More 'Audiophile' bullsh*t
Discussion
StuH said:
No - you're oversimplifying the problem. you're ignoring jitter, clocking and transfer issues. If that were the case all transports would sound the same and they don't.
If you want to gain a deeper understanding of the issues involves, the computer audiophile website has some good articles: http://www.computeraudiophile.com
Please see my previous posts on this thread - jitter, clocking and transfer only come into play if you insist on having an unbuffered stream of data from the laser to the DAC.If you want to gain a deeper understanding of the issues involves, the computer audiophile website has some good articles: http://www.computeraudiophile.com
Except that you don't need that, any £10 CD player can read data with perfect fidelity, and we can transfer it point to point in digital form perfectly as well. You just need a buffer within the DAC to make it all work.
It seems to me that the only reason CD transports sound different is because the current industry is based on the concept of immediate streaming point to point. Originally CD players had their own built-in DACs and transferred the data to the amp in analogue form. That explains the original model but it's no reason to persist with a broken technical architecture. We've been through a number of technical iterations since then (with external DACs etc.) and now the world of hi-fi has kept the original model just with more boxes = more £££ you can extract from willing punters.
While data is kept within the digital domain it is extremely cheap to transfer it from one medium to another, at high speed, with perfect fidelity. We should be doing that with CD audio data so that conceptually the CD transport is like a NAS box with lossless format files - a source of perfectly reproduced digital data.
The fact that this *isn't* the current situation just looks to me like a suboptimal architecture that allows the industry to keep on ripping off audiophiles for expensive transports and external clocks. They *shouldn't* be necessary, but they are. For now.
Not sure what transfer issues would be, I only understand jitter and accuracy for clocks!
Having a master clock don't always help though:
http://www.soundonsound.com/sos/jun10/articles/mas...
I just use the internal re-clock on my sample rate converter TBH.
Having a master clock don't always help though:
http://www.soundonsound.com/sos/jun10/articles/mas...
I just use the internal re-clock on my sample rate converter TBH.
PhilboSE said:
Please see my previous posts on this thread - jitter, clocking and transfer only come into play if you insist on having an unbuffered stream of data from the laser to the DAC.
Except that you don't need that, any £10 CD player can read data with perfect fidelity, and we can transfer it point to point in digital form perfectly as well. You just need a buffer within the DAC to make it all work.
It seems to me that the only reason CD transports sound different is because the current industry is based on the concept of immediate streaming point to point. Originally CD players had their own built-in DACs and transferred the data to the amp in analogue form. That explains the original model but it's no reason to persist with a broken technical architecture. We've been through a number of technical iterations since then (with external DACs etc.) and now the world of hi-fi has kept the original model just with more boxes = more £££ you can extract from willing punters.
While data is kept within the digital domain it is extremely cheap to transfer it from one medium to another, at high speed, with perfect fidelity. We should be doing that with CD audio data so that conceptually the CD transport is like a NAS box with lossless format files - a source of perfectly reproduced digital data.
The fact that this *isn't* the current situation just looks to me like a suboptimal architecture that allows the industry to keep on ripping off audiophiles for expensive transports and external clocks. They *shouldn't* be necessary, but they are. For now.
I'm pretty sure you still get jitter sending the digital signal from a media player to a dac. It's to do with the protocols used in SPDIF/ TOSLINK connections. In theory Async USB should sort it out but apparently many of the cheaper Async USB's f'up somewhere.Except that you don't need that, any £10 CD player can read data with perfect fidelity, and we can transfer it point to point in digital form perfectly as well. You just need a buffer within the DAC to make it all work.
It seems to me that the only reason CD transports sound different is because the current industry is based on the concept of immediate streaming point to point. Originally CD players had their own built-in DACs and transferred the data to the amp in analogue form. That explains the original model but it's no reason to persist with a broken technical architecture. We've been through a number of technical iterations since then (with external DACs etc.) and now the world of hi-fi has kept the original model just with more boxes = more £££ you can extract from willing punters.
While data is kept within the digital domain it is extremely cheap to transfer it from one medium to another, at high speed, with perfect fidelity. We should be doing that with CD audio data so that conceptually the CD transport is like a NAS box with lossless format files - a source of perfectly reproduced digital data.
The fact that this *isn't* the current situation just looks to me like a suboptimal architecture that allows the industry to keep on ripping off audiophiles for expensive transports and external clocks. They *shouldn't* be necessary, but they are. For now.
