fæt

Hi there all! Using Max for live 7.3+Ableton Live 9.7. Thanks to the dude837@youtube (Sam Tarakajian) for his

 on programming within max, if you are looking for a way to get involved I highly recommend his series, they're interesting, informative, and entertaining! Through his tutilege and web tutorial (found on youtube AND the internal package manager in MAX "

I've modeled one of Sam's tutorial patches (with slight tweaks to make it useful to Ableton Live, (atleast intending XD), and tweaks of aesthetic graphical preferences) of a jit.world that is responding to incoming audio, in the case of max for live, ableton is generating the incoming audio via "plugin~" audio is coming from ableton, being processed by max, and finally a signal sent back to ableton via "plugout~"

The patch is designed so that rendered graphics are being manipulated real time- as a response to the lower frequencies of the mix created within ableton live.

how do I record the "rendered video" response either with jit.record, or jit.vcr(ideally)?

I think this is a problem due to incorrectly routing matrices, or incorrectly labeling said matrices in a correct manner that jit.record or jit.vcr can coherently understand.

I've tried multiple configurations with no success, and various different failures :X

 my first error message (I don't believe I ever overcame, but gave up on the routing configuration and tried jit.record) "make sure DSP is enabled"?? 

I thought this may be an incompatability issue because of the following but I am unclear.

Ableton is Launching/Authorizing Max inside of Ableton, and while Ableton is streaming audio over to Max, Max reads the audio in "preview mode" thus I'm not sure if .vcr can be used because I don't know how to enable DSP. It's my understanding that with M4L, Max is relying on ableton to "forward" the information of the drivers and audio I/O.

 I managed to get the setup to "write $" and "stop $", but the return file was always <60kb, never what I intended, and never anything of any value, other than to learn from another mistake :D

So I reconfigured (not entirely sure the changes I've made, but I keep re-analyzing it and wondering where I am going wrong)...

 A few attrui's later, and I thought I might have this figured out. nope. I have set preferences (globally and within attrui's) (that never seem to save correctly and are reset when i reload the patch) and now when I try to "write $" with jit.record Max crashes X(

So max tells me to input the jit.matrices for .record

or for .vcr matrices+msp audio for .vcr but i'm wondering if I can feed in ableton sound output instead of msp audio) 

Which matrices do i need to connect to the jit.record or jit.vcr, and what kind of specifics need to be "outputted" for those said "matrices"?

If I set the patch up appropriately in max, do you know of any functionality limitations that would prohibit me from "write $" "stop $" a Jit.rec or jit.vcr from within ableton live, with the end result of getting a video file? 

whenever I try to open the help files (not reference manual) for jit.vcr or jit.record max crashes

 :( + Online information is limited (atleast until it grows more) and hard to sort out (through various versions) for jitter info. 

Hope that was a coherent enough jumble of jarbled jargon XD 

Thx for the tutorial series (@Sam Tarakajian)

Any help, guidance, or suggestions would be much appreciated! 

essentially my end goal is to have video accompaniment (as a movie file) that syncs to audio's that I create/compose in ableton live, and to keep experimnting. 

I realize there are multiple ways of achieviing this end goal (and probably many more ways that I do not realize or understand).

Ableton Live has a video/audio encoder that will render various different formats *but AbletonLive reads only .mov (*from my understanding*) as input video formats, I think the ideal route is to configure .vcr in max to snag audio from ableton/ and the real time video render from max in synch all in one go (opposed to manually aligning audio to video in a video editor or Live), but the question is how?

If you made it this far, and think you can help, have a look at the patch attached below, with a few things to note.

This patch was designed by, Sam Tarakajian

, and I merely remodeled to purpose Ableton Live functionality,

 I take little to no credit for originality.

2) It is a hot MESS. still working on cleaning and organizing everything.

3)If you open the patch, its currently designed for Max 4 Live audio, repurposing for MSP should be pretty simple?

4)To Help streamline things, look for the big white numbers in the patch. 1-2-3-4-5-6, these are positioned near buttons or switches to help guide/initialize/setup the functionality of the patch.

5)"Write $" attached to .record crashes my program due to some incoherency in the whole inccorectly functioning. not entirely sure what, or how yet.

