Compendium of 3D knowledge

[BoardBonobo]

Has anyone thought about putting together a site that explains the differences in approach to 3D rendering that todays cards use? A site that that separates fact from fiction, and truth from lies? And to dispell the FUD about different manufactures implemenations of said features.

With all the debates that occur about differences in FSAA, MSAA, RIP mapping etc etc. It can get very confusing for someone new to the scene to get to grips with. And the majority of articles don't make much sense or contain so many white lies, that the newbie can't help but become confused.

A compendium of Beyond3D knowledge presented in a non biased way with links to more technical documents for those who are really interested. IMHO, it would be a very good idea that may save hours of recursive debate about topics that have been doing the rounds for months (even years in some cases...).

Anybody willing to give it a go?

 

[Dave Baumann]

We keep meaning to do many of these things but in the end it comes down to a question of time and resource. We do have a few articles that we keep threatening to get round to...

 

[MuFu]

That would be great. The amount of knowledge on this board is very refreshing compared to some others!

MuFu.

 

[Ailuros]

Let's see when we'll see that promised Parhelia review at B3D first

Of course would a wide collection of 3D related articles be a superb idea for B3D. Maybe some time you'll find the time guys between all and each's obligations. Someone bug Mr.Beets to do a Multisampling on Tilers article heh.

In the meantime for those who do read german there are a couple of interesting articles over at 3Dcenter.de. Reminds me of how much I miss 3dconcept at times....

 

[BoardBonobo]

I just thought than since Beyond3D has a lot of members, most of whom are well versed in the arcane art of 3D, that it could be made a community effort. We could even have a section here for proof reading so to speak.

It would be quite a fast parallel production line. Mind you it might scare a few companies who would probably prefer people to be uncertain about things like this , watch out for the matrix mafia...

 

[Ailuros]

I guess I didn't understand your initial proposal then correctly. That's an even more excellent idea quite frankly. I just don't know if the available webspace is sufficient, since I wouldn't figure that just a few would turn up with articles if called for.

 

[BoardBonobo]

Couldn't people donate web space from their home accounts? I know I've got a few meg spare that I wouldn't mind donating.

Also I have access to a large number of servers @ work and if I ask really nicely I might be able to squeeze some space and bandwidth out of those...

 

[LittlePenny]

I don't know what kind of information you are seeking specifically. www.pvrdev.com, http://www.ati.com/developer/techpapers.html,http://developer.nvidia.com/view.asp?PAGE=papers, and http://developer.matrox.com/docshome.cfm?CFID=343685&CFTOKEN=98715994&q=3 Have more information than you can shake a stick at. Though I do wish I knew someone with the old 3dfx whitepapers.

 

[pascal]

It could be done like what people do at IETF with RFCs

 

[jvd]

well

i'm not sure how to access it but you can have whatever
room cablevision gave me to use as you see fit .

 

[Simon F]

I don't know what kind of information you are seeking specifically. www.pvrdev.com, http://www.ati.com/developer/techpapers.html,http://developer.nvidia.com/view.asp?PAGE=papers, and http://developer.matrox.com/docshome.cfm?CFID=343685&CFTOKEN=98715994&q=3 Have more information than you can shake a stick at. Though I do wish I knew someone with the old 3dfx whitepapers.

I've got their texture compression white papers.. somewhere....

Back on the orignal topic... there is a lot of info to be found in good textbooks eg:"Computer Graphics Principles and Practice." ,
"Advanced Animation and Rendering Techniques" and, no doubt, the soon to be released 2nd edition of "Realtime Rendering". I have a worry that a lot of web pages get written before the author gets a solid grounding in 'the art'.

Simon
(Edit: Noticed I used the verb practise instead of the noun)

 

[Brutal Deluxe]

"Computer Graphics Principles and Practise."

I plan on taking a couce on computer graphics as part of my CS degree. The cource uses this book in the process, so I figure that cource would be a good place to start? Huh?

 

[BoardBonobo]

Text books are always a good place to go if you want a truly *academic* discussion of 3D principles, but they aren't any good for a comparison of current technologies in practice.

Also they tend to cost a small fortune, I guess my budget per year for text books is in the five to six hundred pound range. Luckily I can get alot through work and earn enough to buy the rest myself.

But text books don't answer the common questions that people raise on forums such as this and, for the main part, aren't written for the casual browser. A site that is accessible for the majority of casual browsers and explains things in plain english and dispels the myths and BS would be a damn fine addition to a forum such as this IMO.

Maybe we should have a vote on it. Just to see how many people would be truly interested in such a beast.

 

[Dave Baumann]

A site that is accessible for the majority of casual browsers and explains things in plain english and dispels the myths and BS would be a damn fine addition to a forum such as this IMO.

There is a note of caution to be had here.

Much of the fundamentals of 3D rendering and their implementation in hardware are well documented and in many cases it this could be done. However 3D hardware is getting both more and more complicated and at the same time more and more general (if that makes sence) - it used to be the casr that hardware vendors would talk to quite a low level of detail about the hardware so it was pretty easy to gain an understanding of what's really occuring, however as the chips get more complex and their functionality gets more general they are telling people, even developers, less and less about whats going on at a hardware level. To this end its going to be increasingly difficult to truely be able to dispell many 'myths' that may be occuring at a hardware level.