[sdiy] SSM2164 panning curve

Tom Wiltshire tom at electricdruid.net
Thu Jul 23 20:33:39 CEST 2009


Ok, cool. Now I'm with you. I think. Famous last words.

Basically you want the sum of the outputs to remain constant. I'm not  
bang up to date on the latest in the panning law debates, but I  
understand that some people prefer the "equal power" version whilst  
other people go for "equal level" or some such distinction.

Let me just think this through aloud...
Presumably the X-axis and the Y-axis are identical? So we can look at  
one of them independently of the other. So if you solve the problem  
for one direction, you just repeat it for the other.

On that basis, let's have a look at the Left/Right axis. On the hard  
left side, you'd want only LHS signal, at unity gain. On the hard  
right, you'd want all RHS signal, again at unity. As you move away  
from these positions, the volume of that channel decreases whilst the  
other increases, following the log response curve of the VCA. In the  
dead centre, both channels will be mixed equally at some level less  
than unity, but which when added gives the same loudness as a single  
signal at unity. Am I still on the right lines?

Given that the SSM2164 does the complicated linear CV to logarithmic  
response part for you, I'd have thought you just want a circuit that  
gives you a linear voltage for the CV, say 0V (unity) to +4V (>120dB  
attentuation). This needs to be inverted and offset for the opposite  
VCA, to give 4V to 0V instead. This isn't difficult (a lot simpler  
than that electromusic schematic). The joystick pot can provide one  
CV directly (with a bit of tweaking) and then the opposite one needs  
an inverting op-amp mixer, fed with the CV and an offset voltage of  
-4V. And then same again for the other channel.

Hope this helps,

On 23 Jul 2009, at 14:21, Derek Holzer wrote:

> Hi Tom,
> sorry don't have an image right now, but let me try someone else's  
> explanation:
>> Some words about panning and crossfading:  If you simply have two  
>> VCAs connected together such that varying the control voltage  
>> brings the gain on one up and the other one down, you have a  
>> rudimentary panner, but the panner has a problem.  At dead center  
>> (both VCAs equal), the loudness will be less than full right or  
>> full left.  You need some flattening of the response curve to make  
>> this complete.  The Serge VCAs provide this equalized left-center- 
>> right response curve, so panning sounds fairly natural.
> From: http://www.serge-fans.com/wiz_vca_mix.htm
> Does that explain my idea a bit better? At center position, all  
> channels should be at unity, and should never get louder than that.
> D.
> Tom Wiltshire wrote:
>> Derek,
>> I'm afraid I don't understand what you're trying to do. I need to  
>> see more diagrams for my visual head to get it. So long as you're  
>> clear about it, that doesn't matter.
>> However, I have played with the SSM2164. 15V is too much for the  
>> control port. The datasheet gives the CV response as -33mV/dB.
>> Hence you need 3.3V for fully off -100dB, and -0.66V for 20dB  
>> gain, with 0V giving you unity gain. I don't know of a way to set  
>> the SSM2164 up for attenuation only. You just make sure that you  
>> only have positive control signals. I don't think just a simple  
>> diode would work that well, since the diode won't conduct until  
>> the CV is significantly negative, which isn't what you want. You  
>> could modify the op-amps in the Electro-music schematic to  
>> precision half-wave rectifiers, which would give you a more  
>> accurate result.
>> Regards,
>> Tom
>> On 23 Jul 2009, at 11:31, Derek Holzer wrote:
>>> I've got a hopefully simple question about using the SSM2164 as a  
>>> joystick controlled quadraphonic spatializer/panner. I have been  
>>> looking at Marc Bareille's Quad VCA/CV-Mixer as a good model for  
>>> this project:
>>> http://m.bareille.free.fr/modular1/quadvca/quadvca.htm
>>> with some additions for the joystick control:
>>> http://electro-music.com/forum/viewtopic.php?t=24705
>>> My concern is this: when the joystick is it dead center the gain  
>>> on all channels should be unity. However, for this particular  
>>> gain/panning curve I don't want more than unity gain, only  
>>> attenuation. I am aware that this nulls the effects of half the  
>>> range of the joystick... however from what I have read this is  
>>> the panning curve used by the Serge and Buchla spatializers, and  
>>> it makes sense acoustically to me.
>>> So let's assume that set up the SSM2164 as in the datasheet,  
>>> where 0V is unity and minus voltage is attentuation. And I wire  
>>> up my joysticks as voltage dividers between +15 and -15 (or  
>>> should it be +/-5V?). But since I don't want any amplification, I  
>>> use a diode to block the positive voltage coming from the  
>>> joysticks, making sure the diode is rated not to break down  
>>> anywhere near 15V (or 5V?). Am I on the right track? Or is there  
>>> a way to set up the SSM2164 itself so that it only attenuates?
>>> Thx+best!
>>> Derek
>>> -- 
>>> ::: derek holzer ::: http://blog.myspace.com/macumbista :::  
>>> http://www.vimeo.com/macumbista :::
>>> ---Oblique Strategy # 52:
>>> "Do nothing for as long as possible"
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at synth-diy.org
>>> http://synth-diy.org/mailman/listinfo/synth-diy
> -- 
> ::: derek holzer ::: http://blog.myspace.com/macumbista ::: http:// 
> www.vimeo.com/macumbista :::
> ---Oblique Strategy # 107:
> "Look at the order in which you do things"

More information about the Synth-diy mailing list