Quartz Composer 4

Syphon and mad mapper.

sophi--e's picture

Hello Kineme people, I'm new to Quartz Composer, MadMapper and Resolume but learning fast. I recently have been playing with QC and have been trying to syphon it into Mad. I have downloaded the plug in and i can save a QC file and it will work if i drag and drop the saved file in ( into the QC area near the movies images etc)

However, there is another option in which to syphon, and it doesn't work. Can i use syphon to change things on QC live and if so am i missing something? How do i do it?

Plus on the syphon of QC in Mad it receives something but not image/animation shows.

Any help would be much appreciated.

Thank you

PS Kineme is amazing

sprite transparency

moorooduc's picture

hello everyone, I have asked this question some time ago in the vidvox forum. I might go into some details at explaining my problem again. I have an image of some birds sitting on a power line. I cut out the birds and saved them as a png file with a transparent background. I want to overlay them to original image. the final composition for vdmx should be able to zoom in to the "cut out-picture. thus, the birds will change size, while the rest of the picture remains just a still. the problem is quite simple: I want to overlay a png or psd file wit a transparent background. I have tried to change the blending of the sprite to OVER. and it doesn't work. Someone has sent me his qtz-file, which worked fine on his laptop, but not on mine. I don't understand why this happens. The only difference was the version of QC. On his system, her was running some older version. I am running QC 4 with snow leopard. Are there known issues with QC4 and snow leopard? Is there some way to go back to QC 3 (but keep snow leopard)? has anyone suggestions for how to solve this problem? best, moorooduc

qc workshop dilema

dust's picture

[edit short ? no rant ]

ok i went on a rant there below so i am rephrasing the question so you don't have to read my rant to give some hopeful feedback.

i'm teaching a qc workshop to 1st year new media students and their professors. i am having a hard-time coming up with a project to assign the students that will take two weeks to complete in a collaborative group setting. i don't think this only me when i say this but there isn't much you can't accomplish in qc overnight with the exception of programming plugins. (they take time). what im saying is something that could take two weeks to code in a group collaboration could easily be done in QC in a few hours. does anyone have any suggestions as to what a lengthy qc project could be ? don't get me wrong it has taken me the better part of a year to get to where im at with QC and i'm still learning on a daily basis, and have only scratched the surface of some of its powerful technologies but i can't say your assignment in the next two weeks is to learn all of quartz composer thats ridiculous.

so i need to come up with something that would be challenging for them but also leave them feeling accomplished i guess would be a good idea. don't worry if it sounds hard, they have spent the last few months doing php, processing, max-msp and the arduino, actually making them all work together in a networked pong game. so im pretty sure they will be happy to learn qc. although that seems like a lot to grasp in the first semester of your first year at college when you thought you where going to be drawing flowers in photoshop. so no networked pong for the above mentioned reasons and other reasons below.

[end edit - short version ]

[workshop dilema rant]

i have an important question. well let me be a bit narcissistic which is totally not my character by the way. narcissism can bother me, its a recursive issue i have with my x wife. so my question is important to me and maybe not you but it is equally important to furthering the interest of quartz composer and possibly creating a bigger user base. so depending on your perspective of things in regards to quartz composer, its users, and how instantiating a greater interest will effect the community as a whole i'm asking for any feedback possible.

i have been invited to instruct a work shop in quartz composer. this is exciting for me because i'm still an undergraduate interdisciplinary student and usually professors and phd students are invited to do these workshops. i know a few of the people on here have done similar workshops because i have watched them. the university that the workshop is held at, has just been awarded a very large grant for the purpose of incubating innovation and developing emerging new media technologies.

besides the construction of buildings and labs, quartz composer has been chosen to be integrated into a course curriculum for new media students. so it is my task to introduce quartz composer to the 100 level students and some faculty. fortunately i have had the opportunity to skip requisite classes like this and take graduate level courses in new media. i won't bother explaining why because its not important plus i don't get credit for classes i was able to skip so it is irrelevant but keep this in mind if you choose to the question i will get to in a second.

i think QC is perfect for beginning students although a bit different than the other traditional motion stuff like AE or animation stuff like Maya. it defiantly has the ease of connectivity that other programs like it do not have. dragging your video clip onto the editor can not really get any simpler is what i'm trying to say. i do think that having some sort of hard coding or scripting knowledge would be highly beneficial to understanding quartz composer. unfortunately new media students usually take computer science as a math requirement and not as pre requisite to an intro new media class.

