First try at audio quantizing - not great
Moderator: James Steele
Forum rules
This forum is for seeking solutions to technical problems involving Digital Performer and/or plug-ins on MacOS, as well as feature requests, criticisms, comparison to other DAWs.
This forum is for seeking solutions to technical problems involving Digital Performer and/or plug-ins on MacOS, as well as feature requests, criticisms, comparison to other DAWs.
First try at audio quantizing - not great
I wanted to take some drum tracks and quantize them to the grid using MOTU's Beat Detection Engine and PureDSP. I recorded 8 tracks of drums straight into DP, played to a click.
Then I made a kick/snare guide track to use as a beat template. Copied the beats from the template track onto the other tracks and they showed up with an offset. Unusable. Talking with MOTU tech support made this problem worse (that's another topic for another day.)
Tried a different session and beat copy worked better. Then tried to quantize audio. 2 hours of fiddling with beats and quantize settings yielded no usable tracks on anything except some "boom, chick" passages. Flamming, phase errors, and gritty cymbals were the norm, not the exception.
If anyone has had better luck in a similar situation I'd love to hear some tips or tricks to make this work as it would be an enourmous time saver for live drums. Would also love to hear comparisons to PT's or Logic's beat detection and elastic audio.
Then I made a kick/snare guide track to use as a beat template. Copied the beats from the template track onto the other tracks and they showed up with an offset. Unusable. Talking with MOTU tech support made this problem worse (that's another topic for another day.)
Tried a different session and beat copy worked better. Then tried to quantize audio. 2 hours of fiddling with beats and quantize settings yielded no usable tracks on anything except some "boom, chick" passages. Flamming, phase errors, and gritty cymbals were the norm, not the exception.
If anyone has had better luck in a similar situation I'd love to hear some tips or tricks to make this work as it would be an enourmous time saver for live drums. Would also love to hear comparisons to PT's or Logic's beat detection and elastic audio.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
- kassonica
- Posts: 5231
- Joined: Sat Oct 30, 2004 11:01 pm
- Primary DAW OS: MacOS
- Location: melbourne, Australia
- Contact:
Re: First try at audio quantizing - not great
Creativity, some digital stuff and analogue things that go boom. crackle, bits of wood with strings on them that go twang
Re: First try at audio quantizing - not great
I actually started my session by reading MagicDave's post.
Since the time compress method didn't work, the next step is to try the slice and fill method and see if that offers any utility.
Since the time compress method didn't work, the next step is to try the slice and fill method and see if that offers any utility.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
Re: First try at audio quantizing - not great
Not much better results with smoothing audio edits. Looks like I'll have to hire Vinnie Colaiuta after all.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
Re: First try at audio quantizing - not great
quick question... what was your sample rate?
- Shooshie
- Posts: 19820
- Joined: Sat Oct 16, 2004 10:01 pm
- Primary DAW OS: MacOS
- Location: Dallas
- Contact:
Re: First try at audio quantizing - not great
We just had a huge thread about that only in the past few weeks. Digital Performer was the originator of this feature, and was widely hailed as the wunderkind of pitch and time stretching. This was a part of DP back in the mid/late 1990's. It has continued to improve over time, but is rather taken for granted, and thus doesn't have a catchy name like "elastic audio." The newer implementations by Pro Tools and more recently Logic each have features that make them attractive, but in essence, they do the same thing that DP's been doing for over a decade.Theodore wrote:Would also love to hear comparisons to PT's or Logic's beat detection and elastic audio.
Since DP7, some users are saying they think that DP's time stretching/pitch transposing has even gotten better. We have nothing to prove that other than some anecdotal stories, as MOTU is very quiet about the things they do to improve DP. In any case, it sounds excellent. You need to know that there can be a bit of an art to doing this. Sometimes the transients simply aren't clear enough in the audio track to identify the exact beat points. Some people overcome this by recording either a metronome click or a simple percussion track (drum or woodblock, etc.) with a single hit on each beat. Using that as their guide, they can then better demark the beats. Then there's the matter of offsetting it a bit to be sure you catch the attacks, not just the peaks. I'd prefer that you read the thread that was here recently. If you look up "Elastic Audio", you'll surely find many references to a single thread. That's probably the one. There are some good tips in that thread.
