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

[2 picodumbs]

In Amir's measurements, we can read it in the picture:
Vrms

I don't see it on the dynamic range chart.
Is that what you are referring to?

 

[Rja4000]

I don't see it on the dynamic range chart.
Is that what you are referring to?

I see it in the dashboard. Amir always measures around 4Vrms for balanced output.

By the way, the load (200kohm) is also there, in the bottom... (As described here)

You may also deduct the fact it's Vrms from the DAC's description :
Its range goes up to +19dBu, which is 6.9Vrms (or 9.8Vpeak) for a sine wave for balanced output.

 

[2 picodumbs]

Thank you.

200k is a very gentle load for a dac output stage, it is not uncommon for a preamp/amp to have an input impedance of 10k.
It would be good to test at both easy and more difficult loads to better differentiate a dacs performance into a wide range of possible preamps/amps.

 

[Fred Jacquot]

Thank you.

200k is a very gentle load for a dac output stage, it is not uncommon for a preamp/amp to have an input impedance of 10k.
It would be good to test at both easy and more difficult loads to better differentiate a dacs performance into a wide range of possible preamps/amps.

One of the "star of the day headamps" has even an input impedance of 2k.

 

[Willem]

But why use a pre amp with the ADI-2? It is effectively a pre amp as well as a DAC and has all traditional functionalities such as tone and balance controls, dynamic loudness, filters etc. In short, for best results just ditch the preamp.

 

[2 picodumbs]

But why use a pre amp with the ADI-2? It is effectively a pre amp as well as a DAC and has all traditional functionalities such as tone and balance controls, dynamic loudness, filters etc. In short, for best results just ditch the preamp.

I said amp/preamp, ie even poweramps commonly have input impedances much lower than 200k.
I agree, I would use the inbuilt preamp functionality.
However if we're going to test, let's truly see what these dacs are capable of. If someone already has a $2000 to $10,000 preamp it's likely they are going to want to make use of it.
Or what if someone has an integrated amp that doesn't have the capability to bypass the preamp stage or simple analogue volume pot. 10k is not an uncommon load at all for quite a few reasons,

 

[Matias]

I heard that the standard is 47k ohms on the inputs of power amplifiers.

 

[2 picodumbs]

I heard that the standard is 47k ohms on the inputs of power amplifiers.

Yeah 47k is generally the most commonly used input impedance on a power amp.
We should still test at lower impedances though as these are not uncommon.

 

[Matias]

And using the 1000 x 1 impedance rule of thumb, an adequate output impedance of a dac or preamp should be 47 ohms.

Now since this is ASR I would like to have more science behind those, if anyone has it I would appreciate to learn it.

 

[2 picodumbs]

And using the 1000 x 1 impedance rule of thumb, an adequate output impedance of a dac or preamp should be 47 ohms.

Now since this is ASR I would like to have more science behind those, if anyone has it I would appreciate to learn it.

Low output impedance of a dac (or separate preamp) is desirable for a couple of reasons.
First, to prevent high frequency roll off due to the formation of an RC filter formed by the output impedance of the dac with parasitic capacitance of coax cables as well as the the input capacitance of the device you're feeding. You either want to reduce the output impedance of the dac or minimise input capacitance of the device your feeding.
.
In addition to that having low output impedance will result in less voltage loss, that is to say the current to drive for example a 10k load will result in a lower voltage drop at the output of the dac.
Eg 2V rms driving a 10k load load draws 200 uA from the dac. If the DAC has an output impedance of 47 Ohms you will have a loss (voltage drop) of 0.0002 x 47 = 0.0094 V rms or 9.4 mV rms.
If the output impedance of the DAC was 1k then the voltage loss driving the same load would be 0.0002 x 1000 = 0.2 V rms or 200 mV rms. Keeping the math simple as possible just to illustrate (the current in the second example would actually be a little less).
Alternatively you could think of it as forming a voltage divider (whatever is easiest for you to visualise). If the output impedance of the dac was 10k and the input impedance of the amp was 10k, then only half your signal would end up at your amp.

 

[Rja4000]

It would be good to test at both easy and more difficult loads to better differentiate a dacs performance into a wide range of possible preamps/amps

Well, if you want proofs that the RME performs similarly under low impedance load, just look at the headphones measurements.
After all, in worst case scenario, one may use the headphone out to drive an amp.

