Digital Output for SY77/SY99

The Yamaha SY99 is a synthesiser combining frequency modulation synthesis (branded as Advanced FM) and sample-based synthesis (branded as Advanced Wave Memory 2) and the direct successor to Yamaha's SY77/TG77

Moderators: Derek, parametric, Fozzer, Saul

mjbrands
Member
Member
Posts: 15
Joined: Sun Aug 25, 2013 3:49 pm

Re: Digital Output for SY77/SY99

Unread post by mjbrands » Sat Sep 14, 2013 2:44 pm

Nice post, I enjoyed reading that. I have a 192 kHz audio interface in storage; I'll fetch it and see if I can create some 96k/192k recordings.

Edit: Darn it, I damaged the glass fiber that connects the interface to my PC and now I'll have to wait for the replacement to arrive...

I'm wondering if audio equipment was able to reproduce a 15 kHz sound (or higher) with any fidelity at the time the SY99 was released. I can imagine consumer audio cassettes didn't have the bandwidth for that, so removing everything above a specific frequency might make sense (maybe it was already done as part of the mastering process).

About 10-12 years ago I was still able to hear up to 18.3 kHz. I was archiving all my audio CDs at that time and I was using CD Paranoia to rip them and LAME to encode them to VBR MP3s (using the 'extreme' setting, I think). I applied a filter to them that rolled off all frequencies above 18 kHz, since a) I couldn't really hear sounds up there well anyway, b) I found that it often contained digital crap left over from mastering (i.e. Alanis Morrisette's Under Rug Swept album is a good example of an album that had clearly audible digital artifacts even on a CD) and c) the compression got slightly better.

My point: maybe it was acceptable for Yamaha to roll of sound starting at say 16 kHz, since it didn't play an important part in recordings at that time. That would also allow them to take some 'shortcuts' (cheaper solutions); i.e. use the mentioned 64 kHz sampling rate for the OP3, instead of say 96 kHz. This was probably also a technical limit; better technology => higher cost (or not even possible yet).
Gear list: Saleae Logic16, Rigol DS1052E (@100 MHz), some Xilinx FPGA boards, lots of screwdrivers and a decent soldering iron

JK1974
Member
Member
Posts: 85
Joined: Tue Oct 02, 2012 1:01 am

Re: Digital Output for SY77/SY99

Unread post by JK1974 » Tue Sep 17, 2013 12:14 am

I just did a recording with the AUX2 input of my Creative Labs Audigy 2 - it says both in the windows audio settings as well as in the settings of the recording software, that this input is able to record with 96 kHz/24 bit.

And here is the result of an AFM sine sweep:
SY99 Sine Wave AFM.png
As you can see, there is a flatline at 24 kHz - which leads to the conclusion that the internal sampling rate is 48 kHz. Furthermore, I see a volume attenuation beginning at around 8 kHz.

Attached, you find a recording of the SY99 sine sweep as lossless compressed flac file.
Attachments
SY99 Sine Wave.zip
(1.65 MiB) Downloaded 47 times

db7

Re: Digital Output for SY77/SY99

Unread post by db7 » Tue Sep 17, 2013 7:38 am

mjbrands wrote:Nice post, I enjoyed reading that.
Thanks!
My point: maybe it was acceptable for Yamaha to roll of sound starting at say 16 kHz, since it didn't play an important part in recordings at that time. That would also allow them to take some 'shortcuts' (cheaper solutions); i.e. use the mentioned 64 kHz sampling rate for the OP3, instead of say 96 kHz. This was probably also a technical limit; better technology => higher cost (or not even possible yet).
Yeah, I presume so. Also, having the cutoff there as opposed to, say, 24 kHz would help them get away with using fewer poles while still providing a good level of attenuation of ultrasonics and whatnot.

As for the sampling rate of the OPS3, I guess higher rates are always good, but if I’m not mistaken, 64 kHz was the highest rate Yamaha had yet used in an FM chip. The previous leader was the OPP (YM2164, identical sound-wise to the OPP/YM2151) in the FB-01, which was clocked at 62.5 kHz from a master clock of 4 MHz. Other synths using the OPP or OPZ(II) were clocked lower, IIRC, due to having master clocks nearer 3.6 MHz. As I’ve explained elsewhere, the OPS3 seems to have descended from the OPZII (V50) and OPSII (DX7 II), the latter clocked at 49096 Hz (at least as the OPS in the the DX7 mk I). So, that trajectory seems like another plausible reason that they didn’t push the sampling rate much further.
JK1974 wrote:As you can see, there is a flatline at 24 kHz - which leads to the conclusion that the internal sampling rate is 48 kHz.
Interesting. Thanks. I suspected that might be more likely than 64 kHz.
Furthermore, I see a volume attenuation beginning at around 8 kHz.
Of what degree? Using a legible scale on the y-axis would have helped here. An attenuation is effectively meaningless if it is inaudible or within acceptable bounds of ripple. All analogue LPFs will induce attenuation right from 0 Hz onwards, but for the vast majority of frequencies between that and the nominal cut-off, the attenuation will be so small as to be ignored. The point at which the attenuation hits 3 dB is typically accepted as the point where it becomes worth worrying about. I presume this occurs around 16 kHz in your recording, just as it did in mine.

