does anyone use Scale Time?

[Shooshie]

Originally posted by Artspoke:
It must be something about me that kills a good thread.

If there can be no solution, then at least be warned: you cannot accurately scale time on audio flies with DP.

By "accurately" above, I mean "sample accurate" which is what is advertised. You may not be able to understand why this could be useful, but MOTU must, or they wouldn't have bothered to mention it. I need it, and would love to hear from anyone who has a suggestion about how I can find a piece of software that works.

Thanks,
-James

P.S., Ideally, I'd like to be able to identify WE beats at a rate of one per musical beat on phase correated multrack recordings, and then quantize the audio to a different tempo map/grid than that of the material - and retain phase correlation between the tracks. It's not even a little interesting to me to hear about how a difference as small as one sample is "ok", nor do I care to be asked what moral reason I have for wanting such a feature. We all actually need it, and somehow no one can see that.

It's not necessarily you. It's more that you changed the direction of the thread, and in doing so you implicated a substantial chunk of Digital Performer, claiming that it's all broken. Doesn't work as advertised anymore, if it ever did. Well, that's news to most of us, so if we're going to participate in the thread, we have to corroborate your story, and that's a big job. We have to see if you're right, or if there is something missing, or if you're just plain wrong. Then we have to figure out whether this is new, or whether it's always been like this, or when it changed.

With no answers to any of the above, it's easier just to be quiet and see if anyone else has been worried about it. Obviously, phase-accuracy is important if you want to stretch (or scale or groove quantize, etc.) two tracks by the same amount, because you certainly don't want to be introducing phase cancellations when doing this. But this is the sort of thing that I have to put on hold until I come across it in my own usage, and then I have to test it.

Not that I don't trust you to identify a problem. At least, not less than myself. It's just that there are a lot of variables, and it won't be easy to corroborate what you're saying until I come across situations in which the errors are obvious.

Shooshie

[Artspoke]

OK,

Here's an easy test that will show you what I mean, and it'll be worth the ten minutes it takes:

1) Open a new session
2) Route the click out to DP busses, and make the inputs of six different mono audio tracks the same busses as the outputs of whatever you set the click's outputs
3) Arm the automatic punch in to bar 2, beat 1, and punch out to bar 10, beat 1
4) Record, one pass at a time, eight bars of click, at 120 BPM (the default), into each of the audio tracks
5) Change the tempo of the sequence to 111 BPM starting at bar 1, beat 1
6) Select all six aoundbite, and Audio -> Adjust Soundbites to Sequence tempo

You wil now see six different results. None will match the new tempo. The six tracks are likely to now each be different lengths - from each other, and none will be eight bars long as they should be.

Since the routing is digital, these should be identical files to begin with, yet DP interprets what to do with each track differently, and never correctly.

Now imagine if each of these tracks included different mics on the same drum set!

No matter what method you use to try and get the clicks (recorded at 120 BPM) to line up at 110, Scale Time, Scale Tempo, Adjust Beats within Soundbites, the sideways hand - none work, which is why I go out on a limb and say, "there's a bug".

Thanks,
-James

[Artspoke]

Naw, it's just me

[danny]

Wow, I'm going to have to try that test.

I've experienced the imprecision of DP's time stretching when creating loops or real drum performances. If the loop was musically exact, I would enter its length in the Set Soundbite Tempo dialog, which then calculates what its tempo must necessarily be, a near-perfect number like 123.98. With the chunk's tempo set at 124, I'd invoke the change soundbites to seq tempo' command and almost never got a perfect result, although I didn't get the wildly varying results reported by the original poster.

[Artspoke]

Yes,

The imprecision you describe is multiplied within commands that utiliaze the time stretch function multiple times over in one "batch". I mean llike when you try to do a Quantize Beats Within Soundbites for example. For this command, DP tries to stretch each section as defined by the spaces between detected beats. The little imprecision problem then happens between every two detected beats, resulting in mayhem. Try it! It's a bug for real, but the interest in it is low, so beware posting. There seems to be a bit of an issue around here throwing around the word "bug".

I agree that every poster should be wary of crying wolf every time s/he has a little problem with DP - which is excellent software, overall. This is an easily reproduced problem. Maybe I should start a thread about why some bugs are more interesting to members of this BBS than others.

There should definitely be an Ethics forum here, where folks can run on and on about the "why" parts of our uses for DP. Then this forum could be reserved for "how" to make MOTU aware of technical issues so they can fix them. There is no workaround for this issue, which makes it a really annoying bug, since we cannot completely control our recordings with respect to time. This is not debateable -- the features regarding time compression/expansion ARE BROKEN!! What else can I say???

Troubled in time,
-James