Analog, Digital and all that

[Empirical Audio]

Tell me about it , what brands do you consider GOOD digital then .
I ll give them a listen next year in munich

I doubt if you will find Digital nirvana at the Munich show. I am of course partial to my own designs, but there are a few notable others out there, including Amarra,(software) Linn Audio (software), SOtM, Lampizator, Metrum Acoustics, Chord and TotalDAC. It's more of a particular component or software than a brand name, and it's of course a system thing always, so everything else involved is important as well, from the source software to the digital cables used and the means for transmission, whether S/PDIF, USB or Ethernet. Simply everything matters and buying all of it from one manufacturer is not advisable. No one company excels at everything. If you are lucky enough to start with a low jitter signal and deliver it to a low-distortion resolving DAC using good cables, then the magic happens. Anything in the system that is not optimized or not selected carefully can screw this up. I've been reducing jitter with proprietary techniques for 20 years and if there is one thing I have learned it's that jitter is never low enough to be inaudible, at least not yet. And jitter isn't the only demon. Digital Filtering is #2 on my list. I try to eliminate it as much as possible. This is why NOS type DACs are so popular now. It's not really the difference in conversion method IMO, but the difference in digital filtering that makes these so appealing.

One metric to determine if you are close to getting the max out of digital is to listen to the same track in FLAC, AIFF, ALAC and .wav formats. If you don't easily hear a difference in each one, you are not there yet. The .wav file should win hands-down.

 

[Empirical Audio]

Simple: that many listeners don't like the truth, but a beautified sound.

Unfortunately, you are right on this one.

 

[Solypsa]

snip- Because of the improved performance of digital over LP, it is possible that more cables will limit the performance or more preamps will limit the performance than with LP system that has lower bandwidth, lower resolution and lower dynamics.

I am not about to get into a cable debate (not even a little), but coming from a mastering studio background and being a pragmatic cable guy I would like to see the measurements that indicate how the bandwidth; specifically outside of vinyl media limitations, and within digital media limitations, is being limited by the cable (given we are talking about normally good stuff)

 

[Empirical Audio]

I am not about to get into a cable debate (not even a little), but coming from a mastering studio background and being a pragmatic cable guy I would like to see the measurements that indicate how the bandwidth; specifically outside of vinyl media limitations, and within digital media limitations, is being limited by the cable (given we are talking about normally good stuff)

Steady-state bandwidth is the same, no effect. It's the dynamics that are affected by dielectric absorption, skin effects, conductor molecular effects and other physical phenomenon.

For instance, I have measurements on a TDT (throughput TDR) showing the effect of poor cryo-treatment of a cable that broke-up the crystal lattice in the Silver conductors compared to the same cable with no treatment. The pulse waveform shows much different response, with more reflections at the broken crystal lattice boundaries. The L, R and C measurements were identical before and after the treatment at all frequencies from 10Hz to 100kHz. Also, I had to throw this cable in the trash. It was unlistenable. This demonstrates that the crystal lattice is important for dynamics, but not steady-state waveforms.

Green trace is the untreated cable and red is the treated cable.



These were analog cables, RCA terminated.

The effect is even more catastrophic if the cable is a S/PDIF digital, USB or Ethernet cable.

 

[jespera]

I doubt if you will find Digital nirvana at the Munich show. I am of course partial to my own designs, but there are a few notable others out there, including Amarra,(software) Linn Audio (software), SOtM, Lampizator, Metrum Acoustics, Chord and TotalDAC. It's more of a particular component or software than a brand name, and it's of course a system thing always, so everything else involved is important as well, from the source software to the digital cables used and the means for transmission, whether S/PDIF, USB or Ethernet. Simply everything matters and buying all of it from one manufacturer is not advisable. No one company excels at everything. If you are lucky enough to start with a low jitter signal and deliver it to a low-distortion resolving DAC using good cables, then the magic happens. Anything in the system that is not optimized or not selected carefully can screw this up. I've been reducing jitter with proprietary techniques for 20 years and if there is one thing I have learned it's that jitter is never low enough to be inaudible, at least not yet. And jitter isn't the only demon. Digital Filtering is #2 on my list. I try to eliminate it as much as possible. This is why NOS type DACs are so popular now. It's not really the difference in conversion method IMO, but the difference in digital filtering that makes these so appealing.

One metric to determine if you are close to getting the max out of digital is to listen to the same track in FLAC, AIFF, ALAC and .wav formats. If you don't easily hear a difference in each one, you are not there yet. The .wav file should win hands-down.

Is the problem a) jitter as in time variation of the digital signal OR b) outright noise (switching and ringing) on the digital signal?

If you measure for a) on b) you’ll pick up something but the two things are not exactly the same.

Jesper

 

[andromedaaudio]

But between tape recordings the differences are substantial as well.
I have now tapes from about 20 different kinds of manufacturers .
I have some for which i paid 300 euros a reel for and the quality is not even better then CD s .
See with reel to reel live recordings its essential you place the mikes correct .
I think quite high off the ground and close enough to just not overload the mike , but also not further.
To pick up maximum pianosound and less ambient sound .