Some things may seem more difficult in DP at first, but once you learn them they go very quickly. It gives the user a lot of latitude so that the machine isn't overriding your creative impulses. Anyway, don't despair that your first attempt was not a big success. Learn how DP works, then figure out exactly what you want it to do. With so many people here raving about its abilities in that regard, I can't imagine that it would take you too long to find a workable method for yourself.
Shooshie
|l| OS X 10.12.6 |l| DP 10.0 |l| 2.4 GHz 12-Core MacPro Mid-2012 |l| 40GB RAM |l| Mach5.3 |l| Waves 9.x |l| Altiverb |l| Ivory 2 New York Steinway |l| Wallander WIVI 2.30 Winds, Brass, Saxes |l| Garritan Aria |l| VSL 5.3.1 and VSL Pro 2.3.1 |l| Yamaha WX-5 MIDI Wind Controller |l| Roland FC-300 |l|
Re: First try at audio quantizing - not great
It was 48k, 24 bit.mrgkeys wrote:quick question... what was your sample rate?
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
Re: First try at audio quantizing - not great
By how much did you try to stretch?
I find that drums are by far the hardest source material for time stretching, worse than full mixes. Transients, cymbals and drum room sound are very unforgiving when it comes to artifacts, and in drums those three are pretty much all you got, with nothing else left to mask...
With DP I've had mixed results with audio quantize, probably about 50/50 between times when it worked and times when it created too many artifacts. (Haven't used it in DP7 yet...) BIAS Peak 6 sometimes can stretch where DP can't.
Of course regardless which algorithm you use, you can only stretch by a relatively small percentage. Usually I find I can compress (speed up) by a good bit more than I can stretch (slow down) before things get ugly.
I find that drums are by far the hardest source material for time stretching, worse than full mixes. Transients, cymbals and drum room sound are very unforgiving when it comes to artifacts, and in drums those three are pretty much all you got, with nothing else left to mask...
With DP I've had mixed results with audio quantize, probably about 50/50 between times when it worked and times when it created too many artifacts. (Haven't used it in DP7 yet...) BIAS Peak 6 sometimes can stretch where DP can't.
Of course regardless which algorithm you use, you can only stretch by a relatively small percentage. Usually I find I can compress (speed up) by a good bit more than I can stretch (slow down) before things get ugly.
Re: First try at audio quantizing - not great
I was working with a pretty good drummer so we are talking ticks, not 16th notes as far as how much stretching is going on. I have found DP's stretching algorithm to be kind of unforgiving in the past but I am encouraged that Shooshie finds DP7 to be an improvement in this area.Kubi wrote:By how much did you try to stretch?
I find that drums are by far the hardest source material for time stretching, worse than full mixes. Transients, cymbals and drum room sound are very unforgiving when it comes to artifacts, and in drums those three are pretty much all you got, with nothing else left to mask...
With DP I've had mixed results with audio quantize, probably about 50/50 between times when it worked and times when it created too many artifacts. (Haven't used it in DP7 yet...) BIAS Peak 6 sometimes can stretch where DP can't.
Of course regardless which algorithm you use, you can only stretch by a relatively small percentage. Usually I find I can compress (speed up) by a good bit more than I can stretch (slow down) before things get ugly.
When quantizing audio across tracks it really sounded as if the stretching was not phase coherent across all tracks - flamming would be worse in some areas than others. If the algorithm is not consistent across all tracks simultaneously then no amount of improvement in fidelity will help this problem.
Shooshie, thanks - I'll look for that thread. I'd really like to hear from someone who's been successful at this.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
- FMiguelez
- Posts: 8266
- Joined: Sun Oct 24, 2004 10:01 pm
- Primary DAW OS: MacOS
- Location: Body: Narco-México Soul/Heart: NYC
Re: First try at audio quantizing - not great
.
If you really care about your production, and the drummer's playing is decent, I would do all the time corrections by hand.