As for the CD spec, it's what 30 years old? CD players were only 1x when the spec was drawn up, and RAM was far too expensive to have any meaningful buffer.There have been a number of failed optical audio technologies which offered improvements over CD, why would anyone want to try another format without any real advantage to 99.9% of users who use all in one CD players.
PhilboSE said:
Please see my previous posts on this thread - jitter, clocking and transfer only come into play if you insist on having an unbuffered stream of data from the laser to the DAC.
Except that you don't need that, any £10 CD player can read data with perfect fidelity, and we can transfer it point to point in digital form perfectly as well. You just need a buffer within the DAC to make it all work.
It seems to me that the only reason CD transports sound different is because the current industry is based on the concept of immediate streaming point to point. Originally CD players had their own built-in DACs and transferred the data to the amp in analogue form. That explains the original model but it's no reason to persist with a broken technical architecture. We've been through a number of technical iterations since then (with external DACs etc.) and now the world of hi-fi has kept the original model just with more boxes = more £££ you can extract from willing punters.
While data is kept within the digital domain it is extremely cheap to transfer it from one medium to another, at high speed, with perfect fidelity. We should be doing that with CD audio data so that conceptually the CD transport is like a NAS box with lossless format files - a source of perfectly reproduced digital data.
The fact that this *isn't* the current situation just looks to me like a suboptimal architecture that allows the industry to keep on ripping off audiophiles for expensive transports and external clocks. They *shouldn't* be necessary, but they are. For now.
Yes, I've read all your posts and I simply don't agree with you. It has nothing to do with fanciful conspiracy ideas of a greedy industry trying to extract money from gullible punters - who genuinely believes that rubbish? Except that you don't need that, any £10 CD player can read data with perfect fidelity, and we can transfer it point to point in digital form perfectly as well. You just need a buffer within the DAC to make it all work.
It seems to me that the only reason CD transports sound different is because the current industry is based on the concept of immediate streaming point to point. Originally CD players had their own built-in DACs and transferred the data to the amp in analogue form. That explains the original model but it's no reason to persist with a broken technical architecture. We've been through a number of technical iterations since then (with external DACs etc.) and now the world of hi-fi has kept the original model just with more boxes = more £££ you can extract from willing punters.
While data is kept within the digital domain it is extremely cheap to transfer it from one medium to another, at high speed, with perfect fidelity. We should be doing that with CD audio data so that conceptually the CD transport is like a NAS box with lossless format files - a source of perfectly reproduced digital data.
The fact that this *isn't* the current situation just looks to me like a suboptimal architecture that allows the industry to keep on ripping off audiophiles for expensive transports and external clocks. They *shouldn't* be necessary, but they are. For now.
It's quite the contrary, with a whole new "cottage" industry having sprung up in recent years, with small outfits developing DACs and transports for enthusiasts who just want the best bang for their audio buck, and unlike in the bad old days, of outrageously expensive CD players, the playing field is now very level. If you read around a bit at sites like computer audiophile and gain an understanding of things like async USB and all the research work that is being done by some very clever people into the effects of why jitter has such an large effect on preceivied audio quality, and all the other factors that are still poorly understood into why two seemingly identical digital signals can sound so different?
I've been into computer audio for several years now and like many was initially sceptical about how digital audio (those simple ones and zero's
)can be susceptible to such a huge variety of factors before ending up as the analogue audio that we hear at the end of the chain. In my main system I run an Aurender S10 frontend - http://www.computeraudiophile.com/content/426-aure... my office system is Mac/Amarra/Async USB/DAC, rest of teh house ZP90's - all sound VERY different despite playing the same bit-perfect digital tracks into an identical DAC. How is this so?Edited by StuH on Tuesday 8th January 14:11
Mr_Yogi said:
I'm pretty sure you still get jitter sending the digital signal from a media player to a dac. It's to do with the protocols used in SPDIF/ TOSLINK connections. In theory Async USB should sort it out but apparently many of the cheaper Async USB's f'up somewhere.
As for the CD spec, it's what 30 years old? CD players were only 1x when the spec was drawn up, and RAM was far too expensive to have any meaningful buffer.There have been a number of failed optical audio technologies which offered improvements over CD, why would anyone want to try another format without any real advantage to 99.9% of users who use all in one CD players.
