Odd peek~ behavior using m4l patch in live, works in max...

[pucklermuskau]

so, im having very strange behaviour out of the peek~ object used to write a buffer in a m4l audio rack. I attach the patch as dumbed down as i can make it, the patch works fine when open for editing in max, recording through the buffer normally. But when i close the patch up, the peek~ signal plays fine, but nothing gets recorded into the last tenth of the buffer length (regardless of buffer length). I'm at a loss here, is this a bug or what?

http://lokeymassive.net/misc/peektest.amxd
 

[hanc0ck]

Dear pucklermuskau!

I tried your patch and it worked from me like I think it should.
I put it into an own subpatcher so I could see what's going on. Than I set the number box to 44100 samples, so for me 1000 ms.
And it continuously fills the buffer with incoming audio signal. And it has the length of 1000ms.

I made a screenshot to proove it:

Maybe that helps finding the error.

h.

[pucklermuskau]

thanks for looking into this, but i should note that when i have it open in max, like you have, then all seems to work properly. Its only when i use the amxd inside of live that the buffer seems not to record all the way. Are you able to try it in m4l? Or do you just have max?

[hanc0ck]

You missunderstood me. My test was in normal mode inside ableton and NOT in edit mode.

You used the presentation mode. But as I wanted to see whats going on even if I'm inside ableton I encapsulated your hole patcher and double clicked on it (when it's inside ableton).

Worked perfect for me.

[pucklermuskau]

oh, ok. Thanks. How odd. I'm little stumped why the buffer wouldn't be writing for me in live. The signal going into the peek~ object remains regular, and the faction of the buffer which doesn't record stays constant, despite changing the length of the buffer. Odd...but thanks again for helping me look into this!

[Jabbon]

Hi there
You do not resize the buffer length, so changing the BPM does not actually affect how much you wnat to record (you just change the display length of the waveform~ object). This also means that retriggering the phase of Phasor~ is not aligned.

If you want to change the buffer~ length you will need to change it with "size" + msec.
If you need to sync the recording why not simply use a plugphasor~ or phasor~ (synced to Live) with a >~ 0. and count with the amount of samples? You will need to come up with a simple (rate~ ?) solution to sync everything, but you stay in the signal domain, without the need of bangs to reset phasor´s phase.

<pre><code>
----------begin_max5_patcher----------
839.3oc0XtsaaBCFG+ZxSgEWmkgsgPYWLosWiopImfSBafAgMacqp8Ye3CjR
6BNNmPEEoXYvX96e96D9wYd9qJefx8AeB7Mfm2iy77TWRdAOSeO+BxCqyIb0
v72jWxZJ7mquUUMkSYBhHqj88Z5ZgdpPQwKBlCPgpl.yef6MOUYiHmJD+ohp
GtbRIB+4.+UD1V+8iqhHVuKiss+LuTManHUCLLR0bWuIOKUoyxU+3CHbmN2T
xDLRg584+k5LRN3qk4oc2tcEkwZkjZIB68L7r+pdFn70serZ8qFLRdwmlMS9
2bGQHi96V008tEzGTqM+OBPAAAK7sCIavAGpgCR0DBsxFz4vFzovF30iMBvl
1eq7OpAj6VRXjwDRCqkVgE7laHgOCXUP4bxV5+QKHaPPYEIgZV.09XXzhHYm
CiDXxjx9olv34DAEvJEzeQxanbPA+xvThtAEZyzAFeyMcthbptEQOOHU3YaY
jb6wmMgftSaBYM7LL5coIzPdUxnymmAiITiNmEFq8wFfIgSI2p1fsop.sCgE
WiBa.jcyE7jxSprgkBfKNOKFThwVomgyfb48YlbmJTDiTEoDiu10IZrqhcoL
QHbJUl3KoxJ3.NonJmxunvRlX0PHxJjBdW588ZqrK4CKLdbPWrXRlRFLU4Ma
yXGMq9bmxuaJ5A6RUyKu4wktlTp7mzmAhrs6Df8wCNjch1.ITGSNNwF.Nq5a
vmB.Btd.XUylMz5NDXsXG28pVZBsnZRrVnb3TxmZca1cw0nPYrNCUnUzfmR0
D94mAVNDiSfLv9kKOzGlOkHiLPb0NBur1kfw4YbgKbx3bEa8zdF2JbTJncAv
d6QKpVExq+Z3wKapW2QftOZF7xBIkxEYLUsi8Fj7Dr.v8CZWVZJUc+NAVjkV
UlwDFQ.t+faktpI4Y.060YQSAimlPtvI73pIrCZR9o+i8d2Q0Tv3xIolPGSS
Iiplftvo2.yatlRbYuKdb0TjKbZbkjqdcimjV5fjNQEASRjGJcXf5COPvW5c
gZE6nVGuXVxSB736nirunSw1CGWM4RNP7os0gzlY3n6jMn3v88tPw5TrrQNv
QfCZ5DSCgQZ9En.WGMk8tzjlNtWet3qsySy9G.FEvrH
-----------end_max5_patcher-----------
</code></pre>
a 30 seconds example, so check it out before