jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 20:37:29 GMT
Sorry, I don't understand - are you saying that the output from the end of your "perfect cable" will be different from the digital input? i.e output bits are not identical to input bits? Will it be different, or can it be different? Which are you asking? [moderated] I can accept that you haven't considered the physical representation of digital or it's transmission before as you are an IT guy but please take a moment to consider what you are saying!!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 20:40:30 GMT
I just went to load the dishwasher and a shit load of posts arrived !! How many bits would have been used in that time / post and if one of those bits on one of our computers would have been misintrepreted what would that individual have seen ... a wrong letter... a message of the OS ? Let's be very clear that in ALL cases there is NO altering of bits from the start to the end. The PCM representation is exactly the same and the bits have no allowance other than to be 1 or 0 even in the physical form of the electrical signal representing it. We should stop discussing this issue as it is not important to this dsicussion
|
|
|
Post by dalethorn on Jun 13, 2012 20:42:21 GMT
Will it be different, or can it be different? Which are you asking? [moderated] I can accept that you haven't considered the physical representation of digital or it's transmission before as you are an IT guy but please take a moment to consider what you are saying!! [moderated] The original analog signal once converted to digital no longer exists unless you save it on a tape or whatever. But it's still analog when saved as analog, and therefore not exact. The ADC makes a digital file that is exact and stays that way forever, if not corrupted.
|
|
|
Post by freddypipsqueek on Jun 13, 2012 20:42:47 GMT
John - I am going further back that the cable. It is the differences in identical files that has been open to dispute. You have me entirely on the variations in cables etc once the data is sent on its way.
PCs are littered with data errors - Packet transmission of network data is a good example of them being fixed before they cause a problem. God knows why it hasn't been used more widely in audio. I was pissed off I could not get the network funcion of the Aune S1 working.
I am sure Alex makes some good points about the better ripping of files. For those with years of PC use (and probably abuse) the suggestion that identical files sound different questions the funamentals of computers. It is clearly worng. Once the file is on its way . . . .
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 20:44:58 GMT
How many bits would that have been and if 1 of those bits on one of our computers would have been misintrepreted what would that individual have seen ... a wrong letter... a message of the OS ? Let's be very clear that in ALL cases there is NO altering of bits from the start to the end. The PCM representation is exactly the same and the bits have no allowance other than to be 1 or 0 even in the physical form of the electrical signal representing it. We should stop discussing this issue as it is not important to this dsicussion Franz, the important point is that the electrical waveform can be very different between two identical digital representations. It's what digital is all about - an agreed protocol that can deal with this variance in electrical representation
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 20:45:26 GMT
gentlemen... this is getting nowhere.
The key to this isn't in the digital domain although it appears that way. Bit perfect is not debated.
This is all about the physical bits and their transmission. mechanical, electrical issues NOT software or protocol related.
|
|
|
Post by dalethorn on Jun 13, 2012 20:46:03 GMT
I just went to load the dishwasher and a shit load of posts arrived !! How many bits would have been used in that time / post and if one of those bits on one of our computers would have been misintrepreted what would that individual have seen ... a wrong letter... a message of the OS ? Let's be very clear that in ALL cases there is NO altering of bits from the start to the end. The PCM representation is exactly the same and the bits have no allowance other than to be 1 or 0 even in the physical form of the electrical signal representing it. We should stop discussing this issue as it is not important to this dsicussion I'll assume that people are being honest. The original analog waveform is converted to digital and lost. The digital file is then converted to analog, an approximation of the original analog signal. It is possible for the DAC to make mistakes with any digital file, but if the digital files are identical and played side by side, any mistakes should be random and not favor one file over the other.
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 20:50:23 GMT
Freddy, I'm asking that we step back from Alex's example because this is the most extreme example & we have been down the same road so many times & really there is nothing more to say about it.
What I'm talking about is the title of the thread "Jitter & Software Jitter" Once you accept noise can effect the D/A stage then you have to examine what are the possible sources of noise. It turns out that one of the possible sources are the digital interface itself & the noise that can be riding on this signal.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 20:51:10 GMT
The physical representation is different on every different medium and electrical level and the beauty of digital is that this doesn't matter at all UNTIL it hits a DAC circuit that relies on the exact timing of that signal.