6) Mistakes have been made XD If something does'nt make sense, I would be more than happy to try and clarify my meaning.

7) Lots of params to adjust some savy looks (similar to the big white numbers look for big white V's (is a variable)) 

8)HAVE FUN, Any and all input related to functionality (related to my intended re-design) would be appreciated!

Hi there all! Using Max for live 7.3+Ableton Live 9.7.  Thanks to the dude837@youtube (Sam Tarakajian) for his "Delicious tutorial" series on programming within max, if you are looking for a way to get involved I highly recommend his series, they're interesting, informative, and entertaining!  Through his tutilege and web tutorial (found on youtube AND the internal package manager in MAX "Delicious Tutorial") I have before me a perplexation.  

I've modeled one of Sam's tutorial patches (with slight tweaks to make it useful to Ableton Live, (atleast intending XD), and tweaks of aesthetic graphical preferences) of a jit.world that is responding to incoming audio, in the case of max for live,  ableton is generating the incoming audio via "plugin~" audio is coming from ableton, being processed by max, and finally a signal sent back to ableton via "plugout~"

The patch is designed so that rendered graphics are being manipulated real time- as a response to the lower frequencies of the mix created within ableton live.

Now the part I am stumped at.
 how do I record the "rendered video" response either with jit.record, or jit.vcr(ideally)?

I think this is a problem due to incorrectly routing matrices, or incorrectly labeling said matrices in a correct manner that jit.record or jit.vcr can coherently understand.

I've tried multiple configurations with no success, and various different failures :X

Starting w/ .vcr my first error message (I don't believe I ever overcame, but gave up on the routing configuration and tried jit.record) "make sure DSP is enabled"?? 

I thought this may be an incompatability issue because of the following but I am unclear.
Ableton is Launching/Authorizing Max inside of Ableton, and while Ableton is streaming audio over to Max, Max reads the audio in "preview mode" thus I'm not sure if .vcr can be used because I don't know how to enable DSP. It's my understanding that with M4L, Max is relying on ableton to "forward" the information of the drivers and audio I/O.

With Jit.record I managed to get the setup to "write $" and "stop $", but the return file was always <60kb, never what I intended, and  never anything of any value, other than to learn from another mistake :D

So I reconfigured (not entirely sure the changes I've made, but I keep re-analyzing it and wondering where I am going wrong)...

Where I'm at now with Jit.record.  A few attrui's later, and I thought I might have this figured out. nope. I have set preferences (globally and within attrui's) (that never seem to save correctly and are reset when i reload the patch) and now when I try to "write $" with jit.record Max crashes X(
    
  So max tells me to input the jit.matrices for .record
 or for .vcr matrices+msp audio for .vcr but i'm wondering if I can feed in ableton sound output instead of msp audio)  

Which matrices do i need to connect to the jit.record or jit.vcr, and what kind of specifics need to be "outputted" for those said "matrices"?

If I set the patch up appropriately in max, do you know of any functionality limitations that would prohibit me from "write $" "stop $" a Jit.rec or jit.vcr from within ableton live, with the end result of getting a video file?  

Furthermore an issue is that whenever I try to open the help files (not reference manual) for jit.vcr or jit.record max crashes :( + Online information is limited (atleast until it grows more) and hard to sort out (through various versions) for jitter info.  

Hope that was a coherent enough jumble of jarbled jargon XD   Thx for the tutorial series (@Sam Tarakajian)

 Any help, guidance, or suggestions would be much appreciated!  

essentially my end goal is to have video accompaniment (as a movie file) that syncs to audio's that I create/compose in ableton live, and to keep experimnting.  
I realize there are multiple ways of achieviing this end goal (and probably many more ways that I do not realize or understand).

Ableton Live has a video/audio encoder that will render various different formats *but AbletonLive reads only .mov (*from my understanding*) as input video formats, I think the ideal route is to configure .vcr in max to snag audio from ableton/ and the real time video render from max in synch all in one go (opposed to manually aligning audio to video in a video editor or Live), but the question is how?

