PDA

View Full Version : Confirmation of bitrate?



dhallag
2016-04-16, 10:53
Hi Everyone. As a reminder, I have a squeezebox touch running the enhanced digital out via USB to a Jolida Tube DAC III that process 192/24. Is there an easy way to verify the bitrate my music is playing at? I have plenty of 192/24 96/24 files but I always get thrown by the "Converted to 705kbsp FLAC." In looking at a debug of my conversion log, I do see these errors:

getConvertCommand2 (382) Rejecting - because required capability T not supported
getConvertCommand2 (382) Rejecting - because required capability D not supported

for every FLAC file. But I really have no idea what this means.

Thanks much for your help in advance.

darien

Julf
2016-04-16, 12:22
Is there an easy way to verify the bitrate my music is playing at?

What happened to good old "trust your ears"? :)

Mnyb
2016-04-16, 12:37
The 705kBps is bogus , its a bug LMS just presents a placeholder value , so you can discard that .

Does the conversion actually work ? do you get sound ?

Do you have a complete log for the conversion process , somewhere in there you see the SoX parameters used . But the squeezebox environment is such that the player reports to the server what it can do
and the server provides that, further if you sync players the least capable player decides , but its never lesss than 24/44.1 or 24/48 with a Touch you should get 24/88.2 or 24/96 .

And if your read the spec sheet .http://www.jolida.com/product/glass-fx-tube-dac-iii

USB: 24/96 up sampling to 192 Hz

Toslink and coaxial: 24/192Hz

You see that your DAC supports 24/96 over USB so to avoid transcoding of 24/192 files, use the coaxial output .

IF 24/96 files gives transcoding then i suggest visiting the EDO tread for advice it can be some USB issue ? so try coaxial yet again .
USB implementation vary wildly between DAC's and Touch support is not on par with a full computer , so USB is not always automagically the best choice in every case .

pablolie
2016-04-17, 10:45
Like Mnyb stated - there's a bug in the way LMS reports higher bitrates. Ignore it. Just make sure that under Settings>Advanced>File Types you have your relevant stuff set up to "Native". Then you can be sure your SB is doing it's best to stay true to the original, which is 192-24 with the Touch, and I think it's 96-24 with classic versions. From that point on it's your DAC's capabilities that kick in, LMS and the SB provided an error free digital path. :-)

Mnyb
2016-04-17, 13:23
Like Mnyb stated - there's a bug in the way LMS reports higher bitrates. Ignore it. Just make sure that under Settings>Advanced>File Types you have your relevant stuff set up to "Native". Then you can be sure your SB is doing it's best to stay true to the original, which is 192-24 with the Touch, and I think it's 96-24 with classic versions. From that point on it's your DAC's capabilities that kick in, LMS and the SB provided an error free digital path. :-)

Yes you should basically not change any file type settings unless you have to solve a problem in a special setup , for most people LMS just do the rigth thing rigth out of the box .

I just wonder if OP can test if 24/96 really get transcoded ? it should not in his case .

Or he may have as you say disabled "native" for flac etc :) there are some in whos believe system you should send pcm/wav to the player , which is nonsense . but the effect is transcoding for any flac file .

Julf
2016-04-18, 03:31
there are some in whos believe system you should send pcm/wav to the player , which is nonsense.

"but it is more bits, so has to be better!" :)

garym
2016-04-18, 05:43
it's best to stay true to the original, which is 192-24 with the Touch, and I think it's 96-24 with classic versions. From that point on it's your DAC's capabilities that kick in, LMS and the SB provided an error free digital path. :-)

Agree with your point, but as an aside, just correcting above regarding bit rate max for different players:

Touch - 24/96 (unless have EDO installed, then 24/192)
Radio - 24/48
Transporter - 24/96
Boom, SB Classic, SB3 - 24/48

dhallag
2016-04-18, 06:56
The 705kBps is bogus , its a bug LMS just presents a placeholder value , so you can discard that .

Does the conversion actually work ? do you get sound ?