The point is internet transport, physical storage e.t.c. happens WAY before the DAC and all of it's 'analog' properties have become irrelevant. The findings though suggest something is still stored and removed along that whole path.
THAT mechanism would be responsible IF we accept it is tue and let's assume for arguements sake there is a storage of previous physical storage artifacts and can be transmitted and manged/removed and not improved anymore afterwards.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 20:53:55 GMT
Dale,
It's not the bits nor the described waveform that is the issue. The theory is the physical representation is the culprit and above all this is detrimental in the whole chain, stored in an yet unknown way and when reaching the DAC still is supposed to have the 'analog' artifacts of the physical notation/representation of that signal.
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 20:54:25 GMT
Dale, please take Franz advice & desist. I have said all I need to say & when you re-read this series of posts at a later stage you will see that we are misunderstanding what is meant by my use of "analogue waveform" - I'm referring to the fact that the electrical representation of digital when transmitted is an analogue waveform!!
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 20:59:58 GMT
Yes, Franz, correct but as I repeatedly have pleaded, can we stay away from Alex's examples, please? It will go nowhere!
I gave examples of CD differences, software playback differences, USB cables differences, SPDIF cable differences all resulting in different sound even though they are bit-identical.
I'm proposing that noise on the signal is the cause of this. The diagram shows how a cable can result in these differences & cause jitter.
Can we discuss the thread topic & not keep going back to the contentious example?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 21:00:37 GMT
In all honesty jitter in the signal that is fed to the DAC CAN be of influence of the final described waveform (but only extremely minor) and is not at all interesting as it is very measurable and proven by and to EE's. There is NO mystery to that in anyway and is not interesting at all to discuss so I will leave it at this.
I am more interested in finding out how a file can deteriorate in certain aspects (but not described waveform) by ripping processes, data transport over whatever physical medium and how that info is stored in those signals in a way it alters the described waveform in the end stage.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 21:04:36 GMT
Dale,
As an admin I kindly ask everyone involved to stay on topic and end all PERSONAL attacks or I will simply clean those out. I have deleted the IMO offensive post.
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Jun 13, 2012 21:09:05 GMT
In all honesty jitter in the signal that is fed to the DAC CAN be of influence of the final described waveform (but only extremely minor) and is not at all interesting as it is very measurable and proven by and to EE's. There is NO mystery to that in anyway and is not interesting at all to discuss so I will leave it at this. Ok, so you agree that differences exist in all the areas that I've mentioned (which is a change of your position) but you find it trivial & don't wish to discuss it. Fair enough, I'll not waste my time any more It seems that you are more interested in stirring up this contentious issue even though you know it will go nowhere [moderated]
|
|
|
Post by freddypipsqueek on Jun 13, 2012 21:12:26 GMT
Sorry - Agree with you John.
|
|
|
Post by freddypipsqueek on Jun 13, 2012 21:15:23 GMT
"I am more interested in finding out how a file can deteriorate in certain aspects (but not described waveform) by ripping processes, data transport over whatever physical medium and how that info is stored in those signals in a way it alters the described waveform in the end stage."
Agree with this as well.
I am really interested in evidence to show better rips. Then noting the advantages (disadvantages) or SD readers SSDs etc.
|
|
|
Post by freddypipsqueek on Jun 13, 2012 21:18:50 GMT
There is so much work here. I was really impressed with the latest foobar2000 & Asio driver. The reading of digital data at such a low level is so full pitfalls - now its even down to software . . . .
