[Icenine2]

Storage really shouldn't even matter these days.  It's so cheap.  The only worry is when we switch over to more expensive SS drives.

 

[FlyingBear]

With respect (really!), not quite. FLAC is an audio-aware algorithm: the compression algorithm is designed and tuned to work well with audio data. It doesn't actually "know" that it's audio, but it tends to function effectively, with about 50% compression, on data that turns out to be audio. Audio data has certain characteristics that result from it being something that humans can actually hear; its patterns are limited, and somewhat predictable. FLAC runs even better on spoken word recordings than music. If you try to run the FLAC algorithm on, say, a PDF document, it won't work well on it. The actual FLAC algorithm is very subtle, very clever and surprisingly effective. Wikipedia has a good technical write-up, but the math and terminology require a fairly deep knowledge of compression algorithms.
 
FLAC is also designed to be decoded quickly, and to allow streaming. There are compression algorithms that require the decompressor to see the whole file before being able to decode any of it, for example. With today's disk speeds and memory sizes, and CPU speeds, that's less of an issue, so a FLAC replacement could allow more compression.
 
Compression is cool stuff (for a geek). Just ONE of the features that FLAC uses is run-length encoding. In its simplest form, instead of having 20,000 zero bytes in a silent passage, you can have a handful of bytes saying "the next 20,000 bytes should zero, Mr. Decompressor". Most compression algorithms use this, because it's a quick win that works on all data: JPEG and ZIP are two examples. FLAC uses other, much harder, techniques such as linear prediction and Golomb-Rice coding, for stuff that's harder than "20,000 zero bytes". Wonderful stuff.
 
A WAV file is almost 100% pure audio data, and is in a format that is very little changed from the raw sample values of digital audio.
 
Again, no disrespect intended. I know almost nothing about tubes and Woo amps, which is why I hope to learn from you and others, and why enjoy reading this thread so much. I can't wait to get home and unpack my new Woo amp in 8 days' time! I hope that I can return the favor of your knowledge with some of my own.
 
Quote:

From how I understand it, the way that FLAC and ALAC cut the storage requirements in half from that of AIFF and WAV files without compressing the music is by getting rid of all of the non-audio information during the conversion process. AIFF and WAV let the computer read the audio data more efficiently, but FLAC and ALAC re-code that data to use much less storage space.

 

 

[grokit]

Very interesting FlyingBear, thanks for clearing that up 

 
Seriously, I guess the run-length encoding is what I was referring to as far as eliminating the non-audio data.
 
It sounds like you might have hit on why FLAC may have some superiority over ALAC with the easy-to decode part, as Apple has that extra (unactivated but it's there) DRM data and all.
 
 
Quote:

Storage really shouldn't even matter these days.  It's so cheap.  The only worry is when we switch over to more expensive SS drives.

Spinning platters rule! WD just came out with a new 3tb external My Book drive for $250.

 

[Silent One]

Quote:

Ah, a computer and information science point here: WAV does not have more information than ALAC or FLAC. It occupies more bits, but that's because the density of the information is lower. If you ignore the additional information that an ALAC or FLAC has, purely because the definition of the container (file) for the data allows for tags, the information (music) content of a WAV, ALAC and FLAC is identical. The proof is that you can convert endlessly between all three formats, and the bits representing the music in each of the three formats won't change. No information is lost.
 
You may ask yourself: how can the files always be smaller and still have the same information? The answer is that they can't. There is NO lossless compression method that can take any arbitrary input file and always create a smaller file. Lossy compression? Sure. Lossless? No way. It takes 5 minutes to prove this with simple math on a whiteboard. The 5-second version is that there are fewer possible files if they're smaller (there are 2 to the power n possible files of n binary bits size), so the decompressor wouldn't know which of the many larger files is the right one.
 
FLAC and ALAC work because they're dealing with inherently inefficient data representation in WAV (or AIFF....same thing, different clothes). If you took a set of random bits and threw them at FLAC or ALAC, they'd fail, or "compress" the file to one that's larger than the original. Most compression algorithms are smart enough to declare defeat and just give you back the original file.
 
