[edn4x4]

JRiver 19 memory playback has been enhanced:

http://yabb.jriver.com/interact/index.php?topic=81982.0

Overview
Memory playback now holds decoded data in memory instead of encoded data.

This means playing any form of lossless (WAV, FLAC, APE, ALAC, etc.) will have identical data in memory and identical resource usage.

Benefits
The theoretical benefits of memory playback are that no disk or network I/O occurs while playing and that CPU load during playback could be reduced.


I was hoping to compare playing from memory flac and wav - just haven't had the time - but wanted to make you aware of the change for JR users.
I also have all FLAC and was wondering if I should convert to wav - since the JR 19 updated the playback, I may not have to.
Eric

[holzohr]

Russ, "0" level compression is not "no compression".  Zero is just the lowest level of compression.  FLAC also has (with newer FLAC decoders) a "no compression" option which some say delivers wav quality with the benefit of universal metadaa support, etc...

Again something I still didn´t know. I always was using the "0" level compression. Today, I was ripping this mentioned album http://www.audiocircle.com/index.php?topic=68722.msg1266042#msg1266042 and converted it to FLAC, once with "no compression" and once with "0 level compression", both without folder pic tagged. Hmmm... too early to make a statement though I don´t think I had a bad ear day. I still want to test it with 44.1kHz files and will give WAV a try, too.

About the folder size: "0 compression" = 1.44GB, "no compression" = 3GB 

[gregcss]

I compared FLAC to WAV with only a few songs the difference was minimal if any, the favor going to WAV. Since file size and tags are not an issue for me I blew away all FLAC and re-ripped to WAV. Been listening to various songs not included in my first comparison and seems the music is more alive now. Not a significant difference, but still different. Perhaps its because I want it to sound better. Anyway, glad I converted them. Thanks for making this thread.

[JayM]

My Oppo bdp83 started to play up with its spdif output so had to start using wd tv live media player as my transport. That goes via optical to an old benchmark dac 1 then to foster pm2 actives via balanced silver interconnects.

I had all my library ripped to FLAC. About a month ago, I got a new lot of discs i thought to try ripping to WAV and FLAC and see what if any diff it made. Blow me down, but the WAV had a lower noise floor and seemed to present low level detail that was inaudible or muddled in FLAC. That manifested itself in better micro dynamics, in better high frequencies, most easily heard with cymbals, in how cymbal strikes differ subtly from one to the next, and how those strikes decay in space. The overall gestalt is that I heard more of how the music was played as opposed to just the notes being played.

So my very humble playback system works with WAV better than FLAC. In that context I have started to redo all my rips to WAV. Maybe a better retrieval chain will present sound differently, but that is for everyone else to experience. I wished it were different. I lose meta tagging functionality, have all that reripping to do, but that's where I am.

I believe my experience does not necessarily lead to the conclusion of WAV superiority over FLAC as a format. Patently the method of retrieval can't be isolated from the format of the media in my situation. But if I can obtain a more enjoyable sonic experience from simply reripping my library, that is the way forward for this impecunious audiophile.

Next steps...where can I get a better power supply for my WD TV Live Media player.....it always rains in audio land.

[rbbert]

I'm still baffled by why people are ripping to WAV and storing their music that way.  Don't you read earlier posts in this topic?  rip and store as FLAC, save space, speed access and have good tagging, then convert to WAV and play from memory if you think WAV sounds better??

[gregcss]

I'm still baffled by why people are ripping to WAV and storing their music that way.  Don't you read earlier posts in this topic?  rip and store as FLAC, save space, speed access and have good tagging, then convert to WAV and play from memory if you think WAV sounds better??

I ripped using dbpoweramp and it tagged the files for me. I play them through squeezebox and have no issues browsing my library - it's the exact same as the flac files.

[ted_b]

WAV does not inherently support tags; you are lucky in that your player/LMS does...but those tags go away if you change players. 

Part of the WAV vs FLAC debate includes theories that the on-the-fly conversion of a compressed format might be leading to the sonic differences, so that would not support rribert's post (let alone his tone) that we ought to store in FLAC and convert to WAV when playing.  Unless he meant pre-convert (offline, which is more storage than doing it once) I can't see how on-the-fly conversion gets you away from the debate.

Uncompressed FLAC is another animal, and I have just not had enough time to compare, but I will. 

[rbbert]

decode FLAC (hard drive) to WAV (RAM or SSD) and then play the WAV from memory.  You have the benefits of FLAC, WAV and memory playback.  sorry if the tone seemed offensive; it was a reflection of exasperation on my part.

[dminches]

RB, who wants to have to go thru the hassle of decoding flac files every time you want to listen to something. That eliminates any ability to be spontaneous. Or am I missing your point?

