Bit Transparency

[Darren]

So here's the thing...... I've been Googling around a bit, reading up on my new Squeezebox Touch ( as one does) and several reviews tell me that it is "bit transparent" up to 24/96 material. The same reviews go on to say that while the touch is great it doesn't have quite the same sound quality ( when used to output digital data to a DAC) as a top notch cd transport or a dedicated music server.
I don't understand this. Surely if the Touch is bit transparent it is passing the digital signal undamaged in any way? So surely it doesn't have any sonic signature? and as far as it's digital output goes cannot be improved upon?
Im sure im missing something here. Please help unknot my brain

[CentraL]

So here's the thing...... I've been Googling around a bit, reading up on my new Squeezebox Touch ( as one does) and several reviews tell me that it is "bit transparent" up to 24/96 material. The same reviews go on to say that while the touch is great it doesn't have quite the same sound quality ( when used to output digital data to a DAC) as a top notch cd transport or a dedicated music server.
I don't understand this. Surely if the Touch is bit transparent it is passing the digital signal undamaged in any way? So surely it doesn't have any sonic signature? and as far as it's digital output goes cannot be improved upon?
Im sure im missing something here. Please help unknot my brain

Could you provide a link to that review? or comment you refer to?

It sounds weird to say the least especially the cd - output superiority... did they measure it somehow, or was it only subjective listening?

[Welder]

Getting the zeros and ones to the dac at the same bit depth and frequency as they are on the file or CD is only part of the job.
The most obvious and frequently talked about hurdle is to get those bits from place to place jitter free.
Next you have the way particular players and operating systems retrieve and present the bits. They all seem to have their own way of doing it and trying to find out exactly what a particular player does to the zeros and ones is almost impossible as the developers, if they have a particularly good sounding player, don’t want anyone nicking their ideas.
There is also the way a particular operating system stacks the various driver and audio management layers which apparently can affect the sound.
As far as I can find out, not one OS or music player just steams data from point to point.

Is it all a bit of a case of audio nervosa? Hmm, maybe but I’m pretty skeptical about most thing audio but I can hear a difference between a well sorted music server and a slaved Logitech Touch and its not as negligible as I would like by any means.

[Stratmangler]

You're not missing anything Darren.

The Touch has both wireless and wired LAN connectivity, and the inference is that this degrades the sound, because computer network cannot deliver time aligned data to the player.
It's utter bo**ocks, because the player receives packet data which is stored in a buffer, and the player unpacks and plays that data in a time aligned order , but yer average audiophile is suspicious of such things, and paranoid about degrading sound quality, so it's easier to knock it for knocking's sake.

If you want to get an idea as to how big the buffer is connect the Touch to your serving machine via an ethernet cable, then disconnect the cable whilst music is playing.
Eventually the music will stop.
And that word eventually is key - with a FLAC file ripped from CD it will continue for a least 30 seconds before the music stops.
This means that there is plenty of time for the player to sort the bits out into correct order.

The truth behind all the knocking is that it is getting increasingly difficult to justify spending small fortunes on CD players and what have you, because for a fraction of the cost of these über machines you can obtain a superb bit of CE kit that delivers 99% of the sound quality of these wonder transports.

[Welder]

Well that’s that sorted then :eyebrows:
:lolsign:

[Stratmangler]

Well that’s that sorted then :eyebrows:
:lolsign:

I'm sat here behind the sofa waiting for hostilities to commence :sofa:

[Welder]

Ten metres from behind the sofas then if you will my good man and tomatoes only to be thrown. :lol:

For me at least the whole business is rather interesting. As I said, I’m pretty skeptical by nature but………….

For example, you can take a couple of cables with identical measurements (as best we can measure with the current knowledge a tools availible) and people say they can hear a difference.

If I say, particularly on this forum I can hear a difference between such cables, no one would bat an eyelid……come to zeros and ones………whoa, it’s a deiffent ball game :eek:
Just a thought.

But, if given bit perfect output then every operating system and player should sound exactly the same? :scratch:

[Stratmangler]

I use two of these (http://www.microdirect.co.uk/Home/Product/50228/Belkin-Digital-Coaxial-Audio-Cable-3-6m) to connect my SB Touch and DVD player respectively to my Caiman/Gator DAC, and to me I get a slight improvement in sound quality.

Yet they're there to pass "digital" information between devices.

In reality the digital information is rapidly switched variation between two analogue states, with nominal trigger and threshold values.
So the carrier is of importance.

