I especially find that WMA has a richer, warmer sound than MP3 (yes, even LAME APS MP3s), although WMA chokes on some of the high notes, (the sssss's slur and guitar solos can be watery). WMA has rich, full bass.
That post in http://www.hydrogenaudio.org/forums/index....pic=12436&st=0& (http://www.hydrogenaudio.org/forums/index.php?showtopic=12436&st=0&) made me curious and as I don't want to add another off-topic post I start a new thread here.
*** Mod's/Admins - feel free to move the other WMA-related posts here (and edit the title if necessary) ***
1st thing I've been curious about: Are rumours like "WMA encodes louder than original" (see Chrisgeleven's 1st post in the other thread) or similar which could be a reason for sound perceived as "richer" or "warmer" true?
Test:
I encoded a track ripped from CD (Omara Portuondo - Quizas, Quizas, Quizas) with CEP2.1 to WMA 9 128kbps CBR and WMA 9 pro VBR 75 (closest to 128kbps) and had a look at CEP's statistics and frequency analysis:
Statistics:
Uncompressed
Left Right
Min Sample Value: -32393 -32393
Max Sample Value: 32392 32392
Peak Amplitude: -.1 dB -.1 dB
Possibly Clipped: 0 0
DC Offset: -.008 -.007
Minimum RMS Power: -82.9 dB -83.94 dB
Maximum RMS Power: -7.93 dB -6.43 dB
Average RMS Power: -16.89 dB -16.95 dB
Total RMS Power: -15.98 dB -15.97 dB
Actual Bit Depth: 16 Bits 16 Bits
Using RMS Window of 50 ms
WMA 9 CBR 128 CEP2.1
Left Right
Min Sample Value: -32768 -32768
Max Sample Value: 32767 32767
Peak Amplitude: 0 dB 0 dB
Possibly Clipped: 20 114
DC Offset: -.006 -.006
Minimum RMS Power: -85.96 dB -85.88 dB
Maximum RMS Power: -7.95 dB -6.44 dB
Average RMS Power: -16.88 dB -16.93 dB
Total RMS Power: -15.97 dB -15.95 dB
Actual Bit Depth: 16 Bits 16 Bits
Using RMS Window of 50 ms
WMA 9 pro VBR 75 CEP2.1
Left Right
Min Sample Value: -32768 -32768
Max Sample Value: 32767 32767
Peak Amplitude: 0 dB 0 dB
Possibly Clipped: 29 146
DC Offset: -.01 -.008
Minimum RMS Power: -82.55 dB -82.75 dB
Maximum RMS Power: -7.95 dB -6.44 dB
Average RMS Power: -16.87 dB -16.93 dB
Total RMS Power: -15.96 dB -15.95 dB
Actual Bit Depth: 16 Bits 16 Bits
Using RMS Window of 50 ms
Result: WMA isn't louder - it's slightly quieter (but the difference is way too small to be noticable)
2nd question: Does WMA "equalize" (i.e. change the relative volume of certain frequency bands)?
Test:
I took the files from above and Musepac 1.15r --standard --xlevel as reference and let CEP2.1 perform a frequency analysis. Then I copied the results to Exel and summarized the values to 1/2 octave bands: (CBR: WMA9 128kbps CBR; VBR: WMA9Pro VBR Q75)
Band Original CBR VBR MPC (all in dB)
- 00022 Hz -64.35 -61.60 -61.30 -63.82
- 00030 Hz -51.34 -51.20 -51.13 -51.35
- 00043 Hz -49.23 -49.14 -49.11 -49.23
- 00061 Hz -37.63 -37.62 -37.62 -37.63
- 00086 Hz -30.10 -30.10 -30.11 -30.10
- 00122 Hz -31.73 -31.72 -31.73 -31.73
- 00172 Hz -37.96 -37.95 -37.95 -37.96
- 00244 Hz -34.22 -34.22 -34.21 -34.22
- 00345 Hz -30.85 -30.84 -30.84 -30.85
- 00487 Hz -34.32 -34.31 -34.31 -34.32
- 00689 Hz -34.34 -34.33 -34.34 -34.34
- 00974 Hz -36.43 -36.42 -36.42 -36.43
- 01378 Hz -39.59 -39.57 -39.56 -39.58
- 01949 Hz -43.41 -43.33 -43.33 -43.40
- 02756 Hz -49.36 -49.25 -49.22 -49.34
- 03898 Hz -53.11 -52.97 -52.93 -53.10
- 05513 Hz -58.21 -58.00 -57.97 -58.20
- 07796 Hz -59.70 -59.48 -59.44 -59.71
- 11025 Hz -61.85 -61.60 -61.58 -61.83
- 15591 Hz -66.90 -66.57 -66.54 -66.87
- 22050 Hz -76.98 -102.45 -92.81 -95.46
[/font]
Results/Conclusions:
Between 50 and 1000 Hz the difference is >= 0.01dB for both WMA samples and MPC.
Below 50 Hz WMA starts to become louder with arround 3dB difference for frequencies > 22Hz. MPC is only 0.53dB louder < 22Hz.
Between 1378Hz and 15591 Hz WMA starts to become louder again with a maximum at the 11025-15591Hz band: CBR + 0.33dB, VBR + 0.36dB (MPC +0.03dB). The average for 1949-15591Hz band is: CBR + 0.21dB; VBR + 0.24dB; MPC + 0.01dB.
Having a look at the highest band (15591-22050Hz) doesn't make much sense here because the codecs' lowpasses kick in here.
From these measurements it seems like WMA amplifies very low and high frequencies.
Frequencies < 22Hz can't be played back properly with most equipment and don't occur that frequently in music that they could lead to a general change in sound.
About 2000Hz+ frequencies: AFAIK an overall volume change of 0.2-0.4dB can't be noticed/ABXed, but I have no idea about changing only a part of the spectrum. Anyone?
I think most likely the richer/warmer sound is immagination, but as we know how WMA equalizes the music encoded, this could be simulated using FFT filters/equalizers and ABXed. Volunteers?
edited: clarification, typos
Just a 2c opinion -- often, "richer, warmer" is descriptive of distortion rather than frequency balance. I have a theory that once lossy artifacts become unnoticeable/unABXable (at least with short time intervals), it's possible they can still be perceived "subconsciously" as an overall effect on SQ.
So anyway... perhaps there is something to claims like "this codec sounds warmer" or "that codec sounds thinner" (outside of volume and frequency issues) but it would be difficult to prove claims like that.
I have a theory that once lossy artifacts become unnoticeable/unABXable (at least with short time intervals), it's possible they can still be perceived "subconsciously" as an overall effect on SQ.
So anyway... perhaps there is something to claims like "this codec sounds warmer" or "that codec sounds thinner" (outside of volume and frequency issues) but it would be difficult to prove claims like that.
Why would it be difficult ? If a codec sounds "warmer" or "thinner", why would WinABX remove these feelings ? Do you mean that you'd have to listen to the whole CD for the feeling to appear ?
I have a theory that once lossy artifacts become unnoticeable/unABXable (at least with short time intervals), it's possible they can still be perceived "subconsciously" as an overall effect on SQ.
So anyway... perhaps there is something to claims like "this codec sounds warmer" or "that codec sounds thinner" (outside of volume and frequency issues) but it would be difficult to prove claims like that.
Why would it be difficult ? If a codec sounds "warmer" or "thinner", why would WinABX remove these feelings ? Do you mean that you'd have to listen to the whole CD for the feeling to appear ?
No... but who knows how long it would take for that feeling to come up, and whether it's possible that concentrating (as is often done with WinABX) would again erase those impressions.
I know that concentrating on the sound of something ('trying to hear a difference') often changes the way it sounds... it happens to me all the time. Listening casually without pressure to choose something over another, an impression or emotion arises by itself. I'm sure it happens to you, too.
Anyway... I just said "perhaps this is true" (postulating, not claiming). I hope this doesn't turn into a stupid objectivist/subjectivist thing many here seem to love to argue about. Admittedly, I'm now standing in the middle of the road rather than holding strictly objectivist beliefs... if it offends anyone, I'll refrain from suggesting things like this. If the forum is completely intolerant of "middle path" audio views, I'll leave.
Apologies to rjamorim for the recent edit.
Maybe it's a peculiarity of many HA members that they have no right brain at all... only the left (logical/rational) side, and cerebrospinal fluids filling the area where the right (intuitive/emotional) should be.
Makes sense.
Lots of sense, indeed :B
I think the contradiction here is that if there is enough of a "difference" for someone to clearly label something as "richer" or "warmer", or that it has "firmer bass", or whatever else, that this should definitely show up in an ABX test. If the difference is so subtle that they cannot ABX it at all, then I'd find it very hard to believe that they could even describe whatever it is that they are supposed to be hearing. The fact that they could clearly describe the difference would seem to imply that it is a consciously observable phenomena since it would take a degree of comparison or measurement to come up with an adequate labeling.
I know that concentrating on the sound of something ('trying to hear a difference') often changes the way it sounds... it happens to me all the time. Listening casually without pressure to choose something over another, an impression or emotion arises by itself. I'm sure it happens to you, too.
It's not the sound that actually changes, it's only your perception of the sound. Usually I think that when performing these kinds of tests, the reason it may sound different is because you're listening for more detail in the sound than you might otherwise have done. Think of this as a similar effect that you can observe when focusing your vision on a distant object to attempt to view it in more detail -- it changes the way things appear, but it does not change the way they actually are (this can be supported through repeatable tests and measurements of an empirical nature).
I kind of get the impression that you think of this as a degrading process of some sort, which actually decreases sensitivity. I believe this is wrong, and that it is actually a refining process, increasing sensitivity, just without the excess irrelevant information. The point is to eliminate these "feelings" which are not correlated with sound quality itself, but rather with a state of mind or other non-fidelity related matters. These are totally irrelevant to a third party who is interested in determining your level of fidelity perception, or your ability to discern small differences in the absolute makeup of the sound.
So when someone says:
"WMA sounds richer"
We are only interested in the following:
1. What does "richer" mean? How can it be defined in a universally applicable and measureable sense?
2. Can it be shown that you can actually perceive this with any sort of reliability outside of most external influences, or is this simply an artifact of other psychological processes highly influenced by your immediate environment? (For example, you have the "warm fuzzies" because of your new quantum clip and $5000 power cord). Basically, is this even related to the actual sound?
3. Assuming that this is related to the actual sound, is this an attribute common to other peoples systems of perception, or is this unique to your own hearing? If so, why?
At really no point are we interested in the "feeling" itself, or what this means to you on some sort of mystical, spiritual, or other non-scientific level.
Anyway... I just said "perhaps this is true" (postulating, not claiming). I hope this doesn't turn into a stupid objectivist/subjectivist thing many here seem to love to argue about.
Well, you made the statements.... you also made a little joke/insult about the whole right brain thing. Why would want to start a discussion in this vein and then simply end it?
Given the recent onslaught of subjectivist views attempting to infiltrate Hydrogenaudio (people arguing over rule #8 and other established conventions here), I'm put in a situation where I really can't simply ignore these kinds of things.
Admittedly, I'm now standing in the middle of the road rather than holding strictly objectivist beliefs... if it offends anyone, I'll refrain from suggesting things like this. If the forum is completely intolerant of "middle path" audio views, I'll leave.
