An interesting comment from Cookie Marenco was that she found (comparing digital formats in her studio) PCM to be more accurate to the sound produced by the Mic, but DSD more accurate to the sound in the room.
Why, oh why, does vinyl continue to blow away digital?
- Thread starter godofwealth
- Start date
I would agree with that observation. The sound in the room is “soften” to some extend by the absorption and diffraction, while the microphone feeds sound crisp and extended. I would say that my own views on PCM versus DSD are inline with her views. I prefer DSD but enjoy PCM though HQPLAYER also.
A link would good…
I am assuming cosmology of some sort?
(Which I enjoy)
PM me if that seems more appropriate.
Some of my electronics designs are onboard the International Space Station. They are not trivial but rather crucial in significance. Others were part of telemetry systems on the now retired Space Shuttle fleet.
Last edited:
Sorry but the input is digital and the output is analog…how does that comparison work exactly??
It’s all very simple. You send a pure sine wave, in digital form, to the Dac’s digital inputs and verify how accurately it is reconstructed at the analog outputs. Simple as that.
AI says: It is not possible for an ADC-DAC system to reproduce an analog signal with 100% accuracy. There are inherent limitations and sources of error that prevent perfect reconstruction of the original analog signal.
Some of the key reasons why 100% accurate analog signal reproduction is not achievable with an ADC-DAC system include:
1. Quantization Error:
- The ADC can only represent the analog signal with a finite number of digital levels, based on its bit depth.
- This introduces quantization error, where the digital representation cannot perfectly match the continuous analog value.
- Even with high-resolution ADCs, there will always be some level of quantization error.
2. Sampling Limitations:
- The ADC can only capture the analog signal at discrete time intervals, based on the sampling rate.
- Any information between the sampling points is lost, leading to potential aliasing and distortion.
- Perfect reconstruction requires the sampling rate to be infinitely high, which is not practical.
3. Reconstruction Filter Limitations:
- The DAC requires a reconstruction filter to remove high-frequency artifacts introduced during the digital-to-analog conversion.
- However, the reconstruction filter cannot be an ideal brickwall filter, and it will always introduce some level of frequency-domain distortion.
4. Timing and Synchronization Errors:
- Precise timing and synchronization between the ADC and DAC sampling clocks are crucial for accurate signal reconstruction.
- Any jitter or timing mismatch between the clocks will result in timing errors, leading to signal degradation.
Due to these fundamental limitations, it is not possible to achieve 100% accurate analog signal reproduction using an ADC-DAC system. The best achievable performance will depend on factors like the ADC and DAC specifications, the system design, and the specific application requirements.
In practice, the goal is to minimize the various sources of error and achieve the highest possible fidelity within the constraints of the system. However, some level of error and deviation from the original analog signal will always be present.
Some of the key reasons why 100% accurate analog signal reproduction is not achievable with an ADC-DAC system include:
1. Quantization Error:
- The ADC can only represent the analog signal with a finite number of digital levels, based on its bit depth.
- This introduces quantization error, where the digital representation cannot perfectly match the continuous analog value.
- Even with high-resolution ADCs, there will always be some level of quantization error.
2. Sampling Limitations:
- The ADC can only capture the analog signal at discrete time intervals, based on the sampling rate.
- Any information between the sampling points is lost, leading to potential aliasing and distortion.
- Perfect reconstruction requires the sampling rate to be infinitely high, which is not practical.
3. Reconstruction Filter Limitations:
- The DAC requires a reconstruction filter to remove high-frequency artifacts introduced during the digital-to-analog conversion.
- However, the reconstruction filter cannot be an ideal brickwall filter, and it will always introduce some level of frequency-domain distortion.
4. Timing and Synchronization Errors:
- Precise timing and synchronization between the ADC and DAC sampling clocks are crucial for accurate signal reconstruction.
- Any jitter or timing mismatch between the clocks will result in timing errors, leading to signal degradation.
Due to these fundamental limitations, it is not possible to achieve 100% accurate analog signal reproduction using an ADC-DAC system. The best achievable performance will depend on factors like the ADC and DAC specifications, the system design, and the specific application requirements.
In practice, the goal is to minimize the various sources of error and achieve the highest possible fidelity within the constraints of the system. However, some level of error and deviation from the original analog signal will always be present.
No, I mean you can’t compare the digital file to the analog output directly…they are fundamentally different formats. However, I will address your simple (and wrong) explanation.