Do you have a complete log for the conversion process , somewhere in there you see the SoX parameters used . But the squeezebox environment is such that the player reports to the server what it can do
and the server provides that, further if you sync players the least capable player decides , but its never lesss than 24/44.1 or 24/48 with a Touch you should get 24/88.2 or 24/96 .

And if your read the spec sheet .http://www.jolida.com/product/glass-fx-tube-dac-iii

USB: 24/96 up sampling to 192 Hz

Toslink and coaxial: 24/192Hz

You see that your DAC supports 24/96 over USB so to avoid transcoding of 24/192 files, use the coaxial output .

IF 24/96 files gives transcoding then i suggest visiting the EDO tread for advice it can be some USB issue ? so try coaxial yet again .
USB implementation vary wildly between DAC's and Touch support is not on par with a full computer , so USB is not always automagically the best choice in every case .

thanks for the detail. I guess my question is how do I know if the files are being transcoded?

dhallag
2016-04-18, 07:04
and I haven't changed anything under file types. yes I do get sound and yes it sounds good... so to answer the relevant question of trusting my ears, part of this is my personal experiment to see how different file types sound -- I have the album thriller in 192/24, 96/24, 44.1/16 and 320 MP3... so I just want to be sure that my LMS is doing what it should be doing...

Mnyb
2016-04-18, 09:50
and I haven't changed anything under file types. yes I do get sound and yes it sounds good... so to answer the relevant question of trusting my ears, part of this is my personal experiment to see how different file types sound -- I have the album thriller in 192/24, 96/24, 44.1/16 and 320 MP3... so I just want to be sure that my LMS is doing what it should be doing...

Use coax then you can try all 4 resolutions if it transcodes you see SoX and Flac processer running when you playing and also if you set some ( i don't remener exactly ) parameters in the log settings to "debug" then you can see the settings LMS actually used . And actually if you get that info transcoded to 705 on the player then it is transcoded you should not see that otherwise .

Is this Flac files ? If they are wav AIFF or ALAC the transcoding can happen just to make them Flac , but the actual resolution is like the original .
For wav LMS does that to save bandwith if fr example the players is on wig and Flac has better error correction ( I don't now if LMS or the player use that fact , yes a Flac file can be checked years after ripping ).
For ALAC , the player supports ALAC but not hirez ALAC , same for AIFF .

You can make a file type settings change to stream wav natively if you have a lot of wav files and your network is up for it .
But I recommend to convert those to Flac so that file tags have better support , it looks nicer on most players ( but sounds the same )

Great a test :) suggestions .

Use some software to make your own dowsampled version from the 192/24 version , don't trust the different versions to be mastered the same they usually are not , then you hear what the different versions of the album sounds like ( which afaik is the difference in general ).

Make sure they are level matched , not only same volume but no replay gain tags or other volume manupilation and different versions can be differently loud another reason to not test that way .

Try a blind test , tag them exactly the same and shuffle around

drmatt
2016-04-23, 07:01
Always disable replaygain when doing critical listening as it's done in the digital domain and adds aliasing at the LSB. And have the output volume at 100% too and use an external pre amp.

Julf
2016-04-23, 07:12
Always disable replaygain when doing critical listening as it's done in the digital domain and adds aliasing at the LSB. And have the output volume at 100% too and use an external pre amp.

I am always careful of any advice that includes the word "always". :)

I guess this really belongs down in the "audiophile" section, but most digital volume adjustments are done at more than 16 bit precision, so covered by the self-dither of the inherent noise in the actual music.
In any case, the resulting "lack of precision" manifests itself as (very low-level, way beyond the source material) noise. "Aliasing" is usually used for something else, and has nothing to do with volume adjustment.

drmatt
2016-04-23, 07:28
Well, in my case I have an external DAC so I want the original bits off the CD sent to it at 44k/16b. The volume and replaygain are applied within this precision and are audible.

If your assumption of it being processed at 24 bit is true or you're using the analogue outputs anyway then fair enough the difference will be marginal.

But the question was about audiophile comparisons, so fair comment, no?

