Any Innuos users here?

[AndyU]

While setting up our music server, an Audiostore Prestige 3 with 2TB of SSD storage, we tried a few albums which were ripped to FLAC level 5 (i.e. compressed but lossless), FLAC level 0 (supposed to be uncompressed) and AIFF.

The FLAC versions were ripped using the Prestige server's built in posh TEAC disc drive.
The AIFF versions were a real faff to do. It took a 2011 mac mini with apple superdrive (i.e. not the best drive) running XLD or dBPoweramp software to rip them on to our Synology NAS drive. The ripped albums were then copied across on the network from the NAS drive to the music server. Given this faff, the AIFF versions should have been worse than a straight rip using the Prestige 3's built in TEAC drive.

Rather annoyingly, there was a consistent preference between the three versions:
First AIFF
Second uncompressed FLAC
Third compressed FLAC

The differences were fairly subtle, yet important to us. The FLAC versions sounded slightly duller and flatter than the AIFF version of each track. Those subtle, fine detail cues that give the listener a sense of the acoustics of the space in which the music was recorded were lost. The AIFF versions sounded that bit more alive and engaging.

Why annoying? Because I then had to rip over 1000 cds in order to load all of the AIFF albums on to the Prestige 3 server. It took weeks. Groan.

The rest of the system that made the sound quality differences so evident was Chord M Scaler + DAVE, Naim 52/135 and ART Alnico speakers.
Perhaps the format differences would be less evident on a less resolving system.

Damn, but I still wish to this day that the choice of codec would make no difference...

Best regards, FT

Good reason to avoid an Audiostore Prestige 3 server imo.

BTW you are wrong about FLAC level 0 being uncompressed, it typically reduces filesize by about 25%. And is only a few percent faster to decode. See the table here:

https://en.m.wikipedia.org/wiki/FLAC

and notice where it says “The technical strengths of FLAC compared to other lossless formats lie in its ability to be streamed and decoded quickly, independent of compression level.”

 

[Cereal Killer]

You don't need to re-rip anything, simply use dbpoweramp's batch converter.

 

[AndyU]

Here's something to try on a rainy Sunday afternoon...

Take a WAV file
Strip out any metadata
Create a checksum value from the WAV file
Convert the WAV file to FLAC
Convert the FLAC file back to WAV
Create a checksum value from the second WAV file
Compare the two checksum values

The two checksum values will be identical. This proves that all of the original bits are preserved during the FLAC compression / decompression process. It doesn't prove that any given transport will sound better or worse when playing back FLAC or WAV files.

You don’t need to bother with all that, just use dBpoweramps compare audio feature on the two files - it compares just the audio portions of two files after performing any necessary uncompression. Foobar has a similar feature.

 

[Mullardman]

First impression of WAV/FLAC comparison via the Innuos' onboard DAC reveal no difference that I can put down to anything more than expectation bias.

Disclaimer: My old ears/YMMV etc, etc.

 

[Mullardman]

It will be interesting to hear what difference my DAC makes. Theoretically, a Benchmark DAC1 which cost around £900, ought to sound better than whatever DAC Innuos have slotted into the Zen Mini III as part of an overall £900 ish price.

OTOH, the Benchmark is 12-13 years old and presumably DAC technology has moved on a bit?

Anyway.. I'm not wasting any more of the first half decent weather for weeks farting about with audio... It will soon be raining again...

 

[Julf]

WAV is pointless and counterproductive when it comes to handling metadata. FLAC is the best truly editable format. sound is identical to WAV, both are lossless.

An additional benefit of FLAC is the content integrity checksum that detects file corruption / bit rot.

 

[Julf]

The argument for WAV sounding better is that the CPU power to do the decompression processing adds some small amount of electrical noise which is audible.

More than offset by the CPU having to work harder to read, buffer and write the redundant extra data in a WAV file.

 

[Julf]

The differences were fairly subtle, yet important to us. The FLAC versions sounded slightly duller and flatter than the AIFF version of each track. Those subtle, fine detail cues that give the listener a sense of the acoustics of the space in which the music was recorded were lost. The AIFF versions sounded that bit more alive and engaging.

Sounds like sighted listening. Was it A/B or serial listening (relying on auditory memory)?

 

[Foot Tapper]

Good reason to avoid an Audiostore Prestige 3 server imo.

BTW you are wrong about FLAC level 0 being uncompressed, it typically reduces filesize by about 25%. And is only a few percent faster to decode. See the table here:

https://en.m.wikipedia.org/wiki/FLAC ”

Indeed Andy, which is why I said "supposed to be uncompressed" as I have seen many such erroneous claims. For the albums that we tried, FLAC level 0 files were 30% smaller than the corresponding AIFF ones, which correlates well with the Wikipedia article that you quoted.

Following our little test, the Audiostore is looking into adding AIFF codec to the available ripping options for its servers, which would be wonderful!

Best regards, FT

 

[AndyU]

Indeed Andy, which is why I said "supposed to be uncompressed" as I have seen many such erroneous claims. For the albums that we tried, FLAC level 0 files were 30% smaller than the corresponding AIFF ones, which correlates well with the Wikipedia article that you quoted.

