What's new

Where to start?

There is usually no need to script crossfading, just use modulators.
Ok thanks David ;)

I have Reatune open on the masterbus when I am editing samples --before I export them. I usually try to organize the samples so that each pitch is on a single track so I can solo it and tune them as I go. I try to have them tuned before I put them into Kontakt (because, as far as I know Kontakt has no visual tuner?). It's a pain using the test tone for tuning purposes. In my recording sessions I take audio notes by saying something like "C note vibrato." And then I just play the note and even vary the velocity. So I keep a separate recording for each note and articulation.

Reatune in auto or manual mode?
 
Ok thanks David ;)



Reatune in auto or manual mode?
I actually use Reatune for detecting the pitch and then I tune it via track properties. I'm not sure if it's using the same algorhythms or whatever, but I seem to remember hearing that the built in repitching (in track properties) is pretty advanced. Not sure if it's the same thing as Reatune. So I literally use it as a tuner just to detect the pitch so that I know how much to change it by (if it needs it).
But if reatun's pitch correction capabilities are basically the same then I could save myself a lot of time :)
 
If you have a single mic position it's easy, just stick ReaTune on it (increase the window size for low instruments) and put it in auto-mode (you may need manual mode for some notes though). Reaper uses the same tuning algorithm whether you're using ReaTune or pitch envelopes, you can set the default algorithm in the project properties. I find Elastique efficient gives the best results, I get artifacts with nearly all the others. Sometimes a note doesn't need any tuning, when you have this lucky situation you can bypass ReaTune via automation. If you have a sample that has a steady pitch but it's the wrong pitch (a little bit flat or sharp perhaps) then rather than using ReaTune you could adjust the media item's pitch using the pitch item down/up actions, then you can use the convert pitch to rate script to get higher quality tuning (this script is available in ReaPack I think, it doesn't come with Reaper).

Now multi-mic samples are a different, evil, horrible beast :P You can't just stick an autotune plugin on each track because they all work via FFT and will screw up the phase relationship of all the samples so you'll get some nasty chorusing between the mics. Apparently there's a version of Melodyne that can work with multi-mic recordings but I haven't been able to fit Melodyne into my workflow and I don't really want more proprietary software than necessary. So there are only two solutions I have found and they both require some manual work. X-Raym made an awesome script that is basically auto-tune using item pitch envelopes, and there is a script to copy pitch envelopes between items. So you put the script on one mic, tune the samples, and copy the envelope to the other mics. It still uses FFT but because all of the tuning points will be the same it doesn't create any noticeable chorusing. Although the script is automatic and can be used on multiple media items at once, it isn't 100% accurate and you will need to go through the samples one by one and adjust the pitch envelope as necessary (many won't need any adjustment, or very little). I tuned a few thousand samples this way last month. The script isn't free, you can get it from here - https://www.extremraym.com/en/downloads/reascripts-envelope-based-pitch-corrector/.

Now the other method, that I have yet to try in production: Paolo from Fluffy Audio informed me that you can tune a mic position with Melodyne, then rename the file so Melodyne can't find it, and then point Melodyne to a different file of the next mic position. Melodyne will then apply the tuning it applied to the first file to the second file. So I figured why not do the same thing with ReaTune, and in my brief tests it seems to work. You have to use manual mode, tune one media item (best to do this before cutting the samples) then swap the media item for the next mic position and ReaTune should apply the same pitch adjustment. I only did some quick tests with this but if it does work I think it will be faster than Raymond's script for sustain samples but perhaps not for short samples.
 
Wow David! thanks for the good reply!
I find Elastique efficient gives the best results
Yep! I use Elastique too.
you can tune a mic position with Melodyne, then rename the file so Melodyne can't find it, and then point Melodyne to a different file of the next mic position
Eh, I know...but Melodyne apply a little bit synthy sound...I prefer Waves Tune or Cubase VariAudio. ReaTune seems to be a good solution.
It is a good thing normalize samples before edit?
 
Last edited:
It is a good thing normalize samples before edit?
You should normalize your samples at the very end before exporting. Within Reaper normalization is none destructive so you could normalize at any point. I normalize at the very end so that while I'm editing the samples are at roughly the volume they will be when played back by the sampler. This is important so I don't waste time cleaning up little noises that I'd hear in a normalized sample but not at playback.
 
