Let's explore xentimbres

Today I'm exploring inharmonic timbre creation in New Tonality Lab. Let's make some weird music!
newtonality.ne...
Check out Objective Harmony's video about New Tonality Lab:
• Xentonal synth in your...
/ @new_tonality
Also announcing Scale Workshop 2 which has been rewritten from the ground up by Lumi Pakkanen (and a little help from me):
sevish.com/sca...
---
Support Sevish:
sevish.com/ban...
sevish.com/ko-fi
merch.sevish.com
Listen:
sevish.com/spo...
sevish.com/apple
sevish.com/ban...
sevish.com/deezer
sevish.com/sou...
Other links:
sevish.com/dis...
sevish.com/fedi
sevish.com/twi...
sevish.com/blog
sevish.com

[@blendywave5218]

30:25 yayayayayyaya

[@bb-jy7iu]

Really really cool software

[@pacifist7098]

I've always wondered how xentimbres would sound. Thanks so much for the overview!

[@matj12]

The samples from New Tonality Lab are encoded with Opus and wrapped with Ogg. ❯ ffmpeg -i sample.wav … Input #0, ogg, from 'sample.wav': Duration: 00:00:05.00, start: 0.000000, bitrate: 122 kb/s Stream #0:0: Audio: opus, 48000 Hz, stereo, fltp …

[@LiftPizzas]

That first 5edo "organ" sound reminded me of Tony Dubshot - Psionics. :)

[@maandalen]

The 1114¢ interval in the 3/1 equave timbre is equivalent to 3/2 stretched by the same amount. If you think about it, the familiar JI structures would stay the same, in a relative way, but sound very different. Time to experiment.

[@macronencer]

This is utterly fascinating! The way 7edo sounds with the 7edo partials is a revelation to me. I suspect that 80-90% of the "weirdness" of xenharmonic scales comes from the partials, not the actual intervals. By the way, I would love to hear about your Bitwig work-around for the piano roll please!

[@Ivan_1791]

I started getting familiar with inharmonic timbres a few days ago and it is fascinating.

[@fotgjengeren]

Being able to custom-make the scales is a total game changer. Thanks a ton! I might do weird stuff like combine them like you had with 5 and 10 together but with multiple tunings in the layers 😅

[@stephenweigel]

Gosh I can’t believe I missed this stream

[@hyperthesi6370]

same

[@GizzyDillespee]

Stephen, make sure to listen from 52:00 for the following 15 seconds or so...

[@cubicinfinity2]

when I navigated to the scale workshop earlier this week it was a disorienting surprise.

[@egilsandnes9637]

Neat! I didn't even know you did live streams. I have hoped to get into different tunings one day. Now the intimidating thought of starting seems a less daunting. Gleam is my absolut favourite of your tunes, and among my go to pieces when I don't know what to play.

[@roeesi-personal]

A bit of (my) theory about xentimbres and why they sound the way they do, not as a part of a chord but as one note in isolation: Our ear gagues the pitch of a sound by taking the difference between pairs of parials. That's why, for example, if you have a very low but rich note, even if your speakers can't play the fundamental you still hear the pitch of the note. Now, if you change the partials to be the steps of some edo that approximates the harmonic series quite well, like 5, 7, 12 or 22, the low partials are very similar to the natural ones and their difference is so as well, so the ear doesn't care. However, some of the high partials in this case don't fall close to the harmonic series, so they stand out as some "unnatural" or "metalic" high sound that goes along with the note. In "worse" edos like 8 even low harmonics like 3 aren't close to their natural value so the weird sound is much more prevalent there. In the case of compressed timbres, the differences you have don't match to the fundamental, and therefore the note is perceived as lower in pitch than the fundamental frequency (kind of when you have the lower harmonics of a sound filtered out, in purpose or by mistake), but the difference is not constant so you also have this "scify-spaceship-like" feeling associated with the sound. In the case when you strech the octave just by a little, the low partials are close enough to be harmonic so the ear doesn't bother, but the higher partials deviate from the pattern enough that we categorize them as a new unrelated sound, and that's the "metalic" roughness you hear in the higher frequency range there. When you stretch the octave a little more to 1250 cents, lower partials become these "in between" frequencies so the roughness becomes more prevalent in the sound and less high in frequency, and when you try 1902 there is no set difference between all the partials so the "metalicity" encompasses the whole sound and it becomes more "bell like". Happy to hopefully shed some light, or at least give my 2 cents, Roee Sinai.

[@roeesi-personal]

And about 2400, that means the nth partial is actually at n^2 which is a part of the harmonic series so it sounds more natural again.

[@mashtpot8r]

just now coming across this. i had an idea of this! im glad someone else already did lol

[@mashtpot8r]

actually my idea was to change the rate of the slope of the harmonics, that way instead of each octave having twice as many harmonics, the total would stay the same, or a spectrum in between. this is still coo though

[@mashtpot8r]

i would love know how to make things like this and integrate them into a synth vst

[@adamchristianson4100]

For anyone else struggling to open the files using Audacity here's how I got around it using VLC. Change the file extension to .opus, the open it thru the convert/stream dialog in VLC. Select the Audio - CD profile and under the customization audio codec set the sample rate to 48 khz and one channel. Then just choose a destination and export it and you should be good to go!

[@Irishpineapple97]

That is such an interesting program!

[@KimStennabbCaesar]

This is great! Very useful information.

[@Sillu129]

I don't know why but the nka files do not work in kontakt. I found some free "Pro microtuner" that uses those but it does not work with it. It is just that i want to make sure that i can make adaptive tuning in any number of ratios between an octave.

