Hi,
yesterday I found this nice song from the SunVox author:
This vowel sound at the beginning sounds very beautiful. It’s a vowel synth inside SunVox. Sadly I cannot find any VST/AU that even comes close to this sound. Any ideas how this sound could be built in Renoise?
Kurtz made a pretty amazing formant filter demo here:
I was just thinking about that the other day, Kurtz’s vowel filter is just begging to be made into a doofer.
please do it for the lazy people like me 
Ok dokes, doesn’t take long, just right-click, combine, and then set the 4 parameters to macros. Here you go:
Don’t forget to adjust your keyzone to lay off the low end. As a bonus here’s an XRNI with the doofer in the effects chain already macro’d and keyzone fixed for you. Replace the saw wave with whatever, just make sure to make the keyzone range match.
edit: Also don’t forget you can add the doofer to the DSP list by clicking on the “preset” name area and going to save as.
Thanks lots for the hints! Tried Carbonthief’s doofer and also Kurtz’Easy Formants v2.5 Final. Impressive what is possible with only Renoise internal dsp.
But still the SunVox voxel synth sounds more realistic. So still looking for some good vowel/vocal synth vsti (osx). Any ideas maybe?
This is my favorite formant filter plugin:
http://www.vst4free.com/free_vst.php?plugin=Vocalizer&id=1458
Vowel filters are just filters with the right frequency response mimicking the response of our vocal system. Basically you can extract the “formants” from voice samples (it’s like up to 6-7 different resonators in our vocal tract), or find some table of the formant center frequencies and amplitudes, and put resonating bandpass filters in parallel for each formant’s frequency with the right amplitude. Or use a very close eq-curve. Morphing the vowel means moving the frequencies and amplitudes of the formants. You can get recognisable results with 2 formants with fixed amplitude, but the closer to the natural thing the filter is, the more like real voice it will sound. Boosting resonance will make the formants sound more sharp, intense, artificial etc. Also the vocal tract will put dips in the resulting spectrum, like it is blocking certain frequencies, and emulating these will get things even closer.
Note every person has his/her own formant response and frequencies. Also native language makes a big difference. Transposing the formant frequencies in the right way can shift the impression of how the voice is percieved. A male with a very dark voice has different response than a women with a high voice, or children. The opensource tool “praat” is cool stuff to analyse speech samples!
Also it is important, that the base sound you’re filtering is suitable to such action. Like it needs some “dirt” in the spectrum, and have enough harmonics, to have a satisfying effect.
Note you can also use a vocoder with good quality, and feed it with loops of vowels, or samples of speech, maybe a granular sweeping synth of a recorded vowel transition is nice material for making vowel sweeps.
This is my favorite formant filter plugin:
http://www.vst4free.com/free_vst.php?plugin=Vocalizer&id=1458
Hey thanks for the hint! Sadly it’s no OSX 
Vowel filters are just filters with the right frequency response mimicking the response of our vocal system. Basically you can extract the “formants” from voice samples (it’s like up to 6-7 different resonators in our vocal tract), or find some table of the formant center
Thanks for the detailed info. Knowing that. But look, you will never get such a cool sound just using some filters at special frequencies. Example:
http://tstlab.virtualcreations.de/drumroll-replacement.mp3
I think the SunxVox author (what was his name) did some more than just a special filter construction. Maybe some spectral stuff inside?
I’ve now heard examples of the sunvoc vocal filter, and it sounds like just an usual strong complex iir type filter, applying a frequency response of a vowel to the input sound. In some examples I’ve seen/heard the result also sounds very “synthetic”, this stuff is dependant on the input signal that’s filtered - if it’s very close the the base sound (and accompanying hiss/noise) the vocal chords produce, such filters sound very close to a real vocal, while very synthetic base sounds will mostly be modulated to a more “robot” type result.
How was the example posted above here actually generated? In sunvox, with what kind of synth/fx chain? It sounds very natural at least in the beginning.