so im sure i will have to spend time explaining QC terminology, and basic stuff like data types etc... although they might be familiar with processing by this point so maybe not. lecturing a work shop isn't my problem, i can certainly find lots of things to talk about and have more than enough files of my own and examples from this forum i can show.

my problem is that that i have to assign a project for the class to work on in groups for two weeks. so this may not sound hard to you as i need your input, but i don't think i have ever made a patch that has taken that long let alone had group collaborative support and effort. actually i don't think i have ever worked on patch for longer than a day with the exception of programming plug-in patches which sometimes take a few min. but my last audio component plug-in i debugged for week and put it to the side so i could come back to it refreshed. its because of this reason that i like qc so much (not the debugging part, but the quick dev times)

sure there are instances where a quick javascript patch is quicker than some noodles but if you take the qc context as a whole and all the frameworks used to make the standard patches. it would really take along time to accomplish some of the things we do visually in code. so to me its the rapid development and ability to prototype ideas and sketches very fast in qc. i think something that could take 2 weeks with group effort in coding can be done in QC in one evening. so with that said what could be a good project to keep a class busy with quartz composer that can be accomplished in two weeks by working in groups ?

some previous assignments not related to qc have been to make a networked pong game using processing, php and max-msp. that assignment is cool because half the challenge is just getting max to talk to processing to talk php all in real time and group collaboration is needed or it doesn't work, pong that is. i want to stay away from pong even though there has been a resurgence in the general interest of pong mainly do to the multi-touch community. no matter how great a game of pong can be and how there probably is bunch of pong patches i have not seen, i don't think a group of kids would take 2 weeks in collaboration to make a networked pong game. when the whole class could make an 18 hole golf course in unity 3d's terrain editor that would look better than tiger woods. i'm not knocking tiger woods i actually enjoy a competitive game every now and then.

i'm actually working on a 1st/3rd person multi-touch frisbee golf game in unity with some of the spare time i don't have. plus this workshop dose not have anything to do with unity although using cv-tools facial recognition to fps look at head tracking and using some optical flow to translate the camera in unity i think would be a fun but then again the controlling interface could be easily placed together with developer examples in a few hours leaving two weeks to build 3d asset art which i find is the time consuming task. this however i can not lecture on because the students have not taken any maya courses yet.

blah blah rant rant. if you read this tonight or tomorrow during the day i could use some feedback as i have to start the workshop tomorrow at noon. once again the question is a project idea. the assignment obstructions are it needs to take two weeks in group setting.

honestly it has taken me almost a year to understand quartz composer technologies and everyday it still leaves me more unexplored options to investigate. so i understand that lots of things can not be covered fully in a two week session, in addition quartz composer may be user friendly in some aspects in comparison to the time it takes to the same things with other tools. qc is more of a intuitive thing, and its hard to teach intuition or inspire innovative creativity. all i can to do is foster the students in a crash qc course in hopes that they will choose qc as there weapon and ultimately give them an additional tool that they can choose to use as the journey down the path many others have.

if you ask me its kind of a lot to throw QC, processing, max-msp, php, flash and arduino all at a first year student all in one semester. (maybe it was a good thing i skipped to graduate classes.) in addition they also have to learn the other fundamentals of FCP, motion, DVD pro, CS4 etc.. in other classes. so i'm not expecting much. the professor likes to give hard assignments in short times. i guess its a challenge and he understands that usually the goal is never met. so don't worry if you have an idea that sounds to difficult.

Problem with OpenCL kernel patch

utopianational's picture


I'm trying to create a OpenCL kernel but I cannot get it work well. This program is to convert X/Y mouse position queue to an image. It looks work fine, but it gets gradually slower, and finally crashes MacOSX after several minutes.

__kernel void main(__global const float *X,__global const float *Y,__wr image2d_t output)
   int2   pos = (int2)(get_global_id(0), get_global_id(1));
   float4 color;
   color.x=X[pos.x]; //R
   color.y=Y[pos.x]; //G
   color.z=0.0; //B
   write_imagef(output, pos, color);

I set queue size to 80 in qc, so that it creates 80x1 pixels image.

I found that it gets work well when I continue to click "Sample Now" button in "Advanced Kernel Settings" every minute, but I have no idea to get code work stably. Could you help me?


QC4 Kineme3D Serial Number Not 'Sticking'

toneburst's picture

Ever since updating to Snow Leopard, Kineme3D is asks for a serial number every time I start the QC editor. Anyone else getting this? Any workaround?