Mp3 vs Flac vs CD [Archive]

Chronon

January 5th, 2010, 11:56 PM

marco123

January 5th, 2010, 11:57 PM

Depends how you listen to music, and is therefore very subjective.

On my PC or Laptop I can't tell the difference either. But listening to CDs burned from .mp3s on my Hi-Fi equipment it's night and day. .mp3's have no soundstage or dynamics, and the sound is quite flat. Also a lot of inner detail is lost when you compress to .mp3, and I also miss a lot of the smaller scale dynamics and nuance that make music music. (Hence "flat".)

Marco.

pwnst*r

January 5th, 2010, 11:57 PM

subjective and a lot depends on your speakers/headphones

Chronon

January 5th, 2010, 11:58 PM

I don't know much about FLAC beyond that it's a lossless format, but I suppose quality could vary there too.

All lossless compression schemes can be reversed to recover a perfect copy of the input data. They can vary in how much compression is obtained and in how many CPU cycles are needed to decode the compressed data, but the decoded data will always be a bit-perfect replica of the original data.

PurposeOfReason

January 6th, 2010, 12:33 AM

subjective and a lot depends on your speakers/headphones
Very. I've ABX tested pretty well on flac vs 192bitrate. Not so much after 320 and my vinyls are played for more than sound purposes (though I do feel they have more "depth" to them).

donniezazen

January 6th, 2010, 12:33 AM

At what bitrate did you create the mp3? I don't know much about FLAC beyond that it's a lossless format, but I suppose quality could vary there too.

Most of my mp3 are encoded around 192 kbps or 320 kbps or variable. I use my PC, a Sony music system & iPhone to listen my music. Even on music system there hardly is a difference unless the mp3 is poorly ripped. I am expecting Bose QC 15 hopefully i will catch some difference.

Thanks.

l-x-l

January 6th, 2010, 12:51 AM

Hello,

I just couldn't find much difference in sound quality of a Mp3, flac and CD.....

SK.

It depends. Can you tell the difference between a standard-definition video signal vs a High-Definition signal? On a small 13-inch TV, probably not. But the bigger the TV, the more blurry a low quality regular signal appears. By the time you reach 40 inches you can definitely tell the difference. Much of it depends on the equipment you own.

It's the same thing with music. Can you discern lossless music on a Ipod or cheap computer speakers. Probably not. But the better your playback equipment, the more detail you get to hear from lossless audio. Get it.

koleoptero

January 6th, 2010, 01:03 AM

Hello,

I just couldn't find much difference in sound quality of a Mp3, flac and CD. I have not tried Vinyl.

Do you think that there is huge recognizable difference? Are you an analog listener?

Thanks,
SK.
Get a proper sound card (ie not sound blaster), a good pair of speakers and a decent amplifier, and then you'll notice the big difference in colour and resolution between lossy and lossless formats.

Flac and cd shouldn't have a difference but I've not tested them.

PS: You can also see the difference with a good pair of headphones (much cheaper solution than the spaker+amplifier set).

donniezazen

January 6th, 2010, 01:05 AM

Get a proper sound card (ie not sound blaster), a good pair of speakers and a decent amplifier, and then you'll notice the big difference in colour and resolution between lossy and lossless formats.

Flac and cd shouldn't have a difference but I've not tested them.

PS: You can also see the difference with a good pair of headphones (much cheaper solution than the spaker+amplifier set).

Do you consider Sony home theater a good device?

koleoptero

January 6th, 2010, 01:11 AM

Do you consider Sony home theater a good device?

Depends, which one do you have? Most low-budget home theaters aren't good for music. They focus mostly in good thumping bass and spacious sound, not accuracy and fidelity.

PurposeOfReason

January 6th, 2010, 01:28 AM

Do you consider Sony home theater a good device?
Good is very subjective in the audio world. When I buy headphones, I like bright cans and accurate ones so I am not as fond of the grado series because they are too warm for my liking.

original_jamingrit

January 6th, 2010, 01:35 AM

Mp3 is lossy, and flac is loss-less. The difference in quality is extremely subtle. Apparently, it does indeed register at a subconscious level, and studies have been done to show that some people actually prefer the poorer quality mp3s to lossless formats played at the same bit-rate when they're accustomed to mp3's.

http://radar.oreilly.com/2009/03/the-sizzling-sound-of-music.html

Although obviously if you buy your headphones for a dollar, you won't notice a difference at all.

pwnst*r

January 6th, 2010, 02:14 AM

Xbehave

January 6th, 2010, 02:39 AM

Even on high end systems >320kpbs should be enough for anybody, anybody claiming different is an audio snob, but i doubt they would be able to tell the difference in a blind test.

I always listen to music through decent speakers or a hi-fi but i find 192kbps completely transparent (128/96 is close enough for most music)

It tends to vary by musical genre, classical music (and classical-like music such as certain types of metal) generally needs a higher bitrate for there to be no noticeable loss.

VBR (variable bit rate), allows much smaller files to sound good by saving bits on the booring parts and using them on the complex parts. The lowest vbr-rate on a recent version of lame does result in a noticeable loss, but less than you'd expect given the average is much less than 96.

Ogg (an open lossy format), can get slightly better compression than mp3, particularly if you use the non-standard encoder but just like mp3 your probably looking at 80-150 being good enough and 150+ being almost perfect.

One way to notice how flac/wav/high bitrate files are, is to actually encode a typical music sample at various quality levels and see if can the difference between them.

gutterslob

January 6th, 2010, 02:45 AM

Only CD's for my Stax headphones.
(and even then, 80% the CD's I own don't do it justice)

FLAC when I rip and digitize my fave CD's

...and why isn't there a Vinyl, option?

I make do with mp3's when I'm outside on my netbook/laptop, though.
>320 kbps is fine for most travel headphones

maybeway36

January 6th, 2010, 02:48 AM

I use FLAC because I have so little music I can fit it all on my 8GB player. :P

pwnst*r

January 6th, 2010, 02:49 AM

o.O

phrostbyte

January 6th, 2010, 02:58 AM

Sound is stored in a computer as a series of numbers which signify points in a waveform. This is called PCM (pulse-code modulation).

CD is pure PCM data, uncompressed. FLAC is compressed PCM data with no quantization process applied. Thus it is lossless, and the quality is therefore identical.

MP3 on the other hand is lossy. You'll notice that as you lower and lower the bitrate of an MP3, it gets closer and closer to approximating the sound of a sine wave, and low bitrate MP3s sound "crispy" (or sine-wavey).

This is because MP3 applies a discrete variant of the Fourier transform on the blocks of the PCM data before compressing it. Then using a quantum matrix it eliminates the high frequency coefficients from the transformed data, which in turn increases the waveform's entropy. Once the waveform has greater entropy it can compressed to a much smaller size using similar entropy coding techniques as FLAC. The disadvantage is the removal of high frequency components from the transformed waveform also reduces it's quality, which is how much a untransformed waveform can differentiate itself from being a sine wave.

donniezazen

January 6th, 2010, 03:04 AM

gutterslob

January 6th, 2010, 03:14 AM

I left Vinyl because i have never listened to Vinyl.
That's okay. :P
I'm not really a audio connoisseur (the Stax & amp I got as a gift)... I just like the organic feel, warmth and slight roughness that comes from vinyl.... it probably sounds terrible to most people.

pwnst*r

January 6th, 2010, 03:24 AM

It's a different sound, for sure, but just a matter of taste just like audio hardware and such. Most people here aren't used to a "warm" sound at all.

Gizenshya

January 6th, 2010, 04:20 AM

Do you consider Sony home theater a good device?

That depends, but probably for most purposes... but read on...

A chain is only as strong as its weakest link. Everyone has heard this, and it applies well to audio quality.

You won't get any more quality than that of the worst element in a given system. Some examples of elements are: Speakers, amplifiers, original file quality (whether cd, ogg, mp3, radio or whatever), speaker wires, connection type between audio source and amp (and cable quality, where applicable), RF interference and other interference issues, various aspects of the listening environment, and the list goes on. There is literally no limit to what affects sound quality, but the more one gets hammered out the higher cause percentage goes to each of the other causes. But speakers, amplifiers, and source signal quality are the 3 biggest factors most of the time.

etnlIcarus

January 6th, 2010, 04:27 AM

MP3 192kbps or < vs. FLAC/CD, you can pretty consistently tell the difference. 256kbps or >, you can rarely tell the difference.

Besides your ears, your decoder, amp and speakers, another big factor with MP3s tends to be the original encoder. I find that anything using LAME does a much poorer job than many of the proprietary tools out there.

Exodist

January 6th, 2010, 05:46 AM

FLAC is a format using lossless compression. A FLAC file made from a CD track will decode to identical (exactly the same bits) PCM data as is present on the CD itself.