An an aside, when I was a venture capitalist (halcyon days), I saw a number of companies promising lossless compression of all arbitrary files. One even claimed that you could use their technology to compress a 600MB CD to 60KB, then again to 6 bytes, and then reverse the process losslessly; THAT was an uncomfortable meeting. I turned them all down, because they were wrong, sadly misguided, nuts, or worse. Anyone with college-level information science can do the 5 minute proof mentioned above. And, no, this isn't like the people who "proved" that you can't fly faster than sound; the math on compression is much more solid than that: Claude Shannon, one of the fathers of information theory, was a very smart dude.
 
Lossy compression has its place. The HD movie you watch on your TV is missing over 99% of the bits in the data stream. If you've worked in compression, you can see the compression artifacts (and hence working in compression spoils your enjoyment of TV forever), but it's amazing how good it looks. JPEG performs similar, although less severe, miracles on pictures. Most MP3s throw away 90% of the information, but still sound ok to many people.
 
All the above assumes that the FLAC/ALAC software you're using isn't buggy and doesn't destroy your data. A few tests of compress/decompress and comparing against the original should convince you.
 
Is that enough talk to talk you out of WAV and embrace the joys of tagged files, Mike? And, to repeat a prior post: if WAVs make you more comfortable, and sound better to you, go for it. I'm not snickering behind your back. There are enough unknown variables in how we perceive sound, both psychological and physiological, to make your perceptions real and something that I honor and respect. You can always tag the files with "sidecar" files (same name, different extension) or with music software that tracks all of this in a database, like J River. It's safer and easier to manage if the tags are in the files, but some software doesn't keep all of the tags in the files. For example, iTunes doesn't keep ratings or last played date in the files. You can torture it into doing so with scripts and using/abusing other tag fields, but it's not convenient. As always, no perfect answer, and lots of tradeoffs....
 


 

You're right, this isn't Computer Audiophile yet. But contributions like this should be welcome. Sometimes, Woo Audio amp owners want just enough information to make their 'Woo' do what it do! I'm all for sitting back, enjoying the music. But sometimes the enthusiasts inside us comes out. And we have a desire to understand the systems and music environment so our Woo can be better placed in it. And hopefully perform a a higher level.
 
On a side note, with a moniker like "FlyingBear" did you pilot your way to an IOC meeting or flew commercial?  

 

[jerico]

Ah, a computer and information science point here: WAV does not have more information than ALAC or FLAC. It occupies more bits, but that's because the density of the information is lower. If you ignore the additional information that an ALAC or FLAC has, purely because the definition of the container (file) for the data allows for tags, the information (music) content of a WAV, ALAC and FLAC is identical. The proof is that you can convert endlessly between all three formats, and the bits representing the music in each of the three formats won't change. No information is lost.
 
You may ask yourself: how can the files always be smaller and still have the same information? The answer is that they can't. There is NO lossless compression method that can take any arbitrary input file and always create a smaller file. Lossy compression? Sure. Lossless? No way. It takes 5 minutes to prove this with simple math on a whiteboard. The 5-second version is that there are fewer possible files if they're smaller (there are 2 to the power n possible files of n binary bits size), so the decompressor wouldn't know which of the many larger files is the right one.
 
FLAC and ALAC work because they're dealing with inherently inefficient data representation in WAV (or AIFF....same thing, different clothes). If you took a set of random bits and threw them at FLAC or ALAC, they'd fail, or "compress" the file to one that's larger than the original. Most compression algorithms are smart enough to declare defeat and just give you back the original file.
 
An an aside, when I was a venture capitalist (halcyon days), I saw a number of companies promising lossless compression of all arbitrary files. One even claimed that you could use their technology to compress a 600MB CD to 60KB, then again to 6 bytes, and then reverse the process losslessly; THAT was an uncomfortable meeting. I turned them all down, because they were wrong, sadly misguided, nuts, or worse. Anyone with college-level information science can do the 5 minute proof mentioned above. And, no, this isn't like the people who "proved" that you can't fly faster than sound; the math on compression is much more solid than that: Claude Shannon, one of the fathers of information theory, was a very smart dude.
 
Lossy compression has its place. The HD movie you watch on your TV is missing over 99% of the bits in the data stream. If you've worked in compression, you can see the compression artifacts (and hence working in compression spoils your enjoyment of TV forever), but it's amazing how good it looks. JPEG performs similar, although less severe, miracles on pictures. Most MP3s throw away 90% of the information, but still sound ok to many people.
 