If you made it this far, and think you can help, have a look at the patch attached below, with a few things to note.
1)This patch was designed by, Sam Tarakajian, and I merely remodeled to purpose Ableton Live functionality, I take little to no credit for originality.
2) It is a hot MESS.  still working on cleaning and organizing everything.
3)If you open the patch, its currently designed for Max 4 Live audio, repurposing for MSP should be pretty simple?
4)To Help streamline things, look for the big white numbers in the patch. 1-2-3-4-5-6, these are positioned near buttons or switches to help guide/initialize/setup the functionality of the patch.
5)"Write $" attached to .record crashes my program due to some incoherency in the whole inccorectly functioning. not entirely sure what, or how yet.
6) Mistakes have been made XD If something does'nt make sense, I would be more than happy to try and clarify my meaning.
7) Lots of params to adjust some savy looks (similar to the big white numbers look for big white V's (is a variable)) 
8)HAVE FUN, Any and all input related to functionality (related to my intended re-design) would be appreciated!

Cheers
fæt


Hi there all! Using Max for live 7.3+Ableton Live 9.7.  Thanks to the dude837@youtube (Sam Tarakajian) for his "Delicious tutorial" series o

jit·record--jit·vcr-m4l-functionality!-the-quest-continues!

here's how you can record the output of jit.world with jit.vcr, by reading back to matrix with jit.gl.asyncread.

here's how you can record the output of jit.world with jit.vcr,  by reading back to matrix with jit.gl.asyncread.

Hey Rob, thank you for getting back to me so quickly! That certainly helps, but I've ran into a few new problems.

using the jit.gl.asyncread, and the handy attrui(thingy you attached to the jit.gl.async. what is the proper term for that type of attrui with a drop menu to view multiple readouts?) that interprets data-read outs from the jit.gl.asyncread output matrix?

1) i've set it up to a jit.record, and it will write files of 0kb size.

2) by setting to a .vcr (similar to your setup diagram, posted) I did manage to record a video//audio track feeds simoultaneously, but also achieved some poor fps (10 at it's worste) with an AMD FX 4130 CPU and NVIDIA GTX 950REF I don't understand how I am running into bottlenecks. I set the Jit.world to a relatively small size.

3) I set the jit.world size to 432x384, and jit.gl.asyncread destination dimensions changed in the attributes menu to mirror the jit.world size. Although jit.gl.asyncread reflects that same value (when i open the attributes menu or inspector) the readout attrui attached to the output matrix shows the dimensions as double whatever i change the jit.world size to, (even though the destination dimensions of the actual jit.gl.asyncread say 432x384) the final .mov output file is double that dimensional size, which results in only the bottom left quadrant of the screen actually displaying video (doubling 2 dimensions quadruples dimensional surface area), the other 3 quadrants of the mov are written blacked out (empty?) perhaps this is related to my fps performance bottleneck?

I've tried setting the fps of the .world and .vcr to 50 fps (assuming I am setting a "cap" when doing that?) but the frames still dip low.

~I tried setting the vcr cache size to 2gb (probably way more overhead than necessary for these 16mb files) helped a little bit maybe?

~I tried flattening the mix, and rendering the vcr w/ the wav file to save resources on abletons sound processing and that didnt seem to work either.

Do you know of any parameters I can tweak to enhance render capture performace?

And a somewhat unrelated inquiry. I tried to model the jit.world @drawto GEQ0 and the jit.gl.asyncread GEQ0 but max console error reported "an object in use is already named that"? Is there some back end tinkering or programming that needs to be done in order to correlate those two functions (jit.world @drawto x + jit.gl.asyncread x)?

I ended up copying and pasting the jit.gl.asyncread ctx in, and attributing my jit.world @drawto ctx. Is ctx itself an attribute? I figured it as some sort of destination handle?

Im still wet behind the ears when it comes to max XD, holy moley!

I thank you for your insights as I am starting to humbly realize, there is so much to learn about this program and how it can be used an infinite number of ways. I have much to learn, thank you for what you've shared, and my warm regards!

Here is the updated patch if you feel you can help and are curious to look at how ive set functionalities v.2. Again thank you for your help and kindness!

Aha! 
Hey Rob, thank you for getting back to me so quickly! That certainly helps, but I've ran into a few new problems.