MP3 is a lossy compression format. It throws away some data and the decoded PCM will not be a bit-perfect replica of a CD track used to encode the MP3. However, most people reach a transparency threshhold below 200 kbps (or so) and cannot audibly distinguish the decoded PCM from such an mp3 from the original. What bit rate is necessary for transparency is a totally subjective matter and varies from person to person and codec to codec.
THIS++

Modern encoders encoding at 192kbps are 99.9% of the time to perfect for a average person to notice any change, this is what I use now on all my audio CDs. Back 12+ yea ago, I have seen some some encoders that the user would have to encode the file at like 312k or higher. Back then 256k was the safe bet for compression and quality. They were comparable to 160k now. 128k is still to loose in my opinion, but its got better.

donniezazen

January 6th, 2010, 11:17 AM

A couple of more thing.

Mp3 ripped at aroung 200 kbps is good enough.

Once a CD is ripped into mp3 the sound quality is lost forever and increasing the bitrate of a mp3 is useless.

Thanks.

Chronon

January 6th, 2010, 09:34 PM

Mp3 ripped at aroung 200 kbps is good enough.

"Good enough" is a completely subjective thing and depends on details of how an individual perceives sound. Lossy compression schemes use psychoacoustic (http://en.wikipedia.org/wiki/Psychoacoustics) models and there's not an objectively "best" such model. How we experience sound varies from person to person and some people are much more sensitive to certain types of artifacts (introduced by information loss) than others.

I would say that "in most cases, mp3 ripped at around 200 kbps is good enough."

forrestcupp

January 6th, 2010, 10:14 PM

When choosing a file format, you have to look at your overall needs. Flac and wavs are lossless and sound perfect, but if you ever need to stream them over a network, you might be disappointed because the file size is larger. Mp3's are lossy, but if you rip them at 192 or 256 kbps, it will sound almost perfect and it will stream over a network just fine. Also, if you have a lot of songs, you'll save a lot of space while still having a decent sound.

On my PC or Laptop I can't tell the difference either. But listening to CDs burned from .mp3s on my Hi-Fi equipment it's night and day. .mp3's have no soundstage or dynamics, and the sound is quite flat. Also a lot of inner detail is lost when you compress to .mp3, and I also miss a lot of the smaller scale dynamics and nuance that make music music. (Hence "flat".)

With more modern music, a lot of times dynamics and "flat" sounds are all in the recording process. No matter how good your studio is, a CD can only do 16-bit 44.1 khz. So to get a lot of volume, engineers use a lot of audio compression on the tracks and master mixdown. This makes the CD sound a lot louder, but it ends up being "flat" with not much dynamics. In those cases, you're going to have a flat sound no matter what the file format. Personally, I think it's a crappy technique, even though it's a popular one.

Chronon

January 6th, 2010, 10:52 PM

I agree. The 'loudness war' sucks. Compression is not a reversible process (without extra information).

Cam42

January 6th, 2010, 10:58 PM

Whoever said that MP3s sound crispy, I thank you. I've been looking for a way to describe how my MP3s sound after I discovered that my sansa could play FLAC. With my new stereo, you can tell the difference clearly. FLAC and my CDs sound much fuller.

Miguel

January 6th, 2010, 11:34 PM

Let's see. Most people don't own even a mid-end audio system. Second, most people don't have a trained ear. Third, most music produced today has an "MP3 sound" engineered around it. Fourth, most people don't really care once the bitrate is above 128. I personally rip my CD's to ogg vbr -q6 (~192), and my girlfriend thinks I'm extremely nitpicky. I haven't tested it, but I'm quite sure I wouldn't be able to reliably ABX that and a CD. At least where I usually hear music.

Anyway, in most cases, the upgrade in equipment that most people need to reliably tell the difference between a 192kbps mp3 and flac will cause a much larger change in sound than there is between those two files.

BTW: I love my Grado headphones that PurposeOfReason dislikes. But I understand why. In any case, listening Sultans of Swing live with those cans is a beautiful experience.

EDIT: Here (http://www.listening-tests.info/mf-128-1/results.htm) you will find a comparison from December 2005 between different encoders against lossless at 128 kbps. I'm not sure if all the aoTuV patches from back then have made it into "upstream" ogg vorbis. In any case, you see that most codecs score incredibly close to lossless when tested by a selected group. If you add to that improvements in psychoacoustic models plus a larger bitrate, you know what you will get.

marco123

January 6th, 2010, 11:46 PM

I agree. The 'loudness war' sucks. Compression is not a reversible process (without extra information).

http://en.wikipedia.org/wiki/Death_Magnetic#Criticism_regarding_production Has to be the epitome of the loudness wars. That thing even sounds awful on £10k + CD players.

Groucho Marxist

January 6th, 2010, 11:46 PM

Hello,

I just couldn't find much difference in sound quality of a Mp3, flac and CD. I have not tried Vinyl.

Do you think that there is huge recognizable difference? Are you an analog listener?

Thanks,
SK.

I prefer CD and vinyl for differing reasons. In terms of the former, I can easily and legally import tracks to play on one of the radio stations I work for. The latter of the two I use due to more aesthetically pleasing audio quality on my Crosley Radio. For example, If you ever have the cash and time, I highly recommend picking up a vinyl copy of Dark Side of the Moon, closing your eyes and taking one crazy journey through space and time. Speaking of time and space, nothing says "staying power" more than the fact that consumers have been using vinyl since August 12th, 1877 when Edison invented the phonograph.

pwnst*r

January 6th, 2010, 11:46 PM

When choosing a file format, you have to look at your overall needs. Flac and wavs are lossless and sound perfect, but if you ever need to stream them over a network, you might be disappointed because the file size is larger. Mp3's are lossy, but if you rip them at 192 or 256 kbps, it will sound almost perfect and it will stream over a network just fine. Also, if you have a lot of songs, you'll save a lot of space while still having a decent sound.

With more modern music, a lot of times dynamics and "flat" sounds are all in the recording process. No matter how good your studio is, a CD can only do 16-bit 44.1 khz. So to get a lot of volume, engineers use a lot of audio compression on the tracks and master mixdown. This makes the CD sound a lot louder, but it ends up being "flat" with not much dynamics. In those cases, you're going to have a flat sound no matter what the file format. Personally, I think it's a crappy technique, even though it's a popular one.

Dynamic range compression. Sucks.

donniezazen

January 6th, 2010, 11:48 PM

Let's see. Most people don't own even a mid-end audio system. Second, most people don't have a trained ear. Third, most music produced today has an "MP3 sound" engineered around it. Fourth, most people don't really care once the bitrate is above 128. I personally rip my CD's to ogg vbr -q6 (~192), and my girlfriend thinks I'm extremely nitpicky. I haven't tested it, but I'm quite sure I wouldn't be able to reliably ABX that and a CD. At least where I usually hear music.

Anyway, in most cases, the upgrade in equipment that most people need to reliably tell the difference between a 192kbps mp3 and flac will cause a much larger change in sound than there is between those two files.

BTW: I love my Grado headphones that PurposeOfReason dislikes. But I understand why. In any case, listening Sultans of Swing live with those cans is a beautiful experience.

The purpose should be that you can listen to quality music at home not using a super computer at Goddard NASA space flight center.

pwnst*r

January 7th, 2010, 12:00 AM

The purpose should be that you can listen to quality music at home not using a super computer at Goddard NASA space flight center.

except that it has nothing to do with your computer.

marco123

January 7th, 2010, 12:07 AM

EDIT: Here (http://www.listening-tests.info/mf-128-1/results.htm) you will find a comparison from December 2005 between different encoders against lossless at 128 kbps. I'm not sure if all the aoTuV patches from back then have made it into "upstream" ogg vorbis. In any case, you see that most codecs score incredibly close to lossless when tested by a selected group. If you add to that improvements in psychoacoustic models plus a larger bitrate, you know what you will get.

That looks like a comparison of lossy codecs at 128Kbps?

marco123

January 7th, 2010, 12:12 AM

Anyway, in most cases, the upgrade in equipment that most people need to reliably tell the difference between a 192kbps mp3 and flac will cause a much larger change in sound than there is between those two files.

Equipment? I can tell the difference between 192Kbps mp3 and 320 Kbps mp3 on my £105 iPod Nano! Nevermind Lossless. But then again I can hear the difference between a Chord Cobra 3 and a Chord Chameleon Silver.

pwnst*r

January 7th, 2010, 12:14 AM

marco123

January 7th, 2010, 12:17 AM

let's not even go there with cables.

But I love cables. :(

Half of the fun of HiFi is tweaking the sound with cables!;)

pwnst*r

January 7th, 2010, 12:19 AM

^_^ speakers and amps are one thing for the masses here, but cables are a bit too snakeoil for most.

marco123

January 7th, 2010, 12:20 AM

^_^ speakers and amps are one thing for the masses here, but cables are a bit too snakeoil for most.

How do you connect your components then?:D

Xbehave

January 7th, 2010, 12:30 AM

Equipment? I can tell the difference between 192Kbps mp3 and 320 Kbps mp3 on my £105 iPod Nano! Nevermind Lossless. But then again I can hear the difference between a Chord Cobra 3 and a Chord Chameleon Silver.
Good sir, i have some Ethernet cables i wish to sell you.

