What your mother never told you about graphics development

it'a a Stateless particle system? what about s...

2011-05-05T07:41:25.363+04:00

it'a a Stateless particle system?
what about state-preserving particle system, you need to store last fame particle's property(position, velocity, life...etc) to calculate this frame.
in dx9, need to store the particle's property into render target, and double buffer render target. i still not figure out how to pass the particle's uv with the render target.do you have any idea?

link:http://wiki.gamedev.net/index.php/D3DBook:Dynamic_Particle_Systems#GPU_Based_Particle_Systems_.28D3D9.29

Nice work! I think we are spoilt these days (2011...

2011-03-01T17:59:56.578+03:00

Nice work! I think we are spoilt these days (2011) with large instruction buffers and high level shader languages :)

Well since most of the objects are going to need t...

2010-09-23T22:34:58.235+04:00

Well since most of the objects are going to need their world view projection matrices calculated anyway, then there's little loss, otherwise you have to solve all the planes for the frustum in world space for each object. I think a matrix calculation is probably cheaper than a world space frustum cull test.

Tanks for your fast response and for your attentio...

2010-08-09T17:40:21.061+04:00

Tanks for your fast response and for your attention.

It is very important for me to have some help from a matematician because it is very difficult for me to understand projective spaces.

Sorry also for my confusion ( i am Italian and sometimes i do not understand well the english )

Bye
Livio Cicala

First of all - you're confusing me with the al...

2010-08-09T16:15:20.157+04:00

First of all - you're confusing me with the algorithm author. While I'm happy to try to help you with your issues, I'm not him :)

Your version of ZRotation matrix is indeed slightly more correct. However, the only difference is in the direction of post-projected Y axis, and, since there is a unit cube normalization step, the quality should not differ noticeably. I can not verify this since I can't build the demo, and I do not have access to my own XPSM implementation - I'll post an update here once I can check this.

Taking the absolute value of w does not make sense from the mathematic standpoint; I think that you're just lucky to get bounds that are large enough to accomodate everything that is necessary. If I remember correctly, there were indeed some rare clipping issues even after the TransformWithWClip correction; again, I'll check this and post here.

As for the donation (again, it's not my site, but anyway) the problem is that the domain has expired. I'm not sure if the e-mail specified in the paper works, but you can try it (btw, the most up-to-date XPSM page now is http://xmvlad.110mb.com/).

Dear Arseny, the last message was bad formatted an...

2010-08-09T15:30:03.670+04:00

Dear Arseny,
the last message was bad formatted
and difficult to read...

so I repeat now the two matrices

======================= 1
D3DXMATRIX ZRotation(
unitP.x,unitP.y,0.0f,0.0f,
unitP.y,-unitP.x,0.0f,0.0f,
0.0f,0.0f, 1.0f, 0.0f,
0.0f,0.0f,0.0f,1.0f);

======================= 2
D3DXMATRIX ZRotation(
unitP.x,unitP.y,0.0f,0.0f,
-unitP.y,unitP.x,0.0f,0.0f,
0.0f,0.0f, 1.0f, 0.0f,
0.0f,0.0f,0.0f,1.0f);

In addition i have found that:
with the matrix (2) also the Zacne is more
regular ( exactly the same as the "uniform shadow map" )

In other words:
the zbias regulation is the same and does not changes when changing from "XPSM" to "uniform shadow map"

I do not know if the matrix (2) is correct but
visual results in my implementation are very best looking with this matrix.

----------------------------------------

Regard to the function : "TransformWithWClip"

I have found that:
if I transform "w" to "absolute value of w" before to compare and use it then all the shadow clippings are gone

In every situation, also with the original "TransformWithWClip" function, this ABS produces always shadows without any clipping.

Another time I do not know if my correction is matematically correct, but, in my implementation it works very well.

--------------------------------------

another little problem...

I have tried to send to you some euro ( for now i can contribute only with a little donation ) but the donation page is in russian language...

maybe i can send to you some money with PayPal ?

--------------------------------------

bye
Livio Cicala

Anonymous said... Dear Arseny, YES, i am using m...

2010-08-08T18:20:47.605+04:00

Anonymous said...
Dear Arseny,

YES, i am using my own vector-like container with slightly different operation semantics.
Thanks for your attention.

------------------------------------------

Now i have found another little problem.

With your Zrotation matrix composed as shown in (1) i get the rear side of some object not shadowed and with some shadowed part that comes from the other side.
But if i change the composition as in (2) all the objects are with the rear side completely shadowed and the final visual result is better.

======================= 1
D3DXMATRIX ZRotation(unitP.x,unitP.y,0.0f,0.0f, unitP.y,-unitP.x,0.0f,0.0f, 0.0f,0.0f, 1.0f, 0.0f,
0.0f,0.0f,0.0f,1.0f);

======================= 2
D3DXMATRIX ZRotation(unitP.x,unitP.y,0.0f,0.0f, -unitP.y,unitP.x,0.0f,0.0f, 0.0f,0.0f, 1.0f, 0.0f,
0.0f,0.0f,0.0f,1.0f);

DirectX manual says that Z-rotation is:
cos sin 0 0
-sin cos 0 0
0 0 1 0
0 0 0 1

What do you think of this ?

bye
Livio Cicala (lc@freetw.net)

The term +8 is indeed redundant; however, with con...

2010-08-02T15:18:13.199+04:00

The term +8 is indeed redundant; however, with conforming STL implementation there is no harm in extra space reservation - you are either confusing reserve() with resize(), or you're using your own vector-like container with slightly different operation semantics.

They are just world-space camera basis vectors: ca...

2010-08-02T15:06:45.316+04:00

They are just world-space camera basis vectors: camera_x_axis is the camera right vector, camera_y_axis is the camera up vector.