On the Pro, you may even combine both stereo headphones out into one stereo balanced out.
And, for this one too, performance is top level.

Again, for the Pro R, you may look at my loopback measurements above: RME ADI-2 Pro fs R's input impedance is 18kohm in balanced mode (but both inputs are connected in parallel here).
As you can see, there is nothing to complain about under such load. And you should definitely get better figures with an AP555.

 

[2 picodumbs]

Well, if you want proofs that the RME performs similarly under low impedance load, just look at the headphones measurements.
After all, in worst case scenario, one may use the headphone out to drive an amp.

My understanding is that the RME dac drives headphones by a headphone amp that is separate to the dac output.
Using the RCA output things are likely very different.
These things should be tested to confirm.

I do not expect to see problems with this DAC the problem is we are left to make assumptions, or just hope, that it or any other dac is ok.
I would like to see all DACS tested properly if we're going to test them let's do it thoroughly.

This DAC is of great interest to me, knowing how it performs under all possible scenarios in comparison to other dacs is highly valuable to me.

 

[AidenYoung]

I don't have measurements with AP, but I have some in loopback mode.
Not ideal, but gives you an idea, still.

When you reach over 116dB SINAD for the loopback (DAC and ADC), neither DAC nor ADC can be bad
When the range where Loopback SINAD reaches 113dB or more is covering a wide share (>85%) of the range from -2dBu to 24dBu, you know that's not just 'good'.
View attachment 69730

And I also performed some measurements with the Pro fs R.

Note: This is what I call "measurement mode", which means both outputs and inputs are averaged as a mono signal.
Still, this is better than the ADI-2 Pro fs by several dBs in the same configuration
(Measured with Virtins Multi instrument 3.9 @48kHz)

EDIT: 4Vrms (+14.3dBu) is kind of worst case scenario here.

If according to VintageFlanker suggested to set +13dBu and -4.5db(DAC volume) which is +8.5dBu. By looking at the chart, is it where the green line reached to the 114dB range?

 

[Tks]

One question - Is RME ADI 2 FS EQ capable performing room acoustic correction automatically?

No, it's simply a 5-band Parametric Equalizer that you adjust yourself. There's no interfacing the device with some sort of external 3rd party programs or things of that nature.

 

[Artaois]

Should be the same as these here, as their output stages are the same according to the video above.
As for the AD converter that would be interesting to see.

If I remember correctly I think they stated running the Pro in balanced does make it measure better but it's probably not an audible difference.

 

[james_0075]

No, it's simply a 5-band Parametric Equalizer that you adjust yourself. There's no interfacing the device with some sort of external 3rd party programs or things of that nature.

Thanks. Looks like it is quite difficult for me to use it. I thought probably it has a microphone for automatically adjustment. I am wrong.

 

[SimpleTheater]

Thanks. Looks like it is quite difficult for me to use it. I thought probably it has a microphone for automatically adjustment. I am wrong.

Actually I find this type of eq easier to use. For one reason, eq'ing headphones can be done right on your head.
For room eq, simply play pink noise (or whatever you choose), with a mic and a laptop showing you the real-time response. While that's playing, adjust the eq until you get the levels you want at the mic position.

 

[james_0075]

Actually I find this type of eq easier to use. For one reason, eq'ing headphones can be done right on your head.
For room eq, simply play pink noise (or whatever you choose), with a mic and a laptop showing you the real-time response. While that's playing, adjust the eq until you get the levels you want at the mic position.

Got it. Thank you.

 

[MC_RME]

I did not want to spoil John Siau's thread so I put it here. As far as I know Amir has never measured (or published) crosstalk measurements of the ADI-2 DAC. The manual says better than 120 dB, and we mean it. This is the current ADI-2 DAC V2 measured in a way that is comparable to Amir's published one, and a second view with the lower frequency part also visible. First measurement is at full output level of +19 dBu (balanced), second with volume set to -4.5 dB to get 4 Volts. The AP then registers more noise, so that is not real crosstalk that's added, just noise.

 

[Chuu]

I am wondering, are there any reasons to consider the ADI-2 Pro FS over this unit? I got offered a very good deal on a demo unit, but as far as I can tell the differences are the AD capabilities which do not concern me, and potentially double the output in balanced mode (do any headphones really need more than 1.5W of output?).