Mnyb
2016-04-23, 07:35
Squeezeboxes always put out 24bit data 16 bit material is paddel with 8bits and then the volume control is made in 24bit it's not dithered but the steps are cleverly chosen so and truncation happens at rely low volumes where the volume is low :) so you practically don't hear that .

So both volume and replay gain is added within 24 bits .

Julf
2016-04-23, 08:08
Well, in my case I have an external DAC so I want the original bits off the CD sent to it at 44k/16b. The volume and replaygain are applied within this precision and are audible.

So you have a 16-bit DAC that does the volume processing at 16 bits? Must be a really old one. If it is done in software, it is done at a precision greater than 16 bits and then rounded/truncated back to 16 bits, so you don't "lose" any precision.


If your assumption of it being processed at 24 bit is true or you're using the analogue outputs anyway then fair enough the difference will be marginal.

Not just marginal but more importantly inaudible.


But the question was about audiophile comparisons, so fair comment, no?

Only if "audiophile" means "anything is possible" (which it often does). :)

drmatt
2016-04-23, 08:25
Well I have a 16 bit DAC with an analogue pre-amp and volume control after it. The DAC is 24 bit capable to be fair, but I choose to use the pre amp section for volume control instead of the SB output level as I find it better...

drmatt
2016-04-23, 08:34
Squeezeboxes always put out 24bit data 16 bit material is paddel with 8bits and then the volume control is made in 24bit it's not dithered but the steps are cleverly chosen so and truncation happens at rely low volumes where the volume is low :) so you practically don't hear that .

So both volume and replay gain is added within 24 bits .

Cool, so you should get 48db of attenuation without degradation. Good! Assuming the analogue section post-dac doesn't just drop into its noise floor.

All that said.. Is there any way to find out what bit/kHz PCM signal an SB touch is outputting? My DAC either works or doesn't and doesn't indicate what it's receiving..

Julf
2016-04-23, 08:38
Well I have a 16 bit DAC with an analogue pre-amp and volume control after it. The DAC is 24 bit capable to be fair, but I choose to use the pre amp section for volume control instead of the SB output level as I find it better...

Fair enough, but not something I would advice other people to always do.

drmatt
2016-04-23, 08:42
Fair enough, but not something I would advice other people to always do.
My view was that disabling RG and setting fixed 100% vol will never hurt the quality and it /can/ degrade it. So I figured it was a fair punt!

Julf
2016-04-23, 09:01
My view was that disabling RG and setting fixed 100% vol will never hurt the quality and it /can/ degrade it. So I figured it was a fair punt!

Disabling RG will not hurt quality, but you lose the RG functionality, so there should be a somewhat better reason that "it won't hurt". As to volume control, digital volume control can often be much better than analog, so setting fixed 100% vol is actually harmful to sound quality in many cases.

Mnyb
2016-04-23, 10:45
Well if the dac is a true 16bit design i would not use digital volume, the spdiff standard is a bit weird it is basically 16 bit + 8 extra bits legacy design just discards the extra bits ( backwars compatible ) so you would runcate at the dac interface if using digital
Volume you would never get 16 bits ? On the other hand i would not use such a dac 😊

drmatt
2016-04-23, 11:08
It's a 24b/96k DAC. Have to admit I assumed the SB was sending 16 bits because the source is a 16 bit flac.. Maybe that's not true.

I am however certain that using the digital volume or replaygain on the SB results in poorer, not just quieter, sound...

drmatt
2016-04-23, 11:13
Disabling RG will not hurt quality, but you lose the RG functionality, so there should be a somewhat better reason that "it won't hurt". As to volume control, digital volume control can often be much better than analog, so setting fixed 100% vol is actually harmful to sound quality in many cases.
I said "for critical listening". I am religious about using replaygain at all other times ..

Some types of digital volume controls can be better than some types of analogue ones I have no doubt (this is an obvious statement). In this case, I find the SB digital volume degrades sound, so I spoke as I found..

Julf
2016-04-23, 11:45
I am however certain that using the digital volume or replaygain on the SB results in poorer, not just quieter, sound...

