Closed GoogleCodeExporter closed 8 years ago
[deleted comment]
[deleted comment]
[deleted comment]
[deleted comment]
[deleted comment]
Normal behaviour is for the Red Dog to only flash white when hit (from around
1:12):
http://www.youtube.com/watch?v=4Pt0FXCQXuk
If you look closely the white flash should fade smoothly through grey scales
before the Red Dog reverts back to it's normal colours.
Original comment by matbur...@gmail.com
on 24 Aug 2010 at 3:13
I'll look at it
Original comment by Mr.PsyMan
on 30 Aug 2010 at 8:42
Ok.
I think this is bugs of “intensity vertex”.
Original comment by N2008eo...@rambler.ru
on 30 Aug 2010 at 9:20
Looks like it was an issue with pre-calculated intensity. Please test if r62
broke things that worked with r61.
N2008eogen: You were right! How did you know?
Wish I also knew before tracing the issue... :(
Original comment by Mr.PsyMan
on 31 Aug 2010 at 5:59
Ok.
----About "You were right! How did you know?" -----
This is long story...
.... One good man told me about that problem when I asked him. His nickname is
Wind. He is Demul's autor. And he is first man who fixed this issue on his
emulator!!!
NOTE: I allways known what this problem exist for all emulators of Dreamcast!!!
Very long time ago this bug was not interesting for devs. I've asked many times
but nobody answered me. Once, on the forum, Wind was interested by this
problem. He says what never known about this bug. After testing he just fixed
it! I asked him why this bugs is happen?
He says: “This problem with “intensity vertex”
If I undestand him correctly:
Formula must be just like that:
Color = basic color * some coefficient
But Dreamcast have support 2 colors( diffuse & specular) on vertex--- > From
experiments they found:
ONLY diffuse = basic color * some coefficient AND “specular” must be =
“basic color”
He say, it’s easy to fix!
Wind is great man!
Original comment by N2008eo...@rambler.ru
on 31 Aug 2010 at 6:34
Indeed. My "fix" is wrong by the way. But since I have more info now, it will
be shorten out. You can thank Wind for that... and I have to do the "fix" an
actual fix.
Original comment by Mr.PsyMan
on 31 Aug 2010 at 7:11
[deleted comment]
Well, it looks like that approach was wrong after all... Both color types
should use that "coefficient".
I disabled that "coefficient" for both vertex colors on r62 (which is wrong)
and according to what you said Wind disables that "coefficient" for just one of
the vertex colors (which is also wrong).
In short, the problem is not on “intensity vertex”. The problem is on
"intensity clamping".
Related issues should be fixed on r63. :)
Original comment by Mr.PsyMan
on 31 Aug 2010 at 11:45
Ok! I tested it! Nice work! Wonderful!
Original comment by N2008eo...@rambler.ru
on 1 Sep 2010 at 5:28
Original issue reported on code.google.com by
N2008eo...@rambler.ru
on 2 Jun 2010 at 7:48Attachments: