Closed scottveirs closed 1 year ago
Should this be on orcasite instead of orcanode?
Thinking about this again after using CHANNELS=1
vs CHANNELS=2
to troubleshoot the high-frequency noise on the Orcasound Lab feed...
I just checked in S3 to compare files sizes betwee the successive restarts of the streaming container. All other .env
file variables remained the same.
When CHANNELS=2
the file size is ~195 kB for a 10-second HLS .ts segment.
When Channels=1 we are saving bandwidth. I think ffmpeg is successfully taking a single input channel and mixing it to generate a smaller file size 107 kB (again, for a 10-second HLS .ts segment). So, I conclude the streamed file is single channel (mono), but the Orcasound player (on OSX/Mac/Chrome) presents it in stereo, i.e. the same mono input channel in both left and right ears.
I haven't checked CPU usage on the Raspberry Pi in each mode, but will close this issue for now.
Streaming/archiving only a single channel could be a good idea, e.g.
If the Orcasite player could generate a stereo stream from a mono stream, then it could do so if it