Head-Fi.org › Forums › Equipment Forums › Portable Source Gear › Lossy Audio Codec's Comparison [HUGE amount of pics] [iTunes UPDATE on p.7]
New Posts  All Forums:Forum Nav:

Lossy Audio Codec's Comparison [HUGE amount of pics] [iTunes UPDATE on p.7] - Page 5

post #61 of 225
Quote:
Originally Posted by Sir Nobax View Post
What makes you think every human hears the same thing, if i had a hearing loss between 3khz and 5khz and the Psychoacoustic model used in the codec thinks that a tone of 4khz would overshadow the 2khz tone in a specific music file the psychoacoustic model is flawed in this case. So if i heard the source i would hear the 2khz better then the 4khz one, in the compressed file the codec filtered out the 2khz tone making my experience different from the source.
I'm sorry, but I don't quite understand what you are trying to argue here. If you mean that the psy model can fail, yes of course it can. I said no less than that in my previous post. However, this sort of problem usually crops up at the margins, at lower bitrates (128kbps or below) than at higher bitrates. Masking thresholds are often tuned with some margin of error left in to account for users with greater hearing acuity and ear training. A great deal of information can still be removed without qualitative difference even taking into account physiological variability amongst users which, on a scale quantitatively relative to the source information, is not really that high in general. Still, some codecs do allow for manual tuning for those unusual users with peculiar hearing response.

Quote:
But with the pictures you can't see wich 128kbps encoding of different codecs (without parametric stereo and Spectral band replication) is better sounding to you, you only see how accuratly it reproduces the original sound. But with the 256kbps WMA 10 Pro graph you can clearly see that it is WAY better then the MP3 at 256kbps, if someone could hear the difference is something else, and thats why spectrographs are more accurate then listening tests in my opinion.
Hmm, this argument does not make much sense if taking into account practical and theoretical considerations of psychoacoustics and lossy compression. The datarate is finite and set in a constant-bitrate scheme. The lossless compression applied by several codecs is of similar compression ratio before it hits the filterbank, and thereafter all of them have to lose the same amount of data quantitatively in order to achieve the desired bitrate. The data presented in your sonograms could be interpreted in a variety of ways as a result. I'll try once more to explain this to you. I could look at some of these graphs and conclude, for example, that the psy model is inefficient and dedicating bandwidth to masked data whilst neglecting to allocate sufficient bandwidth to audible portions of the signal. Looking at your sonograms, some of them appear to have quantisation noise and probably other error (such as pre or post echo) that isn't immediately apparent on sonograms of this resolution.

Your paradigm of analysation doesn't allow for this sort of analysis, though, despite it being a very realistic and sensible approach when taking into account psychoacoustics, because your analysis works off of a superficial impression of the data with a particular systemic bias towards how it will be intepreted (which essentially amounts to it looking good to you because it superficially appears more similar to the source sonogram). That, sir, is not a scientific approach; it's pseudo-science and now we've several other posters who are committing the same fallacy in their own evaluation. It is entirely scientific to claim that you need ABX listening tests to evaluate codec performance, because the ultimate basis for evaluation of a codec is qualitative and whether the qualitative loss is minimised; that is the objective of psychoacoustics and lossy audio compression. The object of lossy compression is minimum qualitative loss with maximum quantitative loss. Psy models are tuned with listening tests. After listening tests, you may have data which you can correlate with some measured data for making scientific models and predictions, but you still need the listening test data in order to competently evaluate the efficacy of your psy model. Otherwise, again, you cannot evaluate the qualitative performance, which is a problem since that is precisely what you need to maximise. Thus, your sonograms and their superficial appearance do not host the totality of relevant information or background for making competent evaluations of codecs. Insisting that they do is what is the pseudo-science and conjecture here, and it's unfortunate that this sort of thing is taken prima facie as the basis upon which to choose a codec.

hydrocity - One sonogram with a superficial appearance closer to the other doesn't necessarily ensure maximisation of qualitative performance, particularly when dealing with higher compression ratios. If you're having trouble ABX'ing a compressed file against the source file, that means it's likely you are unable to perceptually distinguish them, and as such the codec is doing its job of minimising qualitative loss.

Note: I'm not saying the graphs aren't interesting; they are. I simply disagree with your methodology for analysing the performance of a codec. The information can still be useful, however, when placed against a backdrop of other data and compression theory.
post #62 of 225
Amazing work. Though none of my files are compress I love to see that you put so much time into trying all of that.

-Alex-
post #63 of 225
too complex for my blood.

just going to stick with my cds..
post #64 of 225
Quote:
Originally Posted by cantsleep View Post
too complex for my blood.

just going to stick with my cds..
Try FLAC =)
post #65 of 225

Good Job Sir Nobax!

Thanx for the hard work! What about some report on the ABX findings? Anylink?
post #66 of 225
Just wondering if you could do this test with the new itunes 224 cbr mp3 and the 192 vbr. There is no more vbr for 224 with the latest itunes
post #67 of 225
Thread Starter 
...oh no, ill have to install itunes now .

No prob, itunes mp3 @ 224 CBR and 192 VBR coming up.
post #68 of 225
oh whoops, i meant itunes 192vbr aac and 224 cbr sice vbr with 224 doesnt work
post #69 of 225
And of course you can't forget Hydrogen Audios current recommended LAME mp3 version 3.97. The only version that got recommended since 3.90.3. Try it at --preset fast extreme and --preset fast standard thats the VBR new settings at V0 and V2
post #70 of 225
reply pls hehehew
post #71 of 225
Just for fun, could you do a wav vs. alac vs. flac vs. ape?
post #72 of 225
Could you do one comparing .WAV to Apple lossess and WMA lossless.

Thks, I'm kinda a noob at this.
post #73 of 225
Thread Starter 
Quote:
Originally Posted by pheonix991 View Post
Just for fun, could you do a wav vs. alac vs. flac vs. ape?
Quote:
Originally Posted by That dude
Could you do one comparing .WAV to Apple lossess and WMA lossless.

Thks, I'm kinda a noob at this.
Yes, i could....

But it would be useless , since they are all lossless, no data has been deleted like the lossy ones i posted. The lossless files look 100% exactly the same as the original WAV file i posted on the first page. The only thing that differ's between the lossless codec's is the size of the encoded file. That however will be different with every song, because of the different techniques and time/accuracy ratio of the used lossless codec.

Quote:
Originally Posted by donunus
And of course you can't forget Hydrogen Audios current recommended LAME mp3 version 3.97. The only version that got recommended since 3.90.3. Try it at --preset fast extreme and --preset fast standard thats the VBR new settings at V0 and V2.
All the "MP3 Lame" files in the opening post are encoded with V3.97, but not with the standard presets, i have the files here and i shall upload the pictures as the first thing tomorrow, when i have access to some higher speed internet connection. (using my phone's GPRS connection ATM :S )
About the AAC files, ill probably have them uploaded tomorrow also, since i dont have iTunes on my computer at the moment and need to download iTunes first.
post #74 of 225
What is the command line of blade encoder for EAC ?
post #75 of 225
Nobax:
They are pretty pictures however useful they are!

Strange results for vorbis. Were you using an old version? The AOTUV encoder generally does very well in listening tests, at least since beta 4. If you haven't used this or a later version, any chance you could chart the autov encoder release 1 say at 256kbs?

Musepack might be interesting too if you feel like it.

Nice work anyway.

PS the graphs are just the frequency spectrum over time are they?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Portable Source Gear
Head-Fi.org › Forums › Equipment Forums › Portable Source Gear › Lossy Audio Codec's Comparison [HUGE amount of pics] [iTunes UPDATE on p.7]