Pro 2 Oscillators

The Dave Smith Instruments Pro 2 Analogue Synthesizer

Moderator: Pym

Post Reply
errissa
Posts: 48
Joined: Tue Apr 08, 2014 3:59 pm

Pro 2 Oscillators

Post by errissa » Fri Sep 05, 2014 3:35 am

When I first received my Pro-2 a couple of weeks ago I recorded the raw oscillators with the intention of posting them here for those interested but I got busy and never did. aiynzahev has now already done that here: http://www.dsiforum.com/viewtopic.php?f ... 140#p53839

However, with all the digital versus analog talk and with the Sub37 released at about the same time as the Pro-2 I thought it would be interesting to compare the two new synths. So, I recorded 3 basic waveforms from the Sub37 and Pro-2 (sawtooth, triangle, and square) so you can compare them yourselves. The recordings are attached.

I recorded at 96KHz/24bit (probably overkill) into Ableton through a UAD Apollo. No further processing. Prior to recording I used a Peterson tuner to tune both synths. The Pro-2 was spot on without any adjustment as one would expect with a digital oscillator; the tuner -- which has an accuracy of +/-0.1 cent -- showed 0 deviation at A440. The Sub37 was a touch flat but I got it to within 0.5 cents with the Fine Tune adjustment at A440. I used the sequencer on both synths to sequence what you hear in the recordings. The Sub37 sequence is a bit shorter because it has 1 fewer As on its keyboard so I ended up taking the Pro-2 an octave higher than the Sub37. I didn't realize that until I'd already recorded and didn't feel like doing it again, so... sorry for that!

Here are the spectra for the Sub37 and Pro-2's triangle and sawtooth waves at A220...
Pro2_Triangle.png
Pro2_Triangle.png (386.58 KiB) Viewed 4332 times
Sub37_Triangle.png
Sub37_Triangle.png (422.93 KiB) Viewed 4332 times
Pro2_Sawtooth.png
Pro2_Sawtooth.png (444.61 KiB) Viewed 4332 times
Sub37_Sawtooth.png
Sub37_Sawtooth.png (467.09 KiB) Viewed 4332 times


I found the spectrograms interesting. On both the sawtooth and square waves (not included here) you can clearly see the Pro-2 rolling off at higher frequencies, presumably to prevent aliasing. By the time you get to 22KHz there's nothing left, whereas on the Sub37 there's harmonic content up to 22KHz. Also, see the hump on the Sub37 spectrograms at 80Hz? That hump is present on pretty much every note above 80Hz in these recordings. Perhaps that contributes to the impression of 'fatness'? Finally, the spectrum of the Pro-2's triangle wave goes higher than that of the Sub37 which I found to be quite noticeable by ear -- the Pro-2's triangle wave has a kind of high frequency crunchiness too it that is not present in the Sub37.

My general impression when listening to the two synths side by side is that the raw oscillators on the Sub37 do seem more 'full' and mellow than the Pro-2 which seems 'thinner' and brighter by comparison. But in 'blind' listening, I found them to be somewhat difficult to distinguish with the exception of the triangle waveform which, on the Pro-2, had an unmistakeable brightness to it.

I'm enjoying both synths very much. I'd say I prefer the raw sound of the Sub37 a bit more, but the Pro-2 has been more fun to explore -- so many more possibilities.
Attachments
pro2_sub37_waveforms.zip
(96.57 MiB) Downloaded 132 times
Last edited by errissa on Fri Sep 05, 2014 7:55 pm, edited 1 time in total.

errissa
Posts: 48
Joined: Tue Apr 08, 2014 3:59 pm

Re: Pro 2 Oscillators

Post by errissa » Fri Sep 05, 2014 7:49 pm

Razmo wrote:This is interresting... because 22Khz is exactly the double of that 11Khz frequency that has been discussied a lot recently.

I still wonder if this 11Khz CV update frequency is in fact 22Khz... to insure a proper tone at 11Khz you'd have to follow the nyquist rate, which is always the double of the intended frequency to reproduce, thus 22Khz.

I'm wondering still, if this update frequency has been raised on Pro 2, compared to P12.

maybe I'm just speculating too much :lol:
The CDM article explains that the 11KHz update rate is on control voltages not the audio signal path per se. In these recordings there's no modulation it's just a single oscillator going straight out so the sound shouldn't be affected by the update rate of the internal CVs.

I'm guessing the DACs in the Pro-2 sample at 44.1 or 48KHz and to prevent aliasing the oscillators are band limited. You can see in this image that there are no harmonics past ~14KHz:
Pro2_A880_96KHz.png
Pro2_A880_96KHz.png (410.83 KiB) Viewed 4332 times
whereas the Sub37 has plenty of harmonic content up into the 30KHz range:
sub37_A880_96KHz.png
sub37_A880_96KHz.png (445.97 KiB) Viewed 4332 times
(Also, notice on the Sub37 the hump at 80Hz. It's always there.)