Ok, I changed idea and I'm starting sample my C Bansuri flute :)
I have one that I bought in India. Today I finished to record staccatos with 4 RR and 3 dynamic layers :shocked:
 
OK well, it might be worth mentioning there are a number of scripted approaches (and by this I mean KSP scripts) that will offer you an alternative to RR's and velocity layers.

Essentially RRs can be seen as the same note with minor timing, EQ and pitch differences (what else could they be?). It is possible to script these changes in Kontakt at play-back, so you dont need to do RRs at all.

Almost the same for dynamics, here you can use the "synth" approach of changing the gain of a single sample and rolling off high end the lower the velocity goes. Of course there can be serious differences between velocity layers that cant be emulated in this way - the "chiff" part of your flute sounds might be a good example. But there are work-arounds for this too. However both these approaches are ways to reduce the amount of sampling (and associated editing) that you need to do. If you want to be a purist then sure: "all-power-to-your-elbow", and dig in for that long sample editing process. Just thought I'd mention it.
 
OK well, it might be worth mentioning there are a number of scripted approaches (and by this I mean KSP scripts) that will offer you an alternative to RR's and velocity layers.

Essentially RRs can be seen as the same note with minor timing, EQ and pitch differences (what else could they be?). It is possible to script these changes in Kontakt at play-back, so you dont need to do RRs at all.
Very good points, I just spent ages editing 3RR samples for an instrument and then once I got it in the player I decided the effect was so subtle I might as well just use the borrowed sample approach - no one can tell the difference :P
 
Very good points, I just spent ages editing 3RR samples for an instrument and then once I got it in the player I decided the effect was so subtle I might as well just use the borrowed sample approach - no one can tell the difference :P

Were the samples recorded in half steps or whole steps? :)
 
If you have a single mic position it's easy, just stick ReaTune on it (increase the window size for low instruments) and put it in auto-mode (you may need manual mode for some notes though). Reaper uses the same tuning algorithm whether you're using ReaTune or pitch envelopes, you can set the default algorithm in the project properties. I find Elastique efficient gives the best results, I get artifacts with nearly all the others. Sometimes a note doesn't need any tuning, when you have this lucky situation you can bypass ReaTune via automation. If you have a sample that has a steady pitch but it's the wrong pitch (a little bit flat or sharp perhaps) then rather than using ReaTune you could adjust the media item's pitch using the pitch item down/up actions, then you can use the convert pitch to rate script to get higher quality tuning (this script is available in ReaPack I think, it doesn't come with Reaper).

Now multi-mic samples are a different, evil, horrible beast :P You can't just stick an autotune plugin on each track because they all work via FFT and will screw up the phase relationship of all the samples so you'll get some nasty chorusing between the mics. Apparently there's a version of Melodyne that can work with multi-mic recordings but I haven't been able to fit Melodyne into my workflow and I don't really want more proprietary software than necessary. So there are only two solutions I have found and they both require some manual work. X-Raym made an awesome script that is basically auto-tune using item pitch envelopes, and there is a script to copy pitch envelopes between items. So you put the script on one mic, tune the samples, and copy the envelope to the other mics. It still uses FFT but because all of the tuning points will be the same it doesn't create any noticeable chorusing. Although the script is automatic and can be used on multiple media items at once, it isn't 100% accurate and you will need to go through the samples one by one and adjust the pitch envelope as necessary (many won't need any adjustment, or very little). I tuned a few thousand samples this way last month. The script isn't free, you can get it from here - https://www.extremraym.com/en/downloads/reascripts-envelope-based-pitch-corrector/.

Now the other method, that I have yet to try in production: Paolo from Fluffy Audio informed me that you can tune a mic position with Melodyne, then rename the file so Melodyne can't find it, and then point Melodyne to a different file of the next mic position. Melodyne will then apply the tuning it applied to the first file to the second file. So I figured why not do the same thing with ReaTune, and in my brief tests it seems to work. You have to use manual mode, tune one media item (best to do this before cutting the samples) then swap the media item for the next mic position and ReaTune should apply the same pitch adjustment. I only did some quick tests with this but if it does work I think it will be faster than Raymond's script for sustain samples but perhaps not for short samples.