[rbbert]

JPlay can do that in the process of loading into memory; given JRiver's habit of adopting JPlay's features, I suspect it will also be able to do this soon.  If you don't care about playback from memory rather than hard drive then none of this matters.  Everyone pretty much does what they want anyway, but if you find that things like WAV vs FLAC and/or memory playback make audible differences, the way I describe is a simple economical way to go.

[JerryLove]

It is possible that the playback unit decompressing the FLAC is not doing it properly. It's also possible that the difference is psycho-somatic. 

FLAC is lossless. If you take a wave, convert it to a FLAC, and then convert that FLAC back into a wave: It is byte-for-byte identical to the original wave.

The "compression level" has nothing to do with accuracy. Higher compression requires more CPU time to compute. So you are trading file-size for compression speed. On modern hardware, I cannot come up with a reason to not run at or near 8.

[ErikMi]

What about Aiff??

[audiventory]

Practical checking WAV vs. FLAC (video)

http://www.youtube.com/watch?v=3jPphh-CsHM

[bdiament]

Not trying to stir the pot, but… anyone else hear a distinct difference between WAV files and FLAC files encoded from the WAV files?...

Hi Russ,

First, deepest thanks to Ted-b for insisting this thread remains civil.  In my opinion, we need more of this on the audio web (not closing threads but excising any sources of incivility).

To your question, I've been saying this since I first tried so-called "lossless" compression: To my ears (and the others who have participated in blind comparisons here), what comes back does not *sound* the same as what went in.  It might not be a "night and day" difference but as far as I'm concerned, if they are at all distinguishable, I'm not interested in the format.  (I know what the theory and the measurements say but I also know what Yogi Berra said about theory.  )  If my library was extraordinarily large, I'd just purchase more storage.

This is one reason why Soundkeeper does not offer downloads.  I see others turning files into FLACs in order to make for practical download times.  Even the files sold as .aif or .wav seem to be sent as FLAC files and expanded by whatever download manager the service uses. 

Personally, I store everything in my music library in .aif format, the same format in which I record, mix (when applicable), and master.  It allows tagging and no extra conversion processes, i.e., compression/expansion, are required.

Just my perspective, of course.

Best regards,
Barry
www.soundkeeperrecordings.com
www.soundkeeperrecordings.wordpress.com
www.barrydiamentaudio.com

[audiventory]

Hi Barry,

Possible trouble in playback system. May be you check for some old FLAC library or versions with bugs.

May be playback part (after decoding) has differences for WAV and FLAC.

Possible many reasons, that need learn.

But digital content WAV and FLAC fully identical theoretically and practically. (I used last on December 2014 flac library version)

Best regards,
Yuri

[asliarun]

I too don't  want to start an argument - just sharing some thoughts. There is a phrase in science and statistics that "correlation does not imply causation". You may well have heard a few FLAC files here and there that were sounding different, maybe even inferior, to WAV.

However, even with all the subjectivity that exists in the audio world, certain basic principles must be considered.

Consider that anyone associated with computers understands the notion of a ZIP file. They understand and expect that a ZIP file will reliably (i.e. 100% of the time) allow the user to convert a zipped file back into a perfect version of the original file.

There is never any debate about how "perfect" the conversion is - it is a given. Of course, there are exceptions where the ZIP compression process sometimes causes file corruption. But it is never ever a matter of debate. Meaning - no organization on earth (to my knowledge) refuses to use ZIP files because it feels that it could end up degrading the original file in some way.

FLAC (again, in my limited knowledge) is a similar lossless compression mechanism. It is expected with 100% certainty and 100% reliability that when the FLAC file is decompressed or deflated, it will give you back a bit perfect copy of the original audio signal (the digital bitstream).

A WAV file by the way is *also* a container format. One can technically store an MP3 file inside a WAV container. Although mostly, people use a WAV container file to store the uncompressed CD bitstream. However, the only difference between the WAV container and the FLAC container is that the FLAC container file has one additional lossless compression step. And is hence typically half the size of a WAV container file storing the same bitstream.

I honestly don't understand why this should ever be a matter of debate. If the argument is a subjective one ("i can hear the difference"), then there are 100 other things I can think of to debate and investigate. Let's see:
1. If I copy paste a WAV file one hundred times between folders, I hear a difference. Perhaps the WAV file suffers from "wear and tear".
2. If I store the WAV file in different digital media (or say, different brands of hard drives), I hear a difference. Perhaps, the hard drive controller makes a difference to the audio quality output.
3. If I play back a WAV file in an Intel PC vs an AMD PC, I can hear a difference. Perhaps, Intel and AMD and ARM CPUs process audio differently

Now, truth be told, a lot of this already exists in the audio world.

