<p dir="ltr">Also very interesting,<br>
thanks Richie and Tom.</p>
<p dir="ltr">/mr<br>
</p>
<div class="gmail_quote">Den 10 maj 2016 1:09 fm skrev "Tom Wiltshire" <<a href="mailto:tom@electricdruid.net">tom@electricdruid.net</a>>:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">I've done similar experiments with envelope sweeps rather than pitch bends, and the quicker you go, the less you hear the steps. Jumps that would be audible done slowly become invisible when done faster. If you think about it, this makes sense, because the error is smaller for a steeper slope when the resolution is limited. To be clear, what I mean precisely is: The sum-total of the difference between a continuous line and a discrete staircase is less when the lines are steeper. There's literally less time for error.<br>
Another way of looking at it is that at steeper slopes, a lower resolution line is a closer approximation than it would be at a shallower gradient.<br>
<br>
I came up against this with my first PIC projects using a 10-bit PWM DAC. This is fine for fast curves, but is at the edge of audibility for slow envelope curves in sensitive applications like highly-resonant filters or VCO pitch. 12-bit is much better, and what I regard as the limit of "human resolution" for this application.<br>
<br>
The other thing that's worth pointing out here is that what's important is not how fast (or how detailed) the MIDI data being sent is, but rather what the receiving instrument does with it. I have a Roland Sound Canvas module that performs smooth bends over an octave when I send MIDI pitch bend from an M-Audio Axiom controller, and yet I know full well that the Axiom only sends data about every 100 msecs or so. The Roland must be doing considerable internal smoothing - as it should, and as we can learn from. This is true for MIDi controllers too - there's nothing wrong with 7-bit controller data (do you *really* need more detailed settings than that?), as long as the receiving instrument has the good sense to filter it and turn it into something better before using it.<br>
<br>
Tom<br>
<br>
On 9 May 2016, at 23:27, Mattias Rickardsson <<a href="mailto:mr@analogue.org">mr@analogue.org</a>> wrote:<br>
<br>
> I confess, I was exaggerating a bit. 0:-)<br>
> But your question is interesting... Would it be possible to hear the side effects of the steps in the sweep? Like having a tiny FM with 1 kHz.<br>
><br>
> /mr<br>
> Den 10 maj 2016 12:16 fm skrev "Richie Burnett" <<a href="mailto:rburnett@richieburnett.co.uk">rburnett@richieburnett.co.uk</a>>:<br>
> Just miss out some of the values if you want to sweep across the same pitch range quicker. Or can you really perceive 1000 discrete pitches every second!?!? ...even when the oscillator pitch is probably less than 1kHz for a large portion of the sweep :-O<br>
><br>
> -Richie,<br>
><br>
> -----Original Message----- From: Mattias Rickardsson<br>
> Sent: Monday, May 09, 2016 11:04 PM<br>
> To: <a href="mailto:rsdio@audiobanshee.com">rsdio@audiobanshee.com</a><br>
> Cc: <a href="mailto:synth-diy@dropmix.xs4all.nl">synth-diy@dropmix.xs4all.nl</a> diy<br>
><br>
> Subject: Re: [sdiy] Vibrato range<br>
><br>
> On 7 May 2016 at 10:31, <<a href="mailto:rsdio@audiobanshee.com">rsdio@audiobanshee.com</a>> wrote:<br>
> Your example seems to prove that 14-bit MIDI Pitch Bend has no problem with unwanted stepping over 10 octaves. If my math is correct, there are 136.5 pitch bend steps per 100 cents when bending 10 octaves. Why do people complain about MIDI being "limited?"<br>
><br>
> Because it takes 15 seconds to do that sweep with those 2^14 pitchbend<br>
> messages. Even more if you want to do something else while waiting.<br>
><br>
> MIDI is slow and limited in temporal sense, not always in precision. :-)<br>
><br>
> /mr<br>
> _______________________________________________<br>
> Synth-diy mailing list<br>
> <a href="mailto:Synth-diy@dropmix.xs4all.nl">Synth-diy@dropmix.xs4all.nl</a><br>
> <a href="http://dropmix.xs4all.nl/mailman/listinfo/synth-diy" rel="noreferrer" target="_blank">http://dropmix.xs4all.nl/mailman/listinfo/synth-diy</a><br>
><br>
><br>
> -----<br>
><br>
> No virus found in this message.<br>
> Checked by AVG - <a href="http://www.avg.com" rel="noreferrer" target="_blank">www.avg.com</a><br>
> Version: 2016.0.7596 / Virus Database: 4568/12201 - Release Date: 05/09/16<br>
> _______________________________________________<br>
> Synth-diy mailing list<br>
> <a href="mailto:Synth-diy@dropmix.xs4all.nl">Synth-diy@dropmix.xs4all.nl</a><br>
> <a href="http://dropmix.xs4all.nl/mailman/listinfo/synth-diy" rel="noreferrer" target="_blank">http://dropmix.xs4all.nl/mailman/listinfo/synth-diy</a><br>
<br>
</blockquote></div>