350 kbps MPC better than MP3 320? Options

[memomai]

Has MPC got better quality in 350 kbps (Q10) than MP3 (Lame 3.98) @ 320 kbps CBR? Or should I go with Nero AAC in this high bitrate area? I convert normal music. i know i can also go with FLAC and so on for best quality but i just want to know if MPC has got a better quality and if Nero AAC would be better than MPC to use in this area.

Thanks

This post has been edited by memomai: Aug 21 2008, 16:51

[pdq]

All of those sound like overkill. How about something like lossy wav?

[senab]

Why do you want to go that high? There is no need to go that high for perceptual lossy encoding. Try all of them at ~128k and see if you can ABX them, if you can then slowly increase the bitrate until you find a bitrate that suits. Anything over ~200k is silly imo for lossy encoding... unless you have "golden ears" of course.

[dutch109]

Why do you want to go that high? There is no need to go that high for perceptual lossy encoding. Try all of them at ~128k and see if you can ABX them, if you can then slowly increase the bitrate until you find a bitrate that suits. Anything over ~200k is silly imo for lossy encoding... unless you have "golden ears" of course.

Even with golden ears is there someone on this planet able to ABX non-killer sample at theses bitrates ?

[guruboolez]

Both formats (encoders) have flaws even at high bitrate. My experience on that subject is that problems are less rare and probably easier to ABX with MP3 (pre-echo with castanets-like samples). Long time ago, I ABXed lame freeformat at 640 kbps with castanets.wav. But MPC wasn't totally transparent too at --quality 10. The tests are old, the encoders now outdated, but for proof:

http://www.hydrogenaudio.org/forums/index.php?showtopic=5060
http://www.hydrogenaudio.org/forums/index....showtopic=12031

Honestly, I wouldn't bother that much about difference in quality. Problems at this bitrate and modern lossy encoders are very rare and they need very specific conditions to be noticed. As a consequence mp3, aac, vorbis, mpc... they all offer the same perceptual quality for music listeners. If you want security, use any lossless encoder.

This post has been edited by guruboolez: Aug 21 2008, 18:36

[jmartis]

Even with golden ears is there someone on this planet able to ABX non-killer sample at theses bitrates ?

If it's ABX-able then it IS a killer sample at such high bitrate, or not?

This post has been edited by jmartis: Aug 21 2008, 18:37

[[JAZ]]

Has MPC got better quality in 350 kbps (Q10) than MP3 (Lame 3.98) @ 320 kbps CBR?

Strictly answering the question, the characteristics of MPC (or MP2 for that matter) make it so that with enough bitrate, it is less prone to problems than other, more complex codecs like MP3 or AAC.
Where is the threshold? I can't tell, but as other have suggested, if you are going to accept bitrates of 300 and 400kbps, probably lossyWAV+FLAC or simply wavpack lossy are going to be a better solution, quality-wise.

[Neasden]

Once we went to the beach and I took with me this AAC data CD, it had like 14 albums and plenty music. The rented room resort had a DVD Player with MP3 support and a decent loudspeaker system, and because of that AAC was out of the question in that "weekend". Result, I never got to play and enjoy the songs I mostly liked for the occasion.

You should choose your lossy formats not because they can be more transparent than another, but how much playback support they have and how much of development they get nowadays. Suppose you need to take your files to somewhere that it's not your own player, that can be very frustrating if it's not supported, as it happened to me. That's why I stick with MP3 (-V0), because there is no use fighting the "world".

[DigitalMan]

OK, so maybe you can hear the difference some times with a lossy encoder, even at high bitrates. So what. I would guess the vast majority of the time you can't hear the difference, and the rest of the time you can just enjoy the music. If you're that worried about transparency just use a lossless codec.

Neasden has a great point - there are other reasons besides perfect transparency to pick a codec and a bitrate.

[weaker]

OT: Hey guru welcome back! Nice to have you here at HA again :-)

[sld]

Has MPC got better quality in 350 kbps (Q10) than MP3 (Lame 3.98) @ 320 kbps CBR? Or should I go with Nero AAC in this high bitrate area?

Thanks

Run your own ABX tests and convince yourself. If you find that your sweet spot is around 128-160 kbps you'll even congratulate yourself for saving so much space.

Why not compare MPC to MP3 at the same bitrate, 320 kbps? I'm ignoring empirical data and the latest in LAME development here, but what I have read and learnt about these 2 formats over the past 6 years is that MPC has much less weaknesses than MP3, pre-echo being a major problem for the latter.

This post has been edited by sld: Aug 22 2008, 11:17

[memomai]

I'm just interested if MPC was better than MP3 in their highest bitrate area. Nothing other recommendations about which codec should be better at that bitrate area, I've asked about MPC and MP3, or if AAC is the best one, tanks.

[Neasden]

I'm just interested if MPC was better than MP3 in their highest bitrate area. Nothing other recommendations about which codec should be better at that bitrate area, I've asked about MPC and MP3, or if AAC is the best one, tanks.

