Sync sweep?

If you set the synced waveform to saw or square, yes, you get that classic sync sound. Won't work well with sine or triangle waves, this might have been your issue.

Both oscillators were set to saw, and it sounded bad... maybe a problem with the unit in the store? I don't know. It sounded nothing like a sync sweep.

Most likely because the "samples" are being synced rather than VA waveforms

Contrary to the diagram on the SH-01, I believe OSC1 modulates OSC2.

When in sync mode, the pitch control of OSC1 will affect the restart period of OSC2. This control needs to be set at a higher value than OSC2's pitch for it to sweep properly. The manual makes mention of this as well.

If OSC1 is not "pitched" properly (set higher than OSC2), it will not sound "good". You're using sync to cause OSC2 to restart its wave before it normally would. So OSC1 needs to be a shorter wavelength: higher pitch.

You can use the LFO to modulate the pitch of OSC1 to create an oscillating sweep. Or use OSC1's pitch envelope to create a one-shot sweep.

Start with an init patch: hit cancel & write simultaneously. Turn on mod sync. Set OSC1 to square wave, max pitch. I found it easier to program a sync sweep by temporarily setting the waveform of OSC2 to noise (the quietest is the green variation). Now I can clearly hear the pitch of OSC1, and concentrate on setting attack and decay of OSC1's pitch env to my liking. I chose around A=2, D=8. Don't forget to give it some envelope depth. Then switch OSC2's waveform back to a saw for the full effect and any final tweaks.

As mentioned previously, it's better to use the saws or squares. PW and PWM is disabled in sync mode. And sadly the Gaia becomes monophonic.

Isn't the oscillator that should have a higher pitch the one that is being SYNCED (slave oscillator) and forced to restart, and not the other way round? According to your explanation, OSC 1 is the master and OSC 2 is the slave, yet you're changing the pitch of OSC 1.
That doesn't make sense to me, I believe it's the other way round, OSC 2 is the master (constant frequency) and OSC 1 is the slave being synced, therefore changing OSC 1's pitch produces the effect. Someone correct me if I'm wrong.

I think you're correct Gil. I studied the diagram on p.31 of the manual and I think I understand the mod sync routing better.

Where I was getting confused is that a higher pitch has a higher frequency. If the frequency is higher, then the period is smaller. I thought that the smaller period would be needed to restart a waveform with a larger period.

In the diagram that they show on p. 31, oscillator 1 waveform is higher pitched -- but they depict it as being greater than one period. I had thought that both waveforms were just one period... that's where I got mixed up.

So yes, I think the diagram was correct all along.

Although I'm still somewhat confused as why the OSC1 filter type is bypassed in sync mod.

I suppose the filter would alter the waveform and maybe screw up the syncing algorithm.. probably would sound bad anyway.

I played the GAIA again today and managed to create a nice sync sweep, turns out you can only get it by using an LFO or pitch envelope, not by manually changing OSC 1's pitch since it doen't go high enough. Which is a pity.
But it sounded great with an LFO, especially with some delay and a flanger on top to fatten it up
Great little synth!