CS-80V TAE anomaly

[bg]

When OSC 1 and 2 are tuned exactly together and a low note is played, I hear what I would describe as "inter-modulation."  It sounds like some kind of random, subtle S/H filter modulation.

Set the CS-80V like so:
All filters, RM, and LFO modulation off.
All VCO, VCF, and TOUCH RESPONSE sliders off.
All VCA sliders off except "VCF LEVEL," "SUSTAIN," and "LEVEL."  Set these three to maximum.
Both octave ranges = 16' [-1 octave]
MIX I/II = 0
DETUNE CH II = 0 (or as close as possible)
Square wave for both lines = ON, all other waveforms off

Now play the lowest C and hold it for several seconds.  You should hear the sound jumping around within the tone.  You can hear a short recording of it   here in the "CS-80V test" folder.  The mp3 is called TAE_oddity.

The behavior also occurs using Sawtooth instead of Square wave, but it seems more pronounced with Square.  If you don't hear the problem at first, keep holding the note for 15 seconds or so.  Also, the lower the pitch the easier it is to hear.

I am using version 1.2 and it does the same thing in Mac OS 9 and OS X.  I have a CS-80V 1.0 demo version on a PC at work and I did not hear the problem on that version/platform.  It was totally smooth.

Hope this helps.

[omissis]

Hi bg
I listened to the example and I'm oriented to say there's no oddity at all. Since tae is designed to achieve the best possible realism in modeling an analog VCO circuit , it reproduces some oddities too : that sort of  SH modulation as you call it comes I think from the model itself : an analog circuitry is subject to flowing tension and all that, this equals to a slight change in the phase and period of the waveform that is a saw tooth for example comes with a short period and the following with a larger period; the audible result is the "oddity" you hear when you tune two oscillators at the same footage, otherwise you would get a "perfect" oscillation with the waves all with the same perfect phases and periods but then the result won't be so much analog. So I wouldn't worry too much about this "defect" ....or not? :wink:  :wink:  :twisted:

[bg]

Apparently, the behavior I describe is designed into TAE.  I tried the same test using the miniMgV demo version and it does the same thing.  I never played a CS-80, but I played a MiniMg for 10 years.  I know the sound of two oscillators tuned together playing a low note.  Whatever anomalies were present on the original, they never upset the consistency and presence of the tone--whether zero-beat tuning or a slight phase.

In the specific condition of low pitched, zero-beat, 2-Osc's, the TAE behavior does not sound analog.   Again, it is not noticeable on higher pitches.  You can zero-beat higher pitches and they sound fine (although I presume the "S/H intermod" is still in effect).  And low pitches that are detuned sound fine.  

Perhaps the jumping around of the tone wasn't that clear in the mp3 I posted.   Try the test yourself, and I think you'll hear the effect is not subtle.

[omissis]

Well bg

I compared the sound coming out of the CS-80V with the output of my Roland JX-10, which isn't a classic analog definitely but it's one of the most analog sounding DCO equipped-subtractive synths of the 80's (and , for the moment the only "analog" I can manage :lol:  :lol:  :mrgreen:) and I didn't notice the "sh" effect on the Roland....well, the let's say, artifact is barely audible  on CS-80V , I don't feel much annoyed by that anyway ; I think that the guys considered this issue....actually the update is in progress for what I know and maybe they will fix this

By the way I presume this oddity has to do with a simulation of  slight drift in the pitch: I selected one-osc in the Templates bank and I noticed sometimes the pitch flows. A CS-80 had 16 VCOs (so 13 more  :wink: ) and a very bad cooling system ( and NO FANS under the board !!!   ) compared to the "perfect" design of the MiniMg-D ...so I guess a MiniMg has more chances to stay tuned some minutes more :wink:  :wink: )