Usually up vector and camera direction are set when you construct a lookat matrix for camera, and right vector can be reconstructed with a cross product from them.

How would you retrieve the Camera_X_Axis and Y_Axi...

2010-08-02T14:33:41.418+04:00

How would you retrieve the Camera_X_Axis and Y_Axis in that first, "less trivial" method que talked about earlier of deriving the positions based on the corners' UVs?

Dear Arseny, I have found a little error in the f...

2010-08-01T11:03:40.871+04:00

Dear Arseny,

I have found a little error in the function: PracticalPSM::BuildXPSMProjectionMatrix()

The following lines are over dimensionaed:
===========================================
shadowCasters.reserve
( m_ShadowCasterPoints.size()*8 + 8 );
shadowReceivers.reserve
( m_ShadowReceiverPoints.size()*8 + 8 );
===========================================

It is better to remove the term "+8"

This error adds some caster and receiver points at 0,0,0 and it degrades the algorithm.
I suspect also that, in some implementations, it can collect garbage values and produce unpredictable results.

Oh dear, here we go again. I could explain in deta...

2010-07-02T08:32:22.792+04:00

Oh dear, here we go again. I could explain in detail why there's no need for a conformant parser for a lot of tasks - pugixml is used by a number of people, and they're loving it - but I won't bother talking to an anonymous person who seems to just want to tease me.

If you want a meaningful discussion, please state who you are so that I know that you're honest about your claims.

it is not a lie, pugixml is not even conformant to...

2010-07-02T03:12:05.814+04:00

it is not a lie, pugixml is not even conformant to be suited for 15% of the task, think this way, if it doesn't do any checking, it certainly would be infinitely fast.. but what use does it have?

добрый день извините, можно попросить у вас посмот...

2010-05-19T15:51:30.097+04:00

добрый день
извините, можно попросить у вас посмотреть в исходник?
это исключительно для ознакомительных целей (учусь программировать), хотел бы разобраться в тонкостях DirectX
буду очень вам признателен
alexpolt [at] gemabank.ru

Thank you for responding! I'm losing hope here...

2010-03-30T09:09:54.361+04:00

Thank you for responding! I'm losing hope here. Could you let me know your settings when you export an animated rig with blendshapes? I will try it with ColladaMaya (by Feeling Software).
Another question... what version of ColladaMaya are you using? This is the only place I can find a free download of ColladaMaya: http://update.multiverse.net/wiki/index.php/Exporting_Models#Exporting_from_Maya

Daniel, I've had no problems with blendshapes ...

2010-03-29T23:25:34.915+04:00

Daniel, I've had no problems with blendshapes export - however I was using Feeling Software exporter rather than OpenCOLLADA. I did not try exporting blend shapes with OpenCOLLADA so can't comment on the bugs you're experiencing. OpenCOLLADA is an almost entirely different codebase rather than Feeling's exporter fork, so it's possible blend shape support is not implemented yet.

Using the OPENCollada Exporter for Maya, I can exp...

2010-03-29T11:18:14.524+04:00

Using the OPENCollada Exporter for Maya, I can export a skeleton animation just fine, but it doesn't let me export any blendshapes. Is there still a bug for exporting blenshapes?
I've tried exporting the animation (skeleton and blendshapes) after doing a geometry cache, but it seems stuck on one frame when the collada file is previewed. Can anyone tell me what's going on? Any help would be great!

reactive on google I'd appreciate a copy as w...

2010-03-02T10:22:43.342+03:00

reactive on google

I'd appreciate a copy as well. I'm especially wary of the unclear licenses as well.

2010-01-11T06:19:50.445+03:00

This comment has been removed by a blog administrator.

Since you're likely to have many many objects ...

2009-12-13T23:06:38.368+03:00

Since you're likely to have many many objects with bounding boxes in world space, and only 6 frustum planes in projection space, wouldn't it be faster in the long run to just bring the planes to world space and to the test there?

pugixml does not claim it's a conformant XML p...

2009-11-23T09:20:46.875+03:00

pugixml does not claim it's a conformant XML parser, it's simply an XML parser suitable for 95% of XML parsing. VTD-XML should not claim it's the fastest parser in the world either, since that's a lie.

Saw you tweets comparing vtd-xml with pugi... pugi...

2009-11-23T05:00:15.630+03:00

Saw you tweets comparing vtd-xml with pugi...
pugi is not a conformant XML parser at all, it does little well-formedness checking
so sigh...

Profiled latest vtd-xml against latest (repo) pugi...

2009-11-18T23:47:53.521+03:00

Profiled latest vtd-xml against latest (repo) pugixml, pugixml is 3x faster on a synthetic test (1 Mb file with lots of markup), at the expense of 25% more memory; on a moderate (several Mb) COLLADA file it's 6x faster at the expense of 2% more memory. Memory management is not overrideable btw, had to hack it.

No, I haven't; I seriously doubt the claims - ...

2009-11-18T09:58:43.102+03:00

No, I haven't; I seriously doubt the claims - the benchmark is only against Xerces; well, guess what - pugixml is much faster and better in terms of memory than Xerces; for some XML files 90-120 Mb/sec is just not good enough. I haven't tested VTD yet though, so I'm not exactly sure. Also it's huge, meaning it's in the same league as Xerces.

All other claims are, uh, weird too - basically, it's stupid to claim you're the worlds *est something unless you deliver concrete proof in terms of benchmarks against latest version of a wide variety of product. Are you VTD-XML developer? If you are, I'd suggest rewriting both claims and benchmarks.

Have you consider vtd-xml for improved perfomrance...

2009-11-18T07:35:58.768+03:00

Have you consider vtd-xml for improved perfomrance and reducing memory usage?

http://vtd-xml.sf.net