Hello all, I have been really considering converting my music to mpc, but I was wondering if it would be beneficial to hold off until sv8 comes out. I can't discern quality all too well, so that shouldn't be the primary factor. If you could post any reasons to start now or wait I would appreciate it.

just start with sv7

1.15r or 1.14 (--xlevel)

sv8 will obviously be more beneficial, but speculation of it's arrival is a bad gamble to say the least

(although, perhaps waiting for 1.2 would be a good idea, i dunno)

if you have a ton of musepack encoded files like i do, waiting is ridiculous, but if you are just getting into the game, it might be worth it


later

With all due respect to Frank Klemm (and I do mean a lot; musepack is an incredible format), I'll believe updates are happening again when I see them. I've waited for so long to see anything, but haven't seen a lot... The last developments I saw in Musepack were near the beginning of my HA experience.

mppenc 1.2 is apparently going to be just 1.15r with a couple tweaks anyhow. You won't gain much by holding off until then.

Besides, it's transparent already. You don't stand to gain much. The one possible benefit is perhaps SV7.5 will come out, which will allow for us to wrap Musepack albums in MKA files, but there's no player out there that uses a "file-as-album" paradigm rather than the "track-as-album" paradigm that's so common. Encoding everything to individual tracks is presently semantically best.

...i'm assuming that the profiles will be better supported in 1.2 in addition to the sv7.5 changes

thats about all
(but great!!!)


later

Dont expect significant improvements to the sound quality very quickly, this is anyhow hard to achieve with respect to MPCs incredible sound quality.

The amin focus, for the time being, is on

- new bitstream format ( but converting tools will be made to convert SV7 files )
- multichannel encoding
- new code structure ( plugins ), allowing other OSS projects to reuse parts of the MPC code ( psy model )

So, i wouldnt wait and start using MPC right now ......

don't forget portable support! also, bitrate peeling would be nice

It would be nice, but it's also very difficult to get bitrate peeling to a level where it outperforms re-encoding (as far as sound quality).

Bitrate peeling has one significant goal, to make smaller 'low storage capacity friendly' files. How you arrive at this has two extremes that may not be mutually exclusive, though they commonly are in reality.

• It should be FASTER than transcoding.
• The quality should be BETTER than transcoding.

Both of these traits rely on the peeler having intimate knowledge of the codec used, unlike current transcoding where no assumptions can be made of the source.

As the resulting quality of transcoding is already suspect, I submit that as long as the quality of bitrate peeling is not vastly inferior, the 'method' to focus on is speed. It would be desireable to just send a playlist from PC to portable in the time it takes to brush your teeth, and have the entire operation be done when you returned. This makes sense for network audio serving as well, because you will be CPU limited there as well.

AAC SSR works like that, but the approach is completely different than Vorbis' peeling. In AAC's case, the stream starts at CELP, at very low bitrates, and successive layers of AAC are added, each one at a higher bitrate, all the way up to lossless (with the recent addition of lossless coding to the MPEG4 portfolio). That way, peeling is very fast (just remove the layers you don't want) and of very high quality.

Unfortunately, it seems there is no SSR encoder publicly available.