Uncompressed v. Lossless Compression

[The Computer Audiophile]

Let's not have a big argument over this one, hence the nice little songbird listening to music. I think we can all benefit from a good conversation about this topic.

 

What format are you guys using? Uncompressed (AIFF or WAV), Apple Lossless, FLAC, WMA, etc...

 

Have you guys tried the other formats? Any likes or dislikes associated with each?

 

This is just a laid back discussion about file formats and music quality. I don't care if you're listening to 8 bit mp3 files, I'm just curious.

 

I'll start this one off. I am using uncompressed AIFF files for all the music I rip from redbook CDs. Whether it is all in my head or not, I have an issue with compression. Over time I get fatigued quickly by compressed music. An AB test reveals no differences to me, but I don't listen to music in an AB way. I had all my music in Apple Lossless, but over the course of about a month I was not happy with it. Call me crazy, some in the industry have, but all that matters to me is what I like the best.

 

[markr]

... I cannot tell the difference between FLAC and uncompressed audio. I do NOT have a lot of material under my belt to be able to make what I would be comfortable with saying was "my opinion" however. I've always chosen to rip CD's to .wav or .aif . I have no experience with vinyl to digital. Yet. I should probably just delete this post and wait for others to chime in, but I would like say here that I would really like to hear what everyone else has to say on this!

 

My dirty little secret: I have over 300 GB of pop music in MP3 format. Most is at 256Kbps or over. I had to have something to travel with me and for listening to at work - though I don't actually get much time to listen there - . Most of this stuff I already have in CD, LP, or mag tape formats. Sniff. Please forgive me! -- Would it help if I said here that I do get 'fatigued' listening to MP3's over long periods?

 

markr

"I've got a google-phonic stereo with a moon rock stylus! -- ..... sounds like SH*T!" - steve martin

 

[Innertuber]

Well, I am trying AIFF but am not really sure why I'm not using .wav for now. I would go with FLAC if iTunes would.

 

[The Computer Audiophile]

Hey Innertuber - I chose AIFF over WAV because of AIFF's support for album art. I have also heard some good things about the inner-workings of AIFF files as opposed to WAV. If they sound the same why not chose the one with more "extras"?

 

FLAC is a serious contender I agree.

 

[The Computer Audiophile]

markr - One of my thoughts has always been "Why rip something less than the whole uncompressed CD if that is what you purchased?" I agree with your choice of wav & aiff.

 

I think everyone, or almost everyone, has some pop songs in their collection. Hell, I have the new Britney Spears album Blackout. It is right in between Jay-Z's The Black Album and Alana Davis' Blame it on Me. It's entertaining and that is fun.

 

[Innertuber]

Yah, I agree. One reason for WAV tho seems that if down the road you want to use a compression or conversion tool WAV will likely be a supported format. AIFF might be. I could be wrong there. At this point, for me to change is just some time since I have all my cds.

 

Seems I thought my iTunes WAV files grabbed all the art. Am I mixed up there?

 

Over the weekend I crammed Clapton's Pretending into iTunes using about 10 different format/bitrate combos. Know what I learned? Me either ... just kidding. I was suprised that even high compression, real high compression didn't sound a whole lot different. Now I wasn't comparing running into a DAC. See other post about my new system which uses a state-of-the-art RCA cord plugged into the headphone jack. So, that's part of the issue. The other thing I believe is the fidelity, or lack thereof certainly depends on the system. I was using a decent stereo receiver, but running through some small infinity bookshelf speakers w/o a sub. I was suprised that I had to get to rediculous compression levels to notice much effect. I tested my wife on all the differences too. I was suprised. I guess my conclusion was that if you have an average system, and run from the headphone jack that your probably not giving up much to some compression, but you get a lot more songs on your iPod.

 

[The Computer Audiophile]

Innertuber: "Seems I thought my iTunes WAV files grabbed all the art. Am I mixed up there?"

 

If iTunes can get the album art from an online source you are good. You just can't add it manually to albums.

 

As far as your compression experience goes, that's cool. Whatever sounds good! No old curmudgeon, arrogant audiophiles around here. Thanks for sharing your findings. Have you tried listening over longer periods of time? That is where I really notice a difference.

 

As I said earlier, if 8 bit music sounds great for someone then who cares.

 

[markr]

Chris wrote:

(snip) "Hell, I have the new Britney Spears album Blackout. It is right in between Jay-Z's The Black Album and Alana Davis' Blame it on Me. It's entertaining and that is fun." (snip)

 

... Jeeeez Chris. When I said "pop music", I meant 'non-classical'. Like Robert Johnson, Cream, Jimi Hendrix or John Mayall !