I've applied parts of that patch
using the jit.gl.asyncread, and the handy  attrui(thingy you attached to the jit.gl.async. what is the proper term for that type of attrui with a drop menu to view multiple readouts?) that interprets data-read outs from the jit.gl.asyncread output matrix?
1)  i've set it up to a jit.record, and  it will write files of 0kb size.

2) by setting to a .vcr (similar to your setup diagram, posted) I did manage to record a video//audio track feeds simoultaneously, but also achieved some poor fps (10 at it's worste) with an AMD FX 4130 CPU and NVIDIA GTX 950REF I don't understand how I am running into bottlenecks. I set the Jit.world to a relatively small size.

3) I set the jit.world size to 432x384, and jit.gl.asyncread destination dimensions changed in the attributes menu to mirror the jit.world size.  Although jit.gl.asyncread reflects that same value (when i open the attributes menu or inspector) the readout attrui attached to the output matrix shows the dimensions as double whatever i change the jit.world size to, (even though the destination dimensions of the actual jit.gl.asyncread say 432x384) the final .mov output file is double that dimensional size, which results in only the bottom left quadrant of the screen actually displaying video (doubling 2 dimensions quadruples dimensional surface area), the other 3 quadrants of the mov are written blacked out (empty?) perhaps this is related to my fps performance bottleneck?

I've tried setting the fps of the .world and .vcr to 50 fps (assuming I am setting a "cap" when doing that?) but the frames still dip low.
~I tried setting the vcr cache size to 2gb (probably way more overhead than necessary for these 16mb files) helped a little bit maybe?
 ~I tried flattening the mix, and rendering the vcr w/ the wav file to save resources on abletons sound processing and that didnt seem to work either.
Do you know of any parameters I can tweak to enhance render capture performace?

And a somewhat unrelated inquiry. I tried to model the jit.world @drawto GEQ0 and the jit.gl.asyncread GEQ0 but max console error reported  "an object in use is already named that"? Is there some back end tinkering or programming that needs to be done in order to correlate those two functions (jit.world @drawto x + jit.gl.asyncread x)?
 I ended up copying and pasting the jit.gl.asyncread ctx in, and attributing my jit.world @drawto ctx.  Is ctx itself an attribute? I figured it as some sort of destination handle?

Im still wet behind the ears when it comes to max XD, holy moley!
 I thank you for your insights as I am starting to humbly realize, there is so much to learn about this program and how it can be used an infinite number of ways. I have much to learn, thank you for what you've shared, and my warm regards!

Here is the updated patch if you feel you can help and are curious to look at how ive set functionalities v.2.  Again thank you for your help and kindness!

-Chris



hi chris, you've run into a bug with jit.gl.asyncread when fsaa is enabled on the context. simply disable fsaa to workaround.

alternatively, you can skip jit.gl.asyncread, and use jit.world @output_matrix 1 instead, which uses a different technique for reading back to matrix. might be slower or might be fine in your particular case.

i'm not sure if you're on windows or mac. on mac jit.record/vcr will probably perform best with avf engine. on windows you must use viddll, and the most performant codec in my experience is the huffyuv codec (which requires writing AVI files instead of MOV, briefly mentioned in the jit.record helpfile under the VIDDLL tab).

i'm not sure if you're on windows or mac. on mac jit.record/vcr will probably perform best with avf engine. on windows you must use viddll, and the most performant codec in my experience is the huffyuv codec (which requires writing  AVI files instead of MOV, briefly mentioned in the jit.record helpfile under the VIDDLL tab). 

Hello again Mr. Ramirez :) You've been immensley helpful again, thank you so much for your graciousness and insight!

So I tried your advice, switched .vcr to @codec huffyuv and i tried messing around w/ applicable qualities from lossless-max-high (what is the actual difference between max/lossless or say max/high? do different codecs utilize different potential quality settings?)

I managed to get 2 output files (in 15+ attempts) that had readable information, and since then the .vcr is now crashing ableton :(

the huffyuv certainly encoded smoother video, but I ran into a few new problems,

1) as mentioned above, I have since had multiple failed attempts to perform "write $" successfully.

"Write $1" command immediately freezes and crashes Live (after the prompt to select a save location and name the file being saved).

