Print

[jonobo]

I have a Track with time-sync problems on .wav-export.

Settings in Milkytracker:
BPM:086
SPD: 06
Mixer Resolution: 44100 Hz
HD Recorder Output: 44100 Hz

Settings in Ardour:
BPM:086
Samplerate: 44100Hz

Strangely the exported .wav is not on time - what could be the reason?

Here are links to the .wav & .xm:
http://ooommm.org/mp3/20100211___mellowdice.xm
http://ooommm.org/mp3/086bpm___mellowdice_3A_FullMix.wav

Just load the .wav into any Editor with BPM-Grid, set it to 86bpm and then try to play Bar/Measure 21 and you immediately see the problem, the bassdrum comes way before the "one" of Measure 21.

What could be the reason for this?

[Saga Musix]

MOD "BPM" are not perfect. They're just very approximate, which is due to their origin - i.e. the timing routines on the amiga, either the CIA or vblank timing. You won't get exact BPM values in classic trackers.

[jonobo]

MOD "BPM" are not perfect. ... You won't get exact BPM values in classic trackers.

:'( *snif* :'(

That breakz my heart.

Zeeeriouuuuuuzly.

Come on - this is Hi-Tek 2010 but still the drummer can't play on time? This can't be true... ...there must be a solution... ...at least a work-around.

I tried to find out the BPM by zooming closely in Ardour and adjusting the Tempo 'til the Grid fits, but am not sure about my findings, because i am not sure if HD-Recorder adds silence to the beginning of the .wav-File. So manually i came to the conclusion that the .wav-File linked above is somewhere around 86.24BPM and labile in Tempo over the course of the track (BPM changes slightly over time - this is hell!).

So the next questions would be:
1. Is there a way to calculate how much exactly Milkytracker is OFF the BPM?
2. Is Milkytracker always the same "bit" off the Beat, or does that vary over time?
3. Are the .wav-Files exported from Milkytracker via the HD-Recorder starting EXACTLY on time, or is there a little silence added in the beginning - and if there is silence added in the beginning, is the amount of added silence constant or does it vary?
4. Is it possible that this only happens in heavy-loaded tracks, so, if i constrain myself to four Tracks i can dodge that?

This is really important to find out for me, because i want to be able to produce clean Instrumentals that are easy to record on later - and that means STABLE speed that only changes if i tell it to do so.

So i am happy for any Tips that get my out of thoughtloops like "OMFSHT how could it ever come to my mind to use a computer to make music, i wasted half my life on this and the fkkking digital drumsukkka iz ztill off-beat and i'm never gonna use a stoopid machine again..." *startshackinglaptopintopieceswithbaseballbatanglegrinderaxeofdeath* 

[Saga Musix]

Come on - this is Hi-Tek 2010 but still the drummer can't play on time? This can't be true... ...there must be a solution... ...at least a work-around.

You got it wrong. Milky is playing perfectly on time, but probably not on the same time as you want. And yes, even in 2010, we tracker developer still make trackers that are compatible to their roots (which is very important in Milky's case for example).
Again, milky keeps the BPM, no matter how many channels you use, but the BPM count in milky is not an integer value. In your example, Milky is playing at about 86.07 BPM. Some trackers like OpenMPT or Renoise have a modern tempo mode, which is not based on the Amiga timing, but working with "real" BPM. Milky does not have this.

[jonobo]

You got it wrong. Milky is playing perfectly on time, but probably not on the same time as you want. And yes, even in 2010, we tracker developer still make trackers that are compatible to their roots (which is very important in Milky's case for example).
Again, milky keeps the BPM, no matter how many channels you use, but the BPM count in milky is not an integer value. In your example, Milky is playing at about 86.07 BPM. Some trackers like OpenMPT or Renoise have a modern tempo mode, which is not based on the Amiga timing, but working with "real" BPM. Milky does not have this.

Yeah!

*celebrate*

*partymode*

Yeeeee-Haaaaa!

Never been so happy to be wrong :[]

Yes, i pray daily to all Tracker-Developers since i stumbled on Milkytracker :!
Including you, since i found your website a few minutes ago :]
And i really appreciate all Tracker-Hackers work and i really want to try to make some good music via tracking, and to be able to do that i have to ask questions :]

So if i understand that right by now Milky is playing perfectly on time.
But my exported .wav IS NOT on 86.07bpm as far as i can see in Ardour (this is making me insane), so
- is there silence added at the beginning of the .wav on HD-Recorder-Export?
- is it somehow possible the export got off Milkytime andor that the exported timing is unstable throughout the .wav?
- or is it maybe an Ardour-Error and my .wav is perfectly right?

*still confused, but on a higher level*

Is there a formula to calculate the closest possible RL-BPM from Milky-BPM?

*feelin' embarrassed to post this but have to do it anyway to prevent me from goin' even more insane than i already am*

[Saga Musix]

- is there silence added at the beginning of the .wav on HD-Recorder-Export?

I don't use milkytracker, but I suppose it wouldn't do something stupid like that.

- is it somehow possible the export got off Milkytime andor that the exported timing is unstable throughout the .wav?

No. Wav Writers are not realtime, they use perfect timing.

- or is it maybe an Ardour-Error and my .wav is perfectly right?

well, i dunno ardour, so i don't know how it works.

Is there a formula to calculate the closest possible RL-BPM from Milky-BPM?

I guess yes, but I don't know it by heart. It's a bit complicated to transform the formula used in OpenMPT into something usable, hehe

[jonobo]

Thanx @ Saga 

Two things seem sure now:

1. MilkyTracker-HD-Recorder is adding silence at the beginning of exported .wav-File (at least as far as i found out - gonna ask others on IRC about that and point them here).

2. I need the formula to calculate the closest possible RL-BPM from Milky-BPM (gonna ask that in IRC too and point people here) (am happy to decipher the formula from any weird Code-Snippets, Pseudo-Code-Poems, Math-Murderers, MagicScrolls - even would drink it as a Potion and channel the results from the great Spaghetti-Monster).

[Saga Musix]

Actually, I just check it out and Milky indeed appears to prepend a few samples of emptiness. It shouldn't be too complicated to get rid of that, though.

[jonobo]

Yeah, i learned a new word: "prepend". Nice Word.

So i'm not totally insane (I AM) - the silence is there. To fix it i gotta learn coding first, back in five minutes... fifty minutes... fivehundred minutes... 5k minutes... 5gogol minutes...

[DasKreestof]

The difference between real BPM and tracker BPM is explained very well here:
http://modarchive.org/forums/index.php?topic=2709.0