I have seen nothing to indicate any form of built-in EQ or otherwise non-standard frequency response from the SY99.

JK1974
Member
Member
Posts: 85
Joined: Tue Oct 02, 2012 1:01 am

Re: Digital Output for SY77/SY99

Unread post by JK1974 » Tue Sep 17, 2013 10:57 pm

You seem to be right concerning the frequency range: I transfered two drumloops and the famous Roland 'Insomnia' sound to the SY99 and recorded it with 48 kHz/16 Bit with my Audigy 2 - and couldn't spot the difference. In the 90ies, I later even used to boost the high frequencies on every sample before I transfered it to the SY99. At that time, I used Creative Labs soundcards that were said to have very linear frequency ranges (only few realized the very bad distortion factor), but after they even didn't manage to support 1:1 SPDIF bitstream I/O, I switched to Terratec EWS64XL - and used this mainly for RAM sample playback. Maybe the Creative cards (SB16, SB32AWE, SB64Gold) simply didn' t have a high-quality input besides to all the other quality flaws.

db7

Re: Digital Output for SY77/SY99

Unread post by db7 » Wed Oct 09, 2013 3:40 pm

A very relevant and equally interesting PM I received from acreil on Gearslutz:
I was wrong about the SY99 AFM sample rate. The FM feedback thing actually oscillates at Fs/3 rather than Fs/4, so the correct sample rate is really 48 kHz, with apparently no sample rate conversions happening anywhere. But this is weird, because it means that there are 128 time slots per sample, but only 48 operators per OPS3. Each operator takes two clocks to calculate, and the rest might be wait states for synchronization to the M3 or something, or it could have even been originally intended to be an 8 operator system. It's kind of weird to apparently waste those clock cycles when the OPS/OPSII calculated one operator per clock.

There is a good reason, though, for the EGM2/OPS3 to use two clocks per operator. The OPS/OPSII both use separate buses for the amplitude and frequency data. On the OPS3 there's only one bus, and it's time-multiplexed. This extra clock cycle may also be used internally to some other benefit, like linear interpolation of the lookup tables or... something.

Sorry for the delay anyway; I didn't want to say anything more about it until I did some digging and read through the service manuals and your threads. I was actually busy analyzing other synths in unnecessary detail.

I'm now working on an Octave script to analyze the feedback FM oscillation thing a little better. The reason that I thought of using this oscillation tone to determine the sample rate to begin with is because at one point I tried to come up with a better alternative to the Tomisawa filter when I was working on an implementation of feedback FM.
So, basically, AFM does not run at a different sampling rate from AWM and then get downsampled: it runs at the same rate as the final output from the synth and presumably from the M3 too. I might comment on this more at another point, when I have time.

I also meant to reply to JK1974 previously, so I hope I find my draft! ;)

swissrolf
Member
Member
Posts: 25
Joined: Wed Apr 02, 2014 6:22 pm
Contact:

Re: Digital Output for SY77/SY99

Unread post by swissrolf » Mon Apr 07, 2014 10:51 pm

Wow, I am impressed about the detailed technical knowledge in here. I just looked into the thread to learn about the audio out of the SY99 I also used some time. Well regarding sample rate conversion - did you find a solution in the meanwhile? I tried with digital filters using FPGAs like mentioned above and found that large FPGA area is needed to perform resampling effectively and most accuratly because of the high number of Filter TAPs for FIR-filtering. I tried with 1024 TAPs and at first got stuck because of the limited FPGA speed.

rtype909
Member
Member
Posts: 22
Joined: Wed May 30, 2012 8:27 pm

Re: Digital Output for SY77/SY99

Unread post by rtype909 » Fri Jun 06, 2014 10:39 am

some more info re PCM56p and digital out here: http://www.diyaudio.com/forums/digital- ... 56p-2.html

User avatar
scottrod
Member
Member
Posts: 19
Joined: Sat Jul 11, 2009 12:00 am

Re: Digital Output for SY77/SY99

Unread post by scottrod » Wed Jun 17, 2015 4:44 pm

I wonder if the EXDG01 digital out card for the EX5 might be of help to see how it could be done. All schematics are in the EX5 service manual.

Would love to have a digital out on the SY99...

Post Reply

Return to “Yamaha SY99 Forum”

Who is online

Users browsing this forum: No registered users and 1 guest