There is another major factor to take into consideration - the feelgood factor.
With these cables in circuit my feelgood factor is immense, 'cos there are some folk out there who have paid many thousands of percents more in the pursuit of audible improvements, and I've spent 2 quid.:ner:

[Tim]

The truth behind all the knocking is that it is getting increasingly difficult to justify spending small fortunes on CD players and what have you, because for a fraction of the cost of these über machines you can obtain a superb bit of CE kit that delivers 99% of the sound quality of these wonder transports.
This is making me smile too .... and that's all I have to say on the subject ;)

[Welder]

So was that a yes or a no then Chris? ;)

[Stratmangler]

Ten metres from behind the sofas then if you will my good man and tomatoes only to be thrown. :lol:

For me at least the whole business is rather interesting. As I said, I’m pretty skeptical by nature but………….

For example, you can take a couple of cables with identical measurements (as best we can measure with the current knowledge a tools availible) and people say they can hear a difference.

If I say, particularly on this forum I can hear a difference between such cables, no one would bat an eyelid……come to zeros and ones………whoa, it’s a deiffent ball game :eek:
Just a thought.

But, if given bit perfect output then every operating system and player should sound exactly the same? :scratch:

OS sounding identical ?
No !
OS drivers and their interaction with respective media players will make things sound different.

Squeezebox Server operates on a number of platforms.
Because there are no audio drivers required to run the server side of things in theory a Squeezebox should sound the same whether it's being run from a Mac, Windoze PC, Linux PC, as the Squeezebox itself does not require drivers either.

Be an interesting one to test:eyebrows:

[Welder]

But if the output is bit perfect and that apparently is what matters then why do different OS and players sound different when output is bit perfect? :scratch:

[Welder]

I’ve got some OS and music players to test on my dodgy music server lash up. I think there may be a bit more to it than meets the eye
Meanwhile, I’ll leave this for anyone else who wishes to state that bit transparency is all that matters.

http://www.mediafire.com/imgbnc.php/f4bb379d7b7478fff2a85f093cfc0b99df94365de6ce6a025b 0112fe10d10be72g.jpg (http://www.mediafire.com/imageview.php?quickkey=605o6bdp2q7hpeb&thumb=5)

:)

[Stratmangler]

But if the output is bit perfect and that apparently is what matters then why do different OS and players sound different when output is bit perfect? :scratch:

All of my comments are referring to the use of a Squeezebox (ie hardware) - yours are about using software players.

In theory the hardware I have mentioned, which is controlled by a Thin Server Client, should sound identical no matter which computer platform is being used to run the Server Client.

Software media players are a total minefield, and I for one refuse to get drawn into commenting on them and bit perfection.

[Welder]

Yep, my apologies Chris, I know the angle you were coming from but I’ve been dying to use that picture for ages :)
I’m having a similar debate with a couple of people at the moment who do in fact say that if output is bit perfect then there can be no difference in audio.

If my comments are out of order mate, let me know and I’ll get the pic taken down and move back in behind the sofa :doh:

[WAD62]

The data transfer from the SB server (whatever OS platform) to the SB player (whether via wi-fi or ethernet cable) is always bit perfect, as it's an asynchronous TCP/IP file transfer, which either works perfectly, or not at all.

The SB player is in this instance/context is simply a file transfer recipient, and has a reasonably large data duffer to this end. As Chris stated earlier I can turn off my router and music will continue to play for about 30 secs.

It's after this point that the problems/issues start, transcoding the FLAC, constructing the SPDIF signal, and communications with the DAC. No different to any other computer, an SB player is simply a mini linux computer after all.

The fact that the wi-fi receiver is working inside the same physical box may have an adverse/interference affect on the construction of the SPDIF signal, in the same way as a rubbish power supply may, the processing effort to transcode from FLAC, the onboard clock, the cable being used...etc.

The same with any other computer/player combination.

But the data transfer to the SB device is always bit perfect, that's how asynchronous file transfer must work. :)

Pass the shovel please...;)

[Welder]

Just to draw attention to the OP ;)

“I don't understand this. Surely if the Touch is bit transparent it is passing the digital signal undamaged in any way? So surely it doesn't have any sonic signature? and as far as it's digital output goes cannot be improved upon?”

No one is arguing that the SB isn’t bit perfect (well I’m not anyway)
The point is there are more factors to the quality of sound from file audio than bit perfectness.


(after you with the shovel please Will)

:sofa:

Damn, forgot the custard creams had run out.