Spectral stuff? Well I could imaging (not with the sunvox filter, but with the other free vst) that there can be some work done on the input signal before the filter to modulate it automatically into something closer to vocal chords pulses. Real spectral manipulation via an fft in high resolution sounds like not suitable for realtime audio, the latency would be too big. I’ve sometimes had the impression that certain advanced (commercial) vowel filters could have some fancy method of interresonating or whatever, creating the mentioned “dirt” by themselves, even from very clean signals that would sound very synthetic with a straight filter curve. I could also imagine those more advanced filters just strip/filter the input signal to the base sound frequencies, and re-add vocal chord/tract like harmonics by themselves in a way or another. Also it’s not just the formants, the way the base frequencies sound can also help with the voice impression.
Using a vocoder for these type of sounds - vocoders apply a rough frequency curve, but the usually peaks won’t move, just scale roughly, so the sound would only be close to natural if the vocoder had very many bands, or formant tracking.
Another oldschool way to create choir type stuff is using a wavetable of looped wavecycles of vowels. But this could sound too static, after all it’s the motion/modulation that creates the stronges suggestion of listening to something close to a real voice. Also when repiching vocal wavecycles, the formant peaks will move with the note frequency, natural formants stay roughly in their positions independent of the pitch that’s sung/spoken.
And yes, vocal tract is not just a few bandpass filters with the right freqs. On my todo-list is experimenting with a mic, praat, and the 10 band parametric eq in renoise to try to model vocal filter curves. I’m curious what kind of results this could provide.
Very short answer to the original question: get a good vowel filter vst, and learn to use it right. Or some synth sporting these kind of filters.
Another cheap trick I’ve thought about, would be to track the bandpassed amplitude of the carrier signal in the range of each formant peak, and amplify the formant filter output of thise ranges after filtering so the final output matches the spectrum of real voice more closely. This way one probably wouldn’t have to think too hard about getting the carrier signal spectrum right.
How was the example posted above here actually generated? In sunvox, with what kind of synth/fx chain? It sounds very natural at least in the beginning.
I simply made it using this SunVox song, selecting the vowel synth and using the incredible simple “Touch Theremin”. Very cool XY-Pad!
Spectral stuff? Well I could imaging (not with the sunvox filter, but with the other free vst) that there can be some work done on the input signal before the filter to modulate it automatically into something closer to vocal chords pulses. Real spectral manipulation via an fft in high resolution sounds like not suitable for realtime audio, the latency would be too big. I’ve sometimes had the impression that certain advanced (commercial) vowel filters could have some fancy method of interresonating or whatever, creating the mentioned “dirt” by themselves, even from very clean signals that would sound very synthetic with a straight filter curve. I could also imagine those more advanced filters just strip/filter the input signal to the base sound frequencies, and re-add vocal chord/tract like harmonics by themselves in a way or another. Also it’s not just the formants, the way the base frequencies sound can also help with the voice impression.
Using a vocoder for these type of sounds - vocoders apply a rough frequency curve, but the usually peaks won’t move, just scale roughly, so the sound would only be close to natural if the vocoder had very many bands, or formant tracking.
Thanks for your detailed thoughts and analysis. But here it really seems to be a FFT vowel synth in SunVox. This is written in the Sunvox’ features list:
SpectraVoice (FFT-based synthesizer for warm atmospheric sounds);
I also really like especially this synthetic touch. And I think it sounds so smooth because of FFT. Using normal IIR or FIR, you will get too much overtones which sounds robotic.
Very short answer to the original question: get a good vowel filter vst, and learn to use it right. Or some synth sporting these kind of filters.
I tried to find any good stuff. It’s quite hard to find anything for Mac OSX
I like the SunVox approach very much: It’s simply a synth with some parameters, easy in usage (no complex filter constructs etc). Pity that the author (“NightRadio”) doesn’t make a VST of that.
Greetz
Ya, OK, a “spectral synth” works different to a substractive/additive/whatever one with a vowel filter. It will synthesize a waveform completely from spectrum charakteristics (via fft techniques), no need to filter afterwards. If the technique is good enough, you get something very close to the original charakter from where the spectral information was extracted. Note this is synthesis/generation - a realtime effect based on fft transformations has the downside of introducing more or less latency an being hard to the cpu.
You cannot come close easily with some random synth with a vow filter - you can hear the difference when comparing your output to other stuff made in sunvox with a standard oscillator through the actual vowel filter - it sounds more like vocoder sounds than like a real voice. So you might want to look not for filters, but for spectral synths with the right options to shape the generated sounds voicelike.