I am well down the rabbit hole of ReplayGain, searching for a scanner -- and, also, the most suitable scanning parameters to suit my needs. I know that I prefer tag based volume normalization, as opposed to the EBU R 128 option in Strawberry (no tags), because I use different players on different platforms (Linux, Windows and Apple) and ReplayGain tags work on either Linux or Windows (Apple, unfortunately, has its own proprietary approach.)
All my m4a files, which I use in portable playing environments, have been scanned for ReplayGain in MusicBee. I don't use ReplayGain for critical listening with lossless files.
The question I am wrestling with currently is whether or not to apply clipping protection during the scanning process. MusicBee and foobar2000 (Windows) and DeaDBeeF (Linux) scan ReplayGain without clipping protection, thereby relying on the anti-clipping function of the player to take care of it. The rsgain utility, with which I am very impressed, uses the default of applying clipping protection during the scan, (although that can be easily changed using a configuration profile).
Which approach (during RG scan) would be best for Strawberry?
In Strawberry settings > backend > Audio normalization >ReplayGain there is an option to "Apply compression to prevent clipping".
Is this the same as preventing clipping using peak tags?
The reason I ask is because of the following, from the rsgain README:
*"...not all ReplayGain-compatible players actually implement clipping protection using the peak tags. rsgain has a clipping protection feature that attempts to prevent clipping at scan-time instead of during playback.
...The clipping protection works by adjusting the calculated peak by the calculated gain value (as it would be during playback). If this "new" peak value exceeds the maximum peak (full scale by default), then the gain will be adjusted lower by the excess amount, bringing the "new" peak down to the maximum level."*
If Strawberry uses peak values to prevent clipping during playback, then there should be no problem. If not, I might consider rescanning for ReplayGain using some different parameters.