Maybe you've been spending too much time at head-fi...
I don't think anyone is offended, the fact of the matter is simply that the subjectivist views have no place here. These views have done more to harm and hamper progress, and to foster charlatans in the field of audio than anything else. I don't see how they serve any useful purpose in a debate style mode of communication either, since subjectivist claims are only relevant to the person making them and not to any other single person, unless the point itself is to promote unique, creative, and arbitrary forms of interpretation, which is useful for artistic persuits, but not for scientific ones.
You certainly don't have to leave, but you'd need to remember to be respectful to the doctrine of this forum.
Maybe you've been spending too much time at head-fi...
Maybe. It's been fun, anyway. To me, audio is either a matter of enjoyment, or there are better/more productive things to do. Life is short, and time is limited.
hi everyone. this post made me think.
is there "one" psychoacoustic?
because, when you´re actually abx´ing samples or songs you are very concentrated in listening, hence your psychoacoustic is changed.
When you just do casual listening you´re in another psychoacoustical "state".
Could it be useful for the psychoacoustic models to have that in mind. And does anybody know if our perception is changed (for example different attention to different frequency ranges when not concentrating) by concentrating or "just" listening
my 2cents. dunno if helpful
hi everyone. this post made me think.
is there "one" psychoacoustic?
because, when you´re actually abx´ing samples or songs you are very concentrated in listening, hence your psychoacoustic is changed.
When you just do casual listening you´re in another psychoacoustical "state".
I would agree that such a thing is possible, even likely -- others would feel differently. The conclusion (if any) is really up to you.
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
as i am working as social worker, we know that just by observating a system (familiy, group, ethnic group and so on) we are changing, because as there is the paradigm of communication: you cannot "not communicate".
that was my food for thought. if maybe our hearing changes significantly and measureable so its of any use for developing and tuning psychoacoustic models.
but as said, i´m a social worker, no acoustic einstein.
tanx for the work of all the acoustic einsteins by the way...
Dibrom, I find the word 'doctrine' implies a negative qualification of things. It sounds like there's no freedom of thought allowed or something. Rather, I'd see them as principles. But then, english isn't my native tongue...
Now, on topic (has anyone noticed that things went off-topic almost from the second post on?!):
From these measurements it seems like WMA amplifies very low and high frequencies.
Frequencies < 22Hz can't be played back properly with most equipment and don't occur that frequently in music that they could lead to a general change in sound.
About 2000Hz+ frequencies: AFAIK an overall volume change of 0.2-0.4dB can't be noticed/ABXed, but I have no idea about changing only a part of the spectrum. Anyone?
I think most likely the richer/warmer sound is immagination, but as we know how WMA equalizes the music encoded, this could be simulated using FFT filters/equalizers and ABXed. Volunteers?
Couldn't it be that the 3dB boost in the deepest lows account for the 'warm' sound? That would explain why we don't all hear it - my speakers don't do 30Hz...
It is very cool to see how mpc is so accurate regarding the levels. I'm only just out of my bed (yes, HA addict), so I didn't think yet about how legitimate this is as a measure of quality. But it's nice Thanks for the extensive work Tigre!
Off-topic again: if you say WMP9 Pro, does that mean the 24bit encoder setting? I've been trying to figure out the WM Encoder the other day. As soon as I select 'Professional', all 16bit options are gone.
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
This has nothing to do with psychoacoustics!!!!! On a macro scale, no observer is going to change the waveforms in the room when he is starting to pay attention to what he hears!
You're being like the philosophers that Feynman rants about in his Lectures on Physics, who use relativity theory in a general non-physical meaning. Edit: http://www.lorentz.leidenuniv.nl/vanbaal/r...ative.html#phil (http://www.lorentz.leidenuniv.nl/vanbaal/relative.html#phil) - this is it.
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
This has nothing to do with psychoaccoustics!!!!!
Sorry to disagree with you, but the observer is not fundamentally separate from what he/she observes. Look at something, and you are looking at your own perceptions (same with listening).
Aside from matters of social agreement (which objectivists are always accusing subjectivists of!) there's no fundamental way of determining whether a perception is "subjective" or "objective." Really, it is neither.
@patsoe
not the waveform changes, but how it is perceived. and thats what its all about here, or?
bueno, i´m not a acoustic physician
so, what thoughts on that exist in this community. there is alot of thought towards transparency. but what happens (and what use could that be of) if the way the receiver/perceiver woks (hence different psychomodel) changes.
Sorry to disagree with you, but the observer is not fundamentally separate from what he/she observes.
That isn't what I said (and sorry if I was being rude - I felt you were insulting quantum theory ). I meant that the coupling of observation and the observed is totally different in quantum physics.
In quantum physics, observation changes the state of the observed.
In psychoacoustic effects, observation changes perception of the observed.
Or something like that
Now, let's get back to Tigre's measurements!
Edit:
@caiman: yes, that's all it's about here. So, it has nothing to do with quantum physiscs... and I felt that the whole mentioning of it was trivialising quantum theory.
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
This has nothing to do with psychoaccoustics!!!!!
Sorry to disagree with you, but the observer is not fundamentally separate from what he/she observes.
Aside from matters of social agreement (which objectivists are always accusing subjectivists of!) there's no fundamental way of determining whether a perception is "subjective" or "objective." Really, it is neither.
I'm glad to see you providing some sort of backing to your claims here.....
Seriously, these are the kinds of things that volumes upon volumes of text have been written about by philosophers and physicists alike for literally thousands of years. I fail to see how you can make such sweeping claims as these without providing
any support for them (you didn't even provide examples or expand upon the ideas to any sigificant degree) and just expect people to take your word for it...
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
This has nothing to do with psychoaccoustics!!!!!
