do audio samples stretch at real time tempo change?

[PopStarWannabe]

I have rummaged this forum and the manuals, but couldn't find a decent description/solution to my problem:

I want to bring into Live performances sounds that we used on the album. These were from other sources than the M3 (guitars, vocals, other synths, etc). The only piece of gear on stage will be the M3. No laptops, no instruments or anything else. Don't ask why!

So far so good, but the singer sometimes requires tempo changes during the performance, according to his or the audience's mood. Not more than +/-10 BPM though. So that's my problem - I don't know how to seamlessly achieve this. I don't own the M3 yet.


Questions:

1) How will the audio samples react to real time tempo changes? Will they automatically stretch/shrink to stay in-sync with the other MIDI Tracks?

2) But if I slice them (based on their attack) prior to importing them into the SEQ (hence I bring only the slices to the sequencer) - is this a better approach?

3) How many tracks in the Sequencer can contain MIDI Notes that trigger samples? Is this limited?

4) It reads that M3 has 4 Inputs. Can I record 4 external sources of audio simultaneously? Let's say on 4 different tracks?

5) How can I adjust the volume of the tracks that contain samples relative to the other Midi tracks? Will the volume slider of the track also affect the loudness of the audio samples?


Any feedback will be appreciated.

Thanks.

[r12]

1 - audio samples (like loops) will stay at original tempo...there is no realtime time stretching

2 - yes, but with limitations to decrease

3 - all tracks can trigger samples (16), and if you use program changes 128x(6+7)=1164 (128 programs x 13 banks)

4 - no you can record one track at a time (in ram or usb) but there is no straming for playback

5 - volume slider control samples volume

[kid_nepro_2002]

If you convert your sample loop to a drum track, you can change the tempo without affecting the pitch. Here's how…..

http://www.kidnepro.com/KN/Korg/M3/korg ... ttern.html

Proton
www.kidnepro.com

[PopStarWannabe]

Guys, thanks a lot!!

2) But if I slice them (based on their attack) prior to importing them into the SEQ (hence I bring only the slices to the sequencer) - is this a better approach?

2 - yes, but with limitations to decrease

r12, what did you mean by that? You mean if I decrease the tempo (hence slow down the song)? What happens then? I thought the problems might appear at INCREASEing tempo: the slices might get too close to each other and overlap.

you can record one track at a time (in ram or usb) but there is no straming for playback

you mean limitation at simultaneous playback tracks?

[r12]

no, the samples will be too short to have a continuous groove (like original)

[r12]

for simultaneous playback, if you import each taken .WAV tracks to ram, you can get up to a 16 stereo DAW (but with limitation time recording 320 MB 48KHz recording - mono or stereo)

[Shakil]

If you use the .WAV tracks in a drum kit, then you can have upto 64 X 16 tracks........ of course, limited by 320MB RAM AND polyphony.

For .wav loops, drum kit works better, you can have each loop mapped on to a key.

3) How many tracks in the Sequencer can contain MIDI Notes that trigger samples? Is this limited?

You could have multiple loops within one track using a drum kit. You can have the slices in different key ranges, and then have the MIDI Notes based on your mapping. Use RPPR Pattern to play the slices using the pads or a different key.

The way I work with .wav loops and slices is that I time stretch them on computer to same tempo and normalize to -6db before loading into M3. So all loops and slices can be at 90 or 120 or 140 .. depending on their original tempo..

In Song mode, you can have bpm matching detuning for each track. So you don't need to edit the samples for each song at different tempo. For example, I could have 30 loops loaded at 90 bpm. I could have several songs some at 88, some 95.. etc and M3 would sync the same samples using detuning for each track in the Song mode.

[PopStarWannabe]

If you use the .WAV tracks in a drum kit, then you can have upto 64 X 16 tracks

Ok, I know the Sequencer has 16 tracks. But what's with 64?

You can have the slices in different key ranges, and then have the MIDI Notes based on your mapping. Use RPPR Pattern to play the slices using the pads or a different key

This seems to be the best approach for what I need.