I suspect the band limiting is what gives the oscillators their 'digital sheen' not anything related to the internal update rate of the CVs. If you look at the waveforms you'll see the Pro-2 sawtooth and square waveforms have the characteristic wiggle of a truncated harmonic series. The Sub37 waveforms don't have that.

aiynzahev
Posts: 32
Joined: Mon Jun 04, 2012 7:18 pm

Re: Pro 2 Oscillators

Post by aiynzahev » Sun Sep 07, 2014 1:57 am

Wow this is great thanks.

I thought of looking at the waveforms with an oscilloscope. But your comparison with the Sub37 is valuable.

I've had problems with various synths sounding "slow" or plodding. Others have suggested to me that the high frequency content contributes to this. Attack time is one thing, but perceived snappiness is another.

The fastest sounding synths I can think of are the Nord Lead 4 and Sylenth1. Some slower ones are Diva and the VirusTI and DSI-Pro2.

Again this is not about actual attack time as the attack on the Pro2 is very quick, but the perceived snap. A lack of high end detail may or may not be a part of it.

Pym
Posts: 2197
Joined: Thu Mar 11, 2010 6:24 am
Location: San Francisco, CA
Contact:

Re: Pro 2 Oscillators

Post by Pym » Sun Sep 07, 2014 6:08 am

If you want more snap, you actually need to slow down the envelope, not speed it up. It's strange, but true in most cases. The initial opening, when extremely quick, causes a slight click and a perception that the note starts a slight bit later. Psychoacoustic stuff I believe.

Try that, see if it helps
aiynzahev wrote:I've had problems with various synths sounding "slow" or plodding. Others have suggested to me that the high frequency content contributes to this. Attack time is one thing, but perceived snappiness is another.

The fastest sounding synths I can think of are the Nord Lead 4 and Sylenth1. Some slower ones are Diva and the VirusTI and DSI-Pro2.

Again this is not about actual attack time as the attack on the Pro2 is very quick, but the perceived snap. A lack of high end detail may or may not be a part of it.
Chris Hector
https://soundcloud.com/iampym
Dave Smith Instruments

aiynzahev
Posts: 32
Joined: Mon Jun 04, 2012 7:18 pm

Re: Pro 2 Oscillators

Post by aiynzahev » Sun Sep 07, 2014 11:33 am

Pym wrote:If you want more snap, you actually need to slow down the envelope, not speed it up. It's strange, but true in most cases. The initial opening, when extremely quick, causes a slight click and a perception that the note starts a slight bit later. Psychoacoustic stuff I believe.

Try that, see if it helps
aiynzahev wrote:I've had problems with various synths sounding "slow" or plodding. Others have suggested to me that the high frequency content contributes to this. Attack time is one thing, but perceived snappiness is another.

The fastest sounding synths I can think of are the Nord Lead 4 and Sylenth1. Some slower ones are Diva and the VirusTI and DSI-Pro2.

Again this is not about actual attack time as the attack on the Pro2 is very quick, but the perceived snap. A lack of high end detail may or may not be a part of it.
Oh yeah good point, this is true with some other synths I've used too. I have been varying the attack slightly to hear this effect but almost on auto pilot without thinking about it much. I'll give this a proper test today.

mr_rock
Posts: 8
Joined: Wed Sep 17, 2014 2:22 am

Re: Pro 2 Oscillators

Post by mr_rock » Sun Jan 24, 2016 4:37 pm

Bit of an old thread, but I thought I would post the raw saw and square waves for the Pro2 - and for comparison, the Evolver (analog) and my Nord Lead 3 (va.)

Pro 2:
Pro2_Saw_Sqr.jpg
Pro2 saw and square waves
Pro2_Saw_Sqr.jpg (1.42 MiB) Viewed 3000 times
Mono Evolver Keyboard:
Evolver_Saw_Sqr.jpg
Evolver saw and square waves
Evolver_Saw_Sqr.jpg (1.39 MiB) Viewed 3000 times
Nord Lead 3:
Lead3_Saw_Sqr.jpg
Lead 3 saw and square waves
Lead3_Saw_Sqr.jpg (1.43 MiB) Viewed 3000 times
I would have expected cleaner, much more accurate waveforms from the Pro 2. While the Pro 2 'saw' wave is saw-like, the 'square' wave has little in common with an actual square wave, looking more saw-like than square.
DSI Evo DT #198, Pro 2 #337, MEK #896, Tempest #3425 // Yamaha AN1X, RY30, RM1X // Nord Lead 3 // Virus Classic

My recordings: https://soundcloud.com/fuzzys-dead-mouse

User avatar
nielsd
Posts: 79
Joined: Sun Sep 06, 2015 1:28 pm
Location: Göttingen / DE
Contact:

Re: Pro 2 Oscillators

Post by nielsd » Sun Jan 24, 2016 10:02 pm

many thanks for your work.

