Camelia, the Perl 6 bug

IRC log for #openframeworks, 2013-10-10

| Channels | #openframeworks index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:21 dantheman joined #openframeworks
00:29 DrCode joined #openframeworks
01:07 superFluid joined #openframeworks
01:09 DrCode joined #openframeworks
02:02 pizthewiz joined #openframeworks
03:16 pizthewiz joined #openframeworks
03:43 vade joined #openframeworks
05:09 admsyn joined #openframeworks
05:26 LCID_Fire joined #openframeworks
05:26 LCID_Fire left #openframeworks
05:34 DrCode joined #openframeworks
06:01 DrCode joined #openframeworks
06:42 DrCode joined #openframeworks
07:34 igorline joined #openframeworks
08:54 underdoeg joined #openframeworks
10:00 igorline joined #openframeworks
10:07 igorline joined #openframeworks
10:32 heavyside joined #openframeworks
11:40 igorline joined #openframeworks
12:03 vade joined #openframeworks
12:07 jacres joined #openframeworks
12:54 superFluid joined #openframeworks
13:03 workergnome joined #openframeworks
13:16 workergnome joined #openframeworks
14:01 igorline joined #openframeworks
14:12 jvcleave joined #openframeworks
14:15 vade joined #openframeworks
15:10 harpreet joined #openframeworks
15:14 timius joined #openframeworks
15:17 timius Hi guys, I am trying out this scenario where I need to know the position of a user wrt to a display kept in a certain space. I need to figure out the corners of the display for which I have been doing physical measurements keeping Kinect and screen in the same plane. However, if I rotate the Kinect, are there any suggestion on how to get the corner coordinates of a display screen?
15:19 igorline joined #openframeworks
15:25 stephanschulz joined #openframeworks
15:28 admsyn joined #openframeworks
15:40 McNight joined #openframeworks
15:50 igorline joined #openframeworks
16:04 peejay joined #openframeworks
17:08 pizthewiz joined #openframeworks
17:41 GuidovanPossum joined #openframeworks
17:44 LCID_Fire joined #openframeworks
17:44 laserpilot joined #openframeworks
17:57 jedahan joined #openframeworks
18:23 jedahan joined #openframeworks
18:24 admsyn what's up internet
18:37 laserpilot NOT MUCH
18:37 laserpilot hackin on some audio analysis
18:59 admsyn wicked
18:59 admsyn what's it doing?
19:00 GuidovanPossum I want to see too?
19:01 laserpilot just taking kyle's fft stuff and making it more like gimme the low/mid/high helper functions, std deviation for noisiness,  and get intensity at frequency
19:01 laserpilot kyle said he was planning on taking another pass at his fft addon and would probably start adding in helpers like that
19:02 laserpilot also using this awesome app called wavetick/waveclock that does real time BPM detection and can find the beginning of the bar in incoming audio - i have to pipe that in via midi clock though
19:02 admsyn cools
19:02 admsyn how sane is your input?
19:03 GuidovanPossum I've been playing with a program that I made that uses fft to analyze and store the values from an mp3 input
19:04 laserpilot its not too bad - but the installation will have to react to whatever people find in spotify, so we have to account for it all
19:04 laserpilot input isnt too bad, im doing normalizing off the highest value
19:05 admsyn fwiw I tried to do a thing like that before, and found it was waaaayyyy easier to just throw a few EQs on the input than to try and wrangle an FFT
19:05 admsyn 'cause you can twiddle it with a UI to find a range that works well
19:05 GuidovanPossum I've got it where it's sampling 32 different frequency ranges at 60Hz storing the values in memory and then writing them all to text files once the song is over]
19:06 admsyn GuidovanPossum: so like a fingerprinting kind of thing?
19:06 GuidovanPossum then I can post process the data and play with certain smoothing and limits to see how it affects the data, yes
19:07 laserpilot sounds awesome - im doing something similar in another project where im making a 3d mesh from the fft: http://blairneal.com/blog/sonic-prints/
19:08 GuidovanPossum and then I can feed it controlling something that too slow render real time like making a camera and fractal or other images and video dance to the music
19:08 laserpilot yeah twiddling stuff would make sense - if it were me I'd much rather have this be human controlled, but its gotta run for 2 month