What's your certainty based on?

drmatt
2016-04-24, 01:50
My ears. Haha. YMMV.
Obviously you can't measure sound quality can you??

Julf
2016-04-24, 02:21
My ears. Haha. YMMV.

Yes - the key part of that is "YMMV". And "haha". :)


Obviously you can't measure sound quality can you??

Obviously I can. A lot of what makes up good sound quality as an objective criteria can be quantified and measured, and subjective preferences can also be measured (in the meaning of "assignment of a number to a characteristic of an object or event, which can be compared with other objects or events") by properly controlled listening tests (see ITU-R BS.1534 and BS.1116).

How do you think audio gear is designed? Solely by ear?

Should we move this down to the "audiophile" section?

Mnyb
2016-04-24, 04:17
I'm even trying to use small amounts of digital attenuatiin to my beneffit ? Result inconclusive , but one noted side effect is that in practice you dont hear the squeezebox volume control imo at least not small amounts of it .

We had discussion about intersample overshots some over sampling filters are likelybto do when encountering to hotly mastered tracks so dropping -3 to -6dB could actually be better .

But it serms masked by the in general horrible sq of such regordings :)

Well i have completely abandoned the notion that playback must be bitperfect at all cost , with current 24 bit systems you are not likly to hear any difference .

But it can aleays be better like the fully dithered volme i have in my meridian system .

It should be noted that there are no real 24 bit audio anywhere ( except some electronica ) both ADC and DAC is practically limited to about 21 bit in practice a squeezebox touch analog out would be aproxmately 17 bits .

drmatt
2016-04-24, 05:49
Obviously I can. A lot of what makes up good sound quality as an objective criteria can be quantified and measured, and subjective preferences can also be measured (in the meaning of "assignment of a number to a characteristic of an object or event, which can be compared with other objects or events") by properly controlled listening tests (see ITU-R BS.1534 and BS.1116).

How do you think audio gear is designed? Solely by ear?

Should we move this down to the "audiophile" section?

Probably best to just move it to the pedantry section.

Sent from my XT1562 using Tapatalk

drmatt
2016-04-24, 05:53
It should be noted that there are no real 24 bit audio anywhere ( except some electronica ) both ADC and DAC is practically limited to about 21 bit in practice a squeezebox touch analog out would be aproxmately 17 bits .

Indeed, I'm told my DAC manages about 18 bit of clear resolution but I think the pre-amp has quite a high noise floor. (The DAC is in the same box so it can't be measured separately..)

drmatt
2016-04-24, 05:58
Meanwhile, still would like to find out how to find a live value of the in-use bitrate on the spdif outputs..

Sent from my XT1562 using Tapatalk

Mnyb
2016-04-24, 06:20
Meanwhile, still would like to find out how to find a live value of the in-use bitrate on the spdif outputs..

Sent from my XT1562 using Tapatalk

If it does not transcode it is what the file is + 8 zeroes on the spdiff out if it's 16 bit material or something else if you use the volume ( but it is bitperfect at 100% volume ), if it transcodes you can set up the log to debug and observe the chosen transcoding parameters .

To my knowledge there is no way for LMS to presents live data in this regard . But the info on the player at least tells if it's transcoded or not .

But is it not good enough to know that if its not transcoding and you use 100% volume you get bit prefect output ? In other words what the file is ?

Julf
2016-04-24, 09:30
Probably best to just move it to the pedantry section.

Indeed. Factual accuracy is a true sign of pedantry.

Sent from my ZQ864e36 using Firefox.

pablolie
2016-04-24, 10:15
due to sheer convenience i dial my Dac+preamp at a medium-high listening level, and fine-tune the volume with the LMS remote on my iPad. i can not ever hear a difference in sound quality, even with 192/24 recordings (not that i can tell or care to tell a difference between 24/192 and 16/44 of very well recorded material, either).

drmatt
2016-04-24, 12:22
If it does not transcode it is what the file is + 8 zeroes on the spdiff out if it's 16 bit material [...]