All the above assumes that the FLAC/ALAC software you're using isn't buggy and doesn't destroy your data. A few tests of compress/decompress and comparing against the original should convince you.
 
Is that enough talk to talk you out of WAV and embrace the joys of tagged files, Mike? And, to repeat a prior post: if WAVs make you more comfortable, and sound better to you, go for it. I'm not snickering behind your back. There are enough unknown variables in how we perceive sound, both psychological and physiological, to make your perceptions real and something that I honor and respect. You can always tag the files with "sidecar" files (same name, different extension) or with music software that tracks all of this in a database, like J River. It's safer and easier to manage if the tags are in the files, but some software doesn't keep all of the tags in the files. For example, iTunes doesn't keep ratings or last played date in the files. You can torture it into doing so with scripts and using/abusing other tag fields, but it's not convenient. As always, no perfect answer, and lots of tradeoffs....
 


 

Outstanding post, Flying Bear! Thanks!

 

[Icenine2]

FWIW  I rip almost everything in wav.  Some FLAC

 

[Icenine2]

Whoa.  Check out the giveaway on Twitter posted here!!!  LCD's and an ALO amp.  http://www.head-fi.org/forum/thread/561847/your-chance-to-win-a-set-of-lcd-2-s-and-the-alo-rx-amp-from-gizmodo

 

[Xcalibur255]

Quote:

If you're using the analog outputs of your sound card, I'd expect it to sound different in a new PC: different electrical noise. A PC is a bad place to have analog stuff going on; it's really electrically noisy inside the Faraday cage that most PCs surround themselves with. If you're using the digital outputs, I would also expect drivers and overall PC hardware to have an effect, especially on timing. The bits are almost certainly the same, but they may be arriving on a different schedule. This isn't computeraudiophile (yet), but there's real electrical science here to explain the issues, versus, ahem, subspace manifold verteron particle interference

.
 
A well-clocked and isolated USB DAC, perhaps of the asynchronous variety, should sound really **** good. It's largely a question then, I think, of the difference between your USB DAC and the DAC that lives inside or next to your CD transport. It's SO hard to do an A/B comparison. I've tried it with my W4S DAC-2, comparing the same CD played from a Theta David II (which can tolerably claim to deliver the right bit on the right schedule) through the coax PCM input and a ripped version playing from Pure Music through the USB input. There's no obvious difference to me, but there certainly might be to others. Even level matching accurately is a challenge, and we all know that louder often sounds better. Harmless fun to try all this out though, isn't it?

Yeah, I could understand if it were analog, in fact I would accept that very easily.  But to attribute things such as changes in tone to jitter and re-clock issues has always been a stretch for my imagination.  The system, as of present is:  J River Media Center 16 playing FLAC files > ASIO direct feed to sound card, no-resampling, no processing > optical digital output to CIA VDA-2 DAC > analog output to the WA6.  It is as direct as I can make it without a USB DAC implementation.  The reason I'm stymied at present is because the device sending signal to the DAC is the same one that was doing so before.  I literally yanked this soundcard out of the old computer and put it into this one, but I'm getting different sonic results now.  All other settings and modes are identical to what they were on the old PC.  I quadruple checked.  I mean....... the purpose of ASIO is direct communication with the hardware so I don't see how Windows7 could be affecting this vs. WinXP before, but this is the only thing that is really different.  It's just....... exasperating.  Between this and another issue I'm having (which I will withhold because it is not relevant to rant about in this topic), I'm finding little joy in music at the moment.

 

[mikemalter]

 
Quote:

Ah, a computer and information science point here: WAV does not have more information than ALAC or FLAC. It occupies more bits, but that's because the density of the information is lower. If you ignore the additional information that an ALAC or FLAC has, purely because the definition of the container (file) for the data allows for tags, the information (music) content of a WAV, ALAC and FLAC is identical. The proof is that you can convert endlessly between all three formats, and the bits representing the music in each of the three formats won't change. No information is lost.
 
You may ask yourself: how can the files always be smaller and still have the same information? The answer is that they can't. There is NO lossless compression method that can take any arbitrary input file and always create a smaller file. Lossy compression? Sure. Lossless? No way. It takes 5 minutes to prove this with simple math on a whiteboard. The 5-second version is that there are fewer possible files if they're smaller (there are 2 to the power n possible files of n binary bits size), so the decompressor wouldn't know which of the many larger files is the right one.
 
