I can’t find a way to pitch an audio event using a ratio. I’m assuming that all pitch changes are done according to equal temperament? I don’t want to pitch by equal temperament.
I want to be able to pitch by exactly 3:2 or 4:3 for instance. Is this possible somehow in Cubase?
I feel like Cubase is the DAW of choice for classically musicians, so you’d think they’d have something like this…
PS (vari audio should also have different scales/temperaments built in)
What would be the purpose of this feature – I mean what is the end result you would hope to achieve, and how would you envision this to this work? I ask because ratios are used to tune two pitches in relation to one another, not an entire audio event with many pitches.
Well, for now, I have a long single note and I want to create a duplicate of it which is exactly a perfect fifth above without any phasing/side-beats. Currently, it’s impossible to do with the pitch system as it is.
There’s all sorts of reasons you can want this… Another reason I just noticed is that when you use the pitch shift, it distorts the audio slightly every time, unless you’re using the octave. I believe this is because it doesn’t use exact ratios, so there are some parts of the audio that might get rounded away or something. I’ve made sure this is not the result of sample rate reduction, although that will happen if you pitch up and then down… but if you pitch down and then back up, there shouldn’t be a loss in quality from sample rate, because you’re not maxing out the sample rate and then losing data in between.
So, if we could pitch by perfect ratios, then we could pitch as many times as we’d like and not have to worry about weird audio distortion/crackle… I’ll make a video in a minute to show you what I mean.
Are you saying that tuning by integers will get rid of artifacts?
I can’t say 100%, but seeing as how there is 0 distortion with the octave pitching, it seems like a reasonable assumption that if the other intervals were also perfect ratios, they, too, would not suffer from artifacts/distortion.
Also, I’ve recorded a test of pitched sine waves to compare pitch shifting at different intervals, and I’ve compiled an audio demo. Each interval was pitched down 4 times and then up 4 times at the given interval. I’ve added a “High Shelf II” EQ boost of 23db at 1kHz (in Cubase 11.0.41 Build 448), just to make the distortion more audible.
The smaller intervals seem to have less distortion in general, with the greater intervals having more. After Aug4/Dim5, the distortion is relatively the same, with just slightly more perhaps on P5. The WORST interval seems to be the perfect fifth and the BEST interval seems to be the major 3rd.
Hm… maybe. but there are other acoustic phenomena in this area. Plus a sine wave which has no overtones will have different results when tuned than one rich in overtones.
The fact that the major 3rd had less noise makes it appear that it’s not a tuning ratio that would improve the problem.
Anyway, after watching your video, I don’t see how using just intonation to transpose pitches would have any effect on the problem you’re trying to solve, but I’m not an acoustic physicist, just a fan of microtonality and tunings.
Maybe, maybe not. But even if it doesn’t help with the artifacts caused by pitch shifting, that’s just one issue. The main issue I cited as an example is phasing. I’d like to generate a harmonic which is perfectly in tune/harmonious and causes no phasing over time.
Ultimately, I wouldn’t say this needs to be a top priority, but it would be very nice, and I don’t think it should be too complicated.
Phasing with other notes?
Shoot, I just realize it didn’t record the audio (I forgot to switch my ASIO driver to Voicemeeter Banana to record ASIO lol)… BUT, you don’t really need to hear it to see the phasing.
I wasn’t asking for proof, just if you meant phasing when played in chords with other notes or not?
Anyway, if this is strictly about just vs tempered intonation you can tune the individual notes. For a more automated manner of achieving this maybe the request shoyuld be for Hermode Tuning in Variaudio, or for audio tracks, more generally speaking?
I can tune notes obviously… but I can’t tune them perfectly. I just want to have that precision. The math is very simple, obviously, for tuning to an exact ratio. I think when we tune by semi-tones / cents, there’s a lot more rounding involved.
I looked into the Hermode Tuning and I honestly don’t see/hear any differences. It definitely doesn’t affect the pitch shift function/algorithm.
Tuning by a ratio should simply be another option on the pitch shift panel, in the same way there are multiple options on the time stretch panel. Heck, while we’re at it, we should be to tune by changing the audio clip length, essentially having it be time stretch without the algorithms to keep the pitch intact.
All of the math is super easy for all this. The hard math is already done, we should also have the easy math too lol
Hermode tuning does indeed affect the tuning. Test it with some major thirds, which have the most obvious difference, and compare 12 tone equal temperament to the results with Hermode turned on.
Also, I’m curious to understand – when you use a ratio in this scenario, what two values shall the ratio represent?
I’ve tried thirds and fifths etc… I’ve bounced the MIDI to audio and there’s literally no difference. Do I have to restart Cubase after switching it on or something? Or does it only work with built-in Cubase vsti’s? I’m using Serum right now, so maybe that would make sense, because I can’t see Cubase controlling the pitch mapping of Serum, especially when you can load .tun files in Serum… it would be weird if Cubase overrides Serum’s mappings,… I’ll try with a Padshop or something in a sec…
In response to your second point/question, it would work the way intervals always work. If you use a 5:4 ratio in the pitch shift module, the resulting pitch would be 5:4 the pitch of the original. In other words, it would be 5/4 (1.25x) faster, (which I believe would be 4/5 the total length). This would work no differently from how the pitch shift already works, except it would tune based on more perfect math, instead of being based on equal temperament.
But it would also be cool to change the length of an audio clip and have it change pitch accordingly… That should be super easy to do also, but maybe I’ll save that for another thread.
Tried Hermode tuning with Padshop using the Init preset; literally no difference still. I bounced the MIDI to audio to verify… I feel like maybe we’ve had this talk before…
I just looked up the video on Hermode tuning and noticed I also have to select it in the MIDI modifiers tab, not just in the project settings window… I did that now… and still no difference using the Init patch in Padshop… Maybe I’ll try Retrologue as shown in the video… not that it would be a solution to the subject matter on this thread, but I’m just curious to see if I can get it to work
OK, nvm. Hermode tuning is working with Padshop and Retrologue, I see there is a difference in the waveform and I’ve inverted the phase of the duplicate. If it’s not working, then they’d perfectly cancel each other out, which they’re not… but I’ve also looked at the Serum one again and tried phase inverting the duplicate Serum audio, and it perfectly cancels out, meaning it’s doing nothing with Serum… so basically, it seems I was right in the fact that Hermode tuning only works for built-in Steinberg plugins.