But is it not good enough to know that if its not transcoding and you use 100% volume you get bit prefect output ? In other words what the file is ?

Thanks for that, so that is what we expect to see, but.. there's more.

Behind this question is an observation. Last night I was listening and found the sound pretty lame. Shouldn't be lame. Nice flac ripped from my own CD using full on cdparanoia. Sounded awesome last week. Thought I had fixed this a few weeks back but no.. It's back. So then about half an hour in and I'm scratching my head still the SBT crashes and reboots.

After reboot it comes up fine, I start playing the same stuff. Still sounds lame. Get off my ass and unplug the coax cable to the DAC, and plug it back in and restart playback.. And lo and behold it sounds awesome again. Wtf?, says I.

So. My suspicion is I have a dodgy spdif output chip that is dropping output stream resolution at times, but on a warm relink to the DAC goes back to full res. That, is why I would love to get data from the driver to tell me what it's outputting. At no point was there evidence that the flac was being transcoded to anything else. This change is purely between SBT and DAC. The DAC, fwiw, is a Naim Supernait.

I will of course swap the SBT and see if another one is more reliable.. Just hoping to figure out a way to troubleshoot further.. For example, does anyone know what driver is used for the sound card in the SBT? Can it be loaded with debug? Does it have any sense state in /sys from the kernel?

drmatt
2016-04-24, 12:23
Indeed. Factual accuracy is a true sign of pedantry.

Everything you said was obvious and not "news" to anyone. You still can't measure to tell the difference between the type of sound I prefer and the type you prefer.

Mnyb
2016-04-24, 12:39
Everything you said was obvious and not "news" to anyone. You still can't measure to tell the difference between the type of sound I prefer and the type you prefer.

But the point is when there is no difference a human could hear ? then taste does not come into to it at all . We are there with most digital sources of reasonable quality .

Most if not all audiophile quibles is about inaudible things imagined to be audible due to lack of controlled listening test . I dbt IS a subjective test using ones hearing but with as much bias as possible removed .
Most "issues" these days are of such smal influence that human bias are magnitudes bigger .

There are of course tastes like tube gear with readily available artefacts that are clearly audible and preferred by some (instead of hearing the source material as intended by the artist ) .

I have quite wide musical taste i can not choose equipment by "taste" it has to be as neutral as possible , that makes some music sound terrible due to the fact that the recording actually sounds like that :/ but it helps greatly when the material has the right qualities .

Taste comes into speakers as they are currently always very coloured especially combined with room acoustics ,thtas the last frontier, here we have tp pick our poison .

Julf
2016-04-24, 13:06
My suspicion is I have a dodgy spdif output chip that is dropping output stream resolution at times, but on a warm relink to the DAC goes back to full res.

Rather unlikely. I would suspect your DAC getting into some funky mode, and resynchronizing when you unplug and replug the cable.

Julf
2016-04-24, 13:10
You still can't measure to tell the difference between the type of sound I prefer and the type you prefer.

You keep making claims without any factual support. We don't know that we can't measure to tell the difference between the type of sound you prefer and the type I prefer until we try. It might be possible that our preferences differ based on easily measurable parameters.

drmatt
2016-04-24, 13:44
Rather unlikely. I would suspect your DAC getting into some funky mode, and resynchronizing when you unplug and replug the cable.
Also possible. Annoying if so cos trust me I do not have a spare..

drmatt
2016-04-24, 13:51
You keep making claims without any factual support. We don't know that we can't measure to tell the difference between the type of sound you prefer and the type I prefer until we try. It might be possible that our preferences differ based on easily measurable parameters.
How do you suggest I go about measuring my perception that the sound wasn't as good? For the sake of a throwaway comment on an internet forum I'm going to wheel out some pro recording gear and analysis equipment? Nope. Maybe it could be measured but it ain't going to be, so I said what I thought and you tried to educate me. Well, good. Thanks.

Hope the OP appreciates your efforts to prevent opinions getting in the way of his pursuit of the truth.

Julf
2016-04-24, 13:54
Also possible. Annoying if so cos trust me I do not have a spare..