Tape adds a recorded hiss level , so maximizing /capturing dynamic swings is essential .
But if its done correctly , tape is another level

 

[Empirical Audio]

Is the problem a) jitter as in time variation of the digital signal OR b) outright noise (switching and ringing) on the digital signal?

If you measure for a) on b) you’ll pick up something but the two things are not exactly the same.

Jesper

The effect is reflections, which puts noise on the digital signal. The noise can add to the jitter, depending on when the reflection arrives at the destination.

The really unexpected result is with the analog cable. These effects should be well beyond audibility, but it's not. You can easily hear the difference and the treated cable is really awful. Worse than using a hangar wire for an interconnect I think.

 

[Empirical Audio]

But between tape recordings the differences are substantial as well.
I have now tapes from about 20 different kinds of manufacturers .
I have some for which i paid 300 euros a reel for and the quality is not even better then CD s .
See with reel to reel live recordings its essential you place the mikes correct .
I think quite high off the ground and close enough to just not overload the mike , but also not further.
To pick up maximum pianosound and less ambient sound .

Tape adds a recorded hiss level , so maximizing /capturing dynamic swings is essential .
But if its done correctly , tape is another level

Certainly 30IPS professional tape is superb, given the right playback equipment. The problem is noise reduction of the hiss. It's a lot like RIAA EQ for LP's. If the electronics is not perfect, it will add artifacts. Electronics is never perfect. Also, the speed of the tape is never perfect, just like the speed of the turntable is never perfect. Digital has these beat with stable, accurate crystal oscillators.

 

[jespera]

The effect is reflections, which puts noise on the digital signal. The noise can add to the jitter, depending on when the reflection arrives at the destination.

The really unexpected result is with the analog cable. These effects should be well beyond audibility, but it's not. You can easily hear the difference and the treated cable is really awful. Worse than using a hangar wire for an interconnect I think.

If its just a question about reflections, ie a transmission line problem, then id expect it to be the same whether you use a computer or a cd player as the digital source.

In my experience computers are way harder to get to work well than cd players.

I like the analog cable story because you actually show objective differences of cryo treatment of cables.

 

[Empirical Audio]

If its just a question about reflections, ie a transmission line problem, then id expect it to be the same whether you use a computer or a cd player as the digital source.

In my experience computers are way harder to get to work well than cd players.

I like the analog cable story because you actually show objective differences of cryo treatment of cables.

I always recommend the digital cable be 1.25-1.5m long minimum. This is so that the first reflection comes after the signal transition. This is based on this anaysis:

http://www.positive-feedback.com/Issue14/spdif.htm

This was verified by single-blind testing at an audiophile magazine. They were not convinced after reading my white-paper, so they did their own A/B tests.

I have also discovered over the last 10 years that digital S/PDIF cables should always be terminated to 75 ohm BNC connector designed for that specific type of cable. Terminating to RCA connectors creates large discontinuity and therefore reflections. Bigger discontinuity than RCA to BNC adapters. I have performed a number of jitter plots that show the difference here:

https://www.audiocircle.com/index.php?PHPSESSID=q0se2hsb2hfue81kmjo5nimif0&topic=154425.0

 

[jespera]

But between tape recordings the differences are substantial as well.
I have now tapes from about 20 different kinds of manufacturers .
I have some for which i paid 300 euros a reel for and the quality is not even better then CD s .
See with reel to reel live recordings its essential you place the mikes correct .
I think quite high off the ground and close enough to just not overload the mike , but also not further.
To pick up maximum pianosound and less ambient sound .

Tape adds a recorded hiss level , so maximizing /capturing dynamic swings is essential .
But if its done correctly , tape is another level

If the tapes sound like cds maybe they are just tape copies of cds or computer files.

Most modern lp reissues sound like cds — go figure.

 

[Solypsa]

...Green trace is the untreated cable and red is the treated cable.

View attachment 65239

These were analog cables, RCA terminated.

The effect is even more catastrophic if the cable is a S/PDIF digital, USB or Ethernet cable.

Thanks for the graph. Interesting.

 

[andromedaaudio]

@jespera


No i dont question that , but i do question their recording techniques .

A very good example is Analogue productions
Its a popular artist hugh masekela they brought on tape with great music .
The recording is excellent no tape hiss basically as quit as CD .
Luciano pavarotti / hemiolia is another one .

So its a question of the right technique not the medium

 

[jespera]

Analogue Productions and tape is awesome

 

[Empirical Audio]

Does your DAC sound different playing the same content redbook files when fed by a streamer or a CD transport?

Update: I discovered the settings in my Amazon streaming app were screwed-up. Music now sounds identical played from CD/Synchro-Mesh, local disk or when streaming Amazon Music HD. No difference. I can get 24/96 with most Amazon tracks.