Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

[d2b]

I've met resistance to using FLAC from those that believe that any compressed audio file format is subject to more severe corruption than a linear PCM (i.e., WAV) file when subjected to bit-read errors or "bit rot." While we know that a perfectly-handled FLAC file is bit-for-bit identical to the source file when decoded, what do we know about the severity of problems that might occur if there are missing or corrupted bits in a block. In real life, such errors do occur. Is that not right?

It's said that such errors are inconsequential as they are limited to a single block in which the bit error(s) occurred and thus do not affect preceding or following blocks. Still, errors are errors and if bit rot occurs, then the decoded file cannot match the source file. Just how bad can a block be affected by a single bit error or a short burst of errors?

I doubt that it is possible to predict the consequences of bit errors in any quantified manner. But it seems like it would be good to minimize the consequences of bit rot. For that reason, a linear PCM or WAV implementation with all the other attributes of a normal FLAC file container seems preferable as long as disc space and file transfer times are not of concern. In particular, our application is for a large (100,000+ albums) library in which we want all of the attributes of a FLAC file with its Vorbis comments but the (apparently) better resistance to bit rot.

So my question is this: Why is it not possible for dBpoweramp to create a "pure" FLAC file wherein the actual payload is truly a uncompressed WAV file even with digital zeros or some kinds of noise? By "pure," I mean a payload that is bit-for-bit identical to the WAV file that created it. Clearly the loss of a single bit or even a short burst of bit read errors only affects the sample(s) within which the errors occurred.

"d2b" aka Dennis...

P.S. Perhaps there would be decoding problems?

[mville]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

I've met resistance to using FLAC from those that believe that any compressed audio file format is subject to more severe corruption than a linear PCM (i.e., WAV) file when subjected to bit-read errors or "bit rot.

Who are those you refer to here, can you provide more on who are they?

[Spoon]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

Because no decoder would decode it, the flac format has no provisions for such.

[d2b]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

Who are those you refer to here, can you provide more on who are they?

In the United States, a number of subscribers to the Public Radio Engineers technical forum (Pubtech). They are resistant to any form of audio compression, convinced that FLAC is fragile as a storage format, subject to corruption in ways that are likely far more destructive than WAV and that savings in disk space are not worth the risk. Some agree and some disagree, but in general the trend is to favor linear PCM as the only really reliable (i.e. therefore favorable) technology for audio content storage and application.

For more information, see http://www.pubtech.org

Of course, none of the usual radio automation systems support the FLAC file format, compounding their reasons for their disdain.

Here at our radio station, we have been building a comprehensively-tagged media library based primarily on the CDs that we receive from record companies, agents and independent artists. At last count, we estimate the library to contain about 135,000 files comprising about 11,000 albums. As a long range goal, we hope to see the music industry give up on lossy formats and adopt FLAC as the chosen standard for distribution and streaming. As we see it, it is the logical way to include metadata from the original source and to ensure its accuracy. We wish for an end to optical disc-based music distribution and storage (costly, fragile, not easily searched for content, etc.) and an intelligent transition to a file-based approach wihtout the constraints of major corporations that desire to maximize their customer base.

I might add that we have NEVER had a problem with FLAC regarding audio performance on-air. If there are any problems at all, they are man-made. ;-)

Spoon's response does not surprise me. I rather expected that a solution (if one were justified) lies with the current group of FLAC developers and those of us that wish to advance the use of the format.

"d2b" aka Dennis....

[mville]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

In the United States, a number of subscribers to the Public Radio Engineers technical forum (Pubtech). They are resistant to any form of audio compression, convinced that FLAC is fragile as a storage format, subject to corruption in ways that are likely far more destructive than WAV and that savings in disk space are not worth the risk. Some agree and some disagree, but in general the trend is to favor linear PCM as the only really reliable (i.e. therefore favorable) technology for audio content storage and application.

There is no audio compression in flac, only data compression.

[d2b]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

Yes, the payload is data. But the data comprises audio information. Is there a meaningful distinction that you wish to make? After all, the codec was designed for compressing audio streams, hence the "A" in FLAC.

[mville]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

Yes, the payload is data. But the data comprises audio information. Is there a meaningful distinction that you wish to make?

Only the distinction between audio level compression and digital data compression. It is important to clarify that the 2 are completely different and not related.

[d2b]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

Ah yes, I understand. Of course I was not referring to level compression, but I see the reason for your observation. Thank you for pointing out the possible misinterpretation.

[mville]

Re: Uncompressed FLAC: Why not a linear PCM (i.e. WAV) file in an FLAC container?

I guess there will always be believers and non-believers, based on an understanding of the very sophisticated science involved here.

I'm a believer and I back you 100% in continuing to use flac at your radio station. As you stated earlier, you have never had a problem with flac.

Any problems that you may run into, e.g. PC failures, memory failures, storage media failures etc. affects ALL digital data, flac and wav included.