 

lol - I too am with you: If it sounds good to you, then it IS good. g-r-i-n.

 

markr

 

[The Computer Audiophile]

So you're saying your pop isn't like my pop? :-)

 

[markr]

IS the spice of life! I'm (almost) always up for something new.

 

BTW re: the .wav vs .aif thing - There are software apps out there that will freely convert between the two formats. I have used them successfully in the past - I started my computer audio journey in the early 1990's with an Atari Falcon (68030 based machine). The audio format of choice there was .aif when used with Steinbergs then-new DAW software: Cubase. When Atari died in the mid 90's I HAD to move to PC. I didn't want to take a chance on Apple at that time, even though my known and trusted music software programmers seemed to gravitate mostly towards Apple... - I and my considerable investment in hard and software had just been 'orphaned' by Atari and didn't want that to happen again. The quality audio format of choice for PC is .wav so I needed a way to get stuff out of one format into the other. I was satisfied with the resulting conversion files. If anyone is interested I could look this stuff up again, but it can easily be found by Googling "aif to wav" or something similar.

 

WOW! I just looked at one of the ads on the site here while typing this & what do my eyes behold but "SWITCH" free audio converter software! See? if you hang around this site enough, you will 'get what you need'. Try sometimes...... heheheheh

 

markr

I may be wrong. But I'm not wrong long......

 

[Lord Chaos]

Oh, yes, the format wars. I've heard a lot of this. "You can't hear the difference between AIFF and MP3, so just save space." And on and on.

 

When I started ripping CDs to the computer I already knew I didn't like MP3. I'd heard enough samples. Now I know that MP3 varies all over the place, but that early version of Itunes would rip in AIFF so that's what I did.

 

From what I've read, the data in AIFF and WAV is the same, but they come in different "wrappers." When the time came to move my whole library to the PC I expected Windows Media Player to have one of those formats, but no score. They offered something called "Windows Lossless" and their WMA. I chose the former and it seemed to be just fine. It's Lempel-Ziv type compression, replacing long strings of repetitive data with a more efficient coding. It's just like ZIP compression, which obviously works perfectly for reconstructing programs which have to have every bit in the right place.

 

Then I decided to move my library to a Mac. I found that Itunes could read WMA lossless and write Apple Lossless so I converted a CD. It didn't work, the playback being slow. So, I just reripped all my CDs, taking advantage of Itunes' link with CDDB to correct all the tagging mistakes I'd made the first time around.

 

Digital music files have to be read and processed in order to be used. Expanding the ZIP-type compression is just part of that process and I would expect it to be transparent. That's my ears, though. I'm pleased with what I have, and glad I can fit my library into 420GB instead of 800. I can hear the differences in how CDs are mastered so there's no lack of resolution, but that doesn't get into the fatigue aspect. That's how I tell the difference between a lossy format and lossless: the music is there, basically, in a lossy format but it's dead. No sparkle.

 

And yes, I do have a weakness for some pop music. I like Neil Diamond, Dan Fogelberg, and I even have a Captain and Tennille CD. Right now I'm listening to Dire Straits.

 

[Wavelength]

Gang,

 

If you have a slower computer then uncompressed will yield better sound stage. The great thing is that if you have lossless files you can easily convert them to AIFF or WAV and hear a difference.

 

The good thing is that Lossless means just that. The bad news is that this test may show your system is too slow to uncompress the sound and present it too the dac in time.

 

We tested this at CES and it was pretty apparent on the slower older mac mini. The newer 2.2 Core Duo MacBook was less apparent.

 

You can play flav via MAX and VLC I think on a mac.

 

Thanks

Gordon

 

[markr]

Thank you Gordon. You've just crystalized the true meaning of some of the (minor) issues that I've been seeing with FLAC on the Mac. When I take the songs that are at issue when played back as FLAC in VLC and convert them to AIFF, viola' - no issue. Still great sound though. I really hadn't been thinking that the computer was doing double duty decompressing the FLAC and then pushing the converted file to the DAC. It makes sense now.

 

I still wonder though if VLC, being a work in progress, might still be optimized to better handle FLAC. I suppose that I will have to go look at a couple more of the Mac FLAC players out there for grins and giggles.

 

markr

 

[The Computer Audiophile]

Hey LC - I just finished listening to Dire Straits - On Every Street. Great album and sounds great too.

 

[sputnik]

iTunes gives me an option to convert my apple lossless files to AIFF, which I might try despite the resulting larger file size based on the comments Ive read here.

