Interesting vid... Thanks for posting.
Just tested a bit further. The OB-X8 has a falling saw / reverse saw oscillator, and the pulse is low for the first half, phase locked (at 50% duty). I think Marcus described it opposite in the video (with rising saw), but his point remains.
In the original OB-8, this must be how their phase relationships of the saw/pulse oscillators were.
I guess the part that still confuses me a bit is why the amplitude of the saw seems to be attenuated down when the pulse is added... Is that how the OB-8 VCOs work as well?
On many other synths, I use this type of phase locked saw+pulse combo, but the relationship between the two waves is usually phase inverted for one of them, which is why I was expecting a different result. And some others have full 0-100% PWM control, rather than 50-100%. My preference from a sound design standpoint would be to have the pulse phase inverted when they are combined, but if this is how the OB-8 is configured, I totally get keeping it that way for this recreation.
If possible via firmware to have a menu option to phase invert the Pulse, that would be feature request of mine -- it would elevate the sound design possibilities. The thin pulse added to a saw can create a very nice buzzy saw sound, and there's lots of other interesting timbres as you pwm the pulse.
On the plus side, this current implementation does allow the OB-X8 to create some interesting analog supersaw type of patches. (by using saw+pulse on both oscillators with detune and pwm, and potentially stacked -- You can get the sound of up to 8 sawtooths per note... for reference, Roland's Supersaw used seven I believe) Pseudo analog supersaw. I created a couple patches last night that sound pretty huge.