Sorry to disagree with you, but the observer is not fundamentally separate from what he/she observes.
Aside from matters of social agreement (which objectivists are always accusing subjectivists of!) there's no fundamental way of determining whether a perception is "subjective" or "objective." Really, it is neither.
I'm glad to see you providing some sort of backing to your claims here.....
Seriously, these are the kinds of things that volumes upon volumes of text have been written about by philosophers and physicists alike for literally thousands of years. I fail to see how you can make such sweeping claims as these without providing any support for them (you didn't even provide examples or expand upon the ideas to any sigificant degree) and just expect people to take your word for it...
What could I possibly provide -- an ABX test?
All I can do is invite someone to look into the nature of perception, and of objectivity/subjectivity... same as those philosophers have been doing for thousands of years. What else could anyone do? If one is so inclined, look into it rather than taking things for granted. If you aren't inclined, don't do so. What more is there to say?
P.S. I'm not some kind of radical, for anyone who may be thinking so. For all practical purposes, objective is objective and subjective is subjective. What I'm questioning is the assumed conflict between the two, as if reality were split in half.
@patsoe
"In psychoacoustic effects, observation changes perception of the observed."
that is my point. but you pinpoint it better. so what happens if the perception of the observed is changing not only for a few people but if it´s kind of a psychoacoustic rule, which can be useful to save bits.
edit: typo
This is another topic running out of hand, and in great part it's my fault. I apologize.
Can we all drop the issue and get back to WMA? I promise not to talk about quantum comparisons here any longer...
Restart:
Couldn't it be that the 3dB boost in the deepest lows account for the 'warm' sound? That would explain why we don't all hear it - my speakers don't do 30Hz...
It is very cool to see how mpc is so accurate regarding the levels. I'm only just out of my bed (yes, HA addict), so I didn't think yet about how legitimate this is as a measure of quality. But it's nice Thanks for the extensive work Tigre!
Could it be useful for the psychoacoustic models to have that in mind. And does anybody know if our perception is changed (for example different attention to different frequency ranges when not concentrating) by concentrating or "just" listening.
I believe (and if I'm not wrong scientific literature agrees) that people used to perfom quick switching, time and level aligned, blind tests with reference (ABX and ABC/HR tests for example), are much more sensitive to detecting subtle differences in those kind of tests, than just when doing casual listening. But it is also true that stress and tireness when performing such a test can lead to poor sensitivity.
On the other side, casual listening is usually not blind, so its results are not reliable at all. When people make sighted A/B tests, they don't complain about having to concentrate, I guess because they feel they are not really at test, they feel "safe" about their perceptions. The problem is that those perceptions can be seriously influenced from other things but the actual sound reaching their ears. For example, some people claim they can hear "glaringly obvious" differences that, when repeated blind, magically dissapear.
For people that have such doubts about ABX tests, casual listening and concentration, maybe you could try a long-term ABX test performed in a more "casual" way of listening, over several days or even weeks. For that, you can use my fileABX utility, that will generate several wav files for performing such a test. If some of you do, please report results here, they would be very interesting.
Edit:
but, in my opinion, those problems have not much to do with different psychoacoustic models depending on the observer "way of listening". I think all you can talk about is a greater or lower sensitiveness, or sighted tests being responsable for unexistant differences. According to James Johnston (JJ), one of the greatest experts in this field, the thresholds of human hearing that have been found according to listening tests, always using DBTs of course, are very similar to the thresholds that could be expected analyzing the physiological working of the ear.
And yes, maybe it's time to move this discussion to a new thread.
Edit: some few additions.
What could I possibly provide -- an ABX test?
All I can do is invite someone to look into the nature of perception, and of objectivity/subjectivity... same as those philosophers have been doing for thousands of years. What else could anyone do? If one is so inclined, look into it rather than taking things for granted. If you aren't inclined, don't do so. What more is there to say?
Since I seriously doubt that you are any sort of authority on psychoacoustics, quantum theory, philosophy of science or mind, metaphysics, or epistemology, you could start by providing at least some sort of reference for the basis of your rather "matter of fact" claims, if you expect anyone to consider what you're saying here as meaningful. If you don't care about this though, then I have to question why you're posting in the first place.
If you're just posting here to "bullshit", so to speak, then it is not only out of place, but it is also unwelcome. Myself and the others around here would like to see the discussion on HA stay useful and relevant, and outlandish and unsubstantiated claims made by those who have no interest (on any level) in providing something which can be used by others, do not help this cause. Instead all they end up causing are more flamewars and bitching.
Edit: I find it seriously disturbing that I should be having to explain this sort of basic matter to someone who has over 1000 posts on this board...
Dibrom, I understand your anger, but please drop it.
Get some sleep, all you Americans! I went to sleep before 1AM last night. I can't remember when I last did that. It should be a HA recommended...
What could I possibly provide -- an ABX test?
All I can do is invite someone to look into the nature of perception, and of objectivity/subjectivity... same as those philosophers have been doing for thousands of years. What else could anyone do? If one is so inclined, look into it rather than taking things for granted. If you aren't inclined, don't do so. What more is there to say?
Since I seriously doubt that you are any sort of authority on psychoacoustics, quantum theory, philosophy of science or mind, metaphysics, or epistemology, you could start by providing at least some sort of reference for the basis of your rather "matter of fact" claims, if you expect anyone to consider what you're saying here as meaningful. If you don't care about this though, then I have to question why you're posting in the first place.
No need to doubt, I'll state outright that I'm not an authority on anything.