[@vltraviolet]

brilliant

[@marsensound]

1:16:40 I'm curious about the synth(?) you talk about developing in university. In particular, when you say that it had a slider that would move between inharmonic and harmonic partials, what exactly is happening over the transition? Like, is it crossfading between a set of each, in the sense of mixing in overtones of one while reducing the overtones of the other, or is it taking the same individual partials and modulating their frequencies between the harmonic points at one end of the slider and the inharmonic points at the other? Or something else perhaps I'm not quite picturing right. I'm not an experienced audio software developer but am a MaxMSP & RNBO user and learning C++ programming over the next while in order to get into that end of things. I'm interested in the idea of developing a vst synth to generate these kind of timbres in real time, or at least with the capability of doing so, in combination with more conventional harmonic series timbres. Would be a cool project to learn towards.

[@marsensound]

I watched later in the video, it seems like you explained it was the latter. Hahah!

[@macronencer]

Hypothesis: that initial attack in the output sample that you keep removing in Audacity... maybe all the partials are set to phase 0. It would make sense, given that the partials have irrational relationships, that constructive interference would happen *only* at the beginning of the waveform, and almost never again (theoretically, strictly never again - but we're dealing with finite limits of sample rates etc.). Does that make sense? I've just experimented in New Tonality Lab with very low pseudo octave values such as 78 cents, and it produces some very interesting "wobbly sci fi" noises, which are the most wobbly at the beginning, so this seems to support my suspicions.

[@jsihavealotofplaylists]

tysm

[@ikey5941]

this is amazinggggg

[@ikey5941]

1:22:00

[@Thoracius]

I have programmed a synth that does this, "The Equally-Tempered Clavier": kzbin.info/aero/PL2CusUsZBnzKWqU4xlTyuay0wMrajKUqk

[@GizzyDillespee]

Thanks for reposting... I entered the livestream 1 minute from the end, unfortunately. The final section, where you mess with the partials' relationships... Ive found it works better with complexity than with a simple saw wave. What I do is to take an entire track, and change its pitch up or down, and then do it again using another copy, pitching it in a harmonic relationship with however I changed the first copy. So, for the simplest example, I could take one copy of my audio and pitch it up by 702 cents, and pitch down a 2nd copy by 498 cents. Then sometimes I'll pitch a 3rd copy to be a 5th up from the lower copy... so I guess that's 200 cents up from the original (I'm bleary eyed so I might be off, there?) Then I layer all 3, sometimes swapping the stereo channels of one layer and/or offsetting it by 5-15 samples (at 48k)... The whole point is to layer them and get some of the partials to cancel out and others to reinforce, and create entirely new timbres that way. I usually start with jam using a piano or EP preset using a digital piano. That's a static pitch instrument - no bender or pitch modulation. Another cool side effect of braiding various stretched and compacted copies is that you can get really nice pitch instabilities in addition to unexpected timbres. An example is one I posted a while ago called something like Digital Piano Blues. It's just a piano or EP preset on a Yamaha P-115, but it sounds like it's morphing between a few different instruments and using a pitch bender. But my pre-layered copy just sounds like a straightforward piano, as it should! If I try that with a simple saw patch, the result isn't so interesting. It's much better when there's a lot going on. However, when there's a lot going on, if you stray from harmonic series relationships in your stretching/contracting choices, then the partials from the copies won't line up so well when you layer them, and won't provide as interesting cancelation and reinforcement, and so they'll simply stack in an inharmonic and not so pleasing jumble. So... the more complex the starting sound, the more interesting it is to mess with its partials by stretching and layering... but also it limits what you can do that still sounds musical. Probably about half (or maybe more!) of the "Yamaha P-115" jams I've posted use this technique to some degree. I don't remember their names offhand, though, but the Digital Piano Blues one is easy to hear - anything that doesn't sound like a digital piano timbre, or its normal reverb tail, was shaped by phase interference of layered partials, as I described. I'm obviously not the first person to do this. I just found the term "piano braiding" in Rick Beato's interview of composer Thomas Newman. It's the kind of thing you'd typically discover when messing around in an audio editor. Usually it seems to be used for more subtle and ambient effects, but aaahhh, WTH, YOLO!

[@TheApostleofRock]

When I tried to bring a sample into Audacity, it would not let me. It said for uncompressed files I could File>Import>Raw Data. When I did this with the default settings, it just gave me white noise. And it didn't seem that the file was 5 seconds long at all. Now I've literally never used Audacity...so did I do something wrong or are the files Newtonality makes just broken for now? It definitely isn't white noise and will play in my audio player (it does put like "423:13:24" on the remaining time counter though lol).

[@Sevish]

Apparently you need ffmpeg for Audacity? Someone else came up with this solution. I have no idea, it just worked for me, but I do have ffmpeg installed as a system package, so that might be why I had no problem.

[@adamchristianson4100]

I had this same problem. The solution is to convert it using vlc. Change the file extension to .opus, the open it thru the convert/stream dialog in vlc. Select the Audio - CD profile and under the customization audio codec set the sample rate to 48 khz and one channel. Then just choose a destination and export it and you should be good to go!

[@upsimusic479]

agh, i missed the stream!

[@guessw3rktunes733]

you don't use an ad blocker on youtube? lol. interesting vid though, def want to mess with this thing

[@lunafoxfire]

would be awesome if this was made into a VST rather than having to do the whole sample generation thing. Would be much faster to play around with.

[@axenwald9790]

Hello, what are the synthesizers that allow to use long sample files as if they were oscillator waveforms?