the end....even had the V5-6000 claimed the performance title, it would have been extremely difficuly to sell it based on the relatively low performance numbers. I don't know how much more the 3dfx team would have squeezed out of the thing that the Quantum drivers didn't...At a cost of $600, the numbers were dissapointing, particularly given all the hype surrounding this board.