Even reverting back to using h264 encoder with default attributes (besides fps and cache size) seems to crash live.

2) 1 of the test files (avi huffyuv) that was successfully rendered, would play the audio channel in windows media player, and no video. But in VLC media player it would play the Video, with no audio, I'm not sure what I did wrong here, i deleted the file (wish i would have kept to upload) and haven't been able to replicate the problem since.

~The other file didn't seem to have audio

and attempting to open the help files seems to crash MAX if ableton is open, but i did manage to open them with ableton closed (after launching from within ableton) for some reason?

In the VIDDLL specifics it mentions using "sub dictionaries" and "sub dictionary options" to set audio encoder options. is that referring to creating a sub patch? how can i access or create the "sub dictionary" parameters?

I took a peek at http://ffmpeg.org/ffmpeg.html#Options listed in the help section and the language seemed over my head, I've bookmarked the page to further review it and hopefully it will help illuminate a solution.

I have not been able to get a successfull render in the last 20 renders. crashing leads me to tweaking (and referring to any potential errors in the max console and there are none currently)either the jit.world, or the jit.vcr, and I have no clue how it came to do this, or how to get through this :( here is the latest patch model.

All the best, and thank you again for all your directive! It has been very much appreciated!

Hello again Mr. Ramirez :) You've been immensley helpful again, thank you so much for your graciousness and insight!

I am on windows 7 64bit.
So I tried your advice, switched .vcr to @codec huffyuv and i tried messing around w/ applicable qualities from lossless-max-high (what is the actual difference between max/lossless or say max/high? do different codecs utilize different potential quality settings?)

I managed to get 2 output files (in 15+ attempts) that had readable information, and since then the .vcr is now crashing ableton :(

the huffyuv certainly encoded smoother video, but I ran into a few new problems,
1) as mentioned above, I have since had multiple failed attempts to perform "write $" successfully.
"Write $1" command immediately freezes and crashes Live (after the prompt to select a save location and name the file being saved).
Even reverting back to using h264 encoder with default attributes (besides fps and cache size) seems to crash live.

2) 1 of the test files (avi huffyuv) that was successfully rendered, would play the audio channel in windows media player, and no video.  But in VLC media player it would play the Video, with no audio, I'm not sure what I did wrong here, i deleted the file (wish i would have kept to upload) and haven't been able to replicate the problem since.

~The other file didn't seem to have audio

 and attempting to open the help files seems to crash MAX if ableton is open, but i did manage to open them with ableton closed (after launching from within ableton) for some reason?

In the VIDDLL specifics it mentions using "sub dictionaries" and "sub dictionary options" to set audio  encoder options. is that referring to creating a sub patch? how can i access or create the "sub dictionary" parameters?

I took a peek at http://ffmpeg.org/ffmpeg.html#Options listed in the help section and the language seemed over my head, I've bookmarked the page to further review it and hopefully it will help illuminate a solution.

I have not been able to get a successfull render in the last 20 renders. crashing leads me to tweaking (and referring to any potential errors in the max console and there are none currently)either the jit.world, or the jit.vcr, and I have no clue how it came to do this, or how to get through this :( here is the latest patch model.

All the best, and thank you again for all your directive! It has been very much appreciated!
~Chris



you have @outputmode 0 which i guess is causing problems so get rid of that. you also have a cache_size of 3GB, so hopefully you're launching in 64 bit mode to access that much memory, otherwise crashes might occur. try setting it to 0 and see if it's more stable.

if the huffyuv codec is not playing back as expected, i would try h264. also try modifying your window size to something more standard, like 640x480. there may be a bug with handling window sizes of that dim.

the quality arg is not applicable to the viddll engine.

you have @outputmode 0 which i guess is causing problems so get rid of that.  you also have a cache_size of 3GB, so hopefully you're launching in 64 bit mode to access that much memory, otherwise crashes might occur. try setting it to 0 and see if it's more stable.

if the huffyuv codec is not playing back as expected, i would try h264. also try modifying your window size to something more standard, like 640x480.  there may be a bug with handling window sizes of that dim.

the quality arg is not applicable to the viddll engine. 

jit.record & jit.vcr m4l functionality! The Quest continues!