Maybe, the real question is "where do we reasonably draw the line"?

If you feel that a FLAC file sounds different, the only possible reason could be the extra burden on the CPU to decompress the file before feeding it to the DAC. Even putting aside memory playback which makes this concern irrelevant, the correct recommendation for this concern would be to *upgrade the CPU* or increase RAM or do both.

The wrong recommendation would be to stop using FLACs altogether. Which is why I find it surprising that people reach this conclusion.

[bdiament]

Hi Barry,

Possible trouble in playback system. May be you check for some old FLAC library or versions with bugs.

May be playback part (after decoding) has differences for WAV and FLAC.

Possible many reasons, that need learn.

But digital content WAV and FLAC fully identical theoretically and practically. (I used last on December 2014 flac library version)

Best regards,
Yuri

Hi Yuri,

I have had the same experience when listening on other systems too.  It is certainly possible that all the systems I've listened to have trouble with FLAC.  Or it is possible these systems, which excel at allowing one to hear past them, are just not good enough to obscure the differences.  ;-}

Whatever the reason, I have yet to hear a comparison where I found the two indistinguishable and personally, I don't see any reason to convert the raw PCM files created in the studio to another format.  (Since I prefer .aif, I have no issues with tagging.)

I know some equate FLAC with ZIP for a word processing file.  I don't because Word processing files are not time-based as audio files are. 

Still, I think folks should use what they want to use.  I just get wary when they want me to hear it the same way they do and want me to use what they use.  (I don't like DSD either, though I know many folks who love the format.  I wouldn't take it away from them and would hope they wouldn't want to force it on me.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.soundkeeperrecordings.wordpress.com
www.barrydiamentaudio.com

[leftside]

I'm fairly new to this site and have enjoyed this thread - and learnt quite a few things. Most of my music files have been ripped from my CD's using dBpoweramp CD Ripper to FLAC, and played using JRiver. I'm going to rip a couple of CD's to WAV to see if I can hear a difference. I doubt I'll be able to, but it won't hurt to try.

[*Scotty*]

I suspect that the difference heard between flac and wav files may be due to the playback software used and the fact that the flac files are converted on the fly in real time. This may add jitter to the signal or the playback software may be creating artifacts that are allowing you to differentiate between wav and flac files during playback. I have no doubt that wav and flac converted to wav are a bit perfect match to one another.
 If the flac was converted to wav and stored in a buffer rather than existing as a data stream that was created in realtime there might be no difference in sound between the two formats.
 The Neutron player does this with all losslessly compressed formats before playback. It also sounds markedly superior all other players that I have tried on my Galaxy S4. I don't know how jRiver 20 converts losslessly compressed formats to a data stream. All of the I the music I have exists as wav or AIFF files and what little music I have in the flac format is not duplicated in either wav or AIFF.
Scotty

[audiventory]

Indeed, this topic we can join to groups same topics about difference of sound, as wrote Asliarun:

1. 2 CD manufactured in different plants.
2. 2 file placed in different places HDD.
3. Processors...

i.e. where theoretically must not be any difference.

I don't deny what possible hear difference. Possible hear, possible not.

But as engineer I don't fully trust my ears (it is very unstable tool for development) and know that need concider exactly full system and decompose it:

power source-storage-file format-playback sofware-its DSP-driver-method trasferring audio stream-DAC-amp-room-placement of listener-state of listener-...

Example:
For some cases, buffer, as said Scotty, must eliminate any difference (like comparing of same CD from different plants).

But other hand it depend on system of synchronizing buffer with output realtime stream.

For low CPU power possible (again theoretically) empty buffer - periodically on shortest time (don't perceived as pause, but as decreasing of transparency).

Additional demand in CPU power posible need for decoding of FLAC and lead to empty buffer.



I'm sure only (since I check it practically and anybody can repeat it):

FLAC with FLAC library (v. 1.3.1) built in AuI ConverteR 48x44 on 44 kHz /16 bit fully keep source data (here I exactly know: what inside the converter).

While I don't check same way other sample rates. Possible there can be bugs.

Thus reason of sound difference possible in other parts of the system.

Same check we can perform with any WAV to FLAC converter.

Theoretically, it also must repeat result from video.

If we get different WAV's and FLAC's contents, there possible some bugs (version FLAC library or software - rounding errors, writing/reading calculation buffer, ...) or applyed digital signal processing.

In video I specially turn off any DSP. Otherwise I can't get identical WAV and FLAC.

Barry, as professional in recoding, has trained hearing. If he hear, possible some technical reasons (while hidden) for it, placed out content of FLAC format.

Possible was troubles with DSP or bugs of used converters or players.