There is one minor get-out: when a CD player reads an audio CD it can always rely on interpolation as a last resort if the error correcting codes can't cope with reading errors. A computer system can't do that, so has even better codes to give better data integrity. This issue can be eliminated in tests provided the CD player is honest and reports any interpolation it has to do.
Once the data is in the computer domain there is no interpolation. It is just data to be exactly preserved. The timing with which it is eventually sent to the DAC is a separate issue, but that would show a flaw in playback method not data storage.
Once the data is in the computer domain there is no interpolation. It is just data to be exactly preserved. The timing with which it is eventually sent to the DAC is a separate issue, but that would show a flaw in playback method not data storage.
in fact part of my job was setting up an early online pdf workflow (mid 90's) for our customers, so that they could work on proxy files, on different machines, proofing on different printers and monitors in different rooms, with different lighting and we would still end up with a consistently colour accurate result. this was for high profile advertising and publishing firms on print runs in hundreds of thousands, inconsistent results would be my *** $$$$$$
Last edited:
That happens rarely on discs that don't have gross damage, and when it does, it represents one sample out of several million (i.e., a transient incident lasting microseconds once per hour) so doesn't affect the overall "sound."
We're dealing with urban legends here.
Interesting but totaly incorrect. As part of my day job I actually lay out complex digital based PCB's (and analogue and digital/analoge and power supplies and...) and do signal integrity simulation, especially when you have a DDR interface, funily in all the time I have been doing it I never noticed the digital signal being turned to analogue to determine whether it is a 0 or a 1. Eye diagrams are to show that the overall signal transfer is withing the system tolerences, you generaly set up an eye mask, and if the diagram dose not close up over your mask, you know that the data transfer for that set up will work and alweays be withing the timing budget the system requires (that is when data is present, enable signal data being at the pin at the right time, you can see the data in most digital devfice data sheets such as ethernet.) Ah ethernet, much tighter requirements with high speed ethernet than audio yet we can send data pictures and obscure vodoo views on digital from our computer over the internet with no problems...........
Now if my CAD files suffered from these artifacts I would be in big trouble...
Same with my photos, 25mega pixel files shot in RAW and manipulated in lightroom 4, dont notice any difference between the same image stored on different devices.
I get upset over this sort of pseudo science because it gives the wrong and incorrect information to those wanting to learn and gives a bad name to the hobby of DIY audio repreduction.
"If you can't hear the difference between an original CD and a copy of your CD,
you might as well give up your career as a tester.
The difference between a reconstituted FLAC and full size WAV is much less than that, but it does exist."-Cookie Marenco. cookiemarenco.com/
Cookie Marenco is an award winning Recording Engineer, who is also an an Audio Consultant.
Blue Coast Records are pioneering high resolution DSD downloads and recordings.
She also supplies her DLs as UNCOMPRESSED Zips.
Home | Blue Coast Records
you might as well give up your career as a tester.
The difference between a reconstituted FLAC and full size WAV is much less than that, but it does exist."-Cookie Marenco. cookiemarenco.com/
Cookie Marenco is an award winning Recording Engineer, who is also an an Audio Consultant.
Blue Coast Records are pioneering high resolution DSD downloads and recordings.
She also supplies her DLs as UNCOMPRESSED Zips.
Home | Blue Coast Records
In YOUR opinion!
T.A.S 220 and T.A.S 221 also confirmed what she has reported about .flac and .wav files. Famous Recording and Mastering Engineer Barry Diament has also reported audible differences between pressings made at different CD plants.
Neither should the Sony BluSpec comparison sets sound different either, as they have the same checksums, and it has nothing to do with the CD players error correction abilities.Barry Diament has listened to a Sony BluSpec comparison set with his CD ROM and thought they must have come from different masters, until he ripped them and saw the binary content was identical.
Like your Bybee dissection test results ?
Yes, that's a good example. Crazy stuff like "data" rather than stories that help to sell things.
you missed the point. how digital circuit know 0/1? trigger voltage(we just discuss normal gate, not Schmidt ), trigger voltage is digital? or its a specific analog voltage level? 0->1 or 1->0 will rely on this voltage, so noise affects jitter cause it will make your internal circuit "missunderstand" eye diagram(eye diagram just use for example, audio circuit doesn't use this kind of high speed transfer except CD readin "RF").