FLAC and ALAC work because they're dealing with inherently inefficient data representation in WAV (or AIFF....same thing, different clothes). If you took a set of random bits and threw them at FLAC or ALAC, they'd fail, or "compress" the file to one that's larger than the original. Most compression algorithms are smart enough to declare defeat and just give you back the original file.
 
An an aside, when I was a venture capitalist (halcyon days), I saw a number of companies promising lossless compression of all arbitrary files. One even claimed that you could use their technology to compress a 600MB CD to 60KB, then again to 6 bytes, and then reverse the process losslessly; THAT was an uncomfortable meeting. I turned them all down, because they were wrong, sadly misguided, nuts, or worse. Anyone with college-level information science can do the 5 minute proof mentioned above. And, no, this isn't like the people who "proved" that you can't fly faster than sound; the math on compression is much more solid than that: Claude Shannon, one of the fathers of information theory, was a very smart dude.
 
Lossy compression has its place. The HD movie you watch on your TV is missing over 99% of the bits in the data stream. If you've worked in compression, you can see the compression artifacts (and hence working in compression spoils your enjoyment of TV forever), but it's amazing how good it looks. JPEG performs similar, although less severe, miracles on pictures. Most MP3s throw away 90% of the information, but still sound ok to many people.
 
All the above assumes that the FLAC/ALAC software you're using isn't buggy and doesn't destroy your data. A few tests of compress/decompress and comparing against the original should convince you.
 
Is that enough talk to talk you out of WAV and embrace the joys of tagged files, Mike? And, to repeat a prior post: if WAVs make you more comfortable, and sound better to you, go for it. I'm not snickering behind your back. There are enough unknown variables in how we perceive sound, both psychological and physiological, to make your perceptions real and something that I honor and respect. You can always tag the files with "sidecar" files (same name, different extension) or with music software that tracks all of this in a database, like J River. It's safer and easier to manage if the tags are in the files, but some software doesn't keep all of the tags in the files. For example, iTunes doesn't keep ratings or last played date in the files. You can torture it into doing so with scripts and using/abusing other tag fields, but it's not convenient. As always, no perfect answer, and lots of tradeoffs....
 


 

FlyingBear, I always appreciate your posts because you are so well thought out and it streaches my thinking.
 
If there is no appreciable difference sonically between FLAC and WAV (which I believe), then the decision point moves on to something that will make a format compelling to use.
 
I've been discussing media management in a different thread that specifically asks for sonic consequences to be eliminated from the discussion, and the thinking centered on tags and tag standards.  That is, no standard exists for tags written to WAV files, but there is one for FLAC.  The observation was that a lot of media server management software will allow filtering on standard and user defined tags and the benefit is meaningful sets of media data.
 
So if the real difference is over tags, it's a different conversation and that is what I am thinking about now.  What benefits do tags have for me.  Right now, I'm a play-my-entire-media-library-randomly kind of a guy.  I'm using dBpoweramp to rip, and I know what it gives me, and so far, album art, the name of the album, the genre, the artist, track number and track and album sound level are all I need.
 
What I have observed about FLAC is that the last music I bought is two versions back from the previous version being used by dBpoweramp.  From that version to the current version there are over 25 bug fixes, enhancements to existing features and new features.  If you go back even further, there are orders of magnitude of new features, bug fixes and enhancements.  And, as sure as God made little green apples, suceeding releases of FLAC will have features, bug fixes and enhancements that prior versions do not.
 
At what point will old gear using legacy versions of FLAC stop playing the most recent release of FLAC?  Being burned by standards changing by new tides is not a new concept in the audio world.  I'm still trying to figure out what creative things I can do with my SACD only disks, and the list goes on.  What I like about WAV is that it's a standard used by music professionals and I doubt it's going to change very much.
 
I have not made my decision yet which format to use and I am leaning towards WAV, because it is an uncompressed format and am open to FLAC if I can see what tagging can do for me.
 

