• 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 UFX III Audio Interface review

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,906
Likes
5,263
Location
Liège, Belgium
This is a review and detailed measurements of the
RME UFX III
professionnal Interface
(ADC, DAC, Headphones Amp).

_R9A7858_dxo_FHD-2.jpg


The UFX III is quite recent, as it was launched around march 2023.

Normal price is around 2700€.
I purchased mine new.


Presentation

The UFX III is the current top of the RME interfaces line.

Here is a quick comparison of the different models in the range.
(I collated this list myself, so please forgive me if there is any mistake)

1718727588782.png

(Number within ( ) means the physical input or output is shared with other functionalities)

As one may see, even if the number of analog inputs and outputs is limited, this is a quite powerful interface, allowing up to 94 inputs and 94 ouptuts.

_R9A7859_DxO_noSerial_dxo_FHD.jpg


Rear panel is quite busy.
The only visible difference with the UFX II, the next model down in the range, is the MADI inputs and outputs and the USB III connection.

This is quite important to me, since I wanted to cascade it with the excellent RME 12Mic-D microphone preamp I already own.
With a MADI connection both directions, I can get a total Dante and Totalmix integration.
This allows me to record in parallel all 16 microphone inputs (12+4) on a network-connected PC through Dante and also locally with Durec on a USB drive as a backup.


Measurements

Measurements were performed mainly with my usual RME ADI-2/4 Pro SE.
On occasion, the UFX III itself was used to add gain and mesure low noise values.
I used a mix of AudioPrecision APx 500 Flex 8.1.2 and Virtins MultiInstrument 3.9.9.6 softwares.
Thanks to AudioPrecision for letting me try this software for a few weeks.



Main goal in using an audio interface is to record music.
So, unlike what's the usual order here on ASR, let's start with the most important part for that purpose.


RME UFX III - Inputs (ADC) measurements

Input sensitivity range
One of the first things to know when using an interface or a microphone preamp is its gain range.
For a digital audio interface, we'd rather look at the range of sensitivities, since there is no such thing as a normalized "Gain".

1718549370251.png

EDIT: If you prefer to have X values incerasing from left to right, see in attachment. I kept it here in decreasing value order, since that's the order of gain increase.

The X scale is in dBu.
For reference, I added the RME ADI-2/4 Pro SE - which is a line-level only interface - and the Yamaha AD-8HR - a wide spread model in Pro audio.

We see that the microphone input range is very extended, from 18dBu (@0dBFS at 0dB gain) to -57dBu (@0dBFS 75dB gain).
-57dBu is around 1.1 mVrms. So you may reach the input's maximum level at that gain setting with just 1.1 mV rms.
That's a lot of gain, really.

RME is following EBU (European Broadcasting Union) standard, which recommends 18dBu @0dBFS as the reference for microphone gain values.
(Note that every interface vendor may use his own rule for that. So it's important to refer to voltage sensitivity for comparison)

Instrument is different: It's unbalanced and 21dBu is reached for a "8 db" gain marking, so reference seems to be 29dBu.
Rear Line inputs use 2 discrete range values: 13dBu and 19dBu.


SINAD (Inputs)

SINAD is a good indicator of the device's engineering quality.
You can't reach "excellent" SINAD without proper attention to details.

Of course, here, SINAD at 4 Vrms is less relevant.
(Amir uses 4V because he standardizes for HiFi use. A different focus.)

But let's have a look, still.
Here, with Microphone input at gain 0dB (4v = -3.7dBFS )

UFX III Mic 9 - 0dB Gain - SINAD - Analog scale 4V_crop.png


OK, that's ranking in "Excellent" category in Amir's ADC chart.


But wait. We have a line input too.
At 4V, the line input's SINAD value is almost identical to the Mic input's at 0dB.

With 6V on line input, we get

UFX III Line input 6V_crop.png



Let's compare the different inputs
THD+N (inputs)
(THD+N is exactly -SINAD. So you just change the sign of the value and you have the SINAD value)

2024-05-25 12_26_48-THD+N 48kHz vs Level (Lower better).png


Vertical offset between curves gives us the different noise levels.
We also see that the distorion increases as level approaches full scale (0dBFS) for mic and inst inputs.
However, distortion here will remain inaudible, at -100dB.
(I've seen interfaces, from other reputable brands, distort completely above -10dBFS or so.)
Anyway, good practice recommends recording with an average level around -20 or -18dBFS.


Another interesting plot is THD+N vs Level for Mic preamp at different gain settings

UFX III Mic 9 - THD+N 44.1kHz vs Level (Lower better)_Crop.png


Here we see mainly the difference in noise, as expected.
We see also that this distortion behaviour close to max level we've seen above is only there for levels >15dBu or so (>4V).
Not a typical Microphone level.