19:09 GuidovanPossum wow very neat work just checking it out now!
19:09 laserpilot i plan on releasing it, i just need to have time to revisit it and clean it up
19:09 admsyn looked good at eyeo IIRC :)
19:11 laserpilot thx
19:11 GuidovanPossum I wanted to go so badly
19:11 laserpilot hows your stuff going adam?
19:12 admsyn not too bad, we just did a big install thing for nuit blanche in Toronto, that was pretty fun
19:13 GuidovanPossum are you guys excited about Vuo at all I saw that they are finally releasing a beta in a couple of weeks?
19:14 admsyn GuidovanPossum: I remember something about it that came up RE: licensing or pricing that was sort of off-putting..but I haven't been paying enough attention to it, really
19:15 laserpilot yeah the licensing has turned me off to it too unfortunately
19:16 laserpilot quartz composer took off because it was free i feel like…its going to be much harder to break onto the scene with something that is locked down fromt he start AND a new paradigm to learn
19:16 laserpilot touchdesigner and vvvv are successful because they have those non commercial versions i feel like
19:23 admsyn this is the oF nuit blanche thing we did: http://instagram.com/p/fIZRe5LMoN/
19:24 admsyn proper video is coming in a bit, but this guy got a decent instagram video of it
19:24 laserpilot oh nice
19:56 Guest2967 dantheman: hey =)
19:57 tgfrerer dantheman: hey =) (now properly identified)
20:03 tigbot joined #openframeworks
20:12 dantheman_ joined #openframeworks
20:12 dantheman_ tgfrerer: yeah ok that makes sense
20:12 dantheman_ but its weird how many shaders work in non-programmable mode but don't in programmable mode
20:13 tgfrerer this is mainly to do with attribute bind positions
20:13 tgfrerer anything (and more) should work in programmable
20:13 tgfrerer best thing to do is to check using the opengl profiler
20:14 tgfrerer from the xcode tools suite - which is actually pretty decent
20:14 dantheman_ ahh
20:14 dantheman_ ok
20:14 tgfrerer there, set a breakpoint before any draw call
20:14 tgfrerer then go into the "resources" window
20:14 dantheman_ right
20:14 dantheman_ yeah
20:14 tgfrerer and have a look at your shader programmes
20:15 dantheman_ gotcha
20:15 dantheman_ yeah there are only a couple that don't work well
20:15 tgfrerer there you should see where the attribute bind positions end up at.
20:15 dantheman_ like blur
20:15 dantheman_ I'm trying to figure out why
20:15 tgfrerer cool=)
20:15 dantheman_ yeah
20:15 tgfrerer my bet is that some attribute is not used in the shader and the shader compiler optimises it away, thus jumbling up the attribute bind positions
20:16 dantheman_ hmmm
20:16 dantheman_ interesting
20:16 tgfrerer that's been a real pita
20:16 dantheman_ yeah let me imgur a few more screen caps
20:17 tgfrerer also have a look at the 'warnings' tab in the shader program view in the resources window in the opengl profiler, once your program is halted on 'before draw call' breakpoint
20:18 dantheman_ kk
20:18 dantheman_ will do
20:18 dantheman_ brb meeting
20:18 tgfrerer =)
20:18 tgfrerer i have irc on speakerphone, just say something and i'll be back =)
20:22 admsyn FITTER. HAPPIER. MORE PRODUCTIVE.
20:31 stephanschulz left #openframeworks
21:24 robot joined #openframeworks
21:26 dantheman_ joined #openframeworks
21:26 dantheman_ tgfrerer: well that was a long meeting
21:26 dantheman_ http://imgur.com/a/dErlO#0
21:27 dantheman_ that is the blur tutorial
21:53 tgfrerer hey dan!
21:55 tgfrerer woha, that's quite some interesting glitch art :D
21:57 tgfrerer most probably the texture coordinates i'd say it's got something to do with the vertex attributes and how they get mapped to the shader - check the default shaders in ofGLProgrammableRenderer.cpp to see the default attribute order
21:58 tgfrerer i'm saying that because e.g. image 2 suggests (the diagonal) that the texture coordinates have not been mapped correctly ...
22:45 blabla joined #openframeworks
22:47 blabla left #openframeworks
23:09 superFluid joined #openframeworks
23:37 superFluid joined #openframeworks
23:38 GuidovanPossum joined #openframeworks
23:41 admsyn joined #openframeworks

| Channels | #openframeworks index | Today | | Search | Google Search | Plain-Text | summary