^_^ speakers and amps are one thing for the masses here, but cables are a bit too snakeoil for most.
+1, hell i think there isn't a noticable difference between most hi-fi systems, i only notice hi-fi vs good speakers vs cheap speakers, but i will believe that non-snobs can notice the difference between models, cables on the other hand...

pwnst*r

January 7th, 2010, 12:33 AM

How do you connect your components then?:D

Let me clarify - the "difference" in cabling is snake oil to most.

Miguel

January 7th, 2010, 12:34 AM

Equipment? I can tell the difference between 192Kbps mp3 and 320 Kbps mp3 on my £105 iPod Nano! Nevermind Lossless. But then again I can hear the difference between a Chord Cobra 3 and a Chord Chameleon Silver.

You might as well have an incredibly accurate hearing. If you like music but are nitpicky, then please accept my condolences. And I'm serious.

By the way, do you use VBR or CBR for your 192 and 320 kbps files? I've seen tons of 320 kbps CBR files out there, but that's just plain silly (unless your player doesn't support VBR). In any case, I have a huge curiosity regarding your equipment. Some serious stuff is needed once you go to those levels.

Miguel

January 7th, 2010, 12:37 AM

I will believe that non-snobs can notice the difference between models, cables on the other hand...

I'd really like to perform actual physical sound pressure measurements between identical systems with different cables. I'd really like to.

marco123

January 7th, 2010, 12:42 AM

You might as well have an incredibly accurate hearing. If you like music but are nitpicky, then please accept my condolences. And I'm serious.

By the way, do you use VBR or CBR for your 192 and 320 kbps files? I've seen tons of 320 kbps CBR files out there, but that's just plain silly (unless your player doesn't support VBR). In any case, I have a huge curiosity regarding your equipment. Some serious stuff is needed once you go to those levels.

Nitpicky: that's not even the half of it! My cables and stands alone exceed £1k. (That's not including all the upgrades I've been though.) My wife is the one that needs the condolences I think.:D

pwnst*r

January 7th, 2010, 12:55 AM

I'd really like to perform actual physical sound pressure measurements between identical systems with different cables. I'd really like to.

sound pressure measurements? what will that prove? the db's will be exactly the same.

Chronon

January 7th, 2010, 01:36 AM

Let me clarify - the "difference" in cabling is snake oil to most.

Yes. Apparently some people are willing to pay a lot of money for placebo effect. (I'm not saying all cables are created equal, only that it appears that one quickly reaches a limit of diminishing returns.)

pwnst*r

January 7th, 2010, 01:59 AM

Yes. Apparently some people are willing to pay a lot of money for placebo effect. (I'm not saying all cables are created equal, only that it appears that one quickly reaches a limit of diminishing returns.)

I totally agree.

Gizenshya

January 7th, 2010, 02:30 AM

Let me clarify - the "difference" in cabling is snake oil to most.

I do love those who spend money on expensive optical cables vs the cheapos, and then go around evangelizing about the benefit. Sometimes I wonder if they are actually that stupid or if it is some sort of subconscious, cognitive dissonance-induced defense mechanism.

And yeah, any measurement of SPL would prove nothing. The most reliable way to measure differences would be to isolate the cables and analyze them electrically. Then introduce interference. Non-shielded cables would perform worse, yeah, but unless they are very long, I highly doubt it would be noticeable.

Marco, if you're that picky, put your equipment in a Faraday Cage (http://en.wikipedia.org/wiki/Faraday_cage), and enclose that in magnetically shielded case. That will stop most types of interference that cause audible distortion your system (that originate outside your system, at least). Doing this will certainly have a far more beneficial effect on your listening experience than any cable upgrade. This would also stop external interference affecting your electronics (EQ, amp, source devices, etc.) before it gets to your cables.

marco123

January 7th, 2010, 03:08 AM

I do love those who spend money on expensive optical cables vs the cheapos, and then go around evangelizing about the benefit. Sometimes I wonder if they are actually that stupid or if it is some sort of subconscious, cognitive dissonance-induced defense mechanism.

And yeah, any measurement of SPL would prove nothing. The most reliable way to measure differences would be to isolate the cables and analyze them electrically. Then introduce interference. Non-shielded cables would perform worse, yeah, but unless they are very long, I highly doubt it would be noticeable.

Marco, if you're that picky, put your equipment in a Faraday Cage (http://en.wikipedia.org/wiki/Faraday_cage), and enclose that in magnetically shielded case. That will stop most types of interference that cause audible distortion your system (that originate outside your system, at least). Doing this will certainly have a far more beneficial effect on your listening experience than any cable upgrade. This would also stop external interference affecting your electronics (EQ, amp, source devices, etc.) before it gets to your cables.

My interconnects, speaker cables and mains cables for the CD player, Pre-Amp and Main Amp are all shielded. I use an Isotek mains conditioner, and my component rack provides isolation from external vibration, so I don't think I'll need a Faraday Cage. Looks cool though.

I use analogue interconnects though, not digital.

By the way, I demo any cables or components I'm thinking of upgrading in the listening rooms of my dealer and at home: I wouldn't buy them if they didn't make a difference. I think the argument also goes the other way though, i.e. people who rubbish expensive cables/stands or other accessories have either not tried them, or maybe tried them with a £200 CD player/integrated amp combo. But I wholeheartedly agree about the theory of diminishing returns, because I think I've reached that point with my current setup.

Now I'm off to clean my Rothwell Attenuators. :)

EDit: This guy is a bit more obsessed than me, lol. (He loads the CD twice to make it sound better??) http://www.youtube.com/watch?v=piPI_ByD0Q0

xtaaxtw

January 7th, 2010, 04:11 AM

I totally agree.
wow,nice!:lolflag::lolflag:

—————————
ShaoLin kung fu (http://world-culture-research.org/c.asp?d=15859)

Miguel

January 7th, 2010, 02:09 PM

sound pressure measurements? what will that prove? the db's will be exactly the same.

You didn't understand me. I don't want to measure average SPL. No way. I want to accurately measure P(t), fourier analyse it and perform a differential analysis between the two sources. Is it possible to do this accurately? I have no idea.

Someone has suggested measuring the electrical signal in the cables, but that misses the transmittance to the speakers. Furthermore, good cables are shielded, making it a bit difficult to perform outside measurements without altering the cables..

pwnst*r

January 7th, 2010, 02:21 PM

You said "sound pressure measurements" which to me, says SPL.

I'm not sure what P(t) and "fourier" mean.

Miguel

January 7th, 2010, 02:28 PM

You said "sound pressure measurements" which to me, says SPL.

I'm not sure what P(t) and "fourier" mean.

That's OK. P(t) refers to the evolution of the sound pressure with time, and a fourier transform is a mathematical tool that allows me to see the importance of the various frequencies in a wave (which is what sound is). If the cables have a real impact, they must cause different sound waves, and hopefully that's measurable with some sort of equipment.

pwnst*r

January 7th, 2010, 02:40 PM

It's not measurable or it would have been done already. This is why the interconnects debate is even more subjective than differing components.

marco123

January 7th, 2010, 03:41 PM

It's not measurable or it would have been done already. This is why the interconnects debate is even more subjective than differing components.

I seem to remember that Russ Andrews got into trouble with the ASA (Advertising Standards) here in the UK over the advertising of it's mains cables, and had prove that their products actually did something.

It's a .pdf: www.russandrews.com/downloads/CableTestPremRes.pdf

There's all sorts of graphs and measurements to sate your curiosity.;)

I suppose the same methods could be used with interconnects to measure the noise floor, RFI, THD and IMD.

Marco.

koshatnik

January 7th, 2010, 04:01 PM

Hello,

I just couldn't find much difference in sound quality of a Mp3, flac and CD. I have not tried Vinyl.

Do you think that there is huge recognizable difference? Are you an analog listener?

Thanks,
SK.

Depends on your system. If I play an MP3 through my set up it sounds awful. If you are playing it through some cheapo computer speakers, they will all sound much of a muchness. Its down to how transparent your set up is.

Vinyl is dependent on a decent turntable and tone arm. Most people listen to vinyl on crappy turntables with poor tone arms. A decent tone arm and cartridge will eliminate all crackles and pops and provide wonderful sound reproduction. But there is a downside. If you have too transparent a system, it will show up the poor production values and engineering on certain recordings. For example, I have an Iggy Pop album that I cant play on my current set up because its so badly engineered. Didnt notice it on my old crappy set up. Sounded fine.

subjective and a lot depends on your speakers/headphones

Source/Amp is more important arbiter of sound quality in a system. A good amp will drive average speakers to better performance. Most speakers/headphones sound crap because the amp they are plugged into is crap.

Re: cabling. Yes it makes a difference, esp if you have decent mid range/high end equipment. I switched from cheapo interconnects to kimber tonix's and the difference was noticeable. Changing your interconnects on a cheap system will do nothing.

Speaker cable makes a difference too. Different types of cable have different impedences, some allow high end signals through but are poor at reproducing low and mid range (silver cables for example) whereas others are more neutral tonally (bronze cables) and some are "bottomy" (let mid and low range through but impede high frequency) like copper cables.

