GEM-dev November 2005

gem-dev@lists.iem.at

14 participants
20 discussions

Re: [GEM-dev] Re: GEM shader usage
by Tebjan Halm 22 Nov '05

22 Nov '05

by the way, there is an amazing code to calculate raytraced quaternion julia sets on the GPU. you need a PS 3.0 card (nvidia 6200 or higher) to get it to run. i rendered this video in only 2 hours on a nvidia 6800 card: http://vvvv.meso.net/tiki-index.php?page=TonfilmsPatches a nice detail: http://vvvv.meso.net/tiki-browse_image.php?imageId=1511 and from somebody else a pic: http://vvvv.meso.net/tiki-browse_image.php?imageId=1509 the code can be found here: http://graphics.cs.uiuc.edu/svn/kcrane/web/project_qjulia.html keep on codin' tebjan

3 4

vertices...questions without answers
by IOhannes m zmoelnig 22 Nov '05

22 Nov '05

hi i am currently trying to re-do the vertex stuff in a manner that everybody can live with. i think we agree on following (synching brains): everything is a float! (even color; see matju's thoughts about out-of-gamut colors; and i think it is a very hard restriction for processing if we stick to 256 values (we already see problems with pix-processing)) vertices, color: 4 floats (x,y,z,w; r,g,b,a) normals: 3 floats (x,y,z) texcoords: 2 floats (u,v) data is tightly packed. however, some things are a bit unclear to me, and i'd like to have your advice on this topic. i have 2 base classes for vertex unops and binops. unary operation (excerpt from Base/GemVertexObj.h) <GemVertexObj.h> ////////// // which vertices to manipulate // this is a generic routine for all arrays (whether they have 2, 3 or 4 elements) // note: you should really implement the optimized routines below virtual void processVertex(GLfloat*array, int count, int stride); ////////// // optimized routines // note: the SIMD-optimized routines fall-back to the non-SIMD routines // note: these routines fall back to the generic processVertex // texCoords are only (u,v) virtual void processVertex2 (GLfloat*array, int count); virtual void processVertex2Altivec(GLfloat*array, int count); virtual void processVertex2SSE (GLfloat*array, int count); // normals are (x,y,z) virtual void processVertex3 (GLfloat*array, int count); virtual void processVertex3Altivec(GLfloat*array, int count); virtual void processVertex3SSE (GLfloat*array, int count); // vertices are (x,y,z,w); colors are (r,g,b,a) virtual void processVertex4 (GLfloat*array, int count); virtual void processVertex4Altivec(GLfloat*array, int count); virtual void processVertex4SSE (GLfloat*array, int count); </GemVertexObj.h> so there are specialized functions for processing an array (without a notion of what the data actually represents; however there is a notion about the "array-width"; there are SIMD-optimized functions which fallback to the generic ones. there is an optional super-generic function which normally just prints an error message if it ever gets called. this is just a fallback for more specialized functions (and most likely will never get really implemented) the render() function calls the appropriate processVertex*() function. if the user specified to process just a sub-array of the data, this is handled by render() (so the processVertex*() functions don't know anything about offsets,...) that was the easy part. however i have some problem with the binop: <GemVertexDualObj.h> ////////// // Do the rendering virtual void processVertices(float*larray, int lsize, int lstride, float*rarray, int rsize, int rstride); virtual void processVertices2(float*larray, int lsize, float*rarray, int rsize); virtual void processVertices2SSE(float*larray, int lsize, float*rarray, int rsize); virtual void processVertices2Altivec(float*larray, int lsize, float*rarray, int rsize); virtual void processVertices3(float*larray, int lsize, float*rarray, int rsize); virtual void processVertices4(float*larray, int lsize, float*rarray, int rsize); </GemVertexDualObj.h> so the basic concept is the same (i just stripped some SIMD-function declarations to make it more readable) i thought it would be very nice, to be able to "merge" 2 arrays of different lengths, so there are "lsize" and "rsize". but when i converted the first code snippets, i just started to wonder, whether this is a good idea? most likely SIMD-optimization (especially streaming extensions) will be disabled with such feature turned on. so would it be better to just have a dedicated [vertex_resize] object that does the extrapolation to bigger/smaller arrays (however well this might work), and enforce the 2 arrays to be of the same size?? quite minor question, but i'd like to get some feedback. additionally, what do you think of it in general? as for now, the coder would have to write the functions for each type of arrays. does anybody know of a way how to faciliate this task? it is basically a copynpaste of the same code-snippet with very minor modifications. it would be nice if the function would have to be coded only once and the compiler/preprocessor would expand it to all eventualities. i would hate to use weird preprocessor magic. and the resulting code must be as fast as if it was written explicitely. anybody has any ressources on this? fma.dsr. IOhannes