I do agree communication equipments have tighter requierement than audio equipments, but ......
the final question is how people's ears could distinguish the playback difference? ears better than AP or AP is the best ?
the truth is, single sine signal, AP wins for sure, but complex signal, ears win!
still in audio field, not very good method to evaluate the whole system, AP could just test a few parameters not all of them.
play back quality: ape is the worse wav is the best flac in the middle.
the higher cpu usage, the worse playback quality.
here we go again, complex signals are just a combination of sine waves.
if your CPU load effects the data integrity at your dac, I hold grave concerns at the quality of your entire system. bit perfect is bit perfect, there are no half bits. it is you who miss the point, perhaps its commercially convenient? regardless of the signal being described with voltage, the decision point is very much digital. digital errors do not create subtle artifacts in the soundstage, you cannot have bit-perfectness and error at the same time...
if your CPU load effects the data integrity at your dac, I hold grave concerns at the quality of your entire system. bit perfect is bit perfect, there are no half bits. it is you who miss the point, perhaps its commercially convenient? regardless of the signal being described with voltage, the decision point is very much digital. digital errors do not create subtle artifacts in the soundstage, you cannot have bit-perfectness and error at the same time...
theory is not the fact. digital audio is kinda distortion from the very beginning cause square wave is made of unlimited frequency bandwidth sine wave. so 44.1KHz sample rate cuts the high band, how come "perfect"?
heavy cpu load cause MMU busy then data latency gets bigger, affects jitter ...
jitter does not affect bit perfect, but affect hearing.
PLL's clock system hearing is too heavy in low-mid frequency and not clear in high frequency. VCO brings very high phase noise below 20KHz.(so any model telling phase noise of the crystal related with audio frequency?)
that's why many DAC use 2nd PLL, but still could not solve the problem of PLL.
no it is not fact, the depth of your misunderstanding is amazing. are you the tech at your company, or promotions?
a clock can be read effectively out of a pure sinewave (and in fact they often ARE sinewaves), as long as the decision point is accurate and consistent, it matters not so much the shape of the transmission. at the speeds of todays clocks, the signal doesnt often look much like a squarewave, yet a bit-perfect result is obtained. If what goes in, is what comes out, job done.
this old chestnut about signals being analogue is often trotted out by people trying to sell by confusion. the transmission line is analogue, the signal is digital
if your dac is so effected by input jitter (you should not be using the embedded clock, nobody has done that for years) then I dont have much faith in your devices. the local audio clock is all that matters, masterclock is not even part of the i2s specification...
a clock can be read effectively out of a pure sinewave (and in fact they often ARE sinewaves), as long as the decision point is accurate and consistent, it matters not so much the shape of the transmission. at the speeds of todays clocks, the signal doesnt often look much like a squarewave, yet a bit-perfect result is obtained. If what goes in, is what comes out, job done.
this old chestnut about signals being analogue is often trotted out by people trying to sell by confusion. the transmission line is analogue, the signal is digital
if your dac is so effected by input jitter (you should not be using the embedded clock, nobody has done that for years) then I dont have much faith in your devices. the local audio clock is all that matters, masterclock is not even part of the i2s specification...
I'm an engineer, if you like, just start the discussion one by one, no misunderstanding.
what I mean square wave is the RECORDed square wave signal, none of the business of system clock.
embedded clock is generate by PLL, we never use this clock from the very beginning
jitter is what digital systems need to eliminate, but nobody could do that.
ears accept signal continous while eyes not, so audio playback quality is a lot important than video playback(but the same theory, better DAC and clock, better output result), so we just discuss audio playback only here.
Last edited:
Of course all of electronics is analogue, we just choose to interpret electrical levels in some parts of the circuitry as representing digital information. Which is fine so long as as we only need to consider the "information" meaning, it always stays "digital". Unfortunately, we can't hear digital, so at some point it needs to translate to analogue, which is where the problems start. Analogue signals love to interfere with other analogue signals, especially when they are very high frequency - just what most digital is like - when they strive to squirm their way into almost everything ...
Frank
Frank
Somebody later in this thread makes the point that if the bits are the same and the timing is the same then the sound is the same. Those are two big IFs. If you send your audio file through the sound processing system of any modern operating system, it will no longer be bit perfect and the sound will changed. Getting a jitter free signal to your dac is important and unless the data coming out of a computer (especially the USB port) is buffered and re-clocked, that's not likely. So I hope we're all past these basic points in the conversation, but it's important to make sure those points are understood.
Jim
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.