Following our little test, the Audiostore is looking into adding AIFF codec to the available ripping options for its servers, which would be wonderful!

Best regards, FT

Have you asked them why their FLAC decoding doesn't sound as good as WAV? There is no reason why there should be any difference. Have they made a mistake somewhere?

 

[andrewd]

This bit intrigues me. I've read that WAV 'doesn't like metadata'. I've read in this thread that I would need to be 'sophisticated' to use metadata with WAV. I ran this past my son in law who has a degree in audio production or somesuch and he basically said that there's no issue with metadata and WAV. So far, in my admittedly very small experience, there is no issue.

It’s been many years since I looked into this, so the information might be out of date. My understanding was that WAV files need an additional metadata file, which can get separated from the audio file when copying to other devices. In most other audio files the metadata is incorporated into the main file. If you don't move the WAV files around you should be fine.

 

[Foot Tapper]

Have you asked them why their FLAC decoding doesn't sound as good as WAV? There is no reason why there should be any difference. Have they made a mistake somewhere?

No, as I don't use WAV. Apparently no to your second question.

 

[andrewd]

Here's something to try on a rainy Sunday afternoon...

Take a WAV file
Strip out any metadata
Create a checksum value from the WAV file
Convert the WAV file to FLAC
Convert the FLAC file back to WAV
Create a checksum value from the second WAV file
Compare the two checksum values

The two checksum values will be identical. This proves that all of the original bits are preserved during the FLAC compression / decompression process. It doesn't prove that any given transport will sound better or worse when playing back FLAC or WAV files.

Those that make the case for WAV sounding different (not saying that includes me) have never said it is due to the FLAC file being lossy in any way (which would give a different checksum). It is all about the effect of noise when the files are processed in real time. Saying that this causes audible difference between file formats is taking the argument to the extreme, and probably only matters for those with golden systems and golden ears. But even with my old ears and modest system I can hear a difference when using an external USB/SPDIF converter with a low noise power supply and internal regulation, compared to using a standard USB output swamped with noise.

The old ‘bits are bits’ argument is in my opinion is a gross oversimplification of digital audio.

 

[Strictly Stereo]

Those that make the case for WAV sounding different (not saying that includes me) have never said it is due to the FLAC file being lossy in any way (which would give a different checksum). It is all about the effect of noise when the files are processed in real time. Saying that this causes audible difference between file formats is taking the argument to the extreme, and probably only matters for those with golden systems and golden ears. But even with my old ears and modest system I can hear a difference when using an external USB/SPDIF converter with a low noise power supply and internal regulation, compared to using a standard USB output swamped with noise.

The old ‘bits are bits’ argument is in my opinion is a gross oversimplification of digital audio.

I agree.

 

[AndyU]

Those that make the case for WAV sounding different (not saying that includes me) have never said it is due to the FLAC file being lossy in any way (which would give a different checksum). It is all about the effect of noise when the files are processed in real time. Saying that this causes audible difference between file formats is taking the argument to the extreme, and probably only matters for those with golden systems and golden ears. But even with my old ears and modest system I can hear a difference when using an external USB/SPDIF converter with a low noise power supply and internal regulation, compared to using a standard USB output swamped with noise.

The old ‘bits are bits’ argument is in my opinion is a gross oversimplification of digital audio.

I’ll readily agree that different USB sources might sound different into different DACs, that’s quite reasonable given how noisy computers are and how sensitive some DACs might be. But if I had a source that made FLAC sound different to WAV through the same USB connection I would dump it, or dump that way of connecting it, for it would mean that that the processor and the USB output were interacting unacceptably. It’s not as though the processor is off when WAV is being played. We are talking here about streamers that sell for thousands. If it is the case that such streamers do a worse job with FLAC than WAV then surely you would be better off streaming to the likes of a tiny low powered Squeezebox Touch and connecting that to your DAC via optical.

 

[Julf]

My understanding was that WAV files need an additional metadata file, which can get separated from the audio file when copying to other devices. In most other audio files the metadata is incorporated into the main file. If you don't move the WAV files around you should be fine.

The real problem is that the metadata is not standardized. Some programs do it one way, others a different way - so you might lose all metadata if you move to another system/player software.

 

[Tw99]

In my opinion it is nothing short of insanity to rip to WAV rather than FLAC

 

[Mullardman]

In my opinion it is nothing short of insanity to rip to WAV rather than FLAC

Please stop equivocating and give it to me straight.. What do you really think?

 

[Mullardman]

So anyway... I was happily ripping a 10 CD box set of Mozart Piano Concertos... (Concerti?) by Vladimir Ashkhenazy.. on the Decca 'London' label. Although the Innuos 'Disc Ripper' found the correct artist, track title and length etc., etc.. it managed to apply cover art from the box set to only about four of the discs.. and cover art for Alfred Brendel.. on Phillips.. to the rest. Quite bizarre.

 

[Colin L]

Welcome to computer audio. It can be frustrating. As a recent convert to Roon I’m delighted it has sorted out all my music catalogue issues. It’s indispensable to me now.