Phasing question for super geeks.

Covering the SY22 to SY55 and TG Modules.

Moderators: Derek, parametric, Saul, Fozzer

User avatar
JonWest
Member
Member
Posts: 12
Joined: Sun May 25, 2008 12:00 am

Phasing question for super geeks.

Unread post by JonWest » Sat May 11, 2013 6:40 pm

Please only continue reading if you are a ubergeek. This is pretty obscure stuff.

Got that, OK here goes:

I've been toying with my newly non-zippery SY-35 and noticed an odd thing about the oscillators. I started with an initialized patch (no vectoring, no AM or PM, all OSC levels set to max) and then set both the AWM and FM oscillators to standard sine waves. So I've got 4 OSCs all set to sine playing at the same pitch. What should happen is that we get a single sine wave output with 4 times the volume of a single oscillator.

As I play what is probably the most boring, uninspired patch across the keyboard, an interesting thing happens: There are several areas where the amplitude drops to almost zero for several clusters of adjacent keys. On the SY35 there are 3 of these clusters I found. Weird eh?

I suspected that some kind of phase cancellation is occurring. To test this out I take one of the 4 OSCS and give it a little pitch offset. Sure enough, the amplitude is restored and you hear 2 sine waves.

I then performed the same experiment on a TG33. I found the same thing there, however the location of the clusters where phase cancellation occurs are different.

Note that I also played all the OSCs alone and the sine wave sound is the same across all of them. I can't tell the difference by ear between an FM and an AWM sine.

The questions are, why and how are the SYs doing this? They seem to be reading the sine data from the same tables. There would be at most 1 table for each of the FM And AWM OSCs, although it wouldn't surprise me if both types of oscillators read the same data for sine. It's like there's an intentional phase offset that happens and the offset is dependent on note.

Anyone want to take a swing at this? Maybe someone remembers something being said or documented about it 25 years ago or something......

Thanks all. :?



User avatar
alodk
Member
Member
Posts: 11
Joined: Wed May 01, 2013 8:13 pm
Contact:

Re: Phasing question for super geeks.

Unread post by alodk » Sun May 12, 2013 8:28 am

Do your sy-35 have an option for synchronizing the oscillators on key-down?

If this is off, the oscillators will start at a random "place" of the sine wave. If sync is on, the waves will all start at zero going up(or something like that).


Hardware: MU128v2, plug-ins: plg150AN-VL-DX, Akai EWI4000S, RX21, MX49.

Software: GXWorks V.3.07(XG) :-), Cubase AI6(MX49) :-( , MidiOx(midi-connetions) :-D , Editor for the EWI( +Patchman sounds for the EWI).

User avatar
JonWest
Member
Member
Posts: 12
Joined: Sun May 25, 2008 12:00 am

Re: Phasing question for super geeks.

Unread post by JonWest » Sun May 12, 2013 2:54 pm

Thanks for reading and commetning on my post alodk.

As far as I know the vectoring SYs don't have anything like what you mention. It's interesting because the phase of sine waves is anything but random. It's always seems to depend on the note alone and you wouldn't even notice it unless you get two OSCs with the same single cycle waveform and the cancelllation effects happen, but only for certain notes. Like I mentioned it doens't matter if the OSC is FM or AWM, same thing happens.

Sure would be nice if there was an initial phase parameter for single cycle waveforms on these vector instruments. I'm guessing that the designers decided to vary it across the keyboard to make it sound more interesting and maybe that usually works but you get those notes where the combined amplitude just drops to near zero and that doesn't work.

One thought I keep having about the these Yamaha vector synths is that they seem to be a simple implementation of a complex idea. Almost like there should be a "Pro" version of them and what we have are the consumer versions. Would have been nice to have 4-6op FM, start phase, better effects, better envelopes, better pitch manipulation options, etc.



User avatar
JonWest
Member
Member
Posts: 12
Joined: Sun May 25, 2008 12:00 am

Re: Phasing question for super geeks.

Unread post by JonWest » Sun May 12, 2013 10:01 pm

Little update, played with this some more and noticed that you only get phase cancellation with like OSCs. So it only happens on patches configured to use all four OSCs (ABCD). So with no pitch modulation at all and all OSC's set to Sin 8' you'll hear the phase cancellation between OSCs B&D and A&C at notes C3, G4, Bb5. Again this is on SY35.

Looks like on a 2 OSC patch configuration, one AWM and one FM no phase cancellation occurs.



db7

Re: Phasing question for super geeks.

Unread post by db7 » Mon May 13, 2013 12:00 am

Nice to see someone else who notices and publicises weird little things like this, as I did for the AN1x. In that case, it was nice to learn that my synth was not weird when other users reported the same phenomenon . . . some pages later! :P I like seeing things like this discussed as it will reassure future readers that they are not imagining things. :lol:

Anyway. Your report of the interference only occurring between A/C or B/D suggests the difference in pitch is Element-specific. For the sake of comparison, on the SY77/TG77/SY99, phase-synchronisation is available for Operators within one Element but is not possible between Elements within a multi-layer patch, with the result that Elements exhibit a (roughly?) constant small difference in phase. Granted, that is not a difference in pitch, but it led me to suspect that a difference between Elements rather than individual oscillators is the source of what you have described.

Combining that with something else you said earlier, and what I reported about the AN1x and its apparently hard-wired fine detuning of VCO2, perhaps this is another example of an engineer doing something for some supposedly beneficial effect, in this case adding a small constant difference between elements in layered voices, without thinking to document it anywhere or make it optional . . . [sigh] . . . because such things should be hard-wired rather than depending upon the specific desire of the user, right? :roll:

But hey, at least on the SY35, you can detune it away; on the AN1x, the difference is too small to be countered by the Fine adjustment!



User avatar
JonWest
Member
Member
Posts: 12
Joined: Sun May 25, 2008 12:00 am

Re: Phasing question for super geeks.

Unread post by JonWest » Sun Dec 08, 2013 12:48 am

db7, all good stuff you posted there. Wish I could have replied earlier but my account got locked out and I couldn't get it reactivated until now.

Always fun to dig deep into these toys isn't it?



db7

Re: Phasing question for super geeks.

Unread post by db7 » Sun Mar 09, 2014 12:01 am

Sorry, I only just saw your reply now!

And indeed it is. :D

I’d be interested in asking some more geeky questions about this, so let me know here or by PM when you log in again!

Meanwhile, I have some digging of my own to do, and some fairly big ideas about how to do it… to be continued 8)



Post Reply

Return to “Yamaha SY22, SY35 & SY55”

Who is online

Users browsing this forum: No registered users and 1 guest