Seems really a bit strange to me - i'm really wondering that the "good old" evolver is producing "optically" the "cleanest" signals in both cases...

I just use a software oscilloscope (BlueCat Multi OSC) over a UAD Apollo 2 which shows me a "optically" better signal from the Pro 2 "Pulse" (which i think is "square") - with Signal over Filter 1 which is fully open with 0 Res, Fdbk etc.

But it seems the signal is "cleanest" with around 50% of volume. I've tried pitches around one octave the center of the keyboard with similar results. The "Pulse" is getting (optically) more "edgy" as more as the shape is higher/lower then 1/1.

Are you sure you have no other influence (filter, feedback, delay, FM/AM, OSC mods over matrix or similar) anywhere in the path? Or is a software oscilloscope like my not useful / "true" here? I.e. the "basic program" of the Pro 2 is not "clean" in case you was "going from there" (i did tweaked my own very plain).

many thanks for any hint.

(If i should make some further images from pro 2 and evolver (have no Nord synth) feel free to ask)
Attachments
Pro2-Pulse.png
Pro 2 - Pulse - Shape 0 (1/1)
Pro2-Pulse.png (1 MiB) Viewed 2985 times
Niels Dettenbach
dettenbach audio
music fab
http://dettenbach.de

mr_rock
Posts: 8
Joined: Wed Sep 17, 2014 2:22 am

Re: Pro 2 Oscillators

Post by mr_rock » Mon Jan 25, 2016 1:42 am

I went back and tested again.

Evolver is the top trace, Pro 2 the bottom trace. I started with a mid C, then up 2 octaves, then up another 2 octaves to the highest playable C.
Evo_Pro2_Sqr.jpg
Square wave: Evolver (top) Pro 2 (bottom) 5 Oct spread
Evo_Pro2_Sqr.jpg (597.22 KiB) Viewed 2979 times
Evo_Pro2_Saw.jpg
Saw wave: Evolver (top) Pro 2 (bottom) 5 Oct spread
Evo_Pro2_Saw.jpg (603.71 KiB) Viewed 2979 times
Both started out at about 1 Vpp. As the frequency increases, the Pro 2 waveforms 'clean up', but show a definite roll off of higher end harmonics, especially on the square waves, as well as a drop of about .5 Vpp. The Evolver maintained 1 Vpp with significantly less high end roll off.

Just a bit surprised by the inconsistancy of the Pro 2 oscillator compared to the Evolver.

Not that any of this makes the Pro 2 is any less fun. :)
DSI Evo DT #198, Pro 2 #337, MEK #896, Tempest #3425 // Yamaha AN1X, RY30, RM1X // Nord Lead 3 // Virus Classic

My recordings: https://soundcloud.com/fuzzys-dead-mouse

Bach
Posts: 56
Joined: Fri Oct 28, 2011 12:27 pm

Re: Pro 2 Oscillators

Post by Bach » Thu Jan 28, 2016 4:09 am

Very very strange. I don't know how I feel about this.

soundxplorer
Posts: 23
Joined: Fri Jan 30, 2015 4:30 pm

Re: Pro 2 Oscillators

Post by soundxplorer » Thu Jan 28, 2016 10:44 pm

mr_rock wrote:Evolver is the top trace, Pro 2 the bottom trace. I started with a mid C, then up 2 octaves, then up another 2 octaves to the highest playable C.
Evo_Pro2_Sqr.jpg
Umm, there's something fishy about your Pro 2 waveforms there. Are you sure you completely initialized all parameters in the patch before recording?
I just initialized a patch and recorded a single OSC set to square, then ran it through the scope in Reaper. The picture below is what I see (fuzziness probably due to the high noise floor of my quick recording).

Image

ihearanewworld
Posts: 10
Joined: Tue Feb 17, 2015 8:23 pm

Re: Pro 2 Oscillators

Post by ihearanewworld » Fri Jan 29, 2016 9:40 pm

Bach wrote:Very very strange. I don't know how I feel about this.
Well, how do you feel about how the synth sounds?

Cires78
Posts: 1
Joined: Sat Dec 12, 2015 11:10 pm

Re: Pro 2 Oscillators

Post by Cires78 » Sun Jan 31, 2016 12:43 am

Hi people,
For what it's worth, here's mine. No overdrive, slop, etc.
square.JPG
Square
square.JPG (39.39 KiB) Viewed 2741 times
Saw
Saw.JPG
Saw.JPG (38.14 KiB) Viewed 2741 times
Cires

soundxplorer
Posts: 23
Joined: Fri Jan 30, 2015 4:30 pm

Re: Pro 2 Oscillators

Post by soundxplorer » Sun Jan 31, 2016 2:59 am

Three normal looking waveforms from myself, nielsd, and Cires78.
I think it's safe to assume at this point that your patch had some settings affecting the waveform, mr_rock.

Post Reply