Then do an FFT of the output and look at all the crap the DAC added to your “pure »”digital sine wave. And that’s just one frequency. When you have multiple frequencies you get inter modulation distortion components. Just two frequencies will already generate a whole bunch more. Then you have effects from jitter and effects from negative feedback, pre and post ringing, noise shaping etc. And worse none of this is any way natural so that your ear/brain can ignore or mask it. Pretty soon your “pure” sine wave won’t look so pure.
You missed the fact that is the whole point of the Sine wave reconstruction test. And the FFT will tell you the sonic characteristics of the DAC. Interesting how you said all that without quite understanding the underlying principle of what you were stating.
What you need to keep in mind is that the magnitude of these “digital” errors are orders of magnitude lower than the errors incurred during “analog” playback.
How do you convert digital without the DAC?? Of course to look at the analog sine wave you have to run it through a DAC and analog output stage. Don’t be daft. The only way to see if the sine wave is pure post conversion is to perform an FFT… you certainly can’t see it from the wave on a scope unless it grossly distorts.
The size is not what’s important…it’s tge nature of the distortion.
The reconstructed sine wave at output can be evaluated through FFT on a high quality Oscilloscope with FFT functionality. Take a look at John Atkinson’s sine reconstruction test for some simple understanding. JA uses an1KHz sine wave but to have a comprehensive analysis you need the test from 20Hz to 20KHz.
The FFT shows all the frequency components of a signal…it’s not the waveform itself. What is immediately obvious is that even a single 1khz sine wave will have lots of small distortion components…multiple frequencies will each have this plus the inter modulation components. Then you have all the conversion errors mentioned by Rexp. The pint is that it isn’t pure on the other side of the DAC even though, yes it still looks like a sine wave on the other side(thank god for that at least). Look I like digital just fine but don’t blow sunshine up everyone’s ass saying it is more accurate. In some ways yes and in some important ways no.
What we are after are the magnitude of those components that we don’t expect to be present. Those contributions are the ones that tell the story.
Well it’s even worse Al, it’s the SAME cable for spdif and AES/EBU…same make and model. So, I normalized for the cable. I am either hearing a difference in the digital interface format, the way the Mutec sends out a timing signal in the two formats or how my DAC receives those two formats. I was hoping that when I normalized for the cable the difference would disappear…it did not and appears to be a bigger issue than the cable itself.I agree with you that cable quality is of utmost importance in digital, even though "theoretically" digital cables should make no difference. I have a very expensive AES/EBU cable from Mutec reclocker to DAC (see my signature) for a reason. And I am not claiming that is perfect either.
It is also true, as you say:
"Digital has clearly lower distortion but the type that is there is wholly unnatural and has no basis in nature."
That is why the human ear/brain interface is several orders of magnitude more sensitive to digital jitter than it is to roughly its equivalent, analog wow and flutter. A major problem is that while in wow and flutter all frequencies are affected equally, in jitter the timing error affects different frequencies differently -- and in an inconsistent manner, depending on the music signal from moment to moment.
Digital can be made musical, but this requires reducing all its unnatural distortions to a minimum. To my ears, I have achieved that in my system, making digital a highly engaging and enjoyable experience for me.
Funny though that some of the nasty distortions that I previously might have attributed to digital, or actually did so, were in fact room distortions and distortions resulting from suboptimal speaker set-up in the room. Or distortions that were downstream of the digital related to suboptimal power delivery to components (remedied by competent power cords). Digital isn't to blame for everything.
You have a Mutec, right? Try this experiment if you can.
Agreed. I have mentioned in #384 jitter as an example for digital distortions that, as you say, are unnatural:
The human ear/brain interface is several orders of magnitude more sensitive to digital jitter than it is to roughly its equivalent, analog wow and flutter. A major problem is that while in wow and flutter all frequencies are affected equally, in jitter the timing error affects different frequencies differently -- and in an inconsistent manner, depending on the music signal from moment to moment.
Basing "accuracy" on magnitude of distortions is an inaccurate, far too simple approach to things.
That is also why THD numbers of amplifiers are relevant only to a very limited extent. Very small amounts of 7th or 9th harmonic can be far more detrimental to the listening experience than larger amounts of 2nd or 3rd harmonic.
The problem with high-end audio digital cables is that they are not compliant or certified to any standards. A simple eye-pattern test would go a long way at evaluating their differences.