Francesco Ariis <fa-ml(a)ariis.it> writes: That's what was worrying me. "Slightly" clearer does not seem like a particularly good deal, considering that it loses part of the sound signature that actually allows telling the various parts of the accordion apart. The original version is played by a combo and obviously has no problems keeping the solo and accompaniment parts simultaneously separated and blended. I don't actually get much of an average loudness change with the kind of compression settings I use (either version is ultimately normalized). In contrast, the videocam's own audio track is much much louder. It's also a whole lot trashier but it does demonstrate that I don't really have a handle on using compression for boosting average volume. The reverb gives it a "richer" sound quality particularly when listening on headphones: the recording is pretty dry. Maybe I should do that and just EQ up the bass (which is compromised in volume in a full accordion due to size constraints) and forget about compression which I don't manage to turn into more than a rather mixed blessing. Shooting and uploading? You wish. The lion's share is spent on practising a whole lot, and reshooting and cussing a lot. With my current tool set (Ardour and Shotcut) I found editing/splicing out mistakes to be so much trouble that reshooting tends to be the easier recourse. Surviving post-production tends to take comparatively little time (a lot of post-production tends to take the form of writing down time codes and measure numbers and ratings of good and not so good passages from various takes and putting things together and then deciding I'd better spend another afternoon on new takes). So basically the amount of time I spend on post-production is negligible compared to the work leading up to recording, and uploading has become sort-of a non-issue too since they put out fiber to probably half a mile away last year or so. -- David Kastrup