Here are a few references for anyone who may be interested in investigating this topic further (check out the first in particular):
- http://faculty.virginia.edu/consciousness/ (http://faculty.virginia.edu/consciousness/)
- http://www.nonduality.com/faq.htm#q1 (http://www.nonduality.com/faq.htm#q1)
- http://www.advaita.org/ (http://www.advaita.org/)
- http://www.worldofkenwilber.com/ (http://www.worldofkenwilber.com/)
- http://www.nisargadatta.net/ (http://www.nisargadatta.net/)
- http://groups.yahoo.com/group/consciousnessisall/ (http://groups.yahoo.com/group/consciousnessisall/)
- http://nothingistic.org/library/laotzu/ (http://nothingistic.org/library/laotzu/)
Some of the above has connections with traditional philosophies and religions (more Eastern than Western) and some doesn't -- I trust in the ability of the reader to take what makes sense to you and leave the rest. Personally I'm no believer in religion, but (fortunately or unfortunately) a lot of mystical/new-agey bullshit has been historically wrapped around this kind of philosophy. If you're interested and open minded enough, you'll be able to sort out the philosophy from the idiocy.
WMA has rich, full bass.
Looking at the findings, this statement seems to hold true "Below 50 Hz WMA starts to become louder with arround 3dB", it is ok to now say this on HA?
IMHO, science has been too neglectful about leaving the observer out of the equation.
Not real science.
On the other side, many of the times philosophy has nothing to do with real science. Poor sensitivity of listening tests or unreliable sighted tests are just problems that must be addressed, and philosophy can't help here.
IMHO, science has been too neglectful about leaving the observer out of the equation.
Not real science.
Real science is open to
any revisions based on new evidence. Scientism is the blind belief in "already proven" theorems, and the rejection of the unknown merely on the basis that it is unknown. Afaic, a "real scientist" will approach an issue with an open mind, and without preconceptions. The purpose of science is discovery, not validating what one already knows in order to feel more secure in one's beliefs.
IMHO, your claim that inclusion of the observer isn't "real science" is unscientific, unless you can prove it and no doubt remains (1+1=2).
P.S... for the record, Dibrom wanted me to "back up" what I was saying... otherwise I wouldn't have carried this off-topic conversation so far. Probably this is my last post on the thread, sorry to distract anyone or use up bandwidth. This stuff is available for anyone to investigate thru Google, and has nothing to do with either the sound of .wma files or the general subject matter of the board (i.e. this should be in "off-topic"). Perhaps a topic split?
WMA has rich, full bass.
Looking at the findings, this statement seems to hold true "Below 50 Hz WMA starts to become louder with arround 3dB", it is ok to now say this on HA?
Looking at the data again, we can't say this. There's a much higher level at 30Hz in the original than there is at 22Hz. So, perhaps the wmp psy-model says that this will mask a 3dB increase at 22Hz, and perhaps that's even correct...
We should be trying more samples/music styles before we conclude.
@Tigre: how did you get the wmp output (wav) written to disk? Edit: never mind, I just found out CEP does that. Question two then: is CEP as good as the reference decoder, using 24bit resolution?
WMA has rich, full bass.
Looking at the findings, this statement seems to hold true "Below 50 Hz WMA starts to become louder with arround 3dB", it is ok to now say this on HA?
Are you asking if it's ok to say "WMA has rich,full bass", or "Below 50 Hz WMA starts to become louder with arround 3dB"?
Which do you think is more specific and informative?
"WMA has rich,full bass", or "Below 50 Hz WMA starts to become louder with arround 3dB"? Which do you think is more specific and informative?
The average Joe on the street wouldn't have a clue what 50Hz 3dB means, he would know what the former means.
Real science is open to any revisions based on new evidence.
Did I say otherwise?
Scientism is the blind belief in "already proven" theorems, and the rejection of the unknown merely on the basis that it is unknown.
I don't know much about that term, but that's not science. Real theorems can't be disproved, are mathematically true (such as 2+2=4 or Pithagoras theorem). Theories are just that, theories, and are subject to revision if new contradicting evidence appears, as you say.
Afaic, a "real scientist" will approach an issue with an open mind, and without preconceptions. The purpose of science is discovery, not validating what one already knows in order to feel more secure in one's beliefs.
Does anyone say otherwise? On the other side, if there is plenty of evidence that supports your beliefs or a theory, it must be considered as the most likely, until , again, new evidence can disprove that.
IMHO, on that basis your claim that inclusion of the observer isn't "real science" is unscientific, unless you can prove it and no doubt remains (1+1=2).
I didn't say that, but the opposite. I said that in real science, inclusion of the observer must be taken into account if needed.
This is an area of great interest in quantum physics, in particular -- how the act of observation may change the observed. IMHO, science has been too neglectful about leaving the observer out of the equation.
This has nothing to do with psychoaccoustics!!!!!
Sorry to disagree with you, but the observer is not fundamentally separate from what he/she observes.
Aside from matters of social agreement (which objectivists are always accusing subjectivists of!) there's no fundamental way of determining whether a perception is "subjective" or "objective." Really, it is neither.
I'm glad to see you providing some sort of backing to your claims here.....
Seriously, these are the kinds of things that volumes upon volumes of text have been written about by philosophers and physicists alike for literally thousands of years. I fail to see how you can make such sweeping claims as these without providing any support for them (you didn't even provide examples or expand upon the ideas to any sigificant degree) and just expect people to take your word for it...
What could I possibly provide -- an ABX test?
Yes, please do. There's a principle in law: not guilty unless proven otherwise. So is with your theory. Prove it to be true. Until then it's not, so don't percist on it. You yourself don't even know if it's true or not.
"WMA has rich,full bass", or "Below 50 Hz WMA starts to become louder with arround 3dB"? Which do you think is more specific and informative?
The average Joe on the street wouldn't have a clue what 50Hz 3dB means, he would know what the former means.
So, should the average Joe try to learn it if he wants to read HA, or should HA regulars start to talk like average Joe, very unspecificly?