[Stratmangler]

Just to draw attention to the OP ;)

“I don't understand this. Surely if the Touch is bit transparent it is passing the digital signal undamaged in any way? So surely it doesn't have any sonic signature? and as far as it's digital output goes cannot be improved upon?”

No one is arguing that the SB isn’t bit perfect (well I’m not anyway)
The point is there are more factors to the quality of sound from file audio than bit perfectness.


(after you with the shovel please Will)

:sofa:

Damn, forgot the custard creams had run out.

It's getting beyond the point of shovels now, so perhaps one of these will help......

http://images.mirror.co.uk/upl/m4/oct2008/6/5/2B1AD0B8-E9D7-B16C-8B208B31E9A95DF3.jpg

[Canetoad]

All I know is my Touch through a Mark Grant coax cable to my Caiman/Gator sounds pretty good for what it has cost me. To improve things further conversion from Flac to PCM is done on the PC and I have switched off the Touch screen as well. I plan to run RJ45 cable at some point instead of using wireless at the moment. All to improve the sound by effectively cutting out parts of the Touch's circuitry and, therefore, sources of noise generation.

It seems to me (as a novice) that all electronics generates noise to some degree or another. It's the equipment design that filters as much of this out as possible to keep the purity of the signal. This is where the sound difference comes from in my opinion. In that respect the Touch may not be perfect but, boy, it is good value for money! :)

[Stratmangler]

All I know is my Touch through a Mark Grant coax cable to my Caiman/Gator sounds pretty good for what it has cost me. To improve things further conversion from Flac to PCM is done on the PC and I have switched off the Touch screen as well. I plan to run RJ45 cable at some point instead of using wireless at the moment. All to improve the sound by effectively cutting out parts of the Touch's circuitry and, therefore, sources of noise generation.

It seems to me (as a novice) that all electronics generates noise to some degree or another. It's the equipment design that filters as much of this out as possible to keep the purity of the signal. This is where the sound difference comes from in my opinion. In that respect the Touch may not be perfect but, boy, it is good value for money! :)

Sounds like you've implemented the Soundcheck mods to me.

I've been intrigued by the claims made about them, so I set myself a little experiment.
I ripped one of the my CDs to both WAV and FLAC, and dragged just one of the tracks from each (same track obviously) into a seperate folder, and accessed this folder to play the tracks.

I also went into the settings on Squeezebox Server and set WAV to play natively, as the default settings are convert WAV to FLAC.

So that I could just keep hitting the forward key on the remote to cycle through the tracks I also set the playlist on repeat.

I started playing the tracks, and after around 15 seconds I forwarded to the next track.
A few track forwards later, and I've got no idea whether the track being played is from WAV or FLAC, as I've lost count of the number of times I've hit the forward button.
What's more I am unable to hear any differences either.

I also tried comparing the same tracks with FLAC converted to WAV server side, and again after a few presses of the forward button I'd lost track of which was which and I started listening intently, and after a couple of switches between tracks I had to conclude that the tracks sounded the same to me.

On my part my Audiophilia Nervosa has been satisfied, so I shall just get on with the business of enjoying the music.

[DSJR]

For example, you can take a couple of cables with identical measurements (as best we can measure with the current knowledge a tools availible) and people say they can hear a difference.

If I say, particularly on this forum I can hear a difference between such cables, no one would bat an eyelid……come to zeros and ones………whoa, it’s a deiffent ball game :eek:
Just a thought.

But, if given bit perfect output then every operating system and player should sound exactly the same? :scratch:

I'm not sure that cables with audible differences would measure exactly the same, to be honest, but I have found that the better designed cables (often quite cheap too) sound far more like each other than some of the rather odd "top end" confections from Siltech, Nordost, Kimber and others, as these latter companies' products often do unspeakable things to get the results they want :)

[WAD62]

On my part my Audiophilia Nervosa has been satisfied, so I shall just get on with the business of enjoying the music.

Nice turn of phrase Chris...:lolsign:

[goraman]

It's getting beyond the point of shovels now, so perhaps one of these will help......

http://images.mirror.co.uk/upl/m4/oct2008/6/5/2B1AD0B8-E9D7-B16C-8B208B31E9A95DF3.jpg








A: Do you enjoy the end result from your music player,if so stop reading reviews they just screw with your head...
CLICK BELOW
http://www.swapmeetdave.com/Humor/Workshop/Trencher.htm

[Darren]

A: Do you enjoy the end result from your music player,if so stop reading reviews they just screw with your head...
CLICK BELOW
http://www.swapmeetdave.com/Humor/Workshop/Trencher.htm

The above is so true. Many thanks to everyone who contributed - I learned a lot.