As for the CD spec, it's what 30 years old? CD players were only 1x when the spec was drawn up, and RAM was far too expensive to have any meaningful buffer.There have been a number of failed optical audio technologies which offered improvements over CD, why would anyone want to try another format without any real advantage to 99.9% of users who use all in one CD players.
Globs said:
uberstealth said:
3. Never listen to the system! Listen to the music.
A friend of mine once said "Great music transcends the system". Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
FlossyThePig said:
Audiophiles don't seem to listen to music. They talk about "Presence", "Open", "Bright", etc., never "It felt as if was was in the middle of the concert hall".
Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
I think people have realised that's never going to happen no matter how much you spend.Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
Besides most popular music isn't recorded in a live environment, it is produced. Some people want to hear it as the producer heard it at the mixing deck, but again outside the exact studio with the same equipment and the ears of the producer I don't think that's going to happen either.
Edited by Mr_Yogi on Tuesday 8th January 14:36
FlossyThePig said:
Audiophiles don't seem to listen to music. They talk about "Presence", "Open", "Bright", etc., never "It felt as if was was in the middle of the concert hall".
Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
You can always rely on Pistonheads to deliver these pearls of wisdom - priecless Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
and utter tosh.StuH said:
I've been into computer audio for several years now and like many was initially sceptical about how digital audio (those simple ones and zero's )can be susceptible to such a huge variety of factors before ending up as the analogue audio that we hear at the end of the chain. In my main system I run an Aurender S10 frontend - http://www.computeraudiophile.com/content/426-aure... my office system is Mac/Amarra/Async USB/DAC, rest of teh house ZP90's - all sound VERY different despite playing the same bit-perfect digital tracks into an identical DAC. How is this so?
Different rooms, different acoustics, differences within the DACs themselves? I can fully accept that two identical DACs will transform the digital sample into an analogue signal with minute differences which some people will claim to be able to detect. )can be susceptible to such a huge variety of factors before ending up as the analogue audio that we hear at the end of the chain. In my main system I run an Aurender S10 frontend - http://www.computeraudiophile.com/content/426-aure... my office system is Mac/Amarra/Async USB/DAC, rest of teh house ZP90's - all sound VERY different despite playing the same bit-perfect digital tracks into an identical DAC. How is this so?Edited by StuH on Tuesday 8th January 14:11
However, it is a material FACT that data within the digital domain can be transferred from A to B at high speed with perfect fidelity. If you take data OUT of the digital domain, then of course all bets are off.
Funny how data travelling over ethernet cables isn't subject to data loss. You can even transfer a signal over mains cables at much higher bitrates than CD samples. Or even down a telephone wire, with who knows how many different types of cables and routers that signal goes through. And yet the fidelity is maintained and jitter-free. How is this so?
PhilboSE said:
Different rooms, different acoustics, differences within the DACs themselves? I can fully accept that two identical DACs will transform the digital sample into an analogue signal with minute differences which some people will claim to be able to detect.
However, it is a material FACT that data within the digital domain can be transferred from A to B at high speed with perfect fidelity. If you take data OUT of the digital domain, then of course all bets are off.
Funny how data travelling over ethernet cables isn't subject to data loss. You can even transfer a signal over mains cables at much higher bitrates than CD samples. Or even down a telephone wire, with who knows how many different types of cables and routers that signal goes through. And yet the fidelity is maintained and jitter-free. How is this so?
Packet based transactionsHowever, it is a material FACT that data within the digital domain can be transferred from A to B at high speed with perfect fidelity. If you take data OUT of the digital domain, then of course all bets are off.
Funny how data travelling over ethernet cables isn't subject to data loss. You can even transfer a signal over mains cables at much higher bitrates than CD samples. Or even down a telephone wire, with who knows how many different types of cables and routers that signal goes through. And yet the fidelity is maintained and jitter-free. How is this so?
PhilboSE said:
Funny how data travelling over ethernet cables isn't subject to data loss.
Isn't it just. Not.FlossyThePig said:
Audiophiles don't seem to listen to music. They talk about "Presence", "Open", "Bright", etc., never "It felt as if was was in the middle of the concert hall".
Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
The best hi-fi I've ever heard was actually very close to what you describe, it was a demo room at the Bristol HiFi show in a seemingly small, crap room, organised by Path Premier.Originally hi-fi was trying to get as close to being at a concert as possible. That is orchestral (acoustic) not amplified. I think that goal seems to have been forgotten.
