Realtime vs Offline Render (Question for PG)

I have always wondered how the offline render handles external wordclock and jitter.
When playing back something in realtime with a high quality clock (say an Antelope OCXHD) you can hear a difference. (To my ears…)
When bouncing offline does the external clock actually matter?


When bouncing offline does the external clock actually matter

Not at all, because there is no clock at all. The stream rate could vary, stop, accelerate,… the only important thing is not to miss a single sample.


Thanks PG,
I can visualize for example adding +2dB on an EQ at certain frequency going to the CPU and back, with it’s respective latency, etc, lining up to be rendered… What boggles my mind are the non linear FX like a long morphing reverb mixed with a delay with modulation and filters creating saturation and resonances over time. The offline bounce has to output every single sample, calculating all the delay compensations per channel or plug, plus the summing of all the channels and FX.:exploding_head:

There is no difficulty because we are in the digital world: time is quantified, eg. 44100 values per seconds.
The algorithm knows that each sample is 1/44100 th of a second. It can handle these units at any rate.