Re: Russ Andrews - he went too far with his claims and was shown up as a result of it. But cabling does make a difference to the sound system. If it makes no difference to yours, then you must have a cheaper end set up that isnt particularly transparent. My set up is transparent, and changes are noticeable.

marco123

January 7th, 2010, 04:16 PM

Depends on your system. If I play an MP3 through my set up it sounds awful. If you are playing it through some cheapo computer speakers, they will all sound much of a muchness. Its down to how transparent your set up is.

Vinyl is dependent on a decent turntable and tone arm. Most people listen to vinyl on crappy turntables with poor tone arms. A decent tone arm and cartridge will eliminate all crackles and pops and provide wonderful sound reproduction. But there is a downside. If you have too transparent a system, it will show up the poor production values and engineering on certain recordings. For example, I have an Iggy Pop album that I cant play on my current set up because its so badly engineered. Didnt notice it on my old crappy set up. Sounded fine.

Source/Amp is more important arbiter of sound quality in a system. A good amp will drive average speakers to better performance. Most speakers/headphones sound crap because the amp they are plugged into is crap.

Re: cabling. Yes it makes a difference, esp if you have decent mid range/high end equipment. I switched from cheapo interconnects to kimber tonix's and the difference was noticeable. Changing your interconnects on a cheap system will do nothing.

Speaker cable makes a difference too. Different types of cable have different impedences, some allow high end signals through but are poor at reproducing low and mid range (silver cables for example) whereas others are more neutral tonally (bronze cables) and some are "bottomy" (let mid and low range through but impede high frequency) like copper cables.

Re: Russ Andrews - he went too far with his claims and was shown up as a result of it. But cabling does make a difference to the sound system. If it makes no difference to yours, then you must have a cheaper end set up that isnt particularly transparent. My set up is transparent, and changes are noticeable.

Agree totally.

I think Russ Andrews redeemed himself by providing some sound scientific evidence for the Audiophile crowd, though.

Marco.

EDit: The full paper can be found here: http://www.russandrews.com/src/researchpaper09/article-research-papers-researchpaper09.htm

cascade9

January 7th, 2010, 04:18 PM

Depends on your system. If I play an MP3 through my set up it sounds awful. If you are playing it through some cheapo computer speakers, they will all sound much of a muchness. Its down to how transparent your set up is.

Vinyl is dependent on a decent turntable and tone arm. Most people listen to vinyl on crappy turntables with poor tone arms. A decent tone arm and cartridge will eliminate all crackles and pops and provide wonderful sound reproduction. But there is a downside. If you have too transparent a system, it will show up the poor production values and engineering on certain recordings. For example, I have an Iggy Pop album that I cant play on my current set up because its so badly engineered. Didnt notice it on my old crappy set up. Sounded fine.

Agreed 100% on 'cheapo computer speakers'. I've had standup arguments with a few friends because they insisted that '192k/sec MP3 is transparent'. Maybe on the AC97/single cone circa 1990 cheapo speakers they use (and probably helped by the fact they they love nightclubs 'till they get the ring').

As for vinyl- yeah, some of them have got awful engineering. But the reverse also applies- I've seen more than 1 forum thread, etc, saying that some vinyl versions of modern albums are better than the CD version, thanks to the loudness war (image below says all I need to about that)-

http://localareawatch.typepad.com/photos/uncategorized/2007/10/17/swearing.jpg

Red Hot Chili Peppers Stadium Arcadium is the one that I always remember, see here-

http://www.hydrogenaudio.org/forums/index.php?showtopic=42376
http://www.petitiononline.com/RHCPWBCD/petition.html

BTW, no, I'm not a RHCP fan. Well, not the new stuff anyway, the older stuff was better IMO.

phrostbyte

January 7th, 2010, 04:24 PM

You said "sound pressure measurements" which to me, says SPL.

I'm not sure what P(t) and "fourier" mean.

If you don't what Fourier analysis is you shouldn't be talking about signal quality with any authority. Period. Fourier analysis is the heart and foundation of the entire field of signal processing, which itself is the heart of DSPs and sound engineering.

koshatnik

January 7th, 2010, 04:26 PM

As for vinyl- yeah, some of them have got awful engineering. But the reverse also applies- I've seen more than 1 forum thread, etc, saying that some vinyl versions of modern albums are better than the CD version, thanks to the loudness war (image below says all I need to about that)-

There are plenty of CD's with terrible engineering, I wasnt solely implying that vinyl records are particularly prone to bad engineering. ;)

I think the Audiophile is a dying breed. Fewer and fewer people are willing to spend money on a decent hifi it seems. People seem content with their computers or with ipod docks. Its a pity because its hard to get over just how different a recording can sound on a really nice setup.

My mother in law loves opera. She brought down a recording of Purcells Dido & Aeneas to listen on my set up. She couldnt believe what she heard. She'd only ever heard it on an ipod dock before.

i think music is worth the extra investment to hear it properly.

Johnsie

January 7th, 2010, 04:27 PM

I've found that the best way to improve sound quality is to get better hardware... By better I mean better sound, not more expensive. I've bought some cheapish speakers that were pretty good and some expensive speakers that were pretty awful.

The people who engineer the recording can also play a huge part in how good things sound. Yes, the format of the audio is important too, but it's just one of many variables needed to get a good sound.

pwnst*r

January 7th, 2010, 04:47 PM

Depends on your system. If I play an MP3 through my set up it sounds awful. If you are playing it through some cheapo computer speakers, they will all sound much of a muchness. Its down to how transparent your set up is.

Vinyl is dependent on a decent turntable and tone arm. Most people listen to vinyl on crappy turntables with poor tone arms. A decent tone arm and cartridge will eliminate all crackles and pops and provide wonderful sound reproduction. But there is a downside. If you have too transparent a system, it will show up the poor production values and engineering on certain recordings. For example, I have an Iggy Pop album that I cant play on my current set up because its so badly engineered. Didnt notice it on my old crappy set up. Sounded fine.

Source/Amp is more important arbiter of sound quality in a system. A good amp will drive average speakers to better performance. Most speakers/headphones sound crap because the amp they are plugged into is crap.

Re: cabling. Yes it makes a difference, esp if you have decent mid range/high end equipment. I switched from cheapo interconnects to kimber tonix's and the difference was noticeable. Changing your interconnects on a cheap system will do nothing.

Speaker cable makes a difference too. Different types of cable have different impedences, some allow high end signals through but are poor at reproducing low and mid range (silver cables for example) whereas others are more neutral tonally (bronze cables) and some are "bottomy" (let mid and low range through but impede high frequency) like copper cables.

Re: Russ Andrews - he went too far with his claims and was shown up as a result of it. But cabling does make a difference to the sound system. If it makes no difference to yours, then you must have a cheaper end set up that isnt particularly transparent. My set up is transparent, and changes are noticeable.

lol @ cheaper set up. I have fairly high-end equipment also (GR Research Alpha LS as mains for instance) so I know what I'm talking about and have been involved with home theater/audio for 20 years.

If your kimber cable made a difference to you, that's what counts right? IMO, most of the super high end cables are all about the placebo once you've paid out the nose for them.

pwnst*r

January 7th, 2010, 04:50 PM

If you don't what Fourier analysis is you shouldn't be talking about signal quality with any authority. Period. Fourier analysis is the heart and foundation of the entire field of signal processing, which itself is the heart of DSPs and sound engineering.

who said I was an authority in measuring audio frequencies? learn to read. I let my ears decide. You can measure all you want with cabling and if it looks better on paper, you're telling me for sure that that cable will sound better to you than the others? If you're saying that, then YOU have no idea what it's like to listen.

pwnst*r

January 7th, 2010, 04:52 PM

Yes, the format of the audio is important too, but it's just one of many variables needed to get a good sound.

Except that you can't polish a turd (although Mythbusters did). So if the source is crap, no hardware is going to fix that.

koshatnik

January 7th, 2010, 05:02 PM

If your kimber cable made a difference to you, that's what counts right? IMO, most of the super high end cables are all about the placebo once you've paid out the nose for them.

I agree completely. But I just hate this luddite reaction to cabling - "its all snake oil", yadda yadda. Ok, attach some 39p a meter bellwire to your hifi and see how great it sounds now. :D Most people that claim cables make no difference have obviously never used a demo pack of cables and tested them on their set up to see what difference it can make. I have, and there were clear differences.

I admit that £12,000 a meter speaker cable is stupid, but back in the real world, better cables make a tangible difference on certain systems. If they don't then cool, send them back. Its all about your own ear.

pwnst*r

January 7th, 2010, 05:07 PM

koshatnik

January 7th, 2010, 05:25 PM

Heh, well I never said I cabled mine with lamp wire ^_^ Mid-range interconnects work/sounds just fine compared to interconnects that literally cost 5 times more that I have tried at home.

Anything above that I call "audio jewelery"

people like to measure themselves by the bling they own :D Even beardscratching hifi nerds are not immune to this.