So, should the average Joe try to learn it if he wants to read HA, or should HA regulars start to talk like average Joe, very unspecificly?
The latter would imo be more polite (by the HA regulars). Or was this a rhetorical question?!?
So, should the average Joe try to learn it if he wants to read HA, or should HA regulars start to talk like average Joe, very unspecificly?
The latter would imo be more polite (by the HA regulars). Or was this a rhetorical question?!?
Scientism is the blind belief in "already proven" theorems, and the rejection of the unknown merely on the basis that it is unknown.
I don't know much about that term, but that's not science. Real theorems can't be disproved, are mathematically true (such as 2+2=4 or Pithagoras theorem). Theories are just that, theories, and are subject to revision if new contradicting evidence appears, as you say.
Try to prove that 2+2=4 ! I never had any success... But supposedly you can get no longer than for example Peano's Axioms or such, and those are really only assumptions. There is no way to prove them correct. Maybe some day someone will come up with a better set of axioms and we will have to revise parts of the mathematics??
You can draw a triangle with three straight corners. How? Does Pithagora's theorem still hold?
The philosopher asks: "If a tree falls in a forest, but no one is there to hear it, does it make a sound?"
David replies: "As long as the tree falls on a philosopher, does it matter?"
Subjective vs Objective
Is it possible to hear a difference that cannot be measured? No.
(Think: ears respond to air pressure variations; we can detect, measure, and record those with more than enough accuracy. The analysis of the results is a different matter, but what you heard is somewhere in what was measured)
Is it possible to perceive a difference that cannot be measured? Maybe.
Perception is more than "hearing". Our mood affects our perception. But that's an internal thing.
In my experience, consistent subjective impressions can be traced to objective measureable and ABXable differences. (Unless those consistent subjective impressions are due to sighted comparisons!)
Sighted comparisons are stupid, if all you're interested in is the sound quality. It's as simple as that. Get rid of that placebo!
As for stress, and different types of hearing/listening. Well, it's obvious that our moods effect our perception. Anyone who claimed otherwise would be foolish.
But we're not talking about moods. We're talking about a simple fact: do you know what you're listening to, or not. That doesn't set your mood to cold hard and analytical. Sitting in front of your PC with PCABX on the screen and 13ms of harpsichord music looped may well do, but it doesn't have to be like that... "Hey dude - is that a new stereo you've got in the next room - sounds really cool!" That's blind listening. Not statistically valid yet - it depends how many friends say that before they see it!
But I'll say it again: every reliable subjective comment that I've ever heard and investigated (or seen investigated) has been traced to a hard objective fact. People who say otherwise are trying to sell you something, or justify buying something, or don't understand science enough to speak the right language.
Cheers,
David.
OT: IMO is the discussion about science/scientific methods/etc. very interesting and deserves to take place. OTH for me one of the things i appreciate most about HA is the scientific way auido-related (especially quality-related) topics are (supposed to be) discussed here. A more subjective/"open-minded" point of view *in discussions* is of no use IMO. Example:
1. "sample A and sample B sound identical, noone was able to ABX successfully."
2. "sample A and sample B sound identical, noone was able to ABX successfully, but once when I had a massage while listening to sample A and was totally relaxed I noticed that it moved somtheing in me. I never had this feeling with sample B".
I prefer 1 and I am thankful that the people who run this site spend their precious time on making people keep the rules. IMO it'd be a sign of thankfulness and appreciation for this place and their work just to accept the way things work here. We are all guests, noone is forced to participate.
__________________________
Now back on topic (if this thread is not already hopefully offtopic )
Looking at the data again, we can't say this. There's a much higher level at 30Hz in the original than there is at 22Hz. So, perhaps the wmp psy-model says that this will mask a 3dB increase at 22Hz, and perhaps that's even correct...
We should be trying more samples/music styles before we conclude.
Definitely. Now that I've set up Excel it's not much work to repeat with other samples. I guess next I'll try with some sine sweep or similar, afterwards with the test samples from rjamorim's 64kbps listening test. One more reason why I think it's important to do more tests is, that I don't know how exact CEP's frequeny analysis is (Blackman-Harris window, 8192 Samples used), especially for low frequencies.
CEP as good as the reference decoder, using 24bit resolution?
I don't know. WMA9 "Pro" encoded files are opened by CEP as 16bit here. Is the reference decoder you're talking about publicly available?
CEP FFT is accurate enough for those tasks.
A better way to compare visually frequency response: create a new 16-bit stereo file. Copy left track of one file into left track of new file, and left track of the other file into right track of new file. Both files should be ideally time aligned before doing this. You will have a stereo file where the both channels are the same, but one has been encoded/decoded and the other not.
Then, use frequency analysis and zoom at the Y (amplitude) axis so that 0.2 dB or so differences are clearly detectable. Use a not very long FFT size, 1024 pt can be ok, but you can try other sizes. Then, look at various parts of the song, or use realtime playback, or average parts of the file (scan button at the FFT window), and see if there are parts of spectrum that are always higher amplitude in one of the channels.
A better method could be to plot a FFT of left/righ channels transfer funcion of this file, Spectralab can do this. Also, using pink or white noise can be more revealing than using music.
I can't resist putting in one flippant suggestion. What we need is an ABX test where you don't know for each trial whether you are paying attention or not.
On WMA boosing low bass below your speakers capability: even with speakers that will reproduce 20-30 Hz, if they distort even a few percent then the harmonics can sound louder than the fundamental frequency. This is because your hearing is less sensitive down there.
I think that speakers that won't reproduce very low bass can also put out audible
harmonics as you are still getting cone motion which will have distortion products within the passband of the speaker.
