source file: mills2.txt Date: Mon, 22 Jul 1996 06:26:37 -0700 Subject: VFX tuning From: Steve Curtin >Some observations on this micro-tunable machine (in the used incarnation I >just purchased): > >1. When editing the tuning table, although the nominal resolution is 1 cent, >the true resolution is in fact a few cents, as nominal 1 or 2 cent changes >often result in exactly the same rate of beating. >2. When changing the pitch of a voice, the resolution does appear to be in >fact 1 cent. However, the amount of detuning displayed may be inaccurate. >For example, a perfect match to the third harmonic is found at 1 octave, 7 >semitones, and 00 cents, instead of 1 octave, 7 semitones, and 02 cents. The resolution of the TS/VFX and EPS/ASR is 768 tones per octave, which is 1.5 cents. This does get mapped a little strangely from the front panel display to the internal tuning sometimes. One trick is to increment slightly past where you want to tune in and then try decrementing- because of the mapping you'll be able to get an offset which is between the steps reached by just incrementing. >3. Interpolating often produces strange results. It appears to insist that >all steps be exactly the same number of cents, and takes the liberty of >changing one of your endpoint pitches to make this possible. Often, the >resulting interpolation is backwards, decreasing in pitch rather than >increasing as you go up the keyboard, with one huge step from an extremely >low note to an extremely high note in the middle. The order in which you >specify the endpoints does not seem to matter. I hadn't looked at interpolation that closely but I had tried it out with many of the standard ETs such as 19 etc. Can you give a specific case for the backwards interpolation? Manual Op De Coul's 'Scala' tuning editor is currently being extended to support the VFX and TS. I've sent him the sysex docs and you would be a good test subject for this. Contact him for more info. >4. Extrapolating seems to work fine. The display won't show you assignments >higher than C8, but they're there, up to a point. Some high notes sound >really distorted though -- must be the digital architecture. The distortion is one of the unfortunate charms of sampling and interpolating oscillators. Because it's skipping steps in a wavetable the number of 'real' samples in the wavetable decreases as you transpose up or down. This is why people multisample- that is put a sample in the upper key range that's playing at a higher pitch so the oscillator doesn't have to transpose as far. Aside from these problems, I hope the unit is still useful to you. Ensoniq is the one of the few manufacturers of music synths these days which is producing both affordable and tunable instruments. regards, Steve Curtin Ensoniq Corp Received: from eartha.mills.edu [144.91.3.20] by vbv40.ezh.nl with SMTP-OpenVMS via TCP/IP; Mon, 22 Jul 1996 17:24 +0100 Received: from by eartha.mills.edu via SMTP (940816.SGI.8.6.9/930416.SGI) for id IAA02578; Mon, 22 Jul 1996 08:24:28 -0700 Date: Mon, 22 Jul 1996 08:24:28 -0700 Message-Id: Errors-To: madole@ella.mills.edu Reply-To: tuning@eartha.mills.edu Originator: tuning@eartha.mills.edu Sender: tuning@eartha.mills.edu