Does a truly transparent ADC actually exist?

[manisandher]

But apparently not to share, and instead got insulting immediately.

I'm accused of 'hearsay' and 'telltales' and I'm the one who's insulting? Strange interpretation of events.

I imagine there is a threshold, although it seems given these comments a valid test of two A2D doe seem out of scope for you.

I'm happy to do more tests. It's just a matter of having the time to do them properly.

Mani.

 

[Blumlein 88]

Is this actually a widely held belief? I'm unaware of anyone in my line of work who thinks that.

Actually, one of our more commonly used ADCs is the Burl B2 precisely because the colourations it imparts tend to pair well with the more shrill sounding sources (and there are a LOT of those). Ideally, we'd opt for a transparent converter if possible, but given the options we tend to go for what would pair best with whatever project we're working on.

Why would you expect a colored ADC like the Burl to sound transparent. A key question is how did you level match? If the answer is by ear, then there is your answer. Things will sound perfect in volume matching by ear, but being even off 1/4 db will cause a difference in sound quality.

 

[DonH56]

Higher sampling rates tend to mean higher noise since the effective noise bandwidth is greater. I have not kept up with audio ADCs, but in general the faster you operate, the higher the distortion. But delta-sigma ADCs tend to obviate that issue since they (over) sample at a much higher rate in the front end anyway. Filter choice could influence the results but I have no experience (or no recent experience) playing with the filters. Higher rates, and correspondingly higher input bandwidth, can reduce aliasing from band-limited sources (as any input should be) but as noted above can lead to in-band IMD that may cause audible differences (I'd expect it would have to be pretty bad, however).

I guess to really dig deeper you'd want to know the exact ADCs involved (the chips inside the boxes) and their architecture (e.g. delta-sigma, SAR, etc.) as well as the buffers. If two different ADC boxes have the same ADC chips inside then something else is going on, whether input buffers and filters, or the way the ADC's output is (digitally, internally) processed before being passed out and down the line.

I may have missed it: what are "digital files" that you run through the ADC? There must be a DAC involved, so are you distinguishing an analogue source (e.g. tape) from a digital source file run through a DAC into the ADC? A number of analogue sources like records (LPs, vinyl), tape, and FM signals can have high-frequency content lacking in typical 44/48 kHz source files, and that HF content could be affecting the ADCs. That would also explain why lower rates sound better, assuming when you drop the rate, the input anti-alias filter is also dropped in frequency and so rejects the high-frequency noise and signals.

ADC overdrive (and related level matching) through the chain are big players and one of my prime suspects "from afar", though of course the input buffers of the ADC could be designed to intentionally color the sound, and it is true enough that different ADCs can have different noise and distortion signatures.

Interesting topic.

 

[manisandher]

Over the weekend, I'll capture the analogue outputs of the SMSL SU-10 with the RME ADI-2 Pro and the E1DA (level matched to 0.1dB). I'll link the files so people can hear/test for themselves.

 

[MaxwellsEq]

Over the weekend, I'll capture the analogue outputs of the SMSL SU-10 with the RME ADI-2 Pro and the E1DA (level matched to 0.1dB). I'll link the files so people can hear/test for themselves.

Thanks. Can you do this sampling at 44.1, 48, 96 and 192. What do you plan as the "content" into SU-10? Do you have an analogue source as well? And it would be useful to try the 4 different filters on the AD of the ADI-2.

 

[HarmonicTHD]

Over the weekend, I'll capture the analogue outputs of the SMSL SU-10 with the RME ADI-2 Pro and the E1DA (level matched to 0.1dB). I'll link the files so people can hear/test for themselves.

Good.

I have seen this as I was interested if you had indeed done measurements in the past:
Thread 'A More Useful Way of Measuring the Performance of DACs?'
https://audiosciencereview.com/foru...y-of-measuring-the-performance-of-dacs.39379/

Great effort I have to say. But it got dodgy on the audibility question. So see if you can do better this time.

 

[manisandher]

Can you do this sampling at 44.1, 48, 96 and 192. What do you plan as the "content" into SU-10? Do you have an analogue source as well? And it would be useful to try the 4 different filters on the AD of the ADI-2.