Another effect with wide range drivers is that high excursions at low frequencies will cause
doppler distortion of higher frequencies sent to the same driver. I don't know if this would be considered "warm" by some.
Of course, this low bass boost will have no effect whatsoever if the song you are playing has no content in that range. As a reference, I think the lowest note on a standard tuned bass guitar is 42 Hz. For 20 Hz, aside from electronic enhancement and explosions, you are looking at pipe organs, ambient venue noise like HVAC rumble, and maybe the largest tycho drums.
The philosopher asks: "If a tree falls in a forest, but no one is there to hear it, does it make a sound?"
David replies: "As long as the tree falls on a philosopher, does it matter?"
I agree with you -- nonpractical philosophy is nothing but a useless mental game. That's why I just said that the whole subjective/objective stance is worth each person looking into for themselves (and expressed a few of my views on it)... that's all. I hope people aren't so paranoid that they automatically assume a subjective/superstitious stance is being promoted when someone challenges conventional thinking a little.
Anyway -- I think it's time that I cut my participation here to a bare minimum -- the subject matter no longer interests me much anymore, anyway. Cheers all...
Now back on topic (if this thread is not already hopefully offtopic )
CEP as good as the reference decoder, using 24bit resolution?
I don't know. WMA9 "Pro" encoded files are opened by CEP as 16bit here. Is the reference decoder you're talking about publicly available?
Well, I guess 'reference decoder' is indeed a too elevated name for WMP9
I've been trying to confirm it but can't get WMP9 to write to disk here though (I tried a virtual soundcard but it crashes my system...).
And who knows how WMP works internally. Perhaps it too decodes to 16bit and then ups the output for 24bit cards...
A better way to compare visually frequency response: ...
There could be a missunderstanding: I didn't compare anything visually. Using frequency analysis for the whole file, pressing "Copy to Clipboard" + Ctrl-V in Excel delivers a table with 1/2*(window size) frequenciy bands, for each band one left and one right channel dB value. The small bands belonging to 1/2 octave bands were summarized (= average dB values were calculated) using Excel.
If you think visual comparison can give better results, I'd be glad to know why...
Thanks for the white/pink noise idea. I'll try this next.
Now back on topic (if this thread is not already hopefully offtopic )
CEP as good as the reference decoder, using 24bit resolution?
I don't know. WMA9 "Pro" encoded files are opened by CEP as 16bit here. Is the reference decoder you're talking about publicly available?
Well, I guess 'reference decoder' is indeed a too elevated name for WMP9
I've been trying to confirm it but can't get WMP9 to write to disk here though (I tried a virtual soundcard but it crashes my system...).
And who knows how WMP works internally. Perhaps it too decodes to 16bit and then ups the output for 24bit cards...
1. I don't want to install WMP9; I don't even have a 24bit soundcard, so AFAIK it won't play back at 24bit anyway (Win2k here).
2. At -40dB a difference of 0.2dB is much bigger than what could be caused by truncation from 24bit to 16bit resolution.
Unless CEP's WMA decoder is somehow buggy I don't think here's a reason to worry. With the next sample I try I'll decode MPC without dither to keep things "fair".
According to Fletcher-Mundson curves, at 80 dB, a 22 Hz frequency is perceived 40 dB quieter than a 345 Hz one. Thus, in this analysis, the 22 Hz band being already 33 dB below the 345 Hz one, it must be perceived 40+33=73 dB quieter.
Therefore we're talking about a 3 dB error on -73 dB noise that is below a 0 dB signal ! This might well fall into the "Graphs don't tell anything" category, because the Psychoacoustic model might well have decided that changing this level would allow to save bits without any audible effect.
Is it actually the case ? This is impossible to tell, because the global analysis don't tell anything about the duration and intensity of the phenomenon in the time domain. This +3 dB may actually be an inaudible boost affecting a continuous -40 dB rumble, but it may as well be a full scale artifact occuring during several milliseconds, that would be accounted for -64 dB after being averaged with hundreds of seconds of silence in the same frequency band for the rest of the track.
Thus we can't tell that WMA boosts 22 Hz frequencies by +3 dB. It could as well highpass the file at 30 Hz, while causing 22 Hz artifacts on problem samples. The effect on the numbers would be the same.
_______________________
The observers'problem in quantum mechanics is not relevant here.
It has a special status in quantum mechanics, because the system measured abides laws (quantum laws) that the observer is violating.
The most obvious is the superposition principle : "if a system can be in the state A or in the state B, then it can also be in any linear combination of A and B" (for example A+B).
For example if Schödinger's cat can be alive, or dead, then it can also be alive+dead at the same time. This is true for quantum systems. Interferences building up in the Young Two-slit experiment (http://zebu.uoregon.edu/~js/21st_century_science/lectures/lec13.html) shows that photons that can pass through silt A or slit B can also pass through slit A+B. But this is wrong for the observer. With photon detectors, the observer can see the photon in slit A or in slit B, but not in slit A+B. The observer violates the superposition's principle.
Since the observer sees what happens, he forces the quantum system to violate the superposition principle when it is observed.
In audio, the sound abides the same mechanical laws in the room when there is nobody and when there is someone. Thermodynamics and Acoustic laws are not changed when someone comes in.
Thus, unlike in quantum mechanics, it is perfectly possible to apply these laws to the observer himself. For example, artifical heads are used to measure the frequency response of headphones. Head and body related transfer functions are used to simulate speaker sound in headphones, etc.
_________________________________
I think that the feeling that a given sound can give doesn't depend on the program that plays it, be it Winamp, Foobar2000, or WinABX.
If someone can feel that a sound is warm to his ears, and can tell it, what would prevent him to do so in a blind test ? Stress ? Well, a blind test, in order to be valid must be conducted in the same conditions that the ones under which the phenomenon tested was observed first, that is under non-stressed conditions. It's up to the subject to relax.