 

[Arcamera]

Quote:

Hi everyone! (first post - nervous!) I'm about to buy some cans and an amp for the first time (other than the Grados I use with an iPod). I like the looks of the Hifiman phones (perhaps the HE-5 or 5LE is most in my price range). And I LOVE the looks of the WA6, which I could also afford. Could anyone comment on the combination of a WA6 & HE-5 or WA6 & HE-5LE? (I've also thought about AKG 701 / 702s or Beyerdynamic 990 600ohms)

A little about me... I listen to a wide range of music that does not include classical. Mostly I'm going to put on Indy Rock / well-produced Pop / singer-songwritery stuff. (my favorite artists are John Vanderslice in the US and Quruli in Japan) In my listening room I've got Harbeth C-7 speakers, but with a 5-month old baby girl about to start crawling around, stand-mounted speakers will be effectively banished for a couple of years. Hence the desire to listen to cans!

I'm not much into tinkering, and I predict that I will make one purchase and probably never roll the tubes until they break. (but who knows, really)

So, what do y'all think about the WA6 & HE-5 or 5LE? I really appreciate your thoughts or suggestions for other directions.

Welcome to Head-Fi. And sorry about your wallet, as they say.
 
I'm afraid I can't be of any help here, as I'm still learning about this equipment. Have fun!
 

 

[mikemalter]

Quote:

Yeah, I could understand if it were analog, in fact I would accept that very easily.  But to attribute things such as changes in tone to jitter and re-clock issues has always been a stretch for my imagination.  The system, as of present is:  J River Media Center 16 playing FLAC files > ASIO direct feed to sound card, no-resampling, no processing > optical digital output to CIA VDA-2 DAC > analog output to the WA6.  It is as direct as I can make it without a USB DAC implementation.  The reason I'm stymied at present is because the device sending signal to the DAC is the same one that was doing so before.  I literally yanked this soundcard out of the old computer and put it into this one, but I'm getting different sonic results now.  All other settings and modes are identical to what they were on the old PC.  I quadruple checked.  I mean....... the purpose of ASIO is direct communication with the hardware so I don't see how Windows7 could be affecting this vs. WinXP before, but this is the only thing that is really different.  It's just....... exasperating.  Between this and another issue I'm having (which I will withhold because it is not relevant to rant about in this topic), I'm finding little joy in music at the moment.

X, couple of thoughts; one is that the digital spends some of its time as a signal, and so signal handling will make a difference as that is analog, and two, the actual programming of how those bits behave will change with different drivers.  Remember that digital audio is not the same as digital information going over the internet and your network.

 
 

 

[Xcalibur255]

Appreciate the input guys.  I've been starting to think that, long-term, my solution is going to be upgrading to a new DAC with USB input.  The Wyred4Sound units look very nice in particular, but this will have to be a very long ways down the road.  I bought the VDA-2 with the intention of it serving the needs of whatever amp comes along, figuring I had perched myself right at the point of diminishing returns when it comes to DACs in terms of price.  It's amazing how fast this stuff has been changing and improving.

 

[Silent One]

Computer environment ~
 
There are goblins (well, what else should I call 'em... it's Friday!) residing inside a computer. And goblins airborne, just hovering and waiting to get dem hands on your output. Last summer, I performed the following exercise using my HP 17" notebook/Windows XP SP3; Mac mini with Boot Camp & same Windows config; EMU 0404 USB Interface w/ASIO driver; WireWorld Ultraviolet USB cable; iTunes.
 
Playing the same test tracks in 32-bit mode; ASIO out in the three configurations brought three different results. A computer's processes and its environment can and sometimes will change the sound. The files on the Mac with Boot Camp were closer but different. On the notebook the difference was prominent. Hope this helps scenario helps!

 

[Silent One]

Quote:

Appreciate the input guys.  I've been starting to think that, long-term, my solution is going to be upgrading to a new DAC with USB input.  The Wyred4Sound units look very nice in particular, but this will have to be a very long ways down the road.  I bought the VDA-2 with the intention of it serving the needs of whatever amp comes along, figuring I had perched myself right at the point of diminishing returns when it comes to DACs in terms of price.  It's amazing how fast this stuff has been changing and improving.

You're most welcome. And that Wyred4Sound DAC? It may come faster than you think at a price you can afford! I'll be looking to go from DAC-1 to DAC-2 soon. 
 

 

[MacedonianHero]

Funny you should mention the Wyred 4 Sound DAC-2....just picked one up today (should be here by Monday according to Canada Post). The only reason I picked it up was to feed my WA22 a balanced source; otherwise I was exceedingly happy with my Cary XCiter DAC (both cost the exact same). This place is truly sinister. :evil: