Fast Extend library — Thread 3
Blitz3D Forums/Blitz3D Userlibs/Fast Extend library — Thread 3
<- Previous topic@Sph!nx: I'm not use PayPal, sorry... (Plimus only)
I bought Blitz3D SDK , and was wondering, when will the other libs, be available for this package. I am really interrested.
@LordShion: FastExt lib for Blitz3D only. It is very difficult to make such library for SDK, and demand very low.
And small demo for you from my friend MadMedic -
Pseudo PSSM (shadow mapping). Please, test it... :)
It is required FastExt.dll (trial or retail), place dll nearby PSSM.exe
ohh..interesting..something i was crying for :)
..Mikhail..this just Plain B3D and fastExt?? Man...may I see source..please!...ahh
Well, Plimus does use pay pal. The problem is, all of my email accounts gets rejected.
Gonna try one more email. If not, I will not be able to get the libraries... :-(
Thanks anyway, Mikhail!
EDIT :
Okay, I can probably use my brothers email to order this! Yay! :-D
Does Plimus also handles the email list for future updates or is that something you do yourself? If so, would you be able to change the email I use with purchase to my own?
TY!
OMG shadow mapping loook brilliant!
@Sph!nx
mikhail will give you a link, then you will be able to download your future updates there.
Ah, I understand now! Thanks! Got a correct email now too, so I will order soon!
good work!!!
for animate model??
..Mikhail..this just Plain B3D and fastExt?? Man...may I see source..please!...ahh
Yes, it's native B3D + FastExt. Sources — no, sorry, it's pre-alhpa demo without framework and optimizations, only for testing.
for animate model??
Animate model supported, but as caster only (not receiver). This problem can be solved in the future...
..uhh..cmon man..give it to me.. :) .. is it going to be part of next FastExt update so we can put our dirty hands on it or what?
Hi to all
I'm totally new in this stuff!
so, where can I start?
Something to read, any a tutorial or doc?
I just want to improve blitz, I saw the demos and whatching and reading you, I'm totally lost!
I don't want to waste your time with fool's questions so,
I know that I have to read some (lot I think), but again, where to start??
thanks
Juan
You can start by downloading the trial version of FastExt and looking at examples. Then if you like it, you buy it. It's simple.
Thank's
I did that!
But I ralize that I need some background to understand the whole thing.
Juan
New PSSM demo!Also it is required FastExt.dll (trial or retail), place dll nearby PSSM.exe
Seems to work fine, but I just get around 120 fps, and I have a ATI HD4850 card, which is latest generation mid/hi end card. So, performance doesn't seem to be great.
100 FPS here with a 8800 512mb but you have to add an option to hide blitz text for real results.
94FPS on nVidia 7600GT 256MB VRAM .. niceee .. selfshading over characters will be kicker..
For all: No need to hurry :) This is a syntetic test, acceleration of videocard not used (temporarily), whole work executes processor.
Nice :) Your FastLibs are revolution in b3d scene.
Where can I find the source code of ppssm3 ?
@Barton:
Sources — no, sorry, it's pre-alhpa demo without framework and optimizations, only for testing.
..Mikhail..regarding animated models, is it supported both MD2 and B3D??
Demo has B3D.
MD2 will be megacool..hundreds characters with ease with shadows and stuff..
First time I can see this and looks amazing!!
Do you plan to release this as a new extension or as part of Fastext, Mikhail?
regards!,
It will be a FastExt part + a framework on a Blitz basic...
May be it will turn out nothing, you early ask such questions.
amazing... do you have a date? :P
regards,
really cool!
i have found a bug in fastlibs:
DoF does only work when CameraViewport = Screen-/WindowResolution.
If you make a splitscreen using cameraviewport, all the DoF FX is gone.
When I use a resoultion lower than 800x600, I get strange artifacts like if post process layers are bigger than screen.
DoF does only work when CameraViewport = Screen/WindowResolution.
It's not bug! Viewport of the camera must have a size — width=GraphicsWidth() and height=GraphicsHeight() and use RenderPostprocess function after RenderWorld only for DOF postprocess fx. Whole two restrictions for correct work. This restrictions is connected with DirectX7, sorry.
If you want splitscreen - use CopyRectStretch function for resize.
When I use a resoultion lower than 800x600, I get strange artifacts like if post process layers are bigger than screen.
You can use other quality for Fx, which use blur.
quality% - blurring quality (values 0-3)
0 and 1 - fast method (0 - low quality, 1 - high quality) with artifacts on NVidia videocards
2 and 3 - slow method without any artifacts (2 - low quality, 3 - high quality)
edit: Other postprocess fx can't show artifacts.
I find it strange that Plimus don't take payments that utilise AOL email addresses:
We do not accept email accounts from free email services such as hotmail.com, yahoo.com, etc
Hi!
Another PPSSM test (DirectX7!)
p.s. Thanks for my friend MadMedic :)
Radeon X700 Pro ~540 fps
Radeon HD7850 ~2400 fps!
..Mikhail, are you real? :) Is it safe..ahh..im paranoid :)
OMG D:
GF8800GT -> 880~1034 FPS
700-800 FPS
GF8800GT
Lovely.
2100fps
8800Ultra SLI
Nice:D
So, looking at the dll, is it using dx9 or 7?
520fps
GeForce 8800 GT
Wow, looks like 8800 is popular around here.
Mikhail, wild guess:
Are you rendering the depthmaps from light POV and camera POV, and passing them to a DirectX9 context to process both through shaders and generate the resulting shadowmap? That's really genius, you ROCK so much, dude.
If not, please, explain in depth how are you doing this, because it's really interesting.
Wow! 1600-1800 fps
GF 9800 GT
1150-2000 fps
Radeon HD4850
The shadows look a bit blocky. Is there a way to change the resolution of the generated shadow so we get les jitter going on?
What's supposed to be happening here?
I get a total white man and a totally white scene, on a black background.
Pentium4 1.60 Ghz 768MB Ram, Nvidia GeForce4 Ti4200, DirectX 9
"..Mikhail, are you real? "
+1
970-980 fps
GF 9600 GT
JP
John Blackledge: Your graphics card doesn't support direct X 9 (Geforce TI 4200)
Oh yes it does. DirectX 9.0c running now.
John Blackledge: The Geforce 4 TI series (7 year old GFX card) is a Direct X 8.0 card and does not support most Direct X 9 specific functionalities. This means that some stuff like Shaders model 2 or model 3 are not possible on your graphics card, and some other goodies which are reserved for more recent graphics cards. If you want those to work, you will need to upgrade to a more recent graphics cards, or just avoid TRYING to run Direct X 9 specific programs. It's not because the API is installed that your graphics card automatically gets upgraded. Some functions are simply not gonna run. That is why this demos does not work well on your system.
_33 - understood, thanks.
But it does leave us with a problem.
Since it's well known that developers always use top-notch machines, and the rest of the public lag behind (I've yet to see even one PC in my neighbourhood which is newer than mine), that means that our potential market will be seeing white screens unless we detect which version of the graphics card/directx they are using.
Interestingly, one of the online game engines I saw recently gave me no water reflections, but worked fine on my friend's - obviously detected my card/dx version and compensated. - Better than getting a white screen.
My point is that any of the new dx9 dll's/systems should have this in as a function. Then the developer can choose what to enable for that particular user.
FastExt has a GfxDriverCaps variable so you can check the if the system has the technical needs.
John Blackledge, you can do like Mortiis says and question the system with the GfxDriverCaps. You also have to understand that realtime shadows on a direct X 8 graphics card is not suggested (too slow). I suggest spending for a new card, and get up to speed with the newer technology.
Agreed Mortiis, but _33 you're missing the point.
It's not what I (or my neighbours) are willing to do, it's about what programmers are willing to do to cope with older machines/cards.
- In the same way that we already cope with the user's selection or restriction of screen resolution.
A start-game message saying 'Some effects such as shadows and water reflection may not be visible on older graphics cards' would suffice.
As it was, the whole demo scene above rendered as completely white (no edges whatsoever) on a black background.
One thing that confuses me, and kinda I'm leaning to John Blackledge side, is that the demo says "Another PPSSM test (DirectX7!)", therefore I would have imagined that the demo is using Directx 7 correct? Therefore, it shouldn't matter if the user has a graphic card which supports direct x 9.0c
Sorry if Im completely wrong, wouldn't be the 1st time ;-)
MikhailV: Is it possible to alpha the shadow to say 0.5, and see how it looks? It isn't too natural to have shadows completely black. I would like to choose the level of alpha if I have to use this shadow capability of your lib. Also, the demo does say Direct X 7, which I think is not true since you're using a Direct X 9 api above it, no? Otherwise, I always like your work!!!
nice 1850-2050 fps!
8800gt 512mb Quad6600 2gbRam
Windows XPsp3
Monitor CRT - 1024x768 85hz
Woaa 1850-2050 fps with a 8800GT !!!
Maybe the core2quad !!!
JP
I don't think so.
Quad used at 25% :)
970-980 fps
GF 9600 GT
C2D E6600 at 40%, 3Go ram
JP
PS :
Uber Jase
2100fps
8800Ultra SLI !!!!!!!!
Mortiis
700-800 FPS !!!
GF8800GT
Tab
GF8800GT -> 880~1034 FPS
Your fps are incredible ;-)
My fps was on a quad core qx6850 with vista 64. So maybe quads do make a difference.
After an update of my video drivers ( Nvidia 181.20 ), I get
1850 FPS !!! WOAAA
50% CPU LOAD
JP
Edit :
2150 Fps after O/C with RivaTuner