Some startup help needed

[jacobacci]

The quibble I have with the current solution is that all samplerates are resampled to the same target sample rate, also samplerates that would not have to be resampled.
The reason I would prefer a more differentiated transcoding / resampling behaviour for downsampling is that for a given processor power, resampling quality is best with integer resampling ratios and for a given quality, processor load is lowest for integer ratios. In my case, MinimStreamer runs on a Synology NAS, where processor power is a scarce resource.
I wasn't thinking of introducing a new property that overrides stream.transcode.
Regarding the stream.transcode property, am I correct in assuming that the parameters in the field are interpreted by MinimStreamer and then passed on to ffmpeg in a format ffmpeg understands? If that were the case, one could simply have a specific parameter that goes into the stream.transcode field that characterizes the max_samplerate of the transcoding table to be used and MinimStreamer would then use the appropriate transcoding table when giving transcoding parameters to ffmpeg. Someting like
flac:wav24;96m
flac:wav24;192m or
flac:wav16;44m
For DSF the problem does not exist, as the optimal output samplerates are multiples of 44.1, independent of whether the input file is DSD64, 128 or 256.
Pls don't get me wrong, I am not trying to be difficult, just looking to get optimum quality for given resources. If what I propose is not doable, I am very happy with MinimStreamer being able to use SOX (big step up from ffmpeg).