is PCM=FLAC ?

They won't be. You can't go to analog and back and keeps things perfectly identical.

 

It doesn't work that way, I think my "experiment" re-cpmpressing DTS to Flac demonstrates better than Flac is actually lossless.
By the way, I did the same with an HDCD (Independence Day soundtrack) CD and I got the same result, HDCD decoded perfectly.

 

If one can compare the PCM and reconstructed FLAC files and they are identical, then one can conclude that the difference in subjective sound is the result of some other variable, such as computing power, circuit anomalies, or other factors. Finding the other causes seems to be at the crux of the "problem". I could be wrong, and frequently am.

 

ABX Comparator for fb2k actually creates temporary WAV files for both tracks prior to being tested which defeats the purpose IMO...

 

Well, any player has to convert lossless to PCM (wav) before playing it back.

 

I'm not saying FLAC isn't lossless, I'm saying digital=>analog=>digital isn't lossless.

 

For the record FLAC encoding does not introduce any downsampling or dithering.

 

I know. That's why I suggested that the software he used may be doing this.

 

Or, the difference may be imaginary.

 

I don't know if my opinion can help in this thread.
If i convert a wav file in .flac and in .shn to my ears files .shn sound better.
I use winamp with proper plugins to listen to my files. All my files are .flac, because of tags, but I'm pretty sure that I always prefer .shn that sound "cleaner" than flac.
Really I don't understand why, because, as I quoted, any plugin convert to wav before playing.
Any suggestion? Anyone experiences what I hear?

 

True, however the FLAC uncompression does not happen during playback in ABX Comparator, but while preparing the ABX test.

 

It's your imagination

 

Hi mwheelerk,

This is a personal call.
As you can see, in this thread and in countless places on the Web and in print, there are folks who insist the audible results of listening to both should be the same. To many folks they are the same.
To me, they are not. (While I wouldn't call the differences "night and day", they are there and I hear them in several characteristics I described in my earlier post.)

Some have suggested the software or hardware is at fault. All I can say is I suppose what I'm using isn't "good" enough to obscure the differences. ;-}

So, again, I say it is a personal call. I know folks who hear differences between so-called lossless formats and raw, linear PCM. I know others who don't hear those differences. I know folks who claim .wav sounds "better" than .aif. To me, they are indistinguishable.

My best suggestion is to try it with a few tracks and draw your own conclusions. For my part, all I put on my server are raw .aif files.

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

 

That too. I once imagined I was hearing differences, but realized later that I wasn't. But if someone says they hear differences, i'm not going to tell them they are wrong.

 

I have issues with the reasoning that FLAC sounds different because it takes longer to decode it. It's actually the opposite. It takes less time to read a FLAC file from a hard drive and decode it than it does to read a WAV file. This is because the hard drive is the slowest part of a computer and the WAV files are much bigger than FLAC. (It also takes more CPU usage to play a WAV than a FLAC on a modern computer due to the CPU waiting to receive data from the hard drive.) Now if you're loading everything in RAM first then you might be able to make that argument.

 

That's basically how a computer works.

 

There could be many mechanisms affecting this; however, before investigating those you have to be convinced there's a difference.

Having written strict time critical software myself, it's not just enough to have the right computing power - though it helps.

 

Depends on the program being used. I guess I should have said ram disk and not just ram. Generally you set a buffer size with an audio program and it tries to keep that amount of the file in memory. It generally does not load the entire file into memory before playing it.

 

Will you tell them they might be wrong, and why?

 

That might help; but the issue is whether the processor i/o is pushing the words out to the d/a in a synchronous manner. And who's in control of this (i.e. running the clock).

A lot of this is hidden inside the drivers and the OS for most users (some bypass all this with special drivers or programs, and again many users claim improvements due to these regardless of audio format).

And there are proponents of fast, solid state discs for playing best quality audio...

 

OK, I think I can hear it now, using FooBar carefully (the a-b comparison doesn't work with different formats as it converts to WAV during the randomization). Win XP, ASIO, Hagerman Ripper D-A 16 bits.

A slight difference in vocal sibilant quality and some phase issues affecting the sound stage 'space'. Very similar to differences between bit identical CDs. I uncompressed one revealing file from a CD I know quite well and already have on the computer.

I'm not saying it's there, only that I don't think the on-the-fly playback in foobar is identical. More listening is needed. But I don't think they sound the same on my setup. It's not life changing, yet.

 

I'm not familiar with Foobar (and I probably should be), but while it's possible that certain plugins could be configured slightly differently, I can't imagine that any random event like you describe would create "phase issues affecting the sound-stage space."

 

This is not entirely correct. ABX Comparator in fb2k converts FLACs to 32bit (float) PCM (preserving the original sample rate of each track) during test preparation, not randomization. No format conversion occurs during playback in ABX testing.

 

The randomization is that you don't know which file is which.

The fact that it's no longer decoding the flac as you listen is the issue.