sync phasor~ with count~
I have a master phasor~ that controls synth playback.
I also use this master phasor~ to cue record via [count~], [poke~], [index~]
then trigger count~ index~ to play back
but master phasor~ and count~ playback will not stay in sync
how can I fix this
here is my patch
(my sample rate is 48000)
I think there are too many less effective and unneeded objects between your master or whatever you call it phasor or bang maker, and recording - playback of a loop.
Also reccording is not clear to me, are you overdubbing ?
If not, why feed index into poke while recording ?
You resize buffer to set recording length
but forget that bpm -> sample count is not really the thing to use.
there is allways some rounding involved.
In fact you could use record~ and just tell it to start recording,
it will stop at end if @loop & @append are off.
that would make dirty loop recording without crossfade at loop start/end
but will work, to start with.
Playback part .... as you use sample count I guess you don't want to speed,
reverse or pitch recorded loop.
That again would lead me to better option
use either count~ based sync all the way through,
ot than phasor based.
But not mix both of them.
But not mix both of them.Ok. Thinking about that...
//
Thanks for the math
//
but phasor~ can drive count~ (didn't know that)
one has to look a bit more critical at the progress ...
Now I remeber that phasor~ Pyjamaparty ...
I was suspicious enough to sample your "continuous_phasor"
because even with fixed rest, it looses sync.
it hickups when resetting the ramp count , causing gaps and so it goes out of sync.
recorded together with phasor~ (lower line)
after moving resetter inside gen~ itself (or better say switching looping on/off
it stays perfectly in sync
haha - phasor pyjamaparty - yeah that was the onebang phasor - I string together a bunch of them doing different stuff - like have a lot of conductors conducting bits of the entire piece.
I see what you are saying - took the onebang phasor and had it re-bang itself- losing time
I will look at that
//
...in the last day I went ahead and started making everything with count~
it's messy under construction but sample perfect.
question
when I set buffer using samps it keeps erasing itself as it records
when I use size it is ok
??
both will erase buffer , but size $1 not if size is unchanged.
You see what I mean with ms or samps ....
here is bit modified patch
you use send~ / receive~ stuff , which adds vector size delay.
I would rather wire that directly.
number~ is too slow for anything in sync chain.
I use it only to display stuff sig~ -> float
and count~ as master ... if you use bpm I am afraid
you need to round it to samples, and any conversion
that follows, being ms or ticks, whatever, needs to fix itself to rounded samples.
yes. I have your math for rounded sample
Should have started with that
but end up with explanation- good
//
Send- / receive~ ultimately no choice. - the synths and video stuff is on the other side of the metaphorical planet. No way to hardwire
And once I start making multiple buffer and play modules, they will each live in their on little universe
Only need one recorder.(for now) That can all be hardwired
***maybe I put the players inside poly- can connect then- already have that
**hmmm… might put recorder inside poly as well - what if I want to record the same thing on multiple tacks and effect it differently?
Yes!
I have nothing against send~ / receive~
but if you wire something locally, and send it elswere, you have offset.
You could follow the signal path and compensate.
If you want to access same recording several times, no need to record it multiple times.
You can play same buffer using 100 instances of player.
are you hard switching punch In/out like that ?
if you loop sustaining sound even fade in / out at loop points won't do.
One should fade in on rec start, record end fades out but recording proceeds at the
beginning fo set fade time, to create perfect crossfade.
You could follow the signal path and compensate.Yes.
You can play same buffer using 100 instances of player.I was unclear. My plan is to use pitch detection and aptitude detection that I built a few years ago and play the new synths along with the bass. I should record on separate tracks. No problem.
are you hard switching punch In/out like that ?Since I play upright bass, and don't use a foot switch, and I cannot position a laptop anywhere near, I make basic controls in Mira and put on a music stand. -> touch the count-in then play.
(I have foot switch, but like it better touch and play- maybe change later)
One should fade in on rec start, record end fades outYes many ways to do that - I found a nice trick recently, hmmm.. if I can find it
//
the modified phasorparty~ is excellent. thanks. seems to work well now with count~
...and I just switch it out with the one in poly~ phasor bank that I already built.
main thing in a looper is if you need to overdub or not.
even if not, performing that short crossfade needs a short overdub.
one thing with count~ - poke~ based recording is that
even if one stops count~, poke~ keeps overwritting last set sample
position.
Can produce clicks ...
here is very basic looper count~ based with record crossfade and latency compensation
(in case one needs precise overdubbing)
which also allows higher vector I/O size
Many years ago I made doublebase pitch detector, half hardware, half software based.
At that time pitch detection in that low range was really slow.
I shifted bass 2 octaves up using distortion, filters and zerocross detector
LM2907 based.
Together with simple env. follower, signal was then sent to powerbook and
software pitch detector in max , I don't remember which one.
Was much faster then trying to detect pitch using original sound.
and .. it was 4 piezo pickups directly under strings which only served that purpose.
Thanks for the patch.
My 'real' upright bass is in the US.
Using electric upright. Cheap Stagg bass. (should have mentioned that)
it gets the job done - pitch detection down to low E