Doubt I'll have enough time to do all this. Was thinking of just doing the following:

1. Take a 'well-recorded' 24/96 track (original)
2. Playback through SU-10
3a. Capture with ADI-2 (steep filter, level-matched to 0.1dB of original)
3b. Capture with E1DA (linear phase steep filter, level-matched to 0.1dB of original)
4. Link all 3 files here

Do you have an analogue source as well?

I do, but what's your thinking here? I mean, only I'd be able to compare the analogue source to the digital captures, no?

Mani.

 

[manisandher]

So see if you can do better this time.

How about dialling down the condescension?

 

[HarmonicTHD]

How about dialling down the condescension?

It wasn’t meant condescendingly. It is simply the most relevant question. Edit: the audibility I mean.

 

[kemmler3D]

Over the weekend, I'll capture the analogue outputs of the SMSL SU-10 with the RME ADI-2 Pro and the E1DA (level matched to 0.1dB). I'll link the files so people can hear/test for themselves.

Love it! Thanks for this.

 

[617]

Worry about the music. The best music ever was recorded on what would now be considered shit. ADC/DAC is a solved problem.

 

[pablolie]

1) Is it normal for ADCs to colour the sound a certain way?

Absolutely not. Shannon-Nyquist is the truth.

3) Given the above, is it fair to assume that it's the analogue section of these devices that is causing the issue? Or is my hypothesis way off the mark?

The analogue section is what we ultimately hear, but these days it isn't rocket science, the designs are proven and true.

That said - one thing to watch out in a chain is to keep the D-A and A-D conversions to a minimum, ideally just one conversion.

4) I've been told that DSD is better for the purpose - I have used a Tascam that had DSD capability and it was pretty awful. Is this genuinely true and the Tascam is just a poor device?

No advantage in DSD.

4) Is there such a thing as an actual transparent ADC, or am I setting my standards too high?

Shannon-Nyquist guarantees utter transparency in the frequency spectrum, and 16bit coding was established to do the same for the signal - but we know more bits are used in recording and often in the final digital encoded file, too.

I find it funny like in audio everybody is free to question the technical standards established after much research, but none of those people question, say, the digital fuel injection systems in their cars or other digital to analog systems that permeate our lives in many environments.

If anybody has used an ADC that is truly transparent, then please recommend.

Many thanks

 

[MaxwellsEq]

3a. Capture with ADI-2 (steep filter, level-matched to 0.1dB of original

There are 4 ADI-2 filters, two of which are steep. My reason for suggesting testing the filters is to understand why your experience of the E1DA is more positive than the ADI-2, since (admittedly different chips), they should be in the same ballpark of performance. I wondered if you are comparing like-for-like if you've not tried all the different ADC options that both devices offer. Is this, for example why two different, theoretically transparent ADCs are considered to sound different?

Do you have an analogue source as well? And it would be useful to try the 4 different filters on the AD of the ADI-2.

I do, but what's your thinking here? I mean, only I'd be able to compare the analogue source to the digital captures, no?

You propose to use a well recorded 24/96 piece of music. The trouble is that this is already an ADC-safe piece of content because it's been through a commercial ADC. I'm thinking of what @DonH56 said. Some of the differences that the OP raised were to do with analogue source behaviour when hitting the ADC filter. Vinyl replay is a mess of distortion and resonance above 10kHz whilst analogue tape has bias frequencies. Many sources have noise behaviour that increases with f. Are these causing the "character" the OP was describing. Your already-ADC-ed content will not stress your ADCs at all in these areas.

 

[manisandher]

I wondered if you are comparing like-for-like if you've not tried all the different ADC options that both devices offer.

I think using linear phase steep filters in both ADCs is the best way to ensure consistency. (In the RME, I'll use 'Sharp', and not 'SD Sharp'.)

You propose to use a well recorded 24/96 piece of music. The trouble is that this is already an ADC-safe piece of content because it's been through a commercial ADC. I'm thinking of what @DonH56 said. Some of the differences that the OP raised were to do with analogue source behaviour when hitting the ADC filter. Vinyl replay is a mess of distortion and resonance above 10kHz whilst analogue tape has bias frequencies. Many sources have noise behaviour that increases with f. Are these causing the "character" the OP was describing. Your already-ADC-ed content will not stress your ADCs at all in these areas.

