3dc support petition for everquest 2

Why is it that only now everyone is going crazy on normal maps compression, seeing we have had a standard for bloody years (DXT5)...

 

Perhaps if we knew the answer to that "the universe would be instantly replaced by something even more bizarre".

Besides, the compression rate of either is hardly outstanding. <shrug>

 

Hey can we have a petition to never release everquest instead They can keep putting new features in forever and I won't have to pay them a monthly fee ever.

 

I want the Bunny Dimension!

 

[Devil's advocate mode] Which will be supported by how many vendors? Why not ask for dxt5 and then, if the developer has time, 3dc?[/D.A.M.]

 

Because N, "the number of systems supporting DXT5", is very much greater than M, "the number of systems supporting 3DC", while the additional quality achieved by 3DC over DXT5 is not that significant.

Therefore, on average, DXT5 is the better looking system. <shrug>

 

While number of cards supporting SM 3.0 is vastly superior to the number of cards supporting SM2.0 ?

 

All I was arguing was that if you are going to have compressed normals and were going to support only one of either DXT5 or 3DC, then the former is the logical choice.

It is, of course, preferable to have both - development resource permitting - and there'd probably be relatively little extra work to support both <shrug>

 

Hmm, what happend if the developer just compress their original normal maps, which will degrade quality and use 3dc as performance improvement feature instead of quality improvement feature ?

 

I agree that 3dc offers quality improvements but I don't see how it would improve performance. Both, AFAIU, are 8bpp formats, the only difference being that DXT5 produces 4 channels, whereas 3dc produces 2 which are therefore more accurate.

 

I wasn't comparing it to DXT5, I was just comparing it to uncompressed normal maps in term of quality and performance.

My point was if you petition the developer, you know there are chance they would just compress the normal maps that they have in hand and not bother about making normal maps of higher resolution.

But I didn't know that 3dc doesn't offer performance improvement over DXT5, from the preview and info of 3dc, I thought it does. Well you learn something new everyday. Thanks Simon F.

 

Sorry, my misinterpretation.

I guess that rather depends on how the normal map was generated in the first place. If they used the technique of taking a very detailed, high polygon model to produce the normal maps, then it should be simple enough, but obviously there are still data management issues to contend with.

OH! Don't take my word for it. It was just my impression from comments on this forum!!!

 

So in the end this fabled 3Dc is just an overhyped, overstimated, old, used-by-no-one feature... Yay...

 

No, I wouldn't want to imply that at all.

3dc does, indeed, give a better way of storing compressed normals than the DXT5 trick. My point was that if you are going to support 3dc then, for the sake of your customers, you probably should also do the DTX5 version. Alternatively, you could use methods with even higher levels of compression.

 

^^ My point exactly.

 

Don't say that until you've seen my upcoming 3Dc demo.

 

Are you sure about that? Was that really E3 2004?