Past a certain price point, yeah, its impossible to hear a difference really.

pwnst*r

January 7th, 2010, 05:30 PM

Truth be told - I ran Kimber in my car in the 90's :o

marco123

January 7th, 2010, 06:11 PM

people like to measure themselves by the bling they own :D Even beardscratching hifi nerds are not immune to this.

Past a certain price point, yeah, its impossible to hear a difference really.

Would this be the price point you are referring to? http://www.hificables.co.uk/11755/Chord-Indigo-Plus-Phono-Interconnect.html?referrer=froogle&utm_source=google&utm_medium=froogle&utm_campaign=pid11755

Or this: http://www.hificables.co.uk/11782/Atlas-Asimi-XLR---XLR.html

*Hides*

phrostbyte

January 7th, 2010, 06:33 PM

who said I was an authority in measuring audio frequencies? learn to read. I let my ears decide. You can measure all you want with cabling and if it looks better on paper, you're telling me for sure that that cable will sound better to you than the others? If you're saying that, then YOU have no idea what it's like to listen.

Call me an elitst, but I find it irritating when someone goes 'I'm not sure what P(t) and "fourier" mean.' while actively debating sound quality. It's like someone debating bridge engineering and saying "I don't know what calculus mean.". "I'll just let my eyes decide."

Chronon

January 7th, 2010, 06:43 PM

phrostbyte

January 7th, 2010, 06:48 PM

That's a rather bad analogy. The performance of a bridge can be assessed by somewhat objective metrics. The whole point of an audio system is to deliver sound that _you_ will experience. Any performance differences that cannot be perceived are meaningless.

Audio can be assessed by certain objective metrics also. Audio is not voodoo. There is an entire science and theory behind audio, and Fourier analysis is a big part of that.

Chronon

January 7th, 2010, 06:52 PM

We are talking about audio systems for music playback. Please demonstrate how objective metrics (corresponding to imperceptible data) determine your perceptual experience of music. Next, you'll be telling me that one LCD is better than another for image reproduction because it has superior output in the infrared.

phrostbyte

January 7th, 2010, 06:54 PM

Chronon

January 7th, 2010, 06:57 PM

What is imperceptible to you, might not be imperceptible to me. And certainly imperceptible does NOT mean there is no difference.

You are the one attempting to use objective measures for this. I have maintained that perception is an entirely subjective matter. I also never argued that no difference existed. I am saying that if "you" can't perceive it then it shouldn't matter to "you".

pwnst*r

January 7th, 2010, 06:57 PM

phrostbyte

January 7th, 2010, 06:59 PM

You are the one attempting to use objective measures for this. I have maintained that perception is an entirely subjective matter. I also never argued that no difference existed. I am saying that if "you" can't perceive it then it shouldn't matter to "you".

Then what is the point of this debate? If it doesn't matter to you, don't bother with it. But don't start acting like audio quality shouldn't matter to everyone, or there ISN'T objective differences between different audio encodings and other systems which play back audio. Because you'd be wrong either way.

pwnst*r

January 7th, 2010, 07:05 PM

Then what is the point of this debate? If it doesn't matter to you, don't bother with it. But don't start acting like audio quality shouldn't matter to everyone, or there ISN'T objective differences between different audio encodings and other systems which play back audio. Because you'd be wrong either way.

We were debating interconnects, not source material.

Talk about being wrong...

Chronon

January 7th, 2010, 07:05 PM

Then what is the point of this debate? If it doesn't matter to you, don't bother with it. But don't start acting it shouldn't matter to everyone, or there ISN'T objective differences between different audio encodings and other systems which play back audio. Because you'd be wrong either way.

I don't think you are thinking clearly.

Perception is a subjective and personal experience. If a given person cannot distinguish a difference (i.e. the difference is imperceptible) then why should that person care. You are arguing that they should care about something that they cannot even experience and I am interested in why. You haven't explained that bit.

Please don't bring up objective differences if you can't connect them to the present discussion. I never argued about them. You seem to be trying to shoehorn them into the discussion because of the rather inapt bridge analogy.

phrostbyte

January 7th, 2010, 07:10 PM

I don't think you are thinking clearly.

Perception is a subjective and personal experience. If a given person cannot distinguish a difference (i.e. the difference is imperceptible) then why should that person care. You are arguing that they should care about something that they cannot even experience and I am interested in why. You haven't explained that bit.

Please don't bring up objective differences if you can't connect them to the present discussion. I never argued about them. You seem to be trying to shoehorn them into the discussion because of the rather inapt bridge analogy.

If there is no objectivity in this debate, then it's largely absurd and pointless. Because you are going with the assumption that you can hear through everybody's ears.

I do not like to argue the absurd and pointless, I like to learn whatever objectivity there is to learn in such a field. I answered the OP in the most objective way I could, which of course seems to be a far cry from the rest of the thread.

murderslastcrow

January 7th, 2010, 07:11 PM

pwnst*r

January 7th, 2010, 07:13 PM

CDs are not the highest quality to rip from. That's why they're thinking of making audio DVDs and even audio Blurays.

Erm...source?

Chronon

January 7th, 2010, 07:26 PM

If there is no objectivity in this debate, then it's largely absurd and pointless. Because you are going with the assumption that you can hear through everybody's ears.
Huh?

I really don't see how

Perception is a subjective and personal experience. If a given person cannot distinguish a difference (i.e. the difference is imperceptible) then why should that person care. You are arguing that they should care about something that they cannot even experience and I am interested in why. You haven't explained that bit.
could be construed to mean that I am claiming that my perception is universal or that I can hear through others' ears.

I do not like to argue the absurd and pointless, I like to learn whatever objectivity there is to learn in such a field. I answered the OP in the most objective way I could, which of course seems to be a far cry from the rest of the thread.
Apparently, this has been the cause of some confusion. It appeared to me that you were chiming in on the unfolding discussion rather than directing your response to the OP.

If you read my initial response to the OP you will find that it was also objective and only discussed bit comparisons, reversibility, etc. The discussion wandered (as they are prone to do).
--
EDIT: I just checked my first post and I did allow subjectivity to creep in. I mentioned that a lossy file is not equivalent to the source from which it was derived and mentioned "transparency". I apologize for misrepresenting this as objective.

cascade9

January 7th, 2010, 07:38 PM

CDs are not the highest quality to rip from. That's why they're thinking of making audio DVDs and even audio Blurays.

If you get a FLAC from the original recording, that's the best sound value. Then, encoding that into a quality level 10 OGG is the best way to get a high quality lossy audio file. It'll be smaller than the equivalent MP3, also.

There is no (official) MP3 over 320K/sec, and -q10 is (aprox) 500K/sec. Pointless IMO, I have .flacs that are smaller than that bitrate. 320K (MP3 or ogg vorbis) is as far as is worth going.

As for audio DVD, all them I've seen is only in DTS or Dolby (both lossy). There is 'DTS-Master Audio' and 'Dolby TrueHD' that are lossless, but I'm yet to see one.....or even hear of one being released. Possibly they will come out on audio-bluray, but even then the better lossless versions are 'optional'. Besides, knowing bluray, they will be a right pain in the butt to rip, if its even possible.

What audio codecs will Blu-ray support?

Linear PCM (LPCM) - up to 8 channels of uncompressed audio. (mandatory)
Dolby Digital (DD) - format used for DVDs, 5.1-channel surround sound. (mandatory)
Dolby Digital Plus (DD+) - extension of Dolby Digital, 7.1-channel surround sound. (optional)
Dolby TrueHD - lossless encoding of up to 8 channels of audio. (optional)
DTS Digital Surround - format used for DVDs, 5.1-channel surround sound. (mandatory)
DTS-HD High Resolution Audio - extension of DTS, 7.1-channel surround sound. (optional)
DTS-HD Master Audio - lossless encoding of up to 8 channels of audio. (optional)

Please note that this simply means that Blu-ray players and recorders will have to support playback of these audio codecs, it will still be up to the movie studios to decide which audio codec(s) they use for their releases.

http://www.blu-ray.com/faq/#bluray_audio_codecs

phrostbyte

January 7th, 2010, 07:40 PM

Huh?

I really don't see how
could be construed to mean that I am claiming that my perception is universal or that I can hear through others' ears.

Apparently, this has been the cause of some confusion. It appeared to me that you were chiming in on the unfolding discussion rather than directing your response to the OP.

If you read my initial response to the OP you will find that it was also objective and only discussed bit comparisons, reversibility, etc. The discussion wandered (as they are prone to do).

I'm all for subjective debates when they involve some kind of profound point that can't easily be discussed objectively (eg: meaning of life, direction of society, the meaning of reality, etc etc). But it's obvious this topic has quite objective answers, and the subjective aspects are quite pointless when all they involves is the perception of the self.

So it possible to keep the discussion objective then? I'd love to have a real, technical/mathematical discussion at least once in my life about audio compression, as it's an area I have interest in.

Xbehave

January 7th, 2010, 07:41 PM

It can be shown that audio cables do/do not make any difference to the sound coming out of the speakers, there are objective tests.

