I am not a technical guy so the question may sound completely stupid, sorry
but ... for the average newbie like me Lossywav & Wavpack Hybrid seems to share lot of similarities, if not technically at last in the goal they tend to achieve: transparency between 256 & 512Kbps with a lower artefact probability than overkill lossy. (to be short a rationnal alternative to MP3 320Kbps)

I have the felling that nowadays the situation between wavpack hybrid vs lossywav is the same as the situation between wavpack 3.0 vs flac 1.0 several years ago ... wavpack 3.0 was a great codec but everyone was using flac 1.0 because wavpack 3.0 was missing fast seeking ... wavpack 4.0 solved this problem ... nowadays lossywav offers you the possibility to use almost any major lossless codec, while wavpack hybrid doesn't.
So I fear that everyone interested in hybrid, will soon use lossywav 1.2 & drop wavpack hybrid in the same way everyone was using flac 1.0 instead of wavpack 3.0 back in the days.

So the question is: is there any chance that in the future wavpack hybrid becomes a wav pre-processor for any lossless encoder ? is it technically possible or are the compression techniques used in wavpack hybrid so lossy that it wouldn't compress more with a lossless encoder ?

In short is wavpack lossy just a wav pre-processor packed in a wv that Bryant could split, or is it so lossy that calling it a wav pre-processor is a non-sense ?

I use wavpack 4.0 but have never used wavpack hybrid so far because it shares the same flaw as lossy for me, you're trapped in a format. I have already been trapped by MP3 & Vorbis, so I switched to 100% lossless instead of Nero AAC ... but now my hard disks are full & the only option I see is lossywav ... i'd like to see some competition in the wav pre-processor area & a mutation of wavpack hybrid seems to me like the only possible thing that could happen.

So I'd rather look stupid & just ask ...

I don;t know about wavpack as a preprocessor but I suppose its easy to merge lossywav into wavpack.. I'll leave the rest for David to comment on.

You could use wavpack lossy on your main drive and backup the additional correction files to your backup drive then you can never be 'locked in' (you are backing up right ?) or use very high quality settings and just transcode to your lossy codec of choice dropping lossless altogether. It will work out great at half the bitrate of lossless or less. With the latest V4.5 encoder around 384k -hx4.. near impossible to abx anything and transcoding will be superb. If you want to go hardcore you can even up it to 500..550k and have an incredible quality headroom and you can probably edit the same source dozens of times without damage. With the latest improvements 320 k is practicaly enough for most common use (listening and single transcode to lossy). Of course there is also lossywav option.

wavPack lossy in it's current form is married to wavPack.
Every lossless codec is based on a predictor scheme. It predicts the value of the next wave sample based on the knowledge of the previous samples. The prediction scheme is known by the decoder, and thus the predicted value itself need not be encoded. The prediction can't be exact of course, and the (usually small) prediction error is encoded losslessly.
What wavPack lossy does is encode the prediction error only with an accuracy of a certain number of bits as the target accuracy. This way wavPack lossy makes use of the specific wavPack prediction scheme.

The integration of a lossy variant into a lossless codec can have advantages for optimizing things.

The separated procedure like it's done with lossyWAV however makes you independent from the final lossless codec.

shadowking's remarks make me share my decision of doings things now:
I don't have a huge collection, but I don't want a lossless plus a lossy production archive any more. I want just 1 productive archive.
What I do is:

a) as a standard procedure I use lossyWAV at a very high quality setting (-q 7.0 --shaping 0.5).
b) for the rare tracks which have a very important meaning to me I use lossyWAV -q 10.0 --shaping 0.8.
c) for all the tracks that can be compressed very effeciently by a lossless codec so that procedure a) resp. b) produces a larger file or a file of nearly the same size I use wavPack normal -x3 (x5 at the moment as I have tak files already so I can tell whether or not lossless coding is efficient). I also encode losslessly in those very rare cases when a track has an extremely important meaning to me.
d) I also have a considerable number of mp3 files. I will continue to use mp3 at very high bitrate for the tracks that have a minor meaning and quality to me.

This way I get what I want qualitywise and i feel safe for the future. For a) to c) final codecs can be exchanged, and mp3 is considered to be pretty much future safe. When the day should come for mp3 transcoding I guess it won't hurt then to transcode to lossless (or very high quality lossy without really sacrifying quality).