2 1

OSX command-line compile fails?
by B. Bogart 22 Nov '05

22 Nov '05

Hey all, I'm trying to compile gemCVS from today on OSX via terminal and running into some problems. I'm not sure, now that I think about it, if I ever got the terminal build to work. Anyhow I tried compiling gem with: cd src/ autoconf #does nothing it seems ./configure # seems ok... make end of the make output looks like: g++ -o Gem.pd_darwin -framework QuickTime -framework Carbon -framework AGL -framework OpenGL Base/*.o Controls/*.o Geos/*.o Manips/*.o MarkEx/*.o Nongeos/*.o Particles/*.o Pixes/*.o openGL/*.o -ldl -lz -lm -lpthread ld: Undefined symbols: _main _canvas_getcurrent _error _gensym _atom_getint _class_addmethod _outlet_anything _outlet_free _outlet_new _inlet_free _inlet_new _clock_delay _clock_getsystime _clock_gettimesince _clock_new _clock_unset _post _atom_getfloat _atom_getsymbol _class_doaddfloat _garray_class _pd_findbyclass _s_float _atom_string _class_addlist _class_addcreator _class_new _class_sethelpsymbol _pd_new _class_addbang _outlet_float _outlet_symbol _outlet_list _s_list _clock_free _s_ _outlet_bang _canvas_makefilename _class_addanything _dsp_add _nullfn _s_signal _garray_getfloatarray _atom_getfloatarg _freebytes _getbytes _s_bang _floatinlet_new _class_addsymbol _resizebytes _startpost _pd_bind _pd_unbind _canvas_getdir _open_via_path _s_symbol _garray_redraw make: *** [Gem.pd_darwin] Error 1 I think think pthread should be in there. ./configure did label a few things as cached but I don't see a config.cache (indeed there is none) in Gem/src/ Any ideas? Anyone know how Erich could compile Gem for G4 on the source-forge compile farm? Thanks! b.

2 1

OpenGL wrapper question
by cyrille henry 16 Nov '05

16 Nov '05

hello again, for the project i'm curently working on, i nead the GEMglBlendEquation object. i was surprised not to find it in the openGL wrapper stuf. is there a good reason not to have it? could it be add? (the attached code was succesfully tested on linux and osX, and should work also on windows as it's only cut&paste from the GEMglBegin object). and one last question for today : there is lot's of nice new stuf in gem CVS that i would like to use, unfortunatly i really want to use only stable releases of pd and Gem as i really nead to run my patch on lot's of diferents computer without compilling stuff everytime. So : is there any plan to have a gem 0.91 soon? thanks cyrille

3 4

camera position
by cyrille henry 14 Nov '05

14 Nov '05

hello, i just had a look at the camera object, i did not fully understand it : should it remplace the view message to gemwin, or should it be complementary? i did not find a way to set the position of the camera : i think it's really stange that it does not offer a setXYZ message, but only rotation velocity (i think accumrotate should be used when you nead to control rotation velocity). i think that this as already been discuss here, but I can't find the thread. I'm sorry to post a new mail about that, but this object look so strange to me regarding other gem objects... cyrille