Erm...source?
There is only so much data you can fit on a CD ~700MB (even though the format for data and audio are completely different, the total information is the same) obviously if you have a bigger medium you can fit more audio data onto the disc.

phrostbyte

January 7th, 2010, 07:44 PM

It can be shown that audio cables do/do not make any difference to the sound coming out of the speakers, there are objective tests.

There is only so much data you can fit on a CD ~700MB (even though the format for data and audio are completely different, the total information is the same) obviously if you have a bigger medium you can fit more audio data onto the disc.

I'll add this to your point:
http://en.wikipedia.org/wiki/DVD-Audio

DVD Audio can have a higher sampling rate, as well as more channels of audio.

cascade9

January 7th, 2010, 07:56 PM

I'll add this to your point:
http://en.wikipedia.org/wiki/DVD-Audio

DVD Audio can have a higher sampling rate, as well as more channels of audio.

True, but its actually not that much bigger than CD-DA.

DTS Digital Surround = 1510k/sec.

CD-DA = 1411k/sec.

The only reason why you can make DTS Digital Surround play on normal CD players is the codec issue (like, it doesnt understand DTS). But the actual physical CD can hold more than enough data to fit albums of DTS Digital. Not understanding the DTS codec why most DVD-A discs are dual layer, layer 1 is red book CD-DA (to play in any CD player) and layer 2 is DVD-Audio.

Using .flac (or .wv) would be easier, they both support stupidly high sampling rates and multichannel.

pwnst*r

January 7th, 2010, 07:57 PM

more channels has nothing to do with audio quality.

measure that.

Chronon

January 7th, 2010, 08:04 PM

I'm all for subjective debates when they involve some kind of profound point that can't easily be discussed objectively (eg: meaning of life, direction of society, the meaning of reality, etc etc). But it's obvious this topic has quite objective answers, and the subjective aspects are quite pointless when all they involves is the perception of the self.
The existence of an objective description (actually an oxy-moron) does not automatically make it superior. The perception, in this case, is of the sound transduced by the audio system. This leads to an assessment of quality (i.e. non-quantitative and subjective). I find objective arguments misguided when they are applied to matters of aesthetics (like which monitor looks best [qualitative] or which sound system sounds best [qualitative]).

Why can't you address this simple issue?

If a given person cannot distinguish a difference (i.e. the difference is imperceptible) then why should that person care. You are arguing that they should care about something that they cannot even experience and I am interested in why.
Why do objective differences matter (to a given individual) if they are subjectively transparent (to that given individual)?

So it possible to keep the discussion objective then? I'd love to have a real, technical/mathematical discussion at least once in my life about audio compression, as it's an area I have interest in.
It is certainly possible for such a discussion to happen. This isn't really my area of expertise so I'm not sure if I can participate if you wish to embark on a tangent focused on the details of audio compression algorithms.

Xbehave

January 7th, 2010, 08:08 PM

more channels has nothing to do with audio quality.

measure that.
a wav isn't a perfect recording of reality, it is a set of samples over time (reality is continuous, well until quantum scales, but that is a different story), more data means more samples/second, means a closer approximation to reality. Think of it as increasing the framerate on your speakers, that needs more data and so a bigger medium than a CD.

pwnst*r

January 7th, 2010, 08:12 PM

a wav isn't a perfect recording of reality, it is a set of samples over time (reality is continuous, well until quantum scales, but that is a different story), more data means more samples/second, means a closer approximation to reality. Think of it as increasing the framerate on your speakers, that needs more data and so a bigger medium than a CD.

but if you don't have the equipment to handle the larger sample rates, what difference does that make?

also, you quoted my post about the audio channels, but replied with something that has nothing to do with said quote.

Xbehave

January 7th, 2010, 08:14 PM

The existence of an objective description (actually an oxy-moron) does not automatically make it superior. The perception, in this case, is of the sound transduced by the audio system. This leads to an assessment of quality (i.e. non-quantitative and subjective). I find objective arguments misguided when they are applied to matters of aesthetics (like which monitor looks best [qualitative] or which sound system sounds best [qualitative]).
Because if a quanatative analysis can show that both monitors are the same, anybody claiming different is an idiot. If a quantative analysis can show that monitor A has a higher resolution than monitor B, then while it can be argued that nobody would notice, anybody claiming monitor B gave a better reproduction of images would be wrong.

Why do objective differences matter (to a given individual) if they are subjectively transparent (to that given individual)?
Because if i paint a car green and tell you it goes faster and you can show me it doesn't, I would be proved wrong about my claim that it goes faster.

Basically while the point at which audio transparency is achieved can vary between people if it can objectively be shown that two audio reproductions are the same then there is no point in arguing about the subjective part because your ears are hearing the same thing.

Chronon

January 7th, 2010, 08:17 PM

It can be shown that audio cables do/do not make any difference to the sound coming out of the speakers, there are objective tests.

Yes, but is your objective to run benchmarks on your equipment and to demonstrate measurable differences or to use it to produce a pleasurable listening experience?

I could design a telescope that efficiently captures light at multiple spectral bands at great expense. This will demonstrably outperform a simple, visible Cassegrain telescope under a variety of benchmarks. However, if I am intending to collect the output from the telescope with my eye then is the added expense worth it? The fancy, expensive telescope will not allow me to see anything more than the simple telescope designed to efficiently collect visible light. Is it reasonable for me to suggest that for the purpose of direct observation of the night sky, the simpler, cheaper telescope is just as good as the fancy, expensive one?

Xbehave

January 7th, 2010, 08:28 PM

I could design a telescope that efficiently captures light at multiple spectral bands at great expense. This will demonstrably outperform a simple, visible Cassegrain telescope under a variety of benchmarks. However, if I am intending to collect the output from the telescope with my eye then is the added expense worth it? The fancy, expensive telescope will not allow me to see anything more than the simple telescope designed to efficiently collect visible light. Is it reasonable for me to suggest that for the purpose of direct observation of the night sky, the simpler, cheaper telescope is just as good as the fancy, expensive one?
I think one of us has misunderstood phrostbyte's earlier posts, I understood his point to be that you can objectively show that cables make no difference*, so it would be like you being able to objectively prove that my telescope wasn't actually any better than yours even if i claim it is. Eitherway I think that being able to objectively show that one product/system is better than another is a prerequisite for it being subjectively better.

If A > B then test A subjectively, else A is subjectively not better than B

*he didn't actually say this but just that such a test would be possible.

CDs are not the highest quality to rip from. That's why they're thinking of making audio DVDs and even audio Blurays.
That is what you asked for citation on, that is what i was explaining, I doubt they will make much of a difference to everyday people but there is a reasn for them existing.

Chronon

January 7th, 2010, 08:32 PM

Because if a quanatative analysis can show that both monitors are the same, anybody claiming different is an idiot. If a quantative analysis can show that monitor A has a higher resolution than monitor B, then while it can be argued that nobody would notice, anybody claiming monitor B gave a better reproduction of images would be wrong.

Nobody has made this claim. You are constructing a straw man.

Because if i paint a car green and tell you it goes faster and you can show me it doesn't, I would be proved wrong about my claim that it goes faster.This is another straw man.

Basically while the point at which audio transparency is achieved can vary between people if it can objectively be shown that two audio reproductions are the same then there is no point in arguing about the subjective part because your ears are hearing the same thing.
I agree but this isn't what we have been discussing. We have been discussing whether, in the case where a particular observer cannot distinguish between two different apparatuses, objective differences in the output of the two apparatuses matters. How much would you be willing to pay for something you will never see or hear? If it's a lot then I have a great deal on an invisible bridge.

phrostbyte

January 7th, 2010, 08:42 PM

True, but its actually not that much bigger than CD-DA.

DTS Digital Surround = 1510k/sec.

CD-DA = 1411k/sec.

The only reason why you can make DTS Digital Surround play on normal CD players is the codec issue (like, it doesnt understand DTS). But the actual physical CD can hold more than enough data to fit albums of DTS Digital. Not understanding the DTS codec why most DVD-A discs are dual layer, layer 1 is red book CD-DA (to play in any CD player) and layer 2 is DVD-Audio.

Using .flac (or .wv) would be easier, they both support stupidly high sampling rates and multichannel.

The standard CD sampling rate is 44.1 kHz. Not sure what you mean by k/sec.

Chronon

January 7th, 2010, 08:44 PM

I think one of us has misunderstood phrostbyte's earlier posts, I understood his point to be that you can objectively show that cables make no difference*, so it would be like you being able to objectively prove that my telescope wasn't actually any better than yours even if i claim it is.
Yes. I did not understand phrostbyte's stance to be this at all. I agree with this point.

Eitherway I think that being able to objectively show that one product/system is better than another is a prerequisite for it being subjectively better.