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 13, 2012 21:20:21 GMT
In all honesty jitter in the signal that is fed to the DAC CAN be of influence of the final described waveform (but only extremely minor) and is not at all interesting as it is very measurable and proven by and to EE's. There is NO mystery to that in anyway and is not interesting at all to discuss so I will leave it at this. Ok, so you agree that differences exist in all the areas that I've mentioned (which is a change of your position) but you find it trivial & don't wish to discuss it. Fair enough, I'll not waste my time any more It seems that you are more interested in stirring up this contentious issue even though you know it will go nowhere [moderated] I have not changed my position and agree that certain aspects may deteriorate the final waveform in SOME specific DAC circuits or complete systems such as CD players as a whole. The audibility of it (the treshold) is the real issue here. And it's not about stirring up contentious issues, it's about finding the mechanism for the reported differences as that mechanism appears to have the same sonic influence, similar to jitter in the final stage) but transmitted somehow inside the digital contents and is what is fascinating and revolutionary and unexplained. Let me be clear.. this is NOT about discrediting Alex nor his findings or showing off knowledge. It's an intersting subject, more so than already investigated and documented phenomena.
|
|
|
Post by dalethorn on Jun 13, 2012 21:20:22 GMT
Dale, It's not the bits nor the described waveform that is the issue. The theory is the physical representation is the culprit and above all this is detrimental in the whole chain, stored in an yet unknown way and when reaching the DAC still is supposed to have the 'analog' artifacts of the physical notation/representation of that signal. Most of the 'unknowns' in the examples of 'Maria' I believe it was, the 3 identical files, can be accounted for by not having the files on the same player at the same time. The physical representation as I've explained previously, i.e. the filenames, the file system storage headers, the drives - all those things *can* make for differences in playback. But if you take those 3 Maria files and copy them to the same folder on the same drive and bring them up in the same player at the same time, the differences disappear. Alex has made it clear that he believes corruption can happen invisibly to a digital file when zipped and unzipped and copied, even though the unzipped WAV file tests identical to the original, and that that corruption will follow that file forever. That's a problem, because it isn't true and it makes progress in finding the real problems impossible.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 14, 2012 5:32:12 GMT
The physical representation as I've explained previously, i.e. the filenames, the file system storage headers, the drives - all those things *can* make for differences in playback. You would have to realise though that even when these files are on different locations even media but on the same PC/player the timing with which the bits are clocked out to the physical DAC is at different physical moments (several clock pulses/instructions later) and not a real time event. Lets call the complete circuit, be it internal or external including the needed parts around it a physical DAC to avoid miscommunication about the CHIP itself. Due to time sharing in the PC (background processes, services and what not also use up processor time at the same moment) the bits are read in a MUCH higher speed than the actual bit stream and is read in bursts. In DAPs also time sharing is going on as the display needs to be refreshed and pointers stored e.t.c. bits are thus read in bursts and temporarely stored in memory (a buffer, but not line buffer type but usually RAM) they are processed (become parallel bytes thus unserialised) when using SATA or SD card or disc and are clocked out again in a continuous speed towards the DAC (in whatever format be it USB protocol, packages (WAN-LAN), spdif, I2S) and ONLY at this stage there is 'room' for jitter to mess up the described waveform by it being ,extremely small, different from the intended described waveform. The severity of this altered waveform depends on the ability of the physical DAC to ignore (jitter reduction) the jitter at the moment it receives it AND the clock it uses. This clock may be derived from the clock in spdif, framerates. Newer DAC's can be VERY tolerant to input jitter and the only jitter that can be present at that point is clock stability and noise on the power supply lines. Even then the influence of that noise on the power supply lines will depend on how the actual DAC chip is immune to power supply ripples. I out chips (those that need an external I-V converter) usually have a high PSSR as a current source can be very noise free even if the PS is noisy. a high PSSR means the garbage on the power supply is NOT entering the audio signal. So when you have a DAC with low self jitter and good jitter reduction the input jitter is trivial. Meaning when you have such a physical DAC the whole USB, I2S, SPdif and PC before it is irrelevant and simply cannot influence the SQ anymore. At least it is designed that way and the real question is... is the theory correct and if so where do the perceived differences come from in