• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

RME ADI-2 FS Version 2 DAC and Headphone Amp Review

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,029
Likes
10,796
Location
São Paulo, Brazil
Bought mine from Nord, now waiting for delivery.
 
Last edited:

eliash

Senior Member
Joined
May 29, 2019
Messages
407
Likes
209
Location
Bavaria, near lake Ammersee
I'm not sure there is a benefit of using the +13dBu range

Here is 800mV input, measured with ADI-Pro fs R (Black) and ADI-Pro fs (Red) ADC, using range +4dBu
(+4dBu-3.7dB = 0.3dBu = -1.9dBV=0.8V) for different scenario
DAC is the ADI-Pro fs R for all. (So when comparing black and red, we are comparing ADCs)

First straight from the DAC, +4dBu output range, with some digital attenuation (RME ADI-2 Pro fs R - new version)

View attachment 67722

Then with DAC set for +13dBu output range, with passive attenuation

View attachment 67723

Finally, with DAC set to +4dBu output range, full output, with a light passive attenuation

View attachment 67724


That illustrates several things, I think
1. ADC is working (slightly) better on "Pro fs R" version than on "Pro fs Non R" version
(That was actually why I was measuring this to begin with)

2. Using +13dBu to generate a signal close to 0dBu is less efficient than using the +4dBu range.

I guess this is true for the ADI-2 "Non Pro" fs too...



Notes:
1. I never "calibrate" the output.
So when you see in the signal generator "Output amplitude" = 1V, that means 0dBFS for the DAC, which is the full range (+4dBu or +13dBu in those examples).
2. Those measurements are all loopback, which accumulates noise and distortions from the DAC and the ADC (and from the passive attenuators where it's used). So 114.5dB SINAD here, while attenuated by 3 7dB, means that both DAC and ADC have SINAD better than that. How much better we can't tell.


The reason for using +13dBu in this case is very simple, +4dBu does not achieve 800mV on one XLR output leg, but -2dBu. So the choice was to use +13dBu, which generates +7dBu on one output leg, or +19dBu, which generates +13dBu respectively (always connected to RCA amp input).
Please note, there is no digital attenuator, but only the vol. pot available in that device (which has about 0.3dB channel balance offset, if used).
In order to keep the opamps´ output currents in best performance spec range and to achieve a low output impedance of the resistor attenuator, I chose the +13dBu setting.
Reading Amir´s test, it may be preferable to use the +19dBu output setting, the advantage may be, next to lower distortion, some less opamp idle noise after attenuation (but margin for oversampling overshoot has to be observed in that calculation*).
To finally explain why I did the whole thing, it is to gain an idea where speaker amp clipping occurs at a certain amp´s vol. pot. setting - in my case it is the 2 o´clock position, where the pre-amp has about 0dB gain, driving the post-amp with 800mVeff @0dBfs (actually 2x800mV on 5m XLR cable), which is its clipping point (of course, personally observed oversampling generates up to 2-3dB higher output peaks though...could be even 6dB in theory, as we learned here in the Forum**).
This general input level adjustment also allows for analog/vinyl conversion precision for the connected Clarity M audio analyser (via the ADI-2 ADC section). Going higher in that reference level is also possible (there is a -6/0/+6dB gain switch inside the preamp), but not to exclude some older analog sources I left it that way, since the overall noise margin allows...


A late remark concerning */** above:
In case, a theoretical oversampling overshoot of 6dB shall be covered without distortion by the DAC´s output stage, +19dBu output setting plus 6dB overshoot correspond to almost +/-10Vpeak output swing on each leg of the XLR output!
 
Last edited:

eliash

Senior Member
Joined
May 29, 2019
Messages
407
Likes
209
Location
Bavaria, near lake Ammersee
That's exactly what RME did with new generation of tge ADI-2 DAC and with the Pro fs R.
That's new design, though.
I don't know if they have a newer version of the ADI-2 fs (non Pro)

...The ADI-2 fs was annouced about a year before its actual availability in last winter, I guess RME and customers (like myself) are happy to have it on the market at all;)...no bad feelings about that:)
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,029
Likes
10,796
Location
São Paulo, Brazil
Does anyone have an idea why the new version has a higher noisefloor on the highs compared to the old version?
Looks like it traded low frequency noise floor to high frequency noise floor.

v1
index.php


v2
index.php
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,682
Likes
4,220
Location
Liège, Belgium
Does anyone have an idea why the new version has a higher noisefloor on the highs compared to the old version?
Looks like it traded low frequency noise floor to high frequency noise floor.
I don't know for the DAC, but here are some plots for the RME ADI-2 Pro fs vs RME ADI-2 Pro fs R