If A > B then test A subjectively, else A is subjectively not better than B

People's subjective preferences often do not correlate well with the outcomes of objective benchmarks. Some people prefer headphones with a lot of low end, even if the frequency response curve is extremely non-flat. Some people consider factors unrelated to performance (like price, warranty, etc.). Better for you is not necessarily better for me, no matter how many benchmarks you show me. :)

phrostbyte

January 7th, 2010, 08:47 PM

It is certainly possible for such a discussion to happen. This isn't really my area of expertise so I'm not sure if I can participate if you wish to embark on a tangent focused on the details of audio compression algorithms.

Tangent? That's the whole point of the thread. All this other stuff is a one giant absurd tangent.

Chronon

January 7th, 2010, 08:50 PM

So, go for it. Tell me something interesting about audio compression methods (preferably lossless so we can avoid transparency issues, which are legitimately part of this thread).

Chronon

January 7th, 2010, 08:54 PM

The standard CD sampling rate is 44.1 kHz. Not sure what you mean by k/sec.

He (or she) meant kb/s (or kbps).

2 channels * 16 bits/sample * 44,100 samples/second = 1411.2 kilobits/second

phrostbyte

January 7th, 2010, 09:01 PM

He (or she) meant kb/s (or kbps).

2 channels * 16 bits/sample * 44,100 samples/second = 1411.2 kilobits/second

DVD audio is up to 5 channels, 24 bit depth @ 192kHz. As you can see, a lot more PCM data per second of audio.

pwnst*r

January 7th, 2010, 09:03 PM

Cool story, but multi-channel audio has nothing to do with quality. I think you missed that.

forrestcupp

January 7th, 2010, 09:07 PM

Multichannel audio has nothing to do with quality, but all the other factors do. Like most digitally recorded music (which is happening a lot more now) is recorded at least at 24-bit 48000 hz or higher. CDs can only do 16-bit 44100 hz. So they have to downgrade the original recording to work with a CD. Audio DVDs don't have that limitation. That's where the quality comes in.

cascade9

January 7th, 2010, 09:11 PM

The standard CD sampling rate is 44.1 kHz. Not sure what you mean by k/sec.

LOL. K/sec, you know, kilobits a second? But if you want to see the maths....

2 channels x 44,100 samples per second per channel × 16 bits per sample = 1,411,200 bit/s (http://en.wikipedia.org/wiki/Bits_per_second) = 1,411.2 kbit/s.

http://en.wikipedia.org/wiki/Compact_Disc_Digital_Audio

Its not just the sample rate, sample size is also important. Bitrate just shows you how much its been compressed, if at all. A 16 bit/44100kHz .flac can have a bitrate less than half the size of the original CD-DA file and be the same, due to compression.

This is why DTS is 'lossy'. Its 6 channels x 96,000 samples per second x 24 bits per sample, so it should be huge (really, really, Really huge). Its actually 1510k/sec because is compressed.......very heavily, and lossy-ly LOL.

phrostbyte

January 7th, 2010, 09:12 PM

So, go for it. Tell me something interesting about audio compression methods (preferably lossless so we can avoid transparency issues, which are legitimately part of this thread).

I already did write a large part of what I know about audio compression early on in the thread in a vain attempt to spur a more technical and worthwhile conversation. Of course that entire post was largely ignored. :D

phrostbyte

January 7th, 2010, 09:15 PM

LOL. K/sec, you know, kilobits a second? But if you want to see the maths....

http://en.wikipedia.org/wiki/Compact_Disc_Digital_Audio

Its not just the sample rate, sample size is also important. Bitrate just shows you how much its been compressed, if at all. A 16 bit/44100kHz .flac can have a bitrate less than half the size of the original CD-DA file and be the same, due to compression.

This is why DTS is 'lossy'. Its 6 channels x 96,000 samples per second x 24 bits per sample, so it should be huge (really, really, Really huge). Its actually 1510k/sec because is compressed.......very heavily, and lossy-ly LOL.

If DVD Audio is compressed, it's compressed in a lossless format only. I think either your 'maths' are way off or you are speaking about a different format.

phrostbyte

January 7th, 2010, 09:18 PM

Cool story, but multi-channel audio has nothing to do with quality. I think you missed that.

Cool story, look up sampling rate and bit depth. Also look up "Fourier" while you are it. ;)

phrostbyte

January 7th, 2010, 09:21 PM

Another thing worth looking up is discrete and continuous, as that ties directly into all of this. This book does a good job of linking the two: http://www.amazon.com/Concrete-Mathematics-Foundation-Computer-Science/dp/0201558025

cascade9

January 7th, 2010, 09:27 PM

If DVD Audio is compressed, it's compressed in a lossless format only. I think either your 'maths' are way off or you are speaking about a different format.

Nope, my maths is fine, and I know my lossy/lossless.

The _only_ Dolby and DTS formats that are lossless is Dolby TrueHD and DTS-HD Master Audio. Evidence? Posted before, but here it is again (and I can find lots of other junk on DTS and Dolby being lossy if you really want)-

Linear PCM (LPCM) - up to 8 channels of uncompressed audio. (mandatory)
Dolby Digital (DD) - format used for DVDs, 5.1-channel surround sound. (mandatory)
Dolby Digital Plus (DD+) - extension of Dolby Digital, 7.1-channel surround sound. (optional)
Dolby TrueHD - lossless encoding of up to 8 channels of audio. (optional)
DTS Digital Surround - format used for DVDs, 5.1-channel surround sound. (mandatory)
DTS-HD High Resolution Audio - extension of DTS, 7.1-channel surround sound. (optional)
DTS-HD Master Audio - lossless encoding of up to 8 channels of audio. (optional)

http://www.blu-ray.com/faq/#bluray_audio_codecs

See? only the 2 versions I mentioned before are marked as 'lossless'. There is PCM, but thats uncompressed, DTS and Dolby are compressed codecs.

phrostbyte

January 7th, 2010, 09:30 PM

Nope, my maths is fine, and I know my lossy/lossless.

The _only_ Dolby and DTS formats that are lossless is Dolby TrueHD and DTS-HD Master Audio. Evidence? Posted before, but here it is again (and I can find lots of other junk on DTS and Dolby being lossy if you really want)-

http://www.blu-ray.com/faq/#bluray_audio_codecs

See? only the 2 versions I mentioned before are marked as 'lossless'. There is PCM, but thats uncompressed, DTS and Dolby are compressed codecs.

I'm talking about DVD Audio. DVD Audio is a specific format that focuses on audio as a replacement for a CD. It offers lossless audio at significantly more PCM bandwidth then a typical CD. So perhaps your math is correct. Thus the latter part of the statement is correct because Blu-ray doesn't factor in this.

Chronon

January 7th, 2010, 09:36 PM

I found your initial post.

MP3 on the other hand is lossy. You'll notice that as you lower and lower the bitrate of an MP3, it gets closer and closer to approximating the sound of a sine wave, and low bitrate MP3s sound "crispy" (or sine-wavey).

This is because MP3 applies a discrete variant of the Fourier transform on the blocks of the PCM data before compressing it. Then using a quantum matrix it eliminates the high frequency coefficients from the transformed data, which in turn increases the waveform's entropy. Once the waveform has greater entropy it can compressed to a much smaller size using similar entropy coding techniques as FLAC. The disadvantage is the removal of high frequency components from the transformed waveform also reduces it's quality, which is how much a untransformed waveform can differentiate itself from being a sine wave.
This seems a bit too oversimplified. You didn't mention psychoacoustic models at all. The art of compressing mp3 (and many other lossy formats) comes down to psychoacoustic modeling of how human perception of sound works. This guides what data can be discarded while introducing artifacts that are imperceptible to most humans and goes far beyond simply discarding high frequency components. (Perhaps you didn't mention this because it has to do with subjective perception.)

The process that increases the entropy of the file also compresses it. They are not separate steps. Maximizing entropy == minimizing size == maximizing uncertainty in the value of each bit in the file (basic cryptology).

pwnst*r

January 7th, 2010, 09:36 PM

If DVD Audio is compressed, it's compressed in a lossless format only.

wrong.

Cool story, look up sampling rate and bit depth. Also look up "Fourier" while you are it.

lol, **** I know what sampling rates and bit depths are. And why should fourier even be a part of the initial discussion? Oh, that's right, because it's not.

Do you go and test your components before you buy them and make a decision strictly based on your measurements?

Didn't think so.

cascade9

January 7th, 2010, 09:39 PM

I'm talking about DVD Audio. DVD Audio is a specific format that focuses on audio as a replacement for a CD. It offers lossless audio at significantly more PCM bandwidth then a typical CD. So perhaps your math is correct. Thus the latter part of the statement is correct because Blu-ray doesn't factor in this.

Yes, I know. But its the same codecs they use on bluray as whats is used on DVD-A.

Just because the sample rate and sample size are huge, doesnt make it lossless.I actually own a few DVD-A discs, I do know what I'm talking about.....

pwnst*r

January 7th, 2010, 09:43 PM

Cascade, out of curiousity, which ones do you own? I have several myself (that I rarely listen to)

phrostbyte

January 7th, 2010, 09:46 PM

Now we are getting somewhere. :D

I found your initial post.

