Hi all,
Replying to my own thread again.
On 7/3/22 8:43 pm, Chris McCormick wrote:
On 7/3/22 7:34 pm, Chris McCormick wrote:
- Make the signal was somewhat robust to volume changes (e.g. if the
source volume was not at max it can still be read). 2. Make the signal robust o quite a saturated spectrum (e.g. if the source is playing high bandwidth music the signal is still recoverable). 3. Make the signal inaudible such that if it goes into a regular mixing desk it doesn't make it sound bad.
Sorry to reply to my own thread but it seems like a "lock-in amplifier" might be what I want:
I've attached a basic implementation of this. The toggle state is transmitted over a noisy audio channel using an 18kHz carrier at 10% volume (inaudible to me but my 11 year old might find it irritating - will test). At the moment there is both lag (10s of ms) and jitter (10s of ms).
I think this can be improved by adding a second detector 90 degrees out of phase:
These two quantities represent the signal as a vector relative to the
lock-in reference oscillator. By computing the magnitude (R) of the signal vector, the phase dependency is removed.
So I'll try that next and see if I can reduce the jitter. If I can get reasonably constant lag it will be good enough for a sync pulse as that can be compensated for on the receiving end.
After building this test I'm optimistic this silly in-band sync messaging can be achieved without interfering with the music itself. \o/
Cheers,
Chris.