i just wonder how good the pitch shift and timetsretch algo of the octatrack works with different kind of
audio material. iam thinking of speech, music (polyphonic and monophonic) sources and unvoiced / noise sounds . over the years i mostly used ircams supervp for max with superb results both for pitch shifting and time stretching. how good is the ot in this respect?
anyone here made direct comparisons with software based algos?
1 Like
First thing come in my mind :
You talk about a software programmed by one of the most prestigious laboratory in sound experimentation on a computer with the power of a computer…
Second thing coming :
A lot of OT user seem to not find this part very useful and somehow a lot deactivate the pitch shift/timestrech when working on the octatrack.
Third thing coming :
I never heard a NICE and precise pitchshifting/timestretch alogorytm on a hardware gear. (it’s JUST OK… no more than that)
So I would say it’s depend from start to Finnish what is the goal but I would say if the goal is precision and values are crazy the Octatrack will probably not handle it very well accordingly with what you already know and the habit you already have with ircams supervp for max.
And to be honest the OT is not build for that it became a beast in the accumulation of all its functions to take materials and make things radically different and completely crazy as completely newborn materials
Regarding precision in Pitch Shift and time stretching it’s difficult to beat Computer and Dedicated Software with the latest technological advances in the field.
9 Likes
thanks for your feedback william,
i was thinking something like that, but hoping you would tell me the opposite.
iam using the digitakt with rk002 cabel doing polyphonic stuff. but since the samples are not pitchshifted,
simultaneously triggered sounds are not time aligned since higher pitched sounds will be played faster than lower pitch sounds etc. so i thought about the pitchhshift function on the ot.
i see that the ot was not made for things like that. but i would like to hear how it does it…
1 Like
I only had the Pioneer Toraiz SP-16 briefly, but I thought it had a decent/nice algorithm… better than most
1 Like
Roland VP-9000 seems pretty amazing imo
Definitely showing it’s age now in the OT
When mk2 was announced thought they would have at least updated the pitchshift / timestretch algos. Maybe an os update they will.
Gets real grainy real fast unfortunately.
It can’t compete with a daw or top plugins
3 Likes
I’m not a live user so I can’t speak for the latest versions, but to my ear the OT timestretch is at least as good as Live 7 (the most recent one I’ve spent time with) but not as good as Roland’s old Elastic Audio from their early 2000s samplers and the V-synth. But Elastic Audio wasn’t a real-time algorithm, and involved a lot of preprocessing and conversion to a proprietary format, so even given the age of the Roland algorithms (they were already about a decade old when the OT was first released) it’s not really a fair comparison.
You can definitely hear the difference if you enable timestretch on a sample even when you aren’t actually changing the pitch or duration, I’d say it’s comparable to the difference between CD audio and a 256kbps mp3 (you probably won’t notice in a complicated track but there’s a noticeable loss of detail and some high frequency rolloff if you A/B). It’s good enough that I have no trouble living with it as part of the OT’s sound but it’s not going to replace Melodyne any time soon.
1 Like
Yeah gonna agree with the other posters. If you absolutely need something stretched for timing reasons then it’ll do the job, but it sounds pretty ugly. I tend to avoid it whenever possible.
Obviously there’s always some beauty in abusing crappy algorithms, etc etc
2 Likes
It bothered me a bit when I first got it but it hasn’t been an issue for me in practice.
I’m pretty sure I said it before but I was listening to Legowelt a couple weeks ago (I forget what, Vaporware Tracks I think) and there were really obvious Ableton timestretch artifacts all over one of the tracks and it still sounded good.
2 Likes
Yeah it’s true, I agree the VP sounds good but his statement about IRCAM stands, I think. It’s not that all stand alone hardware sounds like crap, but the OP did choose the apex to benchmark against. 
I always like when an artist embraces those aspects of technology.
The very obvious time stretch artifacts in this track are a great example. The artifacts become a defining attribute of the track.
3 Likes
I haven’t used the Ircam one, I assume it’s amazing since their stuff usually is.
EDIT: same here with embracing artifacts, I always like that. I guess it’s kind of similar to how most of the best live shows I’ve seen are the ones where something goes wrong technically and the artist or band just rolls with it and turns it in to part of the show.
A few years back a friend gave me an old Ibanez HD1000. It’s a mid-quality pitch shifter from like 1982, and the audio is 8 bits at something like 32kHz. Easily the most musical sounding pitch shift I’ve used,
2 Likes
I just setup my studio focused around OT2 and Sp16. I only had one session but found it perfectly fine at varying tempos with the same files, was really happy with the sound. OT2 sequencing tr8, AS1 and moog LP. Sp16 triggering various samples set to timestretch. Bpms 140/110/80/60 - loved how the tempo changes inspired tweaks to the sounds. I loved it.
1 Like
Yup but it’s always the same problem as the pads are not polyphonic you must use slices (same on the octatrack with the biggest advantage to have the possibility to use long file which is not possible on the SP16), when you playing it’s ok on -4/+4 keys then it’s start to fade to something unnatural to my ears… then, the next step is always slices a file with all the note you need - (multisample original pitch note you need) I always do that. That’s why in past I want the MPC Live for Programs - I let it go because program was to time consuming and in the end I always ending in the same scenario. it’s always more efficient to take what you need in the instant you want to make something with it. You loose the creative moment so you need to record you ideas on poly synth and then replace with sound design you have in mind of find a good sound candidate.
Can’t find a better or more efficient approach… (in Hardware) When I was ITB, Kontakt was very fast to make a program with sample.
For what it’s worth, I love the OT’s pitch shift and time stretch precisely because they can radically change the source material. I bought the OT to mangle audio, not to preserve it. I had more traditional samplers over the years (Emax, MPC’s, Fairlight, and a few others) and I was looking for something that could twist and reshape audio so it bore no relation to the source.
So for me, the OT is perfect in that regard. For others looking to get a single machine to serve multiple purposes, it may not be the best choice.
6 Likes
Works good enough for me with beats, artifacts quickly become noticeable with other kinds of source material. Not always a bad thing.
Amen Brother
Yeah it sounds like crap but its very useable as an effect. I like things that sound crap its my bread anda butter. Any hi end moog gear you could lend me I’d be trying to smear and shit up the sound anyway. Each to their own. Nothing worse than a great time stretch algorithm on software that JUST about manages to do what you want it to do beautifully but not quite. Then its just an artefact you didn’t want but can’t hide. SMEAR IT WITH SHIT!
1 Like
Another thing to say about the OT time stretch is that you can get the effect you hear all over old jungle tunes, early/mid 90s stuff I guess on vocal samples and breaks. Lo fidelity all stars has it all over vocals on their 1st album. You cant get that on ableton, or I never could, and I wanted that sound for so long!
1 Like
1 Like