[magiccarpetride]

The data transfer from the SB server (whatever OS platform) to the SB player (whether via wi-fi or ethernet cable) is always bit perfect, as it's an asynchronous TCP/IP file transfer, which either works perfectly, or not at all.

The above reminds me of the much ballyhooed 'cliff effect' that was touted during the early years of switching to HD TV. When the consumers in North America got upset about the wholesale switch to the HD TV programming, the powers that be tried to appease them by saying that they can continue enjoying the public programs by simply getting a cheapo antenna and pointing it at the relay. The theory went that if you do a good job pointing your antenna, you'll get a perfect, flawless picture. If you're off, you won't get any picture at all. Simple as that. They called it the 'cliff effect', whereby it's very binary -- either it's completely on or completely off.

In reality, however, the situation was not that clear cut. I bought the antenna, and I pointed it to the relay, but what I was getting was very spotty, very sporadic. Part of the screen had picture on it, but parts of the screen had fairly large black squares, who were moving and doing all kinds of weird stuff (all that depending on the weather patterns and god knows what else).

So the theory was wonky, because there apparently isn't such a thing as the all-or-nothing cliff effect.

Same is with this bit-perfect audio -- in theory you are supposed to either get no signal, or if you get ANY signal, then it's guaranteed to be perfect.

In reality, not even close. That's why it pays to invest into perfecting your digital transport. Anyone denying that is fooling themselves.

[WAD62]

The above reminds me of the much ballyhooed 'cliff effect' that was touted during the early years of switching to HD TV. When the consumers in North America got upset about the wholesale switch to the HD TV programming, the powers that be tried to appease them by saying that they can continue enjoying the public programs by simply getting a cheapo antenna and pointing it at the relay. The theory went that if you do a good job pointing your antenna, you'll get a perfect, flawless picture. If you're off, you won't get any picture at all. Simple as that. They called it the 'cliff effect', whereby it's very binary -- either it's completely on or completely off.

In reality, however, the situation was not that clear cut. I bought the antenna, and I pointed it to the relay, but what I was getting was very spotty, very sporadic. Part of the screen had picture on it, but parts of the screen had fairly large black squares, who were moving and doing all kinds of weird stuff (all that depending on the weather patterns and god knows what else).

So the theory was wonky, because there apparently isn't such a thing as the all-or-nothing cliff effect.

Same is with this bit-perfect audio -- in theory you are supposed to either get no signal, or if you get ANY signal, then it's guaranteed to be perfect.

In reality, not even close. That's why it pays to invest into perfecting your digital transport. Anyone denying that is fooling themselves.

What part of 'asynchronous FLAC file transfer via TCP/IP' don't you understand...;)

Do you think the data pixies steal occasional bits right out of the air, do your emails loose occasional characters when you send them, do your attachments grow dim after a while??? :)

The above is merely the transport layer, the data preparation/playback and conversion are entirely different things....

[magiccarpetride]

What part of 'asynchronous FLAC file transfer via TCP/IP' don't you understand...;)

Do you think the data pixies steal occasional bits right out of the air, do your emails loose occasional characters when you send them, do your attachments grow dim after a while??? :)

The above is merely the transport layer, the data preparation/playback and conversion are entirely different things....

I know that, in theory, both Logitech Duet and Logitech Touch deliver bit perfect signal to my Caiman DAC. And yet, everything else remaining equal, Touch sounds way better than Duet.

Also, modded Touch sounds way better than vanilla Touch.

I don't explain them, I HEAR them.

[The Grand Wazoo]

Just to interject a thought here guys - Here are some ones and some zeros.
The values on each line are the same but could you argue that each line is identical?


10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010

[Barry]

Just to interject a thought here guys - Here are some ones and some zeros.
The values on each line are the same but could you argue that each line is identical?


10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010
10101010

That's the beauty of digital techniques. As long as the numerical values are copied exactly and without corruption (i.e. a "1" for a "1" and a "0" for a "0"), it doesn't matter how the ones and zero are conveyed; the message is the same.

Similarly it doesn't matter what typeface is used, or whether a book is a hardback or not, Tolstoy's "War and Peace" reads just the same.

