Hi,

I'd like to use wavpack lossy 512 kbps... (as a unique archive of my music)

Do I need to pay attention to -b -h -n -x4 ...and all those extra things to get high quality wv files ?
or is this bitrate efficient enough to prevent bad tricks ?

Thanks

512kbps seems like overkill. I thought that ~380-450 was expected to be transparent.

You definately need to pay attention to -b, as you will use that switch to set the bitrate.

I would suggest that you take a look at the WavPack docs to decide what other switches you use, but the general answer is: yes, you should pay attention to them. Most won't affect quality (unlike -b) but they will affect compression and speed.

alright thank you

The usage guide says:

So, x3 is recommended while h is, at 384, let alone 512, assumed to be unnecessary (making it slow for little to no gain).

The only thing I would add to what has been said is that I am now recommending the --dns option for all lossy encoding (except for sample rates above 48 KHz). This will be standard in the next release, but for now it's available in the alpha version described here.

I understand that some people may hesitate using an alpha, but that version has been downloaded over 2500 times and I haven't received a single error report. It's probably okay.

I can confirm the alpha version using --dns is fine. I used it for a considerable period with no problems at all and great quality (being very cautious minded I used
-b400 -h -x5, but this is overkill in nearly any situation).

Yup, been using the alpha with no probs here too (my sig's command line plus "--dns " with about 30 CD's), in some cases (Johnny Cash, Corrs, Fishbone, Hairspray 2007 OST) with more polished (compared to the last stable) lossy results.