This is Loopback, so cumulating distortions and noise from DAC and from ADC
(so not to be strictly compared with Amir's measurements)

@+13dBu, 31 tones, 192kHz, Loopback in Mono mode (Sum of Main 1+2)

ADI-2 Pro fs
RME ADI-2 Pro fs @+13dBu 31 tones 192kHz.png



ADI-2 Pro fs R
RME ADI-2 Pro fs R @+13dBu 31 tones 192kHz.png



@+24dBu, 100 tones, 48kHz, Loopback in Mono mode (Sum of Main 1+2)

ADI-2 Pro fs
RME ADI-2 Pro fs @+24dBu 100 tones 48kHz.png


ADI-2 Pro fs R
RME ADI-2 Pro fs R @+24dBu 100 tones 48kHz.png
 
Last edited:

SimpleTheater

Addicted to Fun and Learning
Forum Donor
Joined
Jun 6, 2019
Messages
927
Likes
1,789
Location
Woodstock, NY
Does the parametric EQ work only for headphones or with any of the outputs?

Also, can you save EQ settings (e.g. one for headphones, on for speakers), assuming the answer to the question above that it works for all outputs.
 

VintageFlanker

Major Contributor
Forum Donor
Joined
Sep 20, 2018
Messages
4,940
Likes
19,691
Location
Paris
Does the parametric EQ work only for headphones or with any of the outputs?
For all outputs.
Also, can you save EQ settings (e.g. one for headphones, on for speakers), assuming the answer to the question above that it works for all outputs.
Of course. The EQ is different for HF and speakers anyway. And you can save many for each.
 

Hoary

Member
Joined
Apr 27, 2020
Messages
16
Likes
32
To add:

For all outputs.

And it will remember which EQ setting was used on particular output last time. So, you don't have to change them when you switch output.

Of course. The EQ is different for HF and speakers anyway. And you can save many for each.

One can store up to 20 different EQ settings and give each a meaningful name.
 

OP1M.DR3M

Member
Forum Donor
Joined
Apr 29, 2020
Messages
76
Likes
24
Location
California, USA
Does anyone know if the XLR outputs on the ADI-2 FS Version 2 are differential? Thanks!
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,029
Likes
10,796
Location
São Paulo, Brazil
Does anyone know if the XLR outputs on the ADI-2 FS Version 2 are differential? Thanks!
By definition balanced is differential as far as I know, and the XLR outputs are balanced/differential.
 

eliash

Senior Member
Joined
May 29, 2019
Messages
407
Likes
209
Location
Bavaria, near lake Ammersee
Does anyone know if the XLR outputs on the ADI-2 FS Version 2 are differential? Thanks!

No, a differential output is usually ground-free, but here both output legs are referenced to ground (with opposite polarity, at least for the ADI-2 fs - RME describes this behaviour in their manuals, so you can just look it up there for a specific model...).
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,029
Likes
10,796
Location
São Paulo, Brazil
Amir's measurement of the improved SINAD at higher levels than 4 Volt output inspired me to show the superiority of the ADI-2 DAC's multiple hardware reference levels, that anyone can use without having to think about it by using the default AutoRef setting. What happens is that when you lower the volume every 6 dB (4 x ) the discrete hardware lowers its amplification and the DAC's digital level is ramped up again. So lowering the volume will not make you loose all the THD+N, or SINAD.

View attachment 63801

First I replicated Amir's measurement (pretty much identical), then added the SINAD states that occur when the hardware level is changed. That didn't fit completely into the measurement, and also requires a logarithmic scale on the x axis for the lower levels. Then I set the reference to +19 dBu and schwuppdiwupp had the SINAD that one gets when changing the volume at the unit. Finally I added the ref levels in balanced and unbalanced mode on top as text. I hope this is understandable. Also please don't get confused: this is about manually setting volume digitally at the unit, not about a swept generator with automatic output level action - that would make no sense.

View attachment 63806

So what it shows is that when you use a standard DAC and only look at the SINAD at 0 dBFS you miss it all. How often do you use 0 dB volume? Most of the time people lower the volume, definitely so when using phones. You can take the best DAC ever measured at ASR- it has no chance against the ADI-2 DAC at a real-world volume setting of (for example) -20 dB.
If I may suggest, this 2nd graph is very revealing and promotes your AutoRef technology very very well, so that it should be included in the manual section 31.14 to complement the digital volume control argument.
 
Top Bottom