24/96 (HIRES files for playback...)

[Chronon]

Pardon me, but you were the one who said this:

anyway that post is getting on for being nearly a year old so something could have changed or has no development happened on RockBox for nearly a year   

The implication that if the one feature that you are interested in has not changed then this means that no development has taken place at all seems a bit like trolling to me.  The "eye roll" only accentuates this perception.  Honestly, it would be easy for the people who have worked hard on Rockbox during this time to take offense about that.  However, my comment was not intended to be overly harsh.  I only wanted to draw your attention to the fact that whether or not this feature has changed and whether or not Rockbox has undergone development are two distinct things.

[nc8000]

Well I come here as a new user asking a question that is of interest and all I basically get told is that I'm an idiot for not finding the answer myself. I did not expect anybody to start resarching documentation for me but hoped that perhaps somebody who actually knew the answer (which it does not seem that either of you two do) would read the question and come with a polite answer.

Also I'm not in posesion of measuring equipment that would let me measure what the specs of the signal coming out of the optical output on an iRiver is (and I expect not many have that) and neither do I have a SACD player or any other device that delivers anything better than CD quality so I don't know what to expect and therefore cant compare the sound as I also dont have those specific recordings in ordinary cd version.

I have now poked around a bit more and read the spec sheat for the Phillips ADC/DAC chip that is in the H1XX units and it seems to be able to handle 24/96 (but not 24/192) so it actually seems that it would be a usefull thing at least in these units.

[saratoga]

I did not expect anybody to start resarching documentation for me but hoped that perhaps somebody who actually knew the answer (which it does not seem that either of you two do) would read the question and come with a polite answer.

Getting angry when people don't know things you also do not know seems like a reasonable thing to do.

Also I'm not in posesion of measuring equipment that would let me measure what the specs of the signal coming out of the optical output on an iRiver is (and I expect not many have that) and neither do I have a SACD player or any other device that delivers anything better than CD quality so I don't know what to expect and therefore cant compare the sound as I also dont have those specific recordings in ordinary cd version.

Yes, that was his point.  If you can't tell the difference (which apparently you cannot), then it doesn't matter.  Unless I'm missing something, you just answered your own question?

[nc8000]

I did not expect anybody to start resarching documentation for me but hoped that perhaps somebody who actually knew the answer (which it does not seem that either of you two do) would read the question and come with a polite answer.

Getting angry when people don't know things you also do not know seems like a reasonable thing to do.

Not following you on that one. Why do people who also don't know the answer have to tell me I'm an idiot for asking instead of just finding out for my self, they could just keep quiet

Also I'm not in posesion of measuring equipment that would let me measure what the specs of the signal coming out of the optical output on an iRiver is (and I expect not many have that) and neither do I have a SACD player or any other device that delivers anything better than CD quality so I don't know what to expect and therefore cant compare the sound as I also dont have those specific recordings in ordinary cd version.

Yes, that was his point.  If you can't tell the difference (which apparently you cannot), then it doesn't matter.  Unless I'm missing something, you just answered your own question?

I never said that I couldn't tell the difference, I don't know if I can tell the difference as I have never heard a hires signal before and am not in possion of a device that could play me such a signal, that is why I was interested in whether the RockBoxed iRiver COULD deliver such a signal so I could find out for myself. Say I play the same tune as a 16/44 file and a 24/192 file and don't hear a difference, does that mean that I can't hear the difference or that the unit cant play the full resolution of the 24/192 file and therefore are playing both as 16/44 ?

[Chronon]

Honestly, you would do yourself a big favor to find a way to run an ABX test properly since even moderate bitrate lossy files are transparent for many people.  I have not been able to find evidence of people who can reliably distinguish 24/96 audio from 16/44 but have found accounts of studies that unanimously show the converse -- indistinguishability. 