In Song mode, you can have bpm matching detuning for each track

Do you mean "Detune BPM Adjust"? That's not good for me since this alters the pitch of the loop along with the tempo.

Thanks for your input!

[Shakil]

PopStarWannabe.

64x16 ... you can have 16 kits at a time and each kit can have 64 drums notes... each drum note can play a stereo wave file..

PopStarWannabe:
"My problem is this:
I want to bring into Live performances sounds that we used on the album. These were from other sources than the M3 (guitars, vocals, other synths, etc). The only piece of gear on stage will be the M3. No laptops, no instruments or anything else. Don't ask why Smile And the singer sometimes requires tempo changes during the performance, according to his or the audience's mood. Not more than +/-10 BPM though. So that's my problem - I don't know how to seamlessly achieve this. I don't own the M3 yet. "

This can not be achieved with M3 alone.

The only hardware synth/sampler that allows real time time stretching without changing pitch are Roland Fantom, and MV or SP series workstation/samplers.

So, you can either get M3 and MIDI out to MV or SP Series Sampler/Workstation.. OR get yourself Fantom-X or Fantom-G.

With Fantom-X you get 8 stereo tracks of Audio samples and 512MB max RAM.

With Fantom-G you get 24 stereo tracks of Audio samples and 1GB RAM.

In both cases Audio Tracks are streamed from RAM, not Flash Card or Hard Drive.

Fantom-X loads all samples from the Flash Card... so you can have dedicated flash card for each gig or set of songs which takes about 15 minutes to load completely at startup.. but.... Then you can just call up the song in middle of performance without any delay. Each card can have 7000 samples and up to 128 songs with total size of 512MB.

Fantom-G is project based, with each project having max of 50 songs and 2000 max # of samples total size of 1GB RAM. You can store as many projects as your flash drive can fit. This way you could have a project file for each set of songs or gig...

If you can deal with Fantom-G's other flaws and limitations, for your application, Fantom-G will work best.

[PopStarWannabe]

Shakil, thanks a lot. I knew Roland synths could do real time time stretching, but we have to have the M3 (because of other brilliant features) and the budget is extremely limited.

I will try the workaround described above (with the slices). I might encounter issues with slices overlapping (in case of tempo speeded up too much) or slices ending too soon (tempo slowed down too much).

Do you have any suggestion for that? Is there such thing a thing as "automatic crossfade" in case of overlapping samples/slices?

[Shakil]

PopStarWannabe.....

Just get the small SP series sampler.... SP404SX... it streams samples from the card ...

[Kim Lajoie]

You might want to consider slowing down the audio files using a high-quality algorithm on your computer/DAW, then slicing them and importing the slices to the M3. You wouldn't need to slow the audio down much - only 10 BPM (or whatever is the slowest speed you can imagine your singer asking for). That way you don't need to worry about realtime timestretching, and you can safely slow down the tempo without getting gaps.

-Kim.

[PopStarWannabe]

consider slowing down the audio files using a high-quality algorithm on your computer/DAW, then slicing them and importing the slices to the M3

But if I then playback the song in the Seq at the original tempo (i. e. before slowing down in the DAW), then slices would overlap. The next will start sounding before the previous has finished.

Your solution would work perfect if the singer wanted to ONLY slow down the song. But most of the time, he wants to speed it up...

Thanks anyway, that's a good point to know!

[Kim Lajoie]

consider slowing down the audio files using a high-quality algorithm on your computer/DAW, then slicing them and importing the slices to the M3

But if I then playback the song in the Seq at the original tempo (i. e. before slowing down in the DAW), then slices would overlap. The next will start sounding before the previous has finished.

Your solution would work perfect if the singer wanted to ONLY slow down the song. But most of the time, he wants to speed it up...

Thanks anyway, that's a good point to know!

You can set up the sequence so that the slices are triggered by non-overlapping notes. So instead of overlapping, the slices would just be cut off slightly before they finish.

-Kim.

[Shakil]

That will work only for drums and percussion tracks that have the hits mapped to midi notes.... not any melodic track...