Yes, If one doesn't mind a non-free format and just wants it 'easy' i.e. - One collection, small size, good quality, no transcoding.. LAME -V4 does it.

And MP3 will be entering the public domain this year or the next I believe, making is as free as Vorbis. Combined with universal compatibility, it's a no-brainer.

To be fair halb27, you and I both know that that's what you do today. It isn't what you did a year ago (for obvious reasons) and I'll be really impressed if it's what you're doing a year from now.

It sounds like sauvage78 feels he's in a similar situation, is sick of changing, and wants some knowledge of the future. It would be interesting to know sauvage78 if you felt the need to upgrade each time because of what you heard in your encodings, or because of what you read on line. If the latter, I have a simple solution: stop reading about audio codecs! If you can't resist, you will just have to go lossless and buy another HDD.

The best advice is to pick something and be happy with it. wavpack is great, flac is great, the lossy versions of both are great mp3 is great too. There will be better versions of all of these if you wait a year or two, but what are you going to do - re-encode every year? That way lies madness.

Cheers,
David.

That's it, and that's what I feel is the biggest advantage to me for my decision: I did decide for a certain way of lossy encoding (not talking about the rare exceptions), and I'm half way through with my collection. After encoding I deleted my tak files, so there's no way back.
You're right with respect to my past year, but I can't imagine ripping CDs again, as I'm used to totally ignore my CDs after ripping, and I'm really content with the chosen lossy quality of the overkill kind.

I can only provide advice in my particular circumstance. Most PC's have the space/ability to add hard drives. 512 gig capacity is getting cheap these days. The only reason anymore to eschew a lossless compression codec for an archive is you are so poor you can't afford another hard drive or maybe you have a laptop as your only PC. In that case go through your collection and weed out the less important stuff and turn it into .mp3, .ogg, .mpc or whatever. Keep your important stuff lossless (I use wavpack but thats not critical).

I do have some part of my collection in wavpack lossy 512k, and kind of regret it, but not more than if I had it in .mp3 or any other lossy codec since any transcodings will be 'nearly' perfect.

Any effort toward lossy archiving needs to have a compelling reason and one that I really can't come up with logically given cheap mass storage these days. Lossy has its place in portable players and even that will go away in the (near)future as 128 gig flash memory emerges.

Do yourself a favor and keep your permanent archive lossless. You can do anything you want with your music after that with no regrets.

I want a wavpack or Flac DAP. I want a near lossless ~ lossless free codec .I want out of MP3 & proprietary stuff. I want gapless. I want more choice.

Wait a year, then MP3 will be public domain

bigdog,

If you search this forum, you will find plenty of "why lossless?" threads on HA. This shouldn't become another one.

I think it's enough to say that you don't have to be "poor" to eschew lossless. An improvement that you cannot perceive isn't worth any time or money to most people - it doesn't matter how cheap HDDs get or how fast the files can be transferred, or even how widely they can be used.

I know some people go for lossless in a big way and do buy additional HDDs for all the files. This means they are willing to pay for potential flexibility, to stop them worrying about encoding and transcoding, and mostly for a warm fuzzy feeling.

Meanwhile, I'm sure plenty of millionaires use mp3!


Then there are those of us who think that many modern CDs are of such terrible quality, yet yield such high lossless bitrates, that lossless encoding is tantamount to polishing a turd! Dub it onto a C90, convert it to a 128kbps mp3, and it'll probably sound better!

Cheers,
David.

Really? According to Wiki and http://www.tunequest.org/a-big-list-of-mp3-patents/20070226/ it won't.

And, if this: http://lwn.net/Articles/166346/ is true:


But it's all about USA, not Netherlands.

Anyhow back to the OP, Wavpack and lossywav are nothing alike. WV is a total lossless solution with an optional lossy+correction file mechanism.. Its dead easy to use and the correction files are very well integrated unlike optimfrog. Lossywav is a good option for high bitrate lossy but so is wavpack.

No it shouldn't have

I agree about the quality of many CD's. Point well taken. Funny how ears can be trained. 8-track tapes through cheap car speakers once sounded great to me. One has to wonder if there is a program or something out there to 'train' folks in accepting 'i-tunes' as high fidelity. Not to start another thread...just thinking out loud.


Peace

dog