Edit: Pointing out that you can answer this for yourself is good advice.  And I don't see where Llorean ever called you an idiot.  He just pointed out that you could probably answer this for yourselves much faster by testing than waiting around for an answer.  Many computers have a digital input these days.  This would give you a way to characterize the data stream coming from the digital line out.  I also have a digital TOSlink input on my stereo receiver that accepts 24/192.  Of course, we don't know what equipment you have, but I still think it's good general advice to point out that you can probably find a way to test this yourself.

[nc8000]

Honestly, you would do yourself a big favor to find a way to run an ABX test properly since even moderate bitrate lossy files are transparent for many people.  I have not been able to find evidence of people who can reliably distinguish 24/96 audio from 16/44 but have found accounts of studies that unanimously show the converse -- indistinguishability. 

Now that is a sound comment, the problem is that I'm not going to go buy a SACD player just to test if I can hear the difference which is why it would be fantastic if the RockBoxed H-140 COULD play these files to full resolution. Up until getting a used H-140 all my ripped music was in 320KB aac. From the lineout of my unit into my main home rig I would say that I in an ABX test can spot the full 16/44 uncompressed file 8 out of 10 times a bit depending on the kind of music. Doing the same test from the headphone out of my unit in everyday transport situations I would agree that I can't detect the difference reliably.

Don't get me wrong, I think the RockBox project is an absolutely fantastic achievement and I'm full of respect for the work loads of people have put into it but I did not feel that my entry into this world was greated with any level of respect or helpfulness, just scorn and flippant remarks for being such an idiot that cant find out anything for himself.

[dreamlayers]

SACD isn't even PCM; it's DSD. Essentially that is 1 bit at 2.8224 MHz.  Also there's no convenient way to rip a SACD.  If you somehow managed to get the DSD data, you'd have to convert it to PCM.  DVD-Audio supports up to 24/192 losslessly compressed PCM and it can be ripped.

What does being able to tell the difference between uncompressed 16/44 and 320 kb/s AAC have to do with this?

[Llorean]

Again, I'm sorry you read those as scorn and flippant remarks. Try reading them without imagining intonation and you'll see they're simply suggestions I made.

[nc8000]

I give up.

Some admin please delete my user and posts and forget you ever heard about me and you all go back into your smug little selfasured world where outside idiots really aren't wanted.

Over and OUT, OUT, OUT ...

[Chronon]

 

If you feel the need to leave then you are free to do so.  You can easily delete your own account if you like. 

[saratoga]

I give up.

Some admin please delete my user and posts and forget you ever heard about me and you all go back into your smug little selfasured world where outside idiots really aren't wanted.

Over and OUT, OUT, OUT ...

You've said this before and then come back anyway, so its probably better we just leave your account alone in case you decide to come back and throw yet another fit.

[rmaniac]

Might we consider some work on this for the summer of code?

Rockbox plays back files of higher bit depth and higher sample rates than 16 bits and 44100 Hz just fine right now, but not with full quality. Here is what happens:

• Bit depth: Rockbox will decode all formats to as high a bit depth as possible for both lossy and lossless formats. All internal processing is also done at 32 bits bit depth. However, all current targets truncate to 16 bits at the output. This can be fixed if someone cares enough about it to do some programming, but the DACs in current targets really aren't good enough for this to make much sense.
  
• Sample rate: Rockbox will decode higher sample rate files (like 96 kHz) just fine, but resamples to 44100 at the output, so some quality is lost (especially considering the extremely crappy resampler we use). Again, someone needs to care enough to do some programming to fix this, and it makes more sense to fix this problem than the bit depth problem.
  

[Llorean]

I don't necessarily think it's really a GSoC scale project. At the very least, there are already several proposals that have a wider impact than this one would.

[rmaniac]

I just think it would be nicer than things like.. scummvm. As much as I love it I just don't think it would be the most useful thing on rockbox.

[Llorean]

"Nicer" doesn't necessarily factor into it as much as "is it an appropriate amount of work being proposed by someone capable of doing said amount of work."

All proposals that answer yes to *that* question then basically get sorted on "niceness" or "how much we want"-ness or "how confident we are they'll finish"