Head-Fi.org › Forums › Equipment Forums › Sound Science › FLAC vs. 320 Mp3
New Posts  All Forums:Forum Nav:

FLAC vs. 320 Mp3 - Page 6  

post #76 of 504

I've lost the thread but I was pretty sure that deadlylover had use the ReplayGain plugin for foobar and had the level matched (he posted his entire test log etc), so I was convinced that he did do the test properly.  That said those who does have these golden ears probably make up less than 0.1% of the population.  

 

It's like it is possible for someone to run 100m under ten seconds and for sure someone has done it, but if some random joe comes up to me and say they can ran 100m under 10 I'm naturally not going to believe him until he shows me.  It's the same for mp3/FLAC test here, I'm certain a tiny amount of people can do it, just not 99% of those who claims they can do it, especially not those who says it's "easy and if you can't you must be deaf" kinds of people. wink.gif


Edited by nanaholic - 7/10/12 at 11:26am
post #77 of 504

I don't really know how much the different codecs have improved, but how can anyone tell... if AAC 256 VBR was transparent four years ago for some ears?

 

Well back in 2008 there was an article in Stereophile (from fig. 7 and downwards) about FLAC vs. AAC vs. MP3 that might be of interest.

 

Quote: http://www.stereophile.com/content/mp3-vs-aac-vs-flac-vs-cd-page-2
 
Both MP3 and AAC introduce fairly large changes in the measured spectra, even at the highest rate of 320kbps. There seems little point in spending large sums of money on superbly specified audio equipment if you are going to play sonically compromised, lossy-compressed music on it.

It is true that there are better-performing MP3 codecs than the basic Fraunhöfer—many audiophiles recommend the LAME encoder—but the AAC codec used by iTunes has better resolution than MP3 at the same bit rate (if a little noisier at the top of the audioband). If you want the maximum number of files on your iPod, therefore, you take less of a quality hit if you use AAC encoding than if you use MP3.

 

Back then it supported what majkel had to say about going beyond 320 kbps...

 

 

Quote:
Originally Posted by majkel View Post
 I find full timbral transparency at q8, and full spectral (no treble roll-off) at q10 but it needs very good records/amplification/headphones or speakers to hear it. For headphones like Sennheiser PX100, q8 is the limit.

Edited by Albedo - 7/10/12 at 1:14pm
post #78 of 504
Quote:
Originally Posted by JK1 View Post

The lasers in CD players wear out over time. As they wear out, the laser light output decreases. If the player has had a large amount of use, the light output becomes low enough so that player become much more vulnerable to dirt and scratches on a disk blocking enough light from being reflected from the disc. 

I'll go along with that.  I spent way too much time reading and learning about the under $200 Sony Blu-Ray players this weekend, and have a BDP-S590 in my Amazon cart currently.  My DVP-S9000 has MAAAANY hours on it.

post #79 of 504
I like how everyone who does a careful comparison finds out that there's no audible difference, but then they ask why anyone would play compressed audio on good equipment. Well duh! Because it sounds the same!
post #80 of 504

http://www.stereophile.com/features/308mp3cd/index.html

 

LOL at Stereophile using "objective" data to conclude that something sounds "sonically compromised."  Funny, objective data in Stereophile never seems to prove that two things actually sound the same.

 

What does the objective data reported in that article even show?  That basically the lossy formats reproduced almost all the signals the lossless formats reproduced.  The lossy signals just had a higher noise floor.  Just how high was that lossy noise floor?  About -80db.  Any one know what the noise floor is on a reel to reel analog tape deck?  On a vinyl record?  Just how audible is -80db?  Heck, stereo separation on fancy phono cartridges is less than -30db.  And that was just the 128K files.  The 320K files had noticeably lower noise floors.  Indeed, for the 320K AAC files "the noise-floor components have dropped to below –110dB below 16kHz, and to below –120dB for the lower frequencies."

 

One thing that Stereophile article doesn't do is mention that any one actually listened to the various lossy formats to determine the audibility of any "sonic compromises."  Indeed the only mention I can find of a listening test in that article is in one of the footnotes which refers to a different article from 1995.  In the 1995 article the writer stated "I found the performance of the DTS system in its 240kb/s mode ... phenomenally good."  http://www.stereophile.com/reference/456   The article from 2008 does admit: "the AAC codec produces a result that may well be indistinguishable from CD for some listeners some of the time with some music." 

 

While I agree with the writer's conclusion that "Both MP3 and AAC introduce fairly large changes in the measured spectra, even at the highest rate of 320kbps," nothing in the article itself supports the statement that these changes result in the music being "sonically compromised."  They may well be, but without a proper, double blind listening test we'll never know. 

post #81 of 504
Stereophile was a joke even when I was in high school.
post #82 of 504

Props to Steve Eddy for mentioning this in a different thread.  I'm posting here so he can respond if he wishes.  Thanks, Steve Eddy, for mentioning this!