So worth trying to isolate the issue. Next time your system goes into the weird mode, try turning the DAC on and off to see if it fixes the problem.

Julf
2016-04-24, 14:04
How do you suggest I go about measuring my perception that the sound wasn't as good?

I suggest you check out the two ITU recommendations I mentioned.


Hope the OP appreciates your efforts to prevent opinions getting in the way of his pursuit of the truth.

Nothing wrong with personal opinions when stated as such...

drmatt
2016-04-24, 14:17
But the point is when there is no difference a human could hear ? then taste does not come into to it at all . We are there with most digital sources of reasonable quality .

Most "issues" these days are of such smal influence that human bias are magnitudes bigger .

[...]that makes some music sound terrible due to the fact that the recording actually sounds like that :/ but it helps greatly when the material has the right qualities .

Taste comes into speakers as they are currently always very coloured especially combined with room acoustics ,thtas the last frontier, here we have tp pick our poison .

I dont disagree with you. I'm not going to sit here and tell anyone I think I can hear stuff that isn't there.. I would however say that it's nearly impossible to actually measure the sound that's hitting my ears though. Measuring the electronics is easy of course..

So perhaps something was off with my setup when I did use the digital volume control? Dunno. I noticed it though, and there's plenty of logic that says I want bit-perfect hitting my DAC, so I tried to help the OP by adding to check this out..

The amp/speaker/air interface is totally where the game is. Speakers are pretty imperfect overall and yeah most people choose a speaker they like.. Though I find that after enough exposure to the specific response of a specific speaker I will probably accommodate it..

And I'm totally with you on the awful quality of a lot of recorded music.. Am spoiled by the ones that are good but some of the rest can be pretty painful.

drmatt
2016-04-24, 14:24
So worth trying to isolate the issue. Next time your system goes into the weird mode, try turning the DAC on and off to see if it fixes the problem.
Pretty sure I have power cycled the amp during this phase before and it didn't snap out of it but yes this is obviously one to try. I have certainly switched between digital inputs and switched to/from analogue inputs (the entire DAC board shuts down when this happens) so I really don't think it's the DAC...

drmatt
2016-04-24, 14:26
Fwiw (warning: not measured, for those that care) I have the impression that the background noise level is noticeably higher when the issue is happening...

Gonna swap the SBT..

Mnyb
2016-04-26, 18:55
Fwiw (warning: not measured, for those that care) I have the impression that the background noise level is noticeably higher when the issue is happening...

Gonna swap the SBT..

That's weird ? Assuming you have a normally very quit DAC ? Much gain in the system ?

pablolie
2016-04-26, 21:52
if you think it's the SBT, i personally would have no qualms about a factory reset on the spot. i have done that with several devices over the years. i also like to go for a clean reinstall of LMS (it's why i keep a virginal copy of the original virtual machine) whenever some weird issue appears. i don't spend time fixing, i just go back to something that worked... :-)

drmatt
2016-04-27, 01:22
That's weird ? Assuming you have a normally very quit DAC ? Much gain in the system ?
It's not particularly quiet no, and the gain on the DAC through the pre-amp is very high which makes this worse. Result is noise level is relatively obvious if it steps up a few dB (such as stepping from 24 bit noise floor to 16 bit noise floor would cause!).

One of the foibles of the Supernait. Using just the power amp section (such as when it's in AV mode) the power amp is silent (and superb).

When I don't have this problem, the noise on the DAC inputs is not intrusive either and the sound quality is just incredible - even from a lowly SBT playing flacs over WiFi from a home built server, and running off its bog standard wall wart PSU. :)

Julf
2016-04-27, 02:37
the sound quality is just incredible - even from a lowly SBT playing flacs over WiFi from a home built server, and running off its bog standard wall wart PSU. :)

As it should be in any decently designed DAC - if your DAC is sensitive to source, there is something wrong.

drmatt
2016-04-27, 04:29
As it should be in any decently designed DAC - if your DAC is sensitive to source, there is something wrong.
My point exactly. Bits is bits, so if it isn't awesome, something is wrong.