[Vincent Kars]

As long as we stay inside the digital domain the bits are the bits.
Unfortunately we are forced to leave the digital domain and have to transform our bits to analogue. This DA conversion is exactly the point where the bits are still the same bits but all of a sudden timing comes in to play.
PCM audio is sample + sample rate.
Not only are we in need of bit perfect, we also need a perfect time step
If we focus on the bits only, we leave half of the phenomenon out of the equation
http://www.thewelltemperedcomputer.com/Intro/SQ/Bits.htm

[WAD62]

I know that, in theory, both Logitech Duet and Logitech Touch deliver bit perfect signal to my Caiman DAC. And yet, everything else remaining equal, Touch sounds way better than Duet.

Also, modded Touch sounds way better than vanilla Touch.

I don't explain them, I HEAR them.

You're still missing the point, the FTP mechanism is bit perfect, but from that point on things can differ...Oh I give up :doh:

Oh and P.S. there's no such thing as a duet, what you had is known as a receiver, the duet is a package of the controller and a receiver. :doh::doh:

That's enough head slapping...

If you stick to 16/44.1 FLAC you will notice little difference between the receiver and the touch as a digital source (not analogue), the SB's were designed to handle this, and therefore have appropriate processing speeds and buffer storage for this purpose. In fact for the difference in price you could fit an upgraded clock and it will outperform the touch.

If however you upsample and/or transmit 24/96 PCM you're going outside the parameters of the architecture (you're trying to transmit more than 4 times the data volume), the asynchronous buffer pools are no longer large enough you'll get drop outs, the wi-fi transmission is no longer fast/reliable enough, so you will hear a difference by hard wiring, and you will find the Touch better as it has more grunt/storage...oh and you'll flood your network if you have more than one player, the whole point on a network solution, oh and ethernet cables all over the place.

First rule of asynchronous transmission, compress/transmit/expand, don't flood the bandwidth...

I've been using SoX upsampling on my desktop (XP 3.2Ghz 4Gb dual channel RAM, a beast) with winamp to 24/88.2, and it's taken ages for me to configure my buffers correctly, both software and hardware on the M-Audio 2496, in order to play this resolution sucessfully into a beresford, you're expecting a little SB touch/receiver to be able to do the same...mmmm

FYI This is a runtime snapshot of the work winamp/direct sound/M-Audio 2496 are having to do in order to provide a 24/88.2 SPDIF output to my DAC, look at the size of those buffers missus, both hard and soft ;) and that doesn't include the inter track ones required for gapless playback...and this is all happening in the player (FFSoX is upsampling locally), the FLAC is coming in via TCP/IP from my NAS

Output format: 88200 Hz, 24 bits per sample, 2 channels
Active buffer size: 2000 ms (1058400 bytes)
Device: "02: M-Audio Delta AP S/PDIF"
Mixing: software, primary buffer: active (software)

Buffer playback cursor: 724380 bytes
[==================================#===============]
Buffer write cursor: 704766 bytes
[=================================#================]

Data buffered:
Total: 1993 ms (1054596 bytes)
Async buffer: 30 ms (15810 bytes)

Buffer locks done: 340
Underruns: 0
Time played: 0:11.368 (7070976 bytes)
Time written: 0:13.361 (6016380 bytes)
Total time played: 6:28:30.935
Volume: 0.000000 dB / 0.000000 dB

...then there's the work my M-Audio 2496 PCI soundcard is doing in order to construct the SPDIF out, it all mounts up :)

[magiccarpetride]

If you stick to 16/44.1 FLAC you will notice little difference between the receiver and the touch as a digital source (not analogue), the SB's were designed to handle this, and therefore have appropriate processing speeds and buffer storage for this purpose. In fact for the difference in price you could fit an upgraded clock and it will outperform the touch.

The 'little difference' for you is huge difference for me. I have both boxes feeding into the Caiman via digital coax, and I can rather quickly switch to and fro using the remote controller. Everything else remaining equal, Touch sounds way better, less grain, less glare, more authoritative pace.

[WAD62]

The 'little difference' for you is huge difference for me. I have both boxes feeding into the Caiman via digital coax, and I can rather quickly switch to and fro using the remote controller. Everything else remaining equal, Touch sounds way better, less grain, less glare, more authoritative pace.

You don't transmit FLAC do you?

On many other threads you've stated that you transmit PCM, at high res, so either you're being wilfully ignorant, or you can't/won't read

[magiccarpetride]

You don't use transmit FLAC do you?

On many other threads you've stated that you transmit PCM, at high res, so either you're being wilfully ignorant, or you can't/won't read

You're right, I convert FLAC to PCM on the iMac. Sorry about the confusion, wasn't paying attention.