RME ADI-2 DAC FS - AKM Versus ESS Measurements (DAC, Preamp & Headamp)

[Svperstar]

That's quite wrong. XLR (balanced) output has 6 dB higher level. Using the 'other' +6 dB will cause distortion. It is there only for signals too low in volume.

Turning the volume up all the way on the RME causes it to distort?

 

[MC_RME]

With full level signals - obviously and of course. Not different to anything else in digital.

 

[Music1969]

That's quite wrong. XLR (balanced) output has 6 dB higher level.

So 13dBu mentioned in user manual only refers to RCA unbalanced outpout?

 

[danadam]

So 13dBu mentioned in user manual only refers to RCA unbalanced outpout?

From the manual:

Line Output: Ref Level
Sets the reference level for the analog outputs. Choices are -5 dBu, +1 dBu, +7 dBu, +13 dBu at
the RCA output, referenced to digital full scale level (0 dBFS). The levels at the XLR output are 6
dB higher, +1 dBu, +7 dBu, +13 dBu, +19 dBu.

19.2 Line Out RCA
...
Note: XLR Balanced provides a 6 dB higher output level in all settings, equalling +1 dBu up to
+19 dBu.

30.2 Analog Outputs
XLR
Output level switchable +19 dBu, +13 dBu, +7 dBu, +1 dBu @ 0 dBFS

 

[VintageFlanker]

ADI-2 DAC attenuation is relative to either analog Ref Levels: -5dBu, +1dBu, +7dBu and +13dBu (add 6dB for XLR)

Guys, it's mentioned from post #1...

 

[dped90]

Admittedly, I haven't read this whole thread yet, but did the AKM chips win the shootout over the ESS chips? If yes, is there a MCH DAC that uses those chips?

But were the DAC chips the final determinant in overall sound quality, or were other factors involved, like power supply output stages, etc? Please specify.

 

[Matias]

Please read the first post only. If too long, just skip to the conclusion maybe.

 

[Rednaxela]

If you're really in a hurry, you can even skip the first part of the conclusion.

 

[dped90]

Please read the first post only. If too long, just skip to the conclusion maybe.

Okay, read the first post completely. What is that ESS "bump" that Vintage Flanker was unequipped to measure? Does it have anything to do with that sonic "glare" which some claim they can hear from one or more ESS DAC chips?

Otherwise, at leas the numbers show that both of these stereo DACs may likely sound much the same.

My problem is that RME makes no straight 8 channel DACs. But if they did how likely would they either not sound as good as either of these stereo DACs, unless they were rightfully al lot more expensive?

 

[Rja4000]

Okay, read the first post completely.

Now we can speak.

What is that ESS "bump" that Vintage Flanker was unequipped to measure?

A "hump" or distortion level increase often seen on the early ES9038 days at intermediate levels (-35dBFS, -20dBFS). This has been documented first here, lengthfully studied in multiple threads, and fixes have been found that are now commonly implemented.

Does it have anything to do with that sonic "glare" which some claim they can hear from one or more ESS DAC chips?

We can't say anything about this because this never was objectively proven and qualified.
At that stage, this may as well just be one more of those audio legends that audiophiles like to feed.
We don't chase ghosts here.

(There is one factual difference between AKM and ESS DAC chips though, and that's their capacity to handle intersample overloads. In their implementation, RME did compensate for that difference by voluntarily lowering the digital level by 2.5dB or so before it reaches the ES chip. Both versions of the ADI-2 DAC then behave similarly for intersample overs. All the measurements you see here are with this in place, of course.)

Otherwise, at leas the numbers show that both of these stereo DACs may likely sound much the same.

Indeed. That's what the numbers show.

My problem is that RME makes no straight 8 channel DACs. But if they did how likely would they either not sound as good as either of these stereo DACs, unless they were rightfully al lot more expensive?

The ADI-2 range is a specialized range with 2 or 4 channels output. ("Pro" being 4 channels).
They are meant to be top quality and functionalities level for that purpose and they deliver on that.

If you want 8 channels or more, look at their interfaces or converters range.
I recently measured the UFX III, their top of line interface, and it's performing very well.
The UFX II and 802fs should perform identically, from what I've read.

 

[dped90]

(There is one factual difference between AKM and ESS DAC chips though, and that's their capacity to handle intersample overloads. In their implementation, RME did compensate for that difference by voluntarily lowering the digital level by 2.5dB or so before it reaches the ES chip. Both versions of the ADI-2 DAC then behave similarly for intersample overs. All the measurements you see here are with this in place, of course.)

Good to see that at least one DAC brand took the needed steps to prevent this distortion at the hardware level, as has been discussed at length here. https://gearspace.com/board/mastering-forum/1401406-intersample-clipping-audible.html

Might someone have generated a list of DACs-including multichannel DACs-which also have this overload prevention?

 

[Rja4000]

Good to see that at least one DAC brand took the needed steps to prevent this distortion at the hardware level, as has been discussed at length here. https://gearspace.com/board/mastering-forum/1401406-intersample-clipping-audible.html

Might someone have generated a list of DACs-including multichannel DACs-which also have this overload prevention?

Well, all AKM chips interfaces, to begin with.

Note that Topping D50 III also does something similar, when you select 4V output instead of 5V. I'm not sure how they do it, though. And it's not the case with other models - for what I've seen, with the ESS Topping DACs I own, neither the D90LE/SE nor the G5 do it.

 

[dped90]

Well, all AKM chips interfaces, to begin with.

Note that Topping D50 III also does something similar, when you select 4V output instead of 5V. I'm not sure how they do it, though. And it's not the case with other models - for what I've seen, with the ESS Topping DACs I own, neither the D90LE/SE nor the G5 do it.

The Topping DM7 or Okto DAC 8 Pro, possibly in that order, look to be my top choices to build my 3.3 system. At this technical level my knowledge is sketchy but proceeding: Unfortunately, those DACs won't do intersample overs distortion prevention. But like other players (e.g. Amarra, Roon, Foobar, et al), it looks like JRiver, my player of choice, had long ago implemented an update which “normalizes” program levels, thereby preventing intersample overs distortion. https://yabb.jriver.com/interact/index.php?topic=82023.0

https://tech.ebu.ch/files/live/sites/tech/files/shared/techreview/trev_2010-Q3_loudness_Camerer.pdf

 

[Rja4000]

my top choices to build my 3.3 system

That's a topic for another thread, isn't it ?

 

[Sylvia]

I just read this article, @VintageFlanker thanks a lot for the great work!

 

[VintageFlanker]

I just read this article, @VintageFlanker thanks a lot for the great work!

Thank you ! Glad people are still discovering it more than two years after publishing.

 

[jwmitchell]

Thank you ! Glad people are still discovering it more than two years after publishing.

Your comparison text is very interesting and useful,even though there are no AKM versions available except used. After reading this I purchased the ESS version and have been very happy. Now my adult son, with whom Inshare equipment has his own.Thanks!!

 

[MeatWolf]

Thank you very much for this research, been thinking about which version of RME I should get if I settle on it

 

[Matias]

This is one the most epic reviews in ASR. It should have been a white paper by RME to begin with, but the fact that VintageFlanker did it voluntarily, it's mind-blowing.

 

[VintageFlanker]