 

Interestingly if I take a crap 128kbps track and covert it to AIFF, it also expands the file size immensly. Question is, if the source file is heavily compressed, am I right to assume that such a conversion to AIFF would really serve no purpose at all?

 

[The Computer Audiophile]

sputnik: "Question is, if the source file is heavily compressed, am I right to assume that such a conversion to AIFF would really serve no purpose at all?"

 

Exactly.

 

I suppose you could split hairs here and say that by uncompressing it your computer won't have to work as hard doing the "decompression" and could output better sound than the highly compressed version. But, I don't want to even open that can of worms.

 

Plus, as I always say, "If it sounds good to you, then it sounds good." If Apple Lossless if good to you then no need to convert to AIFF etc...

 

Very popular question around the Internet though.

 

[Lord Chaos]

Converting an MP3 to any other format, at best leaves you with the same file. At worst, you will degrade the sound further. You can never get better than what you start with.

 

Beyond that, I'm not sure. I don't really know how MP3 works. Compressing a sound file to MP3 involves throwing away most of the data, presumably what "they" say we won't hear. I'm not sure what data are thrown away, and I'm equally unsure as to how the loss is made up in playback. Surely they have to pad it out to 44.1kHz, 16-bit sound somehow.

 

The analogy I use is in JPEG compression of images. Never save a JPEG again as a JPEG, as you will throw away more information. If you want to do further processing, save the intermediate steps as TIFF or some other uncompressed format. I suppose the same idea applies to editing an MP3, but I'm not sure of the specific characteristics.

 

[sputnik]

Thanks, the JPEG analogy works nicely in my mind at least.

 

[markr]

Your math is good, by everything my ears tell me ..... eventually - meaning that I do detect 'fatigue' in my hearing after more than an hour or two of listening to 'lossy' compressed audio. There are dependencies and variables that will change that fatigue time frame though.

 

We started out on this thread by wanting to compare Uncompressed Audio sound to the sound of 'Lossless' Compression. We are now on a slightly different tack: converting Lossy 'Compressed' Audio to 'Uncompressed' Audio.

 

The common thread between the slightly different subjects though, would seem to me to remain in the understanding of what all the types of 'Compressed' Audio are. But I'm too lazy (and tired) to type all that right now, SO

 

I hate to keep using 'the wikipedia' because there are so many other sources on the web and in print, but is just easier (especially in this instance), and it is pretty-well kept honest by others with divirgent opinions on the web. Here is a fairly full and accurate comparison of audio compression types. Go there if you care!

 

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

 

markr

"I've got a google-phonic stereo with a moon rock stylus! ..... sounds like SH*T! - steve martin

 

PS: The one reason I've known of to 'support' the conversion of MP3's or other lossy-compression-audio to a 'Redbook-style' CD was this: to allow the play back of songs that were available only in the lossy format to beging with, when all one has available for playback is a machine that plays back only Redbook CD's. It has never been a satisfactory method of listening to a piece for me. I've always preferred the sound of the original MP3 to the converted-to-CD file in these cases.

 

[Lord Chaos]

You wrote "I've always preferred the sound of the original MP3 to the converted-to-CD file in these cases."

 

This is interesting. Given that the MP3 must be converted to "regular" 16/44.1 audio for playback, just as it is when you make it into a Redbook CD. Do you have any ideas about where the perceived difference comes from? Perhaps the DAC in the CD player? Or maybe the direct playback skips the conversion to 16/44.1 and goes straight to analog. Details, details. (we need a smiley to put in here.)

 

Well, I read the Wikipedia article... and it didn't cut through much of the fog. I still don't understand how a compression algorithm can sense when one sound is masking another. I can hear it; I can hear the subtle sound of one instrument below another, but I don't understand how the software does it. The sound wave a microphone sees is a continuous pressure differential.

 

[markr]

... the lossy compressed file has been through "noise shaping" (the link for that is in the text of the link @ wiki I provided), among some other processes during it's conversion to become an MP3. This, of needs, removes actual audio data from the original audio. --- It is GONE into thin air with regard to this MP3 file now --- When you try to reverse the process by trying to return the MP3 data to something resembling the original audio (upsample - sort of....) that WOULD have been in a Red Book CD of that audio, how is that removed data to be known then, to the 'upsampling' program? It cannot actually 'know' what that missing data is. So..... it interpolates data in, that might have been there (this is simplified somewhat) to fulfill the Red Book criteria. More 'noise shaping', just in the opposite direction that was done before. This being done WITHOUT any real point of reference to the original audio data. Now you have audio data that has been changed on 'the way down' and again 'on the way up'. It just never works out IMO. Therefore: MP3--->Red Book CD never equals the original MP3 because it also has 'unnatural' audio data added in to boot. (ICK)