Melodyne tunes stereo tracks based on the left signal alone, meaning that you can do a multimic batch tune based on a single recording as follows:

Track 1: Close mic
Track 2: Decca left
Track 3: Close mic (track 1)
Track 4: Decca center
Track 5: Close mic (track 1)
Track 6: Decca right
etc etc

Then simply use the right audio channel from the stereo tracks, along with a a single track of left audio for that mic and you're done, all kept within the original phase too.
 
Melodyne tunes stereo tracks based on the left signal alone, meaning that you can do a multimic batch tune based on a single recording as follows:

Track 1: Close mic
Track 2: Decca left
Track 3: Close mic (track 1)
Track 4: Decca center
Track 5: Close mic (track 1)
Track 6: Decca right
etc etc

Then simply use the right audio channel from the stereo tracks, along with a a single track of left audio for that mic and you're done, all kept within the original phase too.

I think the same would be true with Reaper's Reatune, if you put all the samples in a multi channel file. From there you could choose which mic channel you wanted to use for tuning. You could even tune more than one channel, actually all of them if you wanted to, however I'm not sure how practicle that would be.
 
I think the same would be true with Reaper's Reatune, if you put all the samples in a multi channel file. From there you could choose which mic channel you wanted to use for tuning. You could even tune more than one channel, actually all of them if you wanted to, however I'm not sure how practicle that would be.
I did actually do this, you have to play around with the routing. I found it a bit messy though.
 
I did actually do this, you have to play around with the routing. I found it a bit messy though.

Yeah, I've never acutally tried it David, but I know if you route all the channels to to their own separate bus, logic tells me it can be done. :)

I also think ReaTune in it's manual mode would work well for it. Unless of course things are way out of tune.
 
Very good points, I just spent ages editing 3RR samples for an instrument and then once I got it in the player I decided the effect was so subtle I might as well just use the borrowed sample approach - no one can tell the difference :P
I'm a huge fan of borrowed RR. It's all about suspending belief. It almost creates an auditory illusion. I think there's probably a good balance to be reached with that method though. If the notes are stretched to high, then those samples will have a quick attack due to a faster play rate. And conversely, the notes being stretched down will have a slower attack. So all in all I think it's about the type of samples. I did a quick little chromatically sampled ukulele about 2 years ago. For fun I tried a 7RR experiment by creating groups up to 3 semitones up and 3 semitones down. It actually worked really well. But I think at a certain point the illusion would be lost and our brains would know we are being fooled. Nothing wrong with experimentation though.
 
I'm a huge fan of borrowed RR. It's all about suspending belief. It almost creates an auditory illusion. I think there's probably a good balance to be reached with that method though. If the notes are stretched to high, then those samples will have a quick attack due to a faster play rate. And conversely, the notes being stretched down will have a slower attack. So all in all I think it's about the type of samples. I did a quick little chromatically sampled ukulele about 2 years ago. For fun I tried a 7RR experiment by creating groups up to 3 semitones up and 3 semitones down. It actually worked really well. But I think at a certain point the illusion would be lost and our brains would know we are being fooled. Nothing wrong with experimentation though.
I always just do 1 up and 1 down. If I do decide to use multiple RR recordings then I will also do the pseudo RR as well to create even more variation.
 
I always just do 1 up and 1 down. If I do decide to use multiple RR recordings then I will also do the pseudo RR as well to create even more variation.
That's probably the best practice :) But if you're in a pinch or if you want to create variation in a smaller sized library, I say whatever works lol. BTW, I first learned about this technique by watching one of your Youtube videos :)
 
Prompted by @Light and Sound and @Tod I decided to revisit the process of routing ReaTune for multi-mic tuning and actually it seems like it might be a really good method, I can't remember why I stopped using it (maybe there is a hidden pitfall I'll stumble into). Anyway, here is a picture of the way I have everything routed. This is setup for 5 mics but will of course work with more.
tuning.png
 
the "chiff" part of your flute sounds might be a good example
But I think that I should spend much time to create a very realistic synth-modeled RR that reproduce perfectly the "chiff" or the tongue sound rather than record a new one and apply some cut and fades, IMHO...
 
Top Bottom