I would like to create a hybrid WavPack with a lossy component of a 128 kbps (or maybe in future a higher bitrate), but whatever I do, the bitrate remains about 200 kbps. Is it possible to create lower-bitrate WavPacks? If yes, how? If not, why?

Many thanks.

Quality doesn't scale well below 250-300kbps. Remember WavPack is a simple lossy encoder, that doesn't use all the fancy compression methods used in formats like AAC, Vorbis, or even MP3...

Does that mean that I should use lossy WavPacks that are about 250-300 kbps?
Will that be okay quality or should I use some other lossy codec (such as Vorbis) for every-day playing?

well, it depends on your hearing, on your equipment, what you understand as "everyday playing"

If your everyday playing is HiFi in livingroom or special hobby-listening room, then go for high quality codecs (eg. MPC, Ogg, MP3-lame) at higher bitrates, or directly Lossless, or as Lossless-Hybrid, wavpack hybrid is fine for this, ie. the lossy file at 200-250k, plus lossless-correction file.

If your everyday playing is in a car eg., or portable (USB-stick or HD-based ?) then go for a lossy codec at lower-mid bitrates. At 128k, I could even recommend MP3 at lame setting -V5 for this. Or Ogg, aac, mpc at pocket-pc.

In the end, you need to find out yourself, which codec is ok for you, and at which bitrate you are pleased, abx,
eg. try out, if already the 200k wavpack hybrid is ok for everyday usage.

Been there, done that. Just needed a quick answer 'cause I got that "crazy" idea about using the hybrid Wavpack so I could use the lossy part for "everyday" use and keep both lossy and correction in the archive. Finally ditched that idea, 'cause I had to use quite high bitrate which I couldn't ABX (about 350-450 kbps).

As for the lossy codecs, nowdays I generally use --aps, --ape, or Ogg Vorbis -q 6 or -q 7.

Thanks for the help everyone.

were 350-450 kbit/s wavpack lossy your limit of transparency ? So, below 300k were clearly abxable, and 200k bad sound ?

I guess that lowest possible bitrate WavPack encodes (generated by -b2) have been proven inferior to those of other lossy codecs?

Wavpack and optimfrog ds don't have artifacts. You are not listening to the same audio from radio, tv, tape & hifi equipment. They also produce the same noise to a degree. The others codecs artifacts are more annoying to my ears.

Still at 200k the other codecs are mostly transparent but wapack is clearly not. The difference is like turning off Dolby NR on a tape deck. At 260k it sounds quite good and differences are only there in quite sections where there are 'gaps' in the music. At 320k it is generaly it is hard to pick these on most samples.

Also test samples are much different that say mp3 or mpc. You only test music with gaps and solo instruments - everything else is handled really well and will never fail. With other codecs any sample can fail.

Another thing that recently occured to me is thoughts of artifatcs in lossy encoding. How many times have you heard weird things you thought were artifacs in your lossy files ? With wavpack and ofs you just tell yourself that its impossible. Hissing is much more difficult to define in the mind.