 

markr

 

OOPS! Missed the actual meaning of this segment in your original post when I responded the first time (sorry) :

 

"This is interesting. Given that the MP3 must be converted to "regular" 16/44.1 audio for playback, just as it is when you make it into a Redbook CD. "

 

Nope. That is not what happens when you play an MP3 back. Let me see if this quote from the article helps here:

 

Because data is removed during lossy compression and cannot be recovered by decompression, some people may not prefer lossy compression for archival storage. Hence, as noted, even those who use lossy compression (for portable audio applications, for example) may wish to keep a losslessly compressed archive for other applications. In addition, the technology of compression continues to advance, and achieving a state-of-the-art lossy compression would require one to begin again with the lossless, original audio data and compress with the new lossy codec. The nature of lossy compression (for both audio and images) results in increasing degradation of quality if data are decompressed, then recompressed using lossy compression. - wikipedia 'audio compression (data)'

 

I hope all that helps.

 

 

[The Computer Audiophile]

I had a nice discussion with a very high end manufacturer last night about Uncompressed v. Lossless Compression and he confirmed my thoughts exactly. He said they have run tests with their equipment and can easily notice differences between the two. He uses uncompressed AIFF for critical listening.

 

[markr]

Yes, but did the equipment express a preference for the sound of one type over the other?

 

markr

 

[rom661]

For what it's worth I finally got around to comparing AIFF to Apple Lossless. A really valid question would be why I didn't do that before ripping all this material to my drive.

 

I'm going to describe the system and this is not about "this is what you should have, my system is bigger than (smaller than in the case of the horn fans) yours" or any of that. I just want to provide some context. I do consider it a very high performance system, as well as being very musical, not always the case, and one capable of revealing small differences. Keep in mind that whether or not some small differences contribute to the musical experience is open to debate. To me some do, some don't. If a gnat flew through the studio during final mix, I don't have a compelling need to be aware of it.

 

The system is a MacBook with 2 500GB external drives, one shadowing the other. The bitstream is sent to a Benchmark DAC1 USB via the provided USB cable. Output of the DAC1 is through a balanced pair of Audioquest's new Signature interconnects to an Audio Research Ref 3 preamp. Then a 7 meter pair of Audioquest Sky balanced interconnects into a pair of Classe Audio CAM-400 monoblocks with Audioquest Everest speaker cables into a pair of Vandersteen 5A speakers. I clean my ears with Johnson and Johnson Q-tips. I find the generic ones to be too stingy with the cotton. That last one was a joke (but true).

 

My CD player varies between an Audio Research CD7 and a Classe CDP-202. These are both excellent CD players and are kind of what I used for a reference. After comparing sources and transport/server quality using the digital outs on the CD players, the Benchmark is definitely the weak link in the system. I think it is an excellent value, just not as good as the other, much more expensive components. I suspect the differences I heard would be more pronounced with a higher resolution DAC.

 

Long lead-in to a short conclusion. AIFF is better. One of the tracks I used was Patricia Barber's cover of The Beat Goes On. The difference that jumped out was the attack on the upright bass. The body was there but that wonderful feel of finger plucking string was diminished to a fairly large degree and I missed it musically. I listened to a number of other recordings and it was this "leading edge" of transients that I found most obvious. Those with phase correct speakers, such as the Vandersteens, and some of the better horns would probably find it more obvious since I feel they both do the wave launch thing better than other types of designs. I also heard fairly small differences in imaging and soundstaging, which is logical if you are talking about very small resolution differences.

 

I suspect it would be less of an issue in some systems, and I do not mean to imply that my impressions are gospel, but it was enough to convince me. I can easily see myself using Apple Lossless for lesser recordings that are already more compromised and compressed and saving the space for AIFF for better recorded stuff. I listened to an old Crowded House recording and although there was a difference, I wouldn't walk across a room for it. It also didn't stop me from dancing to it. This is not a sight you would want to witness.

 

Rick

 

[markr]

.... I guess that I really AM hearing a difference between the FLAC files for the Eagles download and the converted-to-.aif sound files. I think it was Gordon from Benchmark (?) who posted on CA about the fact that lossless files have to be first decompressed, or 'flattened' by the computer (while playback is going on) before the audio can be sent to the DAC. This works the processor more and therefore modifies the bit flow put to the DAC. The slower the computer, the more dramatic the effect. ... he said something to that effect anyway.

 

One thing I have to know though: tell me, do you use hydrogen peroxide or baby oil on your J&J Q-tips? 8^)

 

markr