Voyager HP FIlter? I Think Not!
Posted: Thu Aug 12, 2021 11:53 am
I was showing a friend the Matriarch last week, specifically the independent control you have with each filter and VCA when patching. I then moved over to my Voyager to show him the HP/LP separation and how setting the poles can optimize this separation. Only I was quickly scratching my head wondering why lowering the number of poles on filter 2 made the sound brighter. My understanding was filter 2 is the HP, filter 1 is the LP and they’re connected in series and output to both stereo L/R outputs.
WRONG!
Turns out after much research both filter 1 and filter 2 are LP filters independent of which filter mode you use (LP/LP or HP/LP). The poles selection is for each LP filter. So where’s the HP filter? It doesn’t have one (almost). At least not in the conventional sense we geeks grew up with using capacitors/resistors. What the Voyager has instead is a subtractor circuit. I wasn’t previously familiar with this but it’s pretty simple; subtract the LP filtered signal from the original signal (mixer output) and you’re left with the high frequencies only. I guess this is also a type of filter (cough) although all the control (cutoff, modulation) is done with the low pass filter feeding it. The output of this “subtractor” is then sent to the other LP filter (LP1) and output to both VCAs.
So what’s the problem? Well for one thing in HP/LP mode you have 3 things at play here; the LP filter 2 >> subtractor>>LP1 filter all being controlled by the same cutoff control. The spacing control still does it’s thing of decreasing LP2 cutoff independent of LP1. But who knows what’s doing what when you hear the final sound?
Also, that subtractor works in theory, but only if the signals you want to subtract are the same. The amplitude of the mixer output is much greater than that of the LP2 filter output. Subsequently what you hear is a whole lot of the unfiltered mixer signal. Fortunately, there is a calibration adjustment trimmer to balance the unfiltered mixer signal with the LP filtered output. But on my Voyager, it was set to give the lowest amount of mixer output signal and yet it still dominated the sound. I’ve also read there are phasing issues associated with this subtractor type approach but I don’t think it’s really an issue here as far as I can tell. It is after all a synthesizer.
So how was I able to determine all this? I separated this “HP” filter from the in-series LP1 filter by adding a switch. In the off position the Voyager provide the 2 stock filter modes untouched. In the on position I get (selected with filter mode switch)...
HP/LP1 in parallel where the Left side VCA outputs HP only and the Right side VCA outputs LP1
or…
HP/HP>>LP1 where the left side VCA outputs HP only (still) and the right side outputs HP>>LP1
Not sure how useful this last one is but what the heck.
In HP/LP mode I was able to adjust the mentioned trimmer to give a well-defined HP filter output. I set it so that at the highest cutoff frequency the sound almost disappears. It’s wonderful! But calibrating this way screws up the stock HP/LP mode making the mixer output even more dominant. I ended up adding a second trimmer pot that get switched in (using the same added switch) to optimize the balance for all modes.
Being a fan of HP and other less popular filters this turned out to be a huge improvement for me.
WRONG!
Turns out after much research both filter 1 and filter 2 are LP filters independent of which filter mode you use (LP/LP or HP/LP). The poles selection is for each LP filter. So where’s the HP filter? It doesn’t have one (almost). At least not in the conventional sense we geeks grew up with using capacitors/resistors. What the Voyager has instead is a subtractor circuit. I wasn’t previously familiar with this but it’s pretty simple; subtract the LP filtered signal from the original signal (mixer output) and you’re left with the high frequencies only. I guess this is also a type of filter (cough) although all the control (cutoff, modulation) is done with the low pass filter feeding it. The output of this “subtractor” is then sent to the other LP filter (LP1) and output to both VCAs.
So what’s the problem? Well for one thing in HP/LP mode you have 3 things at play here; the LP filter 2 >> subtractor>>LP1 filter all being controlled by the same cutoff control. The spacing control still does it’s thing of decreasing LP2 cutoff independent of LP1. But who knows what’s doing what when you hear the final sound?
Also, that subtractor works in theory, but only if the signals you want to subtract are the same. The amplitude of the mixer output is much greater than that of the LP2 filter output. Subsequently what you hear is a whole lot of the unfiltered mixer signal. Fortunately, there is a calibration adjustment trimmer to balance the unfiltered mixer signal with the LP filtered output. But on my Voyager, it was set to give the lowest amount of mixer output signal and yet it still dominated the sound. I’ve also read there are phasing issues associated with this subtractor type approach but I don’t think it’s really an issue here as far as I can tell. It is after all a synthesizer.
So how was I able to determine all this? I separated this “HP” filter from the in-series LP1 filter by adding a switch. In the off position the Voyager provide the 2 stock filter modes untouched. In the on position I get (selected with filter mode switch)...
HP/LP1 in parallel where the Left side VCA outputs HP only and the Right side VCA outputs LP1
or…
HP/HP>>LP1 where the left side VCA outputs HP only (still) and the right side outputs HP>>LP1
Not sure how useful this last one is but what the heck.
In HP/LP mode I was able to adjust the mentioned trimmer to give a well-defined HP filter output. I set it so that at the highest cutoff frequency the sound almost disappears. It’s wonderful! But calibrating this way screws up the stock HP/LP mode making the mixer output even more dominant. I ended up adding a second trimmer pot that get switched in (using the same added switch) to optimize the balance for all modes.
Being a fan of HP and other less popular filters this turned out to be a huge improvement for me.