This seems a bit too oversimplified. You didn't mention psychoacoustic models at all. The art of compressing mp3 (and all other lossy formats) comes down to psychoacoustic modeling of how human perception of sound works. This guides what data can be discarded while introducing artifacts that are imperceptible to most humans and goes far beyond simply discarding high frequency components. (Perhaps you didn't mention this because it has to do with subjective perception.)

The process that increases the entropy of the file also compresses it. They are not separate steps. Maximizing entropy == minimizing size == maximizing uncertainty in the value of each bit in the file (basic cryptology).

You are right that it's too oversimplified. If it was complete, it would be many many pages long. :)

I disagree with your second statement. Entropy coding techniques such as (some examples include: Huffman coding or arithmetic coding) are a separate step in any compression algorithm. Usually you have a transformation step, and then an entropy coding step. In lossy algorithms you have a quantization step in the middle, which further increases entropy. This is part of the psychoacoustic model you speak of in the case of MP3.

phrostbyte

January 7th, 2010, 09:50 PM

wrong.

Another thing I'd like you to look up is Meridan Lossless Packing. But only after looking up Fourier transforms. :)

lol, **** I know what sampling rates and bit depths are. And why should fourier even be a part of the initial discussion? Oh, that's right, because it's not.

Do you go and test your components before you buy them and make a decision strictly based on your measurements?

Didn't think so.

It's quite difficult to have an intelligent discussion about audio quality without bringing up bit depth or sampling rates. It's like bringing up a image quality discussion without bit depth or resolution. Let alone the lossy effects of a quantified discrete cosine transform, which despite you not understanding what that is, has a profound impact on image quality.

koleoptero

January 7th, 2010, 10:02 PM

Call me an elitst, but I find it irritating when someone goes 'I'm not sure what P(t) and "fourier" mean.' while actively debating sound quality. It's like someone debating bridge engineering and saying "I don't know what calculus mean.". "I'll just let my eyes decide."
I know quite a lot about fourier analysis and still rely on my ears when I listen to music.

CDs are not the highest quality to rip from. That's why they're thinking of making audio DVDs and even audio Blurays.

If you get a FLAC from the original recording, that's the best sound value. Then, encoding that into a quality level 10 OGG is the best way to get a high quality lossy audio file. It'll be smaller than the equivalent MP3, also.

DVD audio has been deprecated already in favour of the backwards-compatible super-audio CD (SACD) which has a superior sound. (check it here) (http://en.wikipedia.org/wiki/Super_Audio_CD)

But what are you all talking about in this thread? It's gone far away from the "mp3 vs flac vs cd" topic.

phrostbyte

January 7th, 2010, 10:09 PM

I know quite a lot about fourier analysis and still rely on my ears when I listen to music.

Me too. But I don't rely on my ears to debate audio quality. Because I don't own your ears and I am not going to judge what you hear.

pwnst*r

January 7th, 2010, 10:19 PM

Me too. But I don't rely on my ears to debate audio quality. Because I don't own your ears and I am not going to judge what you hear.

We were "debating" interconnects and you come bounding in with your "omg i has measurements and graffs!"

http://i48.tinypic.com/2q2e0rl.jpg

koleoptero

January 7th, 2010, 10:22 PM

Me too. But I don't rely on my ears to debate audio quality. Because I don't own your ears and I am not going to judge what you hear.

Still the best way to compare the quality of music reproduction between formats/speakers/amps/whatever is the double blind test.

And between lossy formats actually you don't need to go that far. Just ask the developers themselves, they'll tell you what's best (if you don't trust your ears that is).

But to compare the output of two different speakers you'd need to sample it, hence microphones, which means more loss. It just doesn't work.

phrostbyte

January 7th, 2010, 10:23 PM

We were "debating" interconnects and you come bounding in with your "omg i has measurements and graffs!"

http://i48.tinypic.com/2q2e0rl.jpg

Yeah I'm sorry to ruin your audiophile cable debate and derail the thread into real scientific discussion. Carry on pwnst*r. :|

phrostbyte

January 7th, 2010, 10:25 PM

Still the best way to compare the quality of music reproduction between formats/speakers/amps/whatever is the double blind test.

And between lossy formats actually you don't need to go that far. Just ask the developers themselves, they'll tell you what's best (if you don't trust your ears that is).

But to compare the output of two different speakers you'd need to sample it, hence microphones, which means more loss. It just doesn't work.

I don't disagree at all.

marco123

January 7th, 2010, 10:30 PM

ratcheer

January 7th, 2010, 11:15 PM

You can already download 24 bit FLAC from specialist retailers, like Linn: http://www.linnrecords.com/catalogue.aspx?format=studio

Hey, what a neat site. I have been a fan of Linn since the mid-80's. I had no idea they were still going strong. Thanks!

Tim

marco123

January 7th, 2010, 11:52 PM

Hey, what a neat site. I have been a fan of Linn since the mid-80's. I had no idea they were still going strong. Thanks!

Tim

Yep, still going strong. They've actually just decided to stop production of CD Players in favour of going down the Hard Disk Player route, with products similar to the Naim HDX. They still make excellent components.

Gizenshya

January 10th, 2010, 08:21 AM

Do you go and test your components before you buy them and make a decision strictly based on your measurements?

Didn't think so.

pwnt. /thread

seriously, though. Can the cables effectiveness be measured? Yeah. I strongly oppose any notion of not isolating the cable... that just introduces all sorts of factors that would only invalidate any results, however measured.

And 2, if you were desperate to do it without isolating the cables, you would have to go through an extremely complicated and comprehensive regression analysis to get the r^2 value for the cable... which would certainly (read: necessarily) have a much larger margin of error than a similar isolated cable test. So it would all be pointless.

So yeah, it can be measured. OK, now what? Does that mean anything? Perhaps. The whole point is perception. I've never heard of any real (valid, reliable, and peer-reviewed) study confirming a significant coeffecient of correlation of the effect of cables on audio quality. Obviously, I mean as judged by humans. I haven't searched too hard though, so if someone finds one, cite it please. I'd like to take a look.

Also, in this thread, one side is saying "it can be measured," while the other is saying, "no, you can't tell the difference!" (In case you missed that, the two sides are talking about different things. ... yet another reason why there can be no meaningful end.) Mostly, anyway. Some are saying that they (or people with trained ears, or blah blah) can tell the difference-- in other words, that humans can tell the difference. which brings me to my next point...

And all the rest of my post means is that, even if they were both talking about the same thing (whether people can perceive the difference), there would still be no possibility of a meaningful end. (Any concensus would be meaningless.) Why? Because, to the best of my knowledge, there isn't any verifiable evidence either way (about perception). It would all come down to who could word things better, or who screwed up first, or who could post the best boobie pics, and any conclusion based on those would be invalid and unreliable (for the purpose of the discussion). Since there is no valid study either way (again, correct me if I'm wrong), neither side could possibly come up with a valid, reliable arguement. So, unless someone actually brings up a study, there isn't really anywhere to go with this discussion. As much as some may want to believe (whatever) is or is not effective... their belief is just that -- belief -- without evidence.

Alternatively, of course, someone here can conduct such a study. Then after peer review and publishing, we can continue the discussion.

That being said, if you want to continue, be my guest. Give me time to heat up another bag of popcorn, though. I finished the last one sometime during page 3. ;)

starcannon

January 10th, 2010, 10:15 AM

Hello,

I just couldn't find much difference in sound quality of a Mp3, flac and CD. I have not tried Vinyl.

Do you think that there is huge recognizable difference? Are you an analog listener?

Thanks,
SK.
The general rule is, the closer you get to the source, the cleaner the music will be. Vinyl trumps all(at the consumer level), truely uncompressed music; down side is, without a great turntable, great needles , and immaculate discs, it can be hissy, scratchy, and poppy.

--Marantz 2270 Receiver/Amplifier, Saunsui Speakers--

Miguel

January 10th, 2010, 11:43 AM

I feel I'm the only one that doesn't see vinyl as the ultimate music format. Some guys at the hydrogenaudio wiki apparently agree: link (http://wiki.hydrogenaudio.org/index.php?title=Vinyl_Myths). Of course, this doesn't mean I'm right. In any case, I do perfectly understand the people that prefer vinyl.

Alternative formats, such as DVD-A and SACD have already been mentioned in this thread. Personally, the only SACD I've had in my hands is Pink Floyd's Dark Side of the Moon... and only listened to the CD part, since I don't know anybody who owns a SACD player. I fear, however, that these high-quality formats will disappear. For one, most CDs today are, from a technical point of view, worse than the CDs produced 15 years ago. Furthermore, the music people buy is in a lossy format.

Is it possible to convince a reasonable amount of teenagers that music has to be experienced the way most of the audiophiles in this thread prefer? Or will they go on with tons of 128kbps MP3s in their iPods with low-end earbuds?

Let's look at the bright side of it: the first time they attend a live concert they'll be blown away. Heck, I'm blown away every time I hear the Bilbao Symphonic Orchestra, a second rate orchestra, play live.