In order to prove the claim that a bit perfect transmission from one drive or digital source should (and probably does) sound no different from another.
Can anyone or has anyone tried be to connect the digital output of one PC CD/ DVD drive to a digital input of a soundcard and compare the checksum of the recorded CD to the ripped CD. Bit perfect should result in the same checksum, assuming the length of the tracks are the same and the audio begins and ends at the same point.
If so, find a different brand drive and hopefully get the same result. Then do an AB comparison between two different brands of PC CD/DVD drives running into the same DAC. They will either sound the same or not.
If they sound the same then theory is proven. If they sound different then we need to go hunting for the reason behind this.
Until this data is presented then this is all speculation based on personal experience and assumptions. At the moment my personal experience says that some transports sound better than others but have no real means to analyze why this is so.
There have been many threads that discuss this, including simulations, that showed that different cable lengths don't effect the signal integrity, proper engineering such as correct terminations do. The sims included perfect lines and lines with impedance mismatches. The shortest digital interconnect is always the best.... Then you have to work out the velocity factor of your cable, not just some rule of thumb figure of 150mm/1ns. Most co-ax cable Vf is in the range of 0.66/0.89, this gives a propagation time range of 1.51s to 1.12s per foot (300ms).
As to the rise times quoted, they are rather slow, usual ones are often 8-1ns! Its a bit of a silly way to try and control reflections really....
Some real signal integrity information...Not based on audiophile myths
https://www.sigcon.com/
Teledyne LeCroy SI Academy
There are plenty more...
And a bit on noise and coupling...
http://www.calex.com/pdf/4ground_shield.pdf
As to the rise times quoted, they are rather slow, usual ones are often 8-1ns! Its a bit of a silly way to try and control reflections really....
Some real signal integrity information...Not based on audiophile myths
https://www.sigcon.com/
Teledyne LeCroy SI Academy
There are plenty more...
And a bit on noise and coupling...
http://www.calex.com/pdf/4ground_shield.pdf
Don't we do this test every time we rip a CD (assuming one of the ripping programs that ensure bit-perfect rips)?
As marce keeps pointing out, jitter is manageable. But my point was that if electrical noise really was an issue, you should be able to determine that by observing the differences between using a toslink connection and an electric one.
And that is not a problem for me. 96K is more than enough for me, and my pet bat doesn't really like my choice of music anyway
Jitter have a wide spectrum. You can not manage only with a PLL loop.
Electrical noise means bad power supplys, bad decoupling etc. This can happen in any poorly designed circuit.
On other hand spdif have galvanic isolation which help a lot.
A limited bandwith mean square waves with low slew-rate. This is translated in high jitter.
Regards,
Tibi
PS. I already feel this discussion is going in a wrong direction, so I'll stop posting.
I have read many times that various "audiophile" playback software has claimed to be bit perfect. I'm fairly sure I have also read of USB>SPDIF converters that claim to be bit perfect. Every different reputable software I have used to play back music sounds different. Each USB>SPDIF converter I tried also sounds different. Obviously I have a preference for certain software and certain soundcards. It still does not answer the age old question of why it sounds different. The test I proposed earlier is probably a good starting point to analyze the reasons behind the differences (if any)
Look at the digital world of signal transmission...
Surprisingly it is very well studied and understood, especially when the correct signal getting through can mean the difference between life or death, or a successful mission or an disaster...
SPDIF by comparison is getting a bit of data from a CD to a DAC, not exactly rocket science.
I bet the differences would disappear in a DBT
We do not, because we are not using the CD ROM to play back directly into a DAC. Yes, we can install the CD ROM and confirm bit perfect. The next step is to use it externally of the computer and play two different CD ROM's into the same DAC and compare the sound.
Not with *only* a PLL loop. But a PLL loop with a buffering mechanism such as an asynchronous sample rate converter will deal with it.
Yes, that is my point - galvanic isolation eliminates all the electric noise.
Back to beginning of the message...
This is not true, to fast a rise time can cause more problems than a sensible rise time.
Curious as to why the discussion is going in the wrong direction, all I and others are saying is it is not an arduous interface to easily engineer a correct solution, then you can worry about the analogue section from the DAC to the speakers
You don't need any sort of "audiophile" magic. A lot of common sense ripping programs either do multiple reads to confirm a match, or check the checksum in on-line databases such as accuraterip, or both.
Obviously. Have you checked if you can hear the difference in a controlled double-blind test?
Look at the digital world of signal transmission...
Surprisingly it is very well studied and understood, especially when the correct signal getting through can mean the difference between life or death, or a successful mission or an disaster...
SPDIF by comparison is getting a bit of data from a CD to a DAC, not exactly rocket science.
I bet the differences would disappear in a DBT
I never suggested SPDIF was rocket science. I really wish you were in Australia because I would invite you over to my house for a DBT. I bet you would hear a difference
Actually, a lot of the more sophisticated modern CD players do buffered reading, where they do read the contents of the CD into a buffer in advance of playing - just like the ripping software does.
Yes, I do this as a matter of course with any new digital source before I will take it seriously. The method I use is to record the SPDIF signal into Reaper on one stereo track, then import the original file on another track. With a bit of fiddling they can be lined up to within 1 sample accuracy. Then I invert the phase of one of them and play back the sum of both tracks.
This is not a perfect test since Reaper uses floating point arithmetic internally, so you get 1 bit of rounding error here and there. But it is the easiest way I've found of doing it. When the tracks are identical, the output meters typically read -Inf or -130dB and there is no audible signal at all. You have to trust your computer to record without dropping samples.
Of course this method doesn't take account of distortion due to jitter. It is only checking the sample values, it doesn't care about their precise times of arrival. If you have an oscilloscope, you can view the eye pattern of your SPDIF link (google it) 30ns of jitter would be fairly obvious even on a cheap scope.
Last edited:
I prefer the free software with annoying interface, and laggy start time. That one sounds best to me. If there was no audible difference I would probably use one of the many convenient softwares with pretty user interface, remote control and album art. But the sound difference is so very noticeably different that I don't need a DBT.
Basically there are two theories with digital audio.
1) Bit perfect all sounds the same.
2) There are many bit perfect players but some sound better than others.
If you are a subscriber to #1 then everybody else is wrong or their gear is broken.
If you are a subscriber to #2 then you don't understand why subscribers to #1 cant hear it!
Last edited:
If I could just get $1 every time someone said that...
Well I feel the same way when someone says all bit perfect sources sound the same. But what are you going to do?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.