I understand.

Say I use a vinyl source then. I'll end up with two 24/96 captures. We can then only determine which is more accurate by comparing to the analogue source, which only I can do. Pretty certain people here wouldn't accept my assessment, even though I have no skin in the game and have no reason to 'promote' the $200 E1DA over the more costly (but far better-equiped) RME.

Mani.

 

[MaxwellsEq]

Say I use a vinyl source then. I'll end up with two 24/96 captures. We can then only determine which is more accurate by comparing to the analogue source, which only I can do. Pretty certain people here wouldn't accept my assessment, even though I have no skin in the game and have no reason to 'promote' the $200 E1DA over the more costly (but far better-equiped) RME

I know, but it would be interesting to see whether you consider the "winner" on the 24/96 source is the same "winner" with vinyl. I have the ADI-2 Pro, and honestly, vinyl I've digitised sounds identical to the LP.

 

[olieb]

Say I use a vinyl source then. I'll end up with two 24/96 captures. We can then only determine which is more accurate by comparing to the analogue source, which only I can do. Pretty certain people here wouldn't accept my assessment, even though I have no skin in the game and have no reason to 'promote' the $200 E1DA over the more costly (but far better-equiped) RME.

But the claim that different ADCs sound different could still be assessed. For this the "original" is not needed. And only if there is a difference at all one version could be "better".

 

[manisandher]

But the claim that different ADCs sound different could still be assessed. For this the "original" is not needed. And only if there is a difference at all one version could be "better".

Sure. But a digital source would be even more useful, as, as well as listening, you could perform nulling against the original, etc.

In any event, I'll see how much I can get through in the short time I'll have.

Mani.

 

[JamesQuorn0]

Take a look at this. some people do prefer vinyl, but technically it's inferior.

It's just a fact that digital is better than analog vinyl in every way. A digital copy can sound like vinyl but vinyl can't sound like digital. There is ALWAYS audible surface noise on a record, and sometimes some nasty clicks & pops. There are often frequency response variations and sometimes distortions. (Vinyl can go higher than CDs... into the inaudible ultrasonic range but digital is flatter over the audible range, and most of the ultrasonics are noise. Or, you can sample at 192kHz to capture the ultrasonics.

Far be it from me to defend vinyl (I find the whole thing a pain, to be honest - and listen digitally 90% of the time), but the 'technicalities you speak of' (presumably THD, SINAD, and so on) don't quite measure up with aural perception. There's a lot of research into this, for instance:

https://hifisonix.com/wordpress/wp-content/uploads/2017/11/Perceptual-Levels-of-distortion.pdf

It's also worth looking into Auditory Scene Analysis, studies displaying the impact of hypersonics using EEGs, even the impact of auditory subliminals. Point being, there's a lot of stuff one DOESN'T consciously hear that impacts overall listening pleasure. It's a fascinating subject - I attended a talk a couple of years back on the subject in Manchester and it was eye (well, ear) opening.

Digital is a technically superior product given the measurables at our disposal today, but does it 'sound' as good? Working on both mediums (tape mainly, as opposed to vinyl) daily for a living, I can tell you that digital audio is certainly more fatiguing (in captures, where mastering is a non-issue) and this is a commonly shared feeling from those working with both. I wish it wasn't so, and is in fact one of the reasons I started this thread in the first place! Clearly, this type of evidence is anecdotal and will remain so until the ability to measure these great unknowns is available to us.

Now, that'll put the cats amongst the pigeons!

 

[DonH56]

Say I use a vinyl source then. I'll end up with two 24/96 captures. We can then only determine which is more accurate by comparing to the analogue source, which only I can do. Pretty certain people here wouldn't accept my assessment, even though I have no skin in the game and have no reason to 'promote' the $200 E1DA over the more costly (but far better-equiped) RME.

Mani.

One thing you could do is capture simultaneously into the two ADCs then use Deltawave or similar SW to compare the digital data, or just run some spectral plots and see if anything "interesting" shows up. You could do this with a test tone (or tones) as well as analog and/or digital sources.

Thanks for doing this!

 

[Blumlein 88]

Yet no answer about how you matched levels. @JamesQuorn0