No, MPC is not. It was once. But don't think so nowadays. MPC has its weaknesses even at -q10, and MPC is not being actively developed in a long time. There may be someone tuning the code or the format features and last thing I know it was that the bistream format quality has not changed in a very long time. MP3/Lame has already caught up with MPC, there are tests with samples in which Lame is stronger than MPC. So choosing a codec because of their high bit-rate is completely irrational. Believe me you will have a very hard time ABX'ing a Lame/V2 file, so that enough is no reason to choose MPC just because MPC is a bit better than MP3. It all boils down to hardware support and development.

If you want *real* (not false) peace of mind, you should acquire a portable FLAC dedicated player or mod an iPod. Can't get better than that.

This post has been edited by Neasden: Aug 22 2008, 15:04

[shadowking]

That mpc isn't developed means little. You can't change the situation through belief. That mpc and mp3 PSY haven't changed. Therefore IMO mp3 320 k is at best as good as MPC --standard but probably not even that.

Cmon, There is a least a sample per month where V2 isn't transparent and 1 every 2 months where 320 k isn't. Not MPC -extreme not even -standard ; mp3. Even back 'in the day' there were only limited problems with mpc.

Mp3 is very good for what it is, but please no need for mpc FUD.

[benski]

Why do you want to go that high? There is no need to go that high for perceptual lossy encoding. Try all of them at ~128k and see if you can ABX them, if you can then slowly increase the bitrate until you find a bitrate that suits. Anything over ~200k is silly imo for lossy encoding... unless you have "golden ears" of course.

Even with golden ears is there someone on this planet able to ABX non-killer sample at theses bitrates ?

http://www.hydrogenaudio.org/forums/index....mp;#entry535132

Granted, this is the exception and not the rule.

[halb27]

As was said I would consider the practical context.
As you're considering using mpc I guess you want to play your music on a pc or notebook system.
In this case the best options for a very quality-aware user are
a) a lossless codec if you allow for pretty large files
b) lossyWAV --standard + a lossless codec if you allow for moderately large file
c) wavPack lossy if you're targeting at ~300 kbps.

For best usability with players you're best off with mp3. AAC gets more and more player support, and there are also a series of DAPs playing Vorbis. FLAC too is supported with some DAPs.
I wouldn't use mpc because of lacking player support. Other than that it's an excellent codec, but other codecs are excellent too.

From the choices you gave I'd choose AAC. Excellent quality and a promising future in the player world.

This post has been edited by halb27: Aug 22 2008, 15:53

[sld]

AAC is very strong in the 32-128 kbps bitrate range, and that's where I encode my music for portable use. Is there actually definitive information on the quality of AAC encodes at bitrates above 300 kbps?

[dutch109]

Even with golden ears is there someone on this planet able to ABX non-killer sample at theses bitrates ?

http://www.hydrogenaudio.org/forums/index....mp;#entry535132

Granted, this is the exception and not the rule.

I don't know where you found the 30s clip, but I guess this kind of killer sample is extremely rare in common music. It is easy for a lot of people to ABX pre-echo killer samples (ie : castanets) encoded with Lame and preset insane, but I was just wondering if we already found someone who can ABX 'normal music' (without looking for special killer samples) with such bitrates.

[Bodhi]

but I was just wondering if we already found someone who can ABX 'normal music' (without looking for special killer samples) with such bitrates.

Exactly!

Come on folks, are you listening to music or looking for faults?

[j7n]

It has been technically feasible for some time to archive uncompressed or lossless, if you need peace of mind to enjoy music. It's simple as that. Stop worrying about whether you made a mistake by encoding your albums to AAC and not Musepack.

This post has been edited by j7n: Aug 22 2008, 18:44

[[JAZ]]

Mp3 is very good for what it is, but please no need for mpc FUD.

While you have your reasons, i believe you took that a bit too hard. You can't compare a codec as commonly used as LAME vs Musepack and say that while the former gets problem samples the latter doesn't.
Musepack used to have its problems and work needed to be done to fix them.
Of course it is a great codec. I was an user in the Buschmann's days (mpegplus, .mp+), and was really happy to have 170kbps avg. files that were transparent.
Later on, Klemm made a really good job improving the speed and later on the quality as well.
Its design, where there are no fixed size blocks (so the bitrate can increase to more than 1mbit) and the avoidance of other common mp3 "characteristics" make it good for its job even today.

Since the question cannot be answered easily (other than what you said, which is not the full story), that's why me and others have said that the question may not even be worth looking into, although the OP is free to test by himself (via ABX, if he's going to post the results)

[benski]

Even with golden ears is there someone on this planet able to ABX non-killer sample at theses bitrates ?

http://www.hydrogenaudio.org/forums/index....mp;#entry535132

Granted, this is the exception and not the rule.

I don't know where you found the 30s clip, but I guess this kind of killer sample is extremely rare in common music. It is easy for a lot of people to ABX pre-echo killer samples (ie : castanets) encoded with Lame and preset insane, but I was just wondering if we already found someone who can ABX 'normal music' (without looking for special killer samples) with such bitrates.

The sample in question was normal music.

[Neasden]

MPC is a great codec, I have used it. I also wished that it became predominant like MP3. Unfortunately, real-world cases completely rule out beliefs and passion over a preferred format.

[dutch109]

The sample in question was normal music.

By 'normal' did you mean that it is not rare, and that similar problems could be found easily in common music ?

Can you post it please ? (just curious)