It was a Mark Levinson CD and Amp (bottom of the range at the time I believe), and Revel M20 speakers, playing a CD by Dean Peer called 'Think...It's All Good'. It sounded like he was in the room and made two things happen 1)me want a Levinson / Revel hi-fi, and 2)we bought a copy of the CD. It also highlighted how good CD can sound, which I suspect has quite a lot to do with how that particular CD was actually recorded.
Mr_Yogi said:
As for the CD spec, it's what 30 years old? CD players were only 1x when the spec was drawn up, and RAM was far too expensive to have any meaningful buffer.There have been a number of failed optical audio technologies which offered improvements over CD, why would anyone want to try another format without any real advantage to 99.9% of users who use all in one CD players.
I'm not suggesting a different specification or technology, I'm suggesting that we treat the data on the CD differently - use the same low level block read functions to extract the raw PCM data into a buffer in the DAC, which can then do its magic.Raw data doesn't have an inherent sample frequency, it's just a series of 1's and 0's and we apply some interpretation to this data to turn it into information (a sound wave). Inside the digital domain we can move this data around the system at will losslessly. So we extract this *data*, move it around the system to the DAC, which then interprets it into information - an analogue soundwave.
Mr_Yogi said:
Packet based transactions
That's one of many reasons. The point is that the data ends up where we want it, in perfect fidelity. It's complete arrival may be non-deterministic, so that's why you buffer it at the other end. And as long as you can write to the buffer faster than you need to read it, you're laughing.So, why can't we use these lossless methods to get data off the CD into the DAC so that it has a perfect data sample to work from?
RedLeicester said:
PhilboSE said:
Funny how data travelling over ethernet cables isn't subject to data loss.
Isn't it just. Not.PhilboSE said:
That's one of many reasons. The point is that the data ends up where we want it, in perfect fidelity. It's complete arrival may be non-deterministic, so that's why you buffer it at the other end. And as long as you can write to the buffer faster than you need to read it, you're laughing.
So, why can't we use these lossless methods to get data off the CD into the DAC so that it has a perfect data sample to work from?
It's not as simple as that you need a method to check the data you receive is the same as what was sent you can't reply on the fact it's 1's and 0's even when programming simple file transfers. the fact ia that there are no protocols for transfering digital audio data. Also DACs are very sensitive to other components and need isolation.So, why can't we use these lossless methods to get data off the CD into the DAC so that it has a perfect data sample to work from?
The other issue with using data access type methods to read an audio CD is that if cd's are in less than perfect condition they can take ages to read the data correctly and interrupt the music. If you have used EAC to rip CD's you will see that some CD's which appear clean, can stop and go through a laborious process where is gets stuck, which would stop the music dead even with a big buffer.
The whole point of red book audio is that it plays pretty well even if the CD is damaged that's to the real time error correction (I forget it's name).
The whole point of red book audio is that it plays pretty well even if the CD is damaged that's to the real time error correction (I forget it's name).
RedLeicester said:
Isn't it just. Not.
And you think we can't detect packet loss and re-send the missing packets?Really, you think that inside the digital domain we can't transfer data with 100% fidelity?
I can rip a CD audio track to a WAV file which is an uncompressed lossless format representing the data stored on the disc. I stick this WAV file on an external hard disk attached to my PC via USB. Let's say I email this WAV file as an attachment to a colleague in another continent, who receives it on his smartphone and plays it back on that device.
How many errors will be in the resultant WAV file when he receives it, as a result of error correction and jitter caused by the USB cable, circuitry inside my PC, the copper telephone wire to the exchange, all of the routers and the fibre optic cable across the Atlantic, and of course the crappy 2G signal to his phone?
Mr_Yogi said:
It's not as simple as that you need a method to check the data you receive is the same as what was sent you can't reply on the fact it's 1's and 0's even when programming simple file transfers. the fact ia that there are no protocols for transfering digital audio data. Also DACs are very sensitive to other components and need isolation.
How many protocols are there for transferring PH forum posts?Answer: none. There are data transfer protocols, and we're standardising on ever fewer (FFS when I was starting out in IT comms we had about 20 protocols just for serial cables running at 300 baud!).
CD audio data is just the informational interpretation we attach to the raw data. There is nothing "magic" or different about it from any other digital data. We can transfer it inside the digital domain from point to point, easily, cheaply and flawlessly.
What the recipient device does with this data is another matter entirely...
Gassing Station | Home Cinema & Hi-Fi | Top of Page | What's New | My Stuff