I found the Astral sample, that I ABXed with MPC standard, in a casual listening. I was not looking for artifacts at all. I just started the track, and poof ! HEY ! What was that ? An MPC artifact or what ? And yes it was. It sounded obvious to my ears in casual listening, while during ABX, I sometimes completely lost track of it. But I was still capable or ABXing it with pauses between the sessions.
My question is:
Has anyone measured the frequency distribution of a tadpole fart? I need some verifiable objective data on this. When I encode my tadpole fart recordings, I don't want to use any codec that may cause a boost in frequecies that may cause a perception of "warm". I want the compressed tadpole fart recordings to be a true to the original as possible, without resorting to lossless.
According to Fletcher-Mundson curves, at 80 dB, a 22 Hz frequency is perceived 40 dB quieter than a 345 Hz one. Thus, in this analysis, the 22 Hz band being already 33 dB below the 345 Hz one, it must be perceived 40+33=73 dB quieter.
Therefore we're talking about a 3 dB error on -73 dB noise that is below a 0 dB signal ! This might well fall into the "Graphs don't tell anything" category, because the Psychoacoustic model might well have decided that changing this level would allow to save bits without any audible effect.
Is it actually the case ? This is impossible to tell, because the global analysis don't tell anything about the duration and intensity of the phenomenon in the time domain. This +3 dB may actually be an inaudible boost affecting a continuous -40 dB rumble, but it may as well be a full scale artifact occuring during several milliseconds, that would be accounted for -64 dB after being averaged with hundreds of seconds of silence in the same frequency band for the rest of the track.
Thus we can't tell that WMA boosts 22 Hz frequencies by +3 dB. It could as well highpass the file at 30 Hz, while causing 22 Hz artifacts on problem samples. The effect on the numbers would be the same.
Thanks for the valuable info, Pio2001! Good point. This means that doning the same with other samples is useless. Analyzing short periods of time won't help either because you can't know from looking at the numbers if an increased band is masked. At best it could help to find places where ABXing could work.
So we are at the starting point again: Who thinks that WMA sounds richer/warmer should ABX it before writing.
________________
Now science vs. open-minded-ness discussion can go on!
Now science vs. open-minded-ness discussion can go on!
Science vs. open-mindedness? Oh lord... I'm truly sorry I ever got onto the conversational track I did last night. Talk about big mistakes...
WMA has rich, full bass.
Looking at the findings, this statement seems to hold true "Below 50 Hz WMA starts to become louder with arround 3dB", it is ok to now say this on HA?
I completely agree that "Below 50 Hz WMA starts to become louder with arround 3dB" is a valid statement, but that "WMA has rich, full bass" is not. The former seems to have evidence backing it up. The latter is a subjective opinion.
I tend to be a control freak myself, and I don't want my encoder messing with equalization any more than it has to. I prefer my encoding step to be as audibly transparent as possible. If I can hear more bass after encoding (and without using any sort of playback DSP/equalization), then the word "muddy" would enter my mind before "rich" or "full". After all, any difference in sound quality, whether perceived as good or bad by the listener, would directly work against the basic tenet of psychoacoustic audio compression, would it not? I want "accurate", not "warm".
If I want more bass, I'll turn it up myself with an EQ during playback. I'll bet I know more about exactly how to "shape" the sound to make
my ears happy than all of Microsoft does with their WMA encoder trying accomplish the same goal.
So when someone says:
"WMA sounds richer"
We are only interested in the following:
1. What does "richer" mean? How can it be defined in a universally applicable and measureable sense?
Geez. I'm sorry I started this mess. I certainly did not mean to. I was just making an innocent comment on why I actually like WMA. I like MPC, AAC, and MP3 as well. I'm not a WMA promoter!
I'm going to try and answer Dibrom's question, since he gave me a warning for making this statement in the 1st place (I still think you were in error to issue me this warning, Dibrom. Out of the respect I have for the great work you did for LAME and creating APS - I will argue this no more), but it is VERY difficult. How do you describe hearing?
Have you ever heard anyone make the statement that vinyl sounds richer than a CD? That is the same perception I have listening to WMA vs. MP3. Is the sound better? I will NOT claim so. Audiophiles in this forum have proven it is not as accurate, that MPC and AAC are much more accurate than WMA at higher bit rates.
When I say richer and warmer, I mean the sound has more bass (I think that has been proven in this thread), with more distinction between the highs and lows. The highs especially are 'fuzzier' sounding than the crip highs produced by LAME APS.
I don't know what else to say. Again, I stress I do NOT make any claim that WMA is better. I hope the vinyl analogy helps. I first heard it from Neil Young who claimed CDs have not the warmth of vinyl, even though CDs are proven to more accurately capture the original recording. He did not care and claimed he preferred the sound of vinyl over digital. Pearl Jam holds similar claims and they still release all recordings, except live hshows, on vinyl. Who knows, maybe Neil, Pearl Jam, and I just have distorted hearing from too many live shows.
ScorLibran, Ezra: three posts above yours it is concluded that there is probably no audible bass boost. The 3dB boost found is allowed by perceptual models. There is nothing more to it.
Now science vs. open-minded-ness discussion can go on!
Science vs. open-mindedness? Oh lord... I'm truly sorry I ever got onto the conversational track I did last night. Talk about big mistakes...
I'm sorry! - Re-reading my post it sounds somehow sarcastic. I didn't want to sound it like this. My English is not that good, especially if I try to talk about non-technical topics. When writing this I was too lazy to browse through previous post and find some better expression. I used the " " because it looks cool and because this thread has become a 2-topic confusion hard to read. As I've written before I find the science vs. "...." discussion very interesting.