that case. Alex has made it clear that he believes corruption can happen invisibly to a digital file when zipped and unzipped and copied, even though the unzipped WAV file tests identical to the original, and that that corruption will follow that file forever. That's a problem, because it isn't true and it makes progress in finding the real problems impossible. I agree with you on this point but I like to start of with the assumption it is still possible by jitter being stored somehow (which I believe is Alex his suspicion). This is based on his findings with tinkering and listening combined with insights. Also it depends on the assumption there MAY be something being stored we are unaware of as we can not measure the responsible phenomen as it is still unknown, similar to us not knowing anything about the universe, brain, science, human body, wildlife you name it. This can be used as compelling evidence we don't know everything. Also it could be we do not have the right 'tools' to measure the deterioration such as voltage, current, frequency, bandwidth, sloping of digital bits, resolution of test gear, proper test protocols e.t.c. IMO we can actually look and investigate MUCH further in the known quantities as our hearing can resolve but I would not rule out this before hand. It's these areas that need adressing, not the described waveform, but the reasons why the final electrical waveform differs from the described one and how much in which way. The second hurdle is HOW much can the electrical waveform be 'off' from the described one so it becomes detectable by the human hearing. Mind you we can measure in noise regions and even in that we can 'see' frequencies, we can measure timing differences and jitter in sub ps while timing differences we seem to be able to perceive in 2-3 micro seconds. So we can measure timing about a million times more accurate than has been shown to be heard with a single tone. With music this doesn't seem possible as our brains are busy interpreting everything that has to do with music (very complex brain activity that alters with fatique, time of day, condition... you name it) But let's not dismiss and perhaps start a friendly converstaion of how and where to start. maybe even in another thread that will be very cliosely monitored to 'decent behavior in language'. Perhaps we learn something from it.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 14, 2012 6:03:05 GMT
Frans A quick reply in case a joint decision is made to kill this thread. In the X-DAC V3 and S.C. DAC for example, it is the Input stage I.C. (DIR9001, 1703E etc.) that mainly governs this, not the DAC chip itself.When an external oscillator is used, a TCXO suppied via a highly stable and VERY low noise supply gives best results in conjunction with very close tolerance components in the Filter cct. 1% or 2% polypropylene capacitors are ideal at that location. Alex Attachments:
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 14, 2012 6:17:11 GMT
That's why I like to differentiate between DAC chip and 'box'. hence... the 'physical DAC' which means the whole assembly from data input interface right up to the DAC chip and also the analog filtering that is essential for certain reasons, in most cases that is. and depends on DAC chip itself and parts directly behind it.
The jitter reduction nowadays is, due to minituralisation, mostly done in the same chip as the actual DAC circuit itself but in a lot, if not most cases still handled by chips that handle incoming signals and supply those to a DAC chip.
As said when a physical DAC is decently build (with proper analog filtering, stable clock and powersupply) and has good jitter reduction circuitery EVERTHING that happened before it reaches that box should be irrelevant and cannot influence the sound quality.
Subjective findings, however, seem to not backup this thesis and the why, how and audibility of it IMO should be explored.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 14, 2012 6:24:14 GMT
Frans If you saw the X-DAC V3# recent posts and photos about the improvement due to a "Cricket"supplying the TCXO, and the improvements being far more obvious with hi res than RB CD. I can assure you that the improvements were not imaginary. TBH, I thought that as the existing 7805 was supplied via a JLH, that there would be S.F.A. change, which is why I put off trying it for several weeks. It was a rainy day, and I was bored, so I got around to it. Alex
|
|
jkeny
Been here a while!
Posts: 463
|
Post by jkeny on Dec 30, 2012 21:13:53 GMT
I hope nobody minds me opening up this thread again - it was the one that I quit this forum over but I think it is probably safe to go back into the water now ? The reason I'm opening it up is because of some new data points: - this one which deals with an easily understood but technically deep treatment of among other things - What is the relationship between small distortion, wow & flutter, & jitter. the level of jitter Audibility of Small Distortions www.madronadigital.com/Library/AudibilityofSmallDistortions.html- And another one link where you can download 3 files of ultrasonic jitter & how it is audible www.whatsbestforum.com/showthread.php?9085-Audibility-of-Small-Distortions&p=160062&viewfull=1#post160062I hope the mods don't mind me opening up this old thread or posting links to another forum? Just remove if inappropriate & I'll resign - haha, joke
|
|