 

This is the Audio DiffMaker:  http://www.libinst.com/Audio%20DiffMaker.htm 

 

I highly recommend downloading it and then trying the "Downloadable DiffMaker 'Dyf' files."  The "Listener Challenge" at the end is a hilarious "ear opener."  How many of the tracks can you identify that have the "difference" in them?   

 

I think this bolsters my point above that -80 db would be very hard to hear.
 


Edited by BlindInOneEar - 7/10/12 at 5:31pm
post #83 of 504

Can we boil down the question of this thread to CD's dynamic range at 96 dB vs. the "instantaneous dynamic range of human audio perception" at 85 dB?

post #84 of 504
Normal listening volume is boosted a bit because our ears aren't very sensitive at very low volumes. In order to hear the quietest parts of a recording with music at -85dB, you would have to raise it 20 or 30 dB to over 100 dB, which is around the point where listening gets pretty uncomfortable. Play it on speakers and you'll have to add a bit to overcome te room's noise floor. Add 20dB more and pain is going to start to set in.

Most recorded music only has around 45-50dB or so, even the most dynamic classical music. Music that goes all the way to te edges of the dynamic range would be unlistenable. You'd be adjusting te volume constantly.
Edited by bigshot - 7/10/12 at 5:46pm
post #85 of 504

I was thinking more of transients and attack, which "is often forgotten in the traditional Fourier analysis based reasoning on sound characteristics".

 

Quote: http://www.zainea.com/physiologicalsound.htm
 
The root of this terminology lies in linear system analysis. There steady‐state means that a clean, often periodic or almost constant excitation pattern has been present long enough so that Fourier based analysis gives proper results. Formally, when exposed to one‐sided inputs (non‐zero only if time is positive), linear systems exhibit output which can be decomposed into two additive parts: a sum of exponentially decaying components which depends on the system and a sustained part which depends on both the excitation and the system. The former is the transient part, the latter steady‐state. Intuitively, transients are responses which arise from changes of state—from one constant input or excitation function to another. They are problematic, since they often correspond to unexpected or rare events; it is often desired that the system spend most of its time in its easiest to predict state, a steady‐state. Because transients are heavily time‐localized, they defy the usefulness of traditional Fourier based methods.

In acoustics and music, the situation is similar in that frequency oriented methods tend to fail when transients are present. Moreover, in music, transients often correspond to excitatory motions on behalf of the performer (plucking a bow, striking a piano key, tonguing the reed while playing an oboe etc.), and so involve
Significant nonlinear interactions (instruments behave exceedingly nonlinearly)
Stochastic or chaotic phenomena (often from turbulence, as when sibilant sounds are produced in the singing voice)
Unsteady vibratory patterns (the onset of almost any note)

Partials with rapidly changing amplitudes and frequencies (as a result of the above).

-

It is kind of funny how little attention time information has received in the classical study, although one of the classic experiments in psychoacoustics tells us what importance brief, transient behavior of sound signals has. In the experiment, we record instrumental sounds. We then cut out the beginning of the sound (the portion before the sound has stabilized into a quasi‐periodic waveform). In listening experiments, samples brutalized this way are quite difficult to recognize as being from the original instrument. Furthermore, if we splice together the end of one sample and the beginning of another, the compound sound is mostly recognized as being from the instrument of the beginning part. In a musical context, the brief transient in the beginning of almost all notes is called the attack, then. For a long time, it eluded any closer inspection and even nowadays, it is exceedingly difficult to synthesize if anything but a throrough physical model of the instrument is available.

-

So the steady‐state part is certainly not the best part to look at if source classification is the issue. The other part of the equation are the neural excitation patterns generated by different kinds of signals—transients tend to generate excitation in greater quantities and more unpredictably. Since unpredictability equals entropy equals information, transients tend to have a significant role in conveying useful data. This is seen in another way by observing that periodic sounds leave the timing pathway of the brain practically dead—only spectral information is carried and, as is explained in following sections, spectra are not sensed very precisely by humans. Kind of like watching photos vs. watching a movie.

post #86 of 504
Quote:
Originally Posted by bigshot View Post

Normal listening volume is boosted a bit because our ears aren't very sensitive at very low volumes. In order to hear the quietest parts of a recording with music at -85dB, you would have to raise it 20 or 30 dB to over 100 dB, which is around the point where listening gets pretty uncomfortable. Play it on speakers and you'll have to add a bit to overcome te room's noise floor. Add 20dB more and pain is going to start to set in.
Most recorded music only has around 45-50dB or so, even the most dynamic classical music. Music that goes all the way to te edges of the dynamic range would be unlistenable. You'd be adjusting te volume constantly.

