SHN, hi-res FLAC support?

[coyote2]

I notice my beloved iAudio 9 doesn't play SHN, or high-res FLAC, files. Do or will any other Cowon player?

[CrowdAds]

[WalkGood]

My i9 plays EAC ripped FLAC 1.1.2-5 just fine but that's not what I normally put on there. I load mine with mp3's so I can carry more tunes ...

[Marvin the Martian]

Originally Posted by coyote2

I notice my beloved iAudio 9 doesn't play SHN, or high-res FLAC, files. Do or will any other Cowon player?

If you mean the 24-bit 96khz stuff, then no, they won't play them.

[WalkGood]

Originally Posted by Marvin the Martian

If you mean the 24-bit 96khz stuff, then no, they won't play them.

Correct, I missed the "hi-res" ... but who would do that on such a small player???

[Marvin the Martian]

Originally Posted by WalkGood

Correct, I missed the "hi-res" ... but who would do that on such a small player???

"audiophiles" aka "true Head-Fi'ers"

[WalkGood]

Originally Posted by Marvin the Martian

"audiophiles" aka "true Head-Fi'ers"

Does not compute

[coyote2]

Originally Posted by Marvin the Martian

"audiophiles" aka "true Head-Fi'ers"

Given that Cowon units appeal to folks seeking better sound quality, I would have thought that hi-res FLAC support would be a FAQ.

I know it would cause me to buy a new unit if that's what it took!

[Satellite_6]

Originally Posted by coyote2

Given that Cowon units appeal to folks seeking better sound quality, I would have thought that hi-res FLAC support would be a FAQ.

I know it would cause me to buy a new unit if that's what it took!

hi-res audio is a misnomer. hi-res does not result in better sound qaulity. So requesting this could only come out of ignorance.

[robdean]

There are cases for the use of hi-res audio in music production. In that situation it allows for repeated compex digital processing of the audio (e.g. EQ, dynamics, reverb, flexible down sampling) without the noticeable generation of digital artefacts.

In theory, high sampling frequencies, were they available, might bring very peripheral benefits to i9 users who are not HiFi purists and who pile on the enhancement effects and EQ, but I'd be brutally sceptical of any real benefit without seeing blind-test proof.
Much though I love my i9, I think its output stage lacks lacks the extreme finesse and fabulous SNR required to make even theoretically worthwhile use of hi-res bit depth.

In reality, the main contribution that such audio formats could make to a player like the i9, beyond the placebo effect, would be to add vastly to power consumption as the DSP chewed on ever more data per second encompassing ever higher frequencies beyond human hearing.

I worked for many years as a tour sound engineer and many more as a BBC radio producer. I have ears good enough to do those jobs, but not golden enough even to reliably tell whether I am listening to v0 lame mp3 or to 44.1kHz 16 bit WAV (when I've been blind tested). Note: the testing must be blind: it's no good to play the mp3 then play a WAV and reflect on whether it sounds better. There's too much room for fooling yourself. Rather, have someone play two recordings and ask you to spot which is the higher fidelity! If you can do much better than 50/50 over a whole series of tries, then that's meaningful...

[saratoga]

Originally Posted by robdean

There are cases for the use of hi-res audio in music production.

I agree, but:

Originally Posted by robdean

In that situation it allows for repeated compex digital processing of the audio (e.g. EQ, dynamics, reverb, flexible down sampling) without the noticeable generation of digital artefacts.

This is not one of them. Such effects are implemented at the machine precision, not the input precision. So if you're got a 32 bit ARM processor, all effects are done on 32 bit sample precision, even if the input is 8, 16, 24, 64 bit. So the input format really doesn't matter compared to the quality of the actual recording and the implementation of the hardware.

I flipped through some of the older Cowon DSP code ages ago (from the D2 IIRC). Mostly they used the ARMv4 32x32=64 multiplier and some 32x32=32 bit stuff. So the internal precision was always much higher then the input precision.

Originally Posted by robdean

Much though I love my i9, I think its output stage lacks lacks the extreme finesse and fabulous SNR required to make even theoretically worthwhile use of hi-res bit depth.

I think its actually about as good as it needs to be. Unless you're listening in an anechoic chamber, and are really good at ignoring the sound of your heart beat, the noise floor on the i9 isn't likely to be the limit.

Originally Posted by robdean

In reality, the main contribution that such audio formats could make to a player like the i9, beyond the placebo effect, would be to add vastly to power consumption as the DSP chewed on ever more data per second encompassing ever higher frequencies beyond human hearing.

Agree 100%. Doubling the amount of work the CPU must do for no reason is not a good idea on battery powered devices.

[robdean]

Excellent point from Saratoga that modern processing routinely uses a greater internal bit-depth for the math, so for practical purposes my comment was ill-considered in the examples I gave. I guess work-in-progress dynamic headroom is the main practical benefit, but I'm sure Saratoga and others here could explain ramifications of bit depth far better than I ever could, and probably have done so already endless times elsewhere on this site!

And sure, the noise floor of the i9 isn't the limit at all: in practice I'm certain 16 bit is everything required for hifi playback to satisfy me on any equipment. In the previous post I was being a bit 'dry', trying (but evidently failing!) to suggest that the i9, wonderful as it is, is flattered by the suggestion that it stretches the limits of 16 bit fidelity let alone anything greater.

BTW, Skip252's links above re. hi-res audio leads to some really clearly written expert exposition which is well worth reading by those who are still on the fence, including a convincing technical case for 24/192 potentially sounding *worse* than 44.1/16 in many real world implementations!

[WalkGood]

I believe that no one has mentioned that the i9:

• Does not have an expansion slot.
• Has limited space - available in 8 and 16GB, yes I know they also made a 4GB model but I’ve not seen that for sale in a long time and now that it’s discontinued I’d guess it’s even harder to find.

Ok the i9 does play FLAC and I’ve tested it on mine, but knowing the facts I’ve just mentioned it’s crazy to use large file sizes on such limited space.

[skip252]

Having a friend play different formats as a test won't have the same reliability as double blind testing. There's any number of ways the tests can go wrong like that. The Clever Hans effect is just one of them. As long as anyone involved knows which file is being played it's possible for that information to be transmitted. The tip off may be absolutely unconscious for all parties involved but it's possible for them to occur.

Double blind testing such as ABX is designed to level the playing field so that there's no way for you to be informed as to which file is being played until the test results are revealed. By hiding the identity of the files, fast switching, level matching the volumes and time syncing you will only be hearing the files themselves with no clues as to which is which. I've known several people that performed ABX testing according to the standards and were surprised that what they thought was a night and day difference was caused by some other factor than the format of the file itself.

You may want to read this article on the audible value of hi-res playback. 24/192 Music Downloads...and why they make no sense I think you may also find A Digital Media Primer for Geeks interesting. There's a section in there as to the value of higher resolutions during production.

[robdean]

I don't know, WalkGood...

...maybe some folks only own one album, and that's why they are so concerned for it to sound really Really REALLY good?


;-)

[WalkGood]

Originally Posted by robdean

I don't know, WalkGood ... maybe some folks only own one album, and that's why they are so concerned for it to sound really Really REALLY good?

Doubtful but funny ... I figure I'd mention because I see it (i9) mentioned in post #1, 28 and 36. I wouldn't bother using FLAC on such limited space.

[robdean]

WalkGood, I roll with -v0 myself, but I was just guessing that at a little over 40 minutes, 'Dark Side of the Moon' would maybe come in around 3Gb as a 24/192 wave file... and you know, more is, like, better and stuff... So wow! Hit me with those ticking clocks, Man... Wait for the chimes... Woahhhh!!!