This way, you can forget about artifacts and undesired things. Just group your tracks and cut and move away.
Lots of work, but the results can be great. Certainly better than just quantizing the audio... (YMMV).
HOPEFULLY it won't be too many edits (providing the drummer can play in the first place).
If you really care about your production, and the drummer's playing is decent, I would do all the time corrections by hand.
This way, you can forget about artifacts and undesired things. Just group your tracks and cut and move away.
Lots of work, but the results can be great. Certainly better than just quantizing the audio... (YMMV).
HOPEFULLY it won't be too many edits (providing the drummer can play in the first place).
Mac Mini Server i7 2.66 GHs/16 GB RAM / OSX 10.14 / DP 9.52
Tascam DM-24, MOTU Track 16, all Spectrasonics' stuff,
Vienna Instruments SUPER PACKAGE, Waves Mercury, slaved iMac and Mac Minis running VEP 7, etc.
---------------------------
"In physics the truth is rarely perfectly clear, and that is certainly universally the case in human affairs. Hence, what is not surrounded by uncertainty cannot be the truth." ― Richard Feynman
Tascam DM-24, MOTU Track 16, all Spectrasonics' stuff,
Vienna Instruments SUPER PACKAGE, Waves Mercury, slaved iMac and Mac Minis running VEP 7, etc.
---------------------------
"In physics the truth is rarely perfectly clear, and that is certainly universally the case in human affairs. Hence, what is not surrounded by uncertainty cannot be the truth." ― Richard Feynman
Re: First try at audio quantizing - not great
+1. I would much rather fix things by hand and it is much faster than trying to use MOTU's convoluted beat detection. It drives me crazy when I spend more time putzing with the UI than making the music happen.FMiguelez wrote:.If you really care about your production, and the drummer's playing is decent, I would do all the time corrections by hand.
Honestly, I found DP's beat detection methodology USELESS. Yes, that's right. I said it... U... S... E... L... E... S... S... USELESS. That one's for you Shooshie. That's not to say that you can't do powerful things with time stretching - you can - but I would never use the beat detection hooey.
I've worked on a number of live projects that were not recorded to a click. Then in post-production they wanted to add loops... nice. So I used the Find Beat tool and went in and defined every beat in the song leaving me with a tempo map in the Conductor track that matched the song. I then merged all the soundbites with the start of the song and then switched the Tempo control to "tempo slider" adjusting the tempo to the relative average of the tempo of the song. Then the magic begins. Select all tracks, and under "Audio" in the menu bar, select "Adjust Soundbites to Sequence Tempo." Bang! DP will timestretch every beat to the grid. You can also do this process for individual or groups of tracks to fix just the drums or a rhythm guitar or a specifically problematic track.
Now I have to confess, it doesn't always work on every track. Sometimes the audio freaks out and just turns to grunge on a particular track. No rhyme or reason to it, it just garbles it. I haven't found a fix for it, but sometimes since I'm posting so much with the track I'll use a portion that worked in another part of the song and rebuild the track. Wouldn't be the first time for a live project!
Anyway, I did it this way before DP had the beat detection thingy - which I thought would make things easier - that requires you to define all the beats on every track because the copy beats function is broken. This was a huge disappointment to me because I had to do this a lot. I just thought the Beat Detection would be much more intuitive and be "smarter" at finding relatively consistant quarter notes or eighths rather that grabbing every transient and then having to go and delete every non-eighth or quarter detected "beat." That's just ridiculous... and USELESS to me. I'll stick with what is fast... and works!