2 1

pix_snap2tex small bug
by cyrille henry 14 Nov '05

14 Nov '05

hello, there is a small bug with pix_snap2tex when destroying the Gem window see the attached patch. i noticed it with linux and windows (not tested on osX), with 0.90 and curent cvs. (any pd version) Cyrille #N canvas 887 307 763 565 10; #X msg 36 243 snap; #X obj 36 371 square 5; #X obj 36 221 t b; #X obj 124 200 gemhead -5; #X obj 36 303 color 0.8 0.8 0.8; #X obj 36 200 gemhead -15; #X obj 141 126 gemhead; #X obj 41 154 gemwin; #X msg 41 33 dimen 400 300 \, create \, 1; #X obj 141 154 sphere; #X obj 36 349 translateXYZ 0.2 0 0; #X obj 36 326 scaleXYZ 1.5 1.1 1; #X msg 62 71 destroy \, create \, 1; #X msg 199 253 400 300; #X text 225 34 <- 1 : create the window : everything work as expected ; #X text 228 279 <- 3 : ctrl-X \, ctrl-Z the pix_object : everything work again; #X obj 36 278 pix_snap2tex 0 0 400 300; #X text 225 75 <- 2 : destroy / create the window : pix_snapt2tex captur only empty (white) window; #X connect 0 0 16 0; #X connect 2 0 0 0; #X connect 3 0 16 0; #X connect 4 0 11 0; #X connect 5 0 2 0; #X connect 6 0 9 0; #X connect 8 0 7 0; #X connect 10 0 1 0; #X connect 11 0 10 0; #X connect 12 0 7 0; #X connect 13 0 16 2; #X connect 16 0 4 0;

1 0

ortho and bollean interaction between objects
by B. Bogart 13 Nov '05

13 Nov '05

Hello all, I want to be able to see the interaction between objects in ortho mode. I have a curve3d in front of a plane. When I move the curve3d control points back I want to see the plane underneath become visible. This all works great, except I don't want the perspective to cause the edges of the curve3d to get smaller as they move from the camera. In ortho mode no matter how far back I move the control point, the place behind it never becomes visible. Any way to use ortho and still being able to keep the object interaction? The effect I'm going for is something like what water looks like as it floods a complex 3D surface, the intersection between the surface and the plane of water being the important aspects. Back _ -/-\----/\-- <- plane / A \/\/ \ <- curve3d Front Here at point "A" the plane is visible in front of the curve3d. Or let me know if you have any other ideas to accomplish this. Thanks all. b.

1 1

Video Codecs Seek performance... *sigh*
by B. Bogart 06 Nov '05

06 Nov '05

Hi all, Now that the smoke has cleared a bit from my installation in Ottawa, I'd like to ask about some performance issues with file seeking. I'm now using mpeg2 (dvd quality or so) encoded videos with single frame GOPs, and seeking them using the table of contents generated by mpeg3toc. The problem is that two 640x480 videos seeking randomly use up 80% of my new AMD64 3200+. This seems to be a bit of a hog... I also endcoded a test file as "photo-jpeg" in a quicktime container. I could also seek that file, but the CPU usage was about the same - 10%. I'm reading the files off a dual disk SATA raid0, so the disk access should not be the problem. (I can play 3 DV videos off the same array using mplayer w/out lost frames.) So the question is: #1 What is the best video codec for random seeking performance? #2 How can I profile pd/gem/diskIO/decoding to find out what is causing the largest bottleneck? Suggest tools? My previous installation used 256x256 photo-jpeg MOVs playing simultaneously on a duron 900 and gem 0.87. These videos were just played and not seeked. Thanks all! I'd really like to be able to get better performance out of this patch. .b.

5 5

gem mingw build
by Thoralf Schulze 02 Nov '05

02 Nov '05

Hi there, I put up a mingw build of gem at http://puredata.info/Members/tfbb/gem_mingw/ . This is an almost completely untested cvs snapshot from the beginning of october. Please don't hold me resonsible if it makes your computer stink or something ... with kind regards, thoralf. ___________________________________________________________ Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com

3 2

pix_image
by Tim Blechmann 01 Nov '05

01 Nov '05

hi all, since i have a gem-patch running on win32, that's displaying only this: |gemhead| | |pix_image| | |pix_texture| | |scaleXYZ 5.33 4 0| | |square| after a few days of changing the image a lot (every 10 seconds or so), it's only displaying a white screen. is this a known issue? related to the pix_image cache? thanks ... tim -- mailto:TimBlechmann@gmx.de ICQ: 96771783 http://www.mokabar.tk latest mp3: kMW.mp3 http://mattin.org/mp3.html latest cd: Goh Lee Kwang & Tim Blechmann: Drone http://www.geocities.com/gohleekwangtimblechmannduo/ After one look at this planet any visitor from outer space would say "I want to see the manager." William S. Burroughs

5 13

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

GEM-dev November 2005