 

This x1000. All the talk I hear about the different dynamic range values between formats is silly. A vinyl record is already capable of producing all the dynamic range you could ever want. Recorded music doesn't go beyond that, and you wouldn't want it to either.

post #87 of 504
I'm not understanding you, Albedo. I'm not good with scientific theories. I only know about the practical application of them.... Digital audio has a 90dB dynamic range which extends downwards. You put your loudest sound... the transient peak... at the top of that range (aka normalizing) and the differences between one dynamic range and another is the difference between noise floors. The transient peaks are the same if you have a 60dB dynamic range or a 110dB dynamic range.

Recording engineers compress sound to maintain a listenable balance. In particular, they compress transient peaks so they can organize sound into an overall level that is comfortable to listen to. They'll allow a certain amount of headroom for dynamic effects and percussion, but in general, it isn't desirable to have huge spikes that go far beyond normal listening levels.

If you wanted to record in a perfectly unfiltered manner, you could certainly set a level that contains the peaks without clipping and digital audio would give you an accurate reproduction of it. But recording and mixing is a creative process that organizes sound to make it clearer and comfortable to listen to without straining. I don't know why you'd want to jettison all of that just for the sake of theory.
Edited by bigshot - 7/10/12 at 6:31pm
post #88 of 504
Quote:
Originally Posted by nanaholic View Post

 

That's not true.  IIRC some forum posters (deadlylover(?)) on this forum had showed that they could ABX 320kbps and FLAC of some tracks with 100% accuracy with a certain setup of his, with some being easier than others and that he has to be in analytical mode to pick up the difference, so some people can certainly tell the difference under certain conditions.    

 

However when someone boosting that they can tell 320kbps and FLAC using leaky earbuds on their smartphone when on the go such that they won't use mp3s I'm going to be very skeptical.  

 

No, no. It's all just your imagination. If you work SOO hard analyzing you get confused and perhaps your brain starts playing games on you. I have an HD 800 which is said to be a very analytical headphone, and a pretty good system built around it. There is no difference. However, there is a difference between 320kbit/FLAC and 160kbit tracks. 

320kbit compression is high enough you are pretty much eliminating all audible artifacts with the file, if the encoding itself is of high enough quality. 


Edited by Lan647 - 7/11/12 at 3:52am
post #89 of 504
Quote:
Originally Posted by Lan647 View Post

 

No, no. It's all just your imagination. If you work SOO hard analyzing you get confused and perhaps your brain starts playing games on you. I have an HD 800 which is said to be a very analytical headphone, and a pretty good system built around it. There is no difference. However, there is a difference between 320kbit/FLAC and 160kbit tracks. 

320kbit compression is high enough you are pretty much eliminating all audible artifacts with the file, if the encoding itself is of high enough quality. 

 

I wish you would actually read my post - it's not me imagining it but some forum poster who actually posted EVIDENCE (which was scrutinized here btw) that he can tell the difference.

 

I personally can't tell the difference and I don't bother with lossless either - just like I can't run 100m under 10seconds and many others can't either, but some gifted individuals can.  The fact is scientifically 320kbps is not completely transparent compared with lossless, even though it may be inaudible to 99.99% of the population, but that doesn't mean that 0.01% of people who can tell the difference don't exists, and that certain people does have better hearing which is also perfectly within reasonable human biological limits.

post #90 of 504
Quote:
Originally Posted by nanaholic View Post

 

I wish you would actually read my post - it's not me imagining it but some forum poster who actually posted EVIDENCE (which was scrutinized here btw) that he can tell the difference.

 

I personally can't tell the difference and I don't bother with lossless either - just like I can't run 100m under 10seconds and many others can't either, but some gifted individuals can.  The fact is scientifically 320kbps is not completely transparent compared with lossless, even though it may be inaudible to 99.99% of the population, but that doesn't mean that 0.01% of people who can tell the difference don't exists, and that certain people does have better hearing which is also perfectly within reasonable human biological limits.


No disrespect, but the fact that a forum poster convinced YOU that he could tell the difference between high bit rate lossy and lossless hardly constitutes "scientific fact" that high bit rate lossy is audibly distinguishable from lossless, or that there is a select group of individuals who can make the distinction.  The incident may be interesting.  It may even be worth investigating further.  What it's not, though, is proof.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
This thread is locked  
Head-Fi.org › Forums › Equipment Forums › Sound Science › FLAC vs. 320 Mp3