MacBook Pro Quad 2.4GHz i7 • 10.12 • 16G RAM • DP 9 • MOTU 896HD Hybrid, Apogee Duet, & MOTU Micro Lite MIDI interface • Waves Platinum, Studio Classics Collection, Abbey Road, etc... • Fabfilter Pro-Q2 • Soundtoys FX • IK Amplitube 3, Ampeg, and TRacks 3 • Altiverb 7 • Slate Digital Everything Bundle • Stylus RMX • Komplete 10 • SampleTank 3 • Arturia V Collection • M-Audio Axiom 49
- Shooshie
- Posts: 19820
- Joined: Sat Oct 16, 2004 10:01 pm
- Primary DAW OS: MacOS
- Location: Dallas
- Contact:
Re: First try at audio quantizing - not great
Wow! Thank you for thinking of me!Splinter wrote:Honestly, I found DP's beat detection methodology USELESS. Yes, that's right. I said it... U... S... E... L... E... S... S... USELESS. That one's for you Shooshie.

|l| OS X 10.12.6 |l| DP 10.0 |l| 2.4 GHz 12-Core MacPro Mid-2012 |l| 40GB RAM |l| Mach5.3 |l| Waves 9.x |l| Altiverb |l| Ivory 2 New York Steinway |l| Wallander WIVI 2.30 Winds, Brass, Saxes |l| Garritan Aria |l| VSL 5.3.1 and VSL Pro 2.3.1 |l| Yamaha WX-5 MIDI Wind Controller |l| Roland FC-300 |l|
Re: First try at audio quantizing - not great
Not exactly useless. I tried a different tack.
Made a Kick/Snare comp track and beat detected that. Copied beats to all drum tracks. Made soundbites from beats. Quantized.
This got me 90% there. Had to clean up a few edits, and had to manually tweak some tom hits and such that weren't on the kick/snare comp track. Also flams (intentional)presented some confusion for beat detector.
All in all, much better than manual, not quite as hip as PureDSP would have been if it had worked. But the most imprtant thing is time is flawless and can be grooved or whatever, and phase coherence is 100%.
Made a Kick/Snare comp track and beat detected that. Copied beats to all drum tracks. Made soundbites from beats. Quantized.
This got me 90% there. Had to clean up a few edits, and had to manually tweak some tom hits and such that weren't on the kick/snare comp track. Also flams (intentional)presented some confusion for beat detector.
All in all, much better than manual, not quite as hip as PureDSP would have been if it had worked. But the most imprtant thing is time is flawless and can be grooved or whatever, and phase coherence is 100%.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02
Re: First try at audio quantizing - not great
Theo:
So you are saying it worked on one song and not another? The Copy Beats tool never worked for me. It would place random transient markers all through the song that had nothing to do with the source or target track. It was very frustrating. Maybe it was just earlier versions of DP and now they've fixed it. Anyone try it in DP7, yet?
So you are saying it worked on one song and not another? The Copy Beats tool never worked for me. It would place random transient markers all through the song that had nothing to do with the source or target track. It was very frustrating. Maybe it was just earlier versions of DP and now they've fixed it. Anyone try it in DP7, yet?
MacBook Pro Quad 2.4GHz i7 • 10.12 • 16G RAM • DP 9 • MOTU 896HD Hybrid, Apogee Duet, & MOTU Micro Lite MIDI interface • Waves Platinum, Studio Classics Collection, Abbey Road, etc... • Fabfilter Pro-Q2 • Soundtoys FX • IK Amplitube 3, Ampeg, and TRacks 3 • Altiverb 7 • Slate Digital Everything Bundle • Stylus RMX • Komplete 10 • SampleTank 3 • Arturia V Collection • M-Audio Axiom 49
Re: First try at audio quantizing - not great
Yeah, when I first tried it the beats would be copied to the other tracks out of sync by more than a quarter note. And like you, sometimes it would just not work. But I went to another session and it worked like a charm. Tech Supp said it was corrupt audio files, but I'm skeptical because it happened even after bouncing the file and reimporting.Splinter wrote:Theo:
So you are saying it worked on one song and not another? The Copy Beats tool never worked for me. It would place random transient markers all through the song that had nothing to do with the source or target track. It was very frustrating. Maybe it was just earlier versions of DP and now they've fixed it. Anyone try it in DP7, yet?
It does work. It's just obviously very unreliable.
And the beat detection engine worked pretty good too, though it was sometimes up to 10 ticks off the start of the transient.
8Core 2.8
10g RAM
DP6.02
10g RAM
DP6.02