We see also the diminushing return of adding gain: the noise benefit between 35dB gain and 75dB gain is not that big.
So you may as well keep a lower gain (and more margin to avoid overloads with peaks) and just push the level later in digital without huge noise penalty.


Dynamic range (Inputs)

First, Mircophone input at 0dB gain
(AES-17 - 20Hz-20kHz - 997hz signal at -60dBFS - CCIR-2k weighting)

Higher is better

UFX III ADC DR Mic 0dB 2k-W_Cropped.png


Note: This is CCIR-2k weighted, to compare with Amir's measurements.

Here are the different values

1718553904998.png


1718554362954.png


1718554405002.png


So we get almost 120dB dynamic range (A weigthed) on Mic input.


EIN - Noise (Mic Inputs)

One of the key performance factors for microphone preamp is the noise.

If we've seen above that the dynamic range at high input level and 0dB gain is high, what about at low level and higher gain?

EIN stands for "Equivalent Input Noise".
EIN measurement is done with a 150 ohm resistor between pin 2 and 3.

I compensate for room temperature fluctuation vs 20°C.
BW is 20Hz-20kHz

As the device heats, the resistor also heats, so values here may fluctuate 0.1dB or even 0.2dB.

We'll measure it for several gain values:
  • Maximum gain (75dB gain here)
  • Gain for sensitivity = 10mV rms @0dBFS (56dB gain here)
  • Gain for sensitivity = 100mV rms @0dBFS (36dB gain here)
  • Minimum gain (0dB gain here(
Lower is better
Black plot is un-weighted noise.
Superposed Red plot is noise with A weighting.


UFX III EIN Mic 75dB Gain_crop.png


130.2dB A-weigthed
128.2dB unweighted
This is state of the art level.

Note that EIN is always best at Max gain.

Noise profile is also quite flat, with low 50Hz mains peak.
(Some preamps have more noise at low or high frequencies.)


UFX III EIN Mic 56dB Gain_crop.png


Above, we see that with a gain giving full scale for 10mV, EIN is almost identical to the maximum gain.
Another excellent achievement.


UFX III EIN Mic 36dB Gain_crop.png


At a gain giving full scale for 100mV, EIN is degraded a bit, which is normal.
It' remains very good, which is the sign of an excellent mic preamp implementation.

As we've said above, in the THD+N vs level section, the noise difference is just 2dB between 75dB gain and 36dB gain.


UFX III EIN Mic 0dB Gain_crop.png


EIN at 0dB is not really relevant.
But we see Noise Level is confirming our previous Dynamic range measurements, at 119.99dB (A).
(You have to reverse sign to compare to DR)


Here is how EIN evolves with gain
When the curve becomes horizontal, there's nothing to be gained in terms of noise by increasing the gain.
From 63dB of gain, the curve is a straight line.

1718555069757.png


And here is a comparison with other mic preamps
Lower is better
(A-weighting values - Reference gain = gain for 100mV @0dBFS = 36dB gain on the UFX III)

1718555349156.png


The graph speaks for itself.
(For the curious, we see here how the Yamaha AD8HR works: it has a very limited effective gain range, 24dB or so, then a pad is engaged - the vertical peak - then 24dB again, then a second pad, and again the 24dB.
The results are not great, but it does provide a very wide range at low cost. If you listen carefully, you can hear the relay engage at both thresholds).



Frequency response (Inputs)

Another key peformance factor is a flat frequency response.

For line level, this is usually perfect.

UFX III Line FR.png



For Instrument input, we see a deviation though.
We see -0.47dB at 20kHz
(Most probably not an issue for instrument pickups, which are typically bandwidth limited anyway.)

2024-06-15 18_35_15-Greenshot_crop.png



For mic inputs, this may also vary with Mic preamp gain

2024-06-15 12_47_49-Greenshot_crop.png


At max gain, we see the highest deviation
(Remember, at 75dB gain, we have a maximum acceptable signal of 1mV.
This is very very low.
And boosting 0.5dB at 20kHz at mixing stage if required won't be noticed, anyway.)

2024-06-15 12_42_17-Greenshot_crop.png


But at a gain that gives 0dBFS with 10mV (56dB gain), this is mostly gone.
Deviation is not significant at that gain and below.

2024-06-15 12_38_54-Greenshot_crop.png


Overall, this performance is excellent.


Linearity (Inputs)


2024-05-20 07_44_53-Linearity (ideal = 0 dB)_crop.png



Intermodulation distortion (Inputs)

2024-06-12 17_58_37-Intermodulation Distortion + Noise (lower better)_crop.png



Multitone (Inputs)

2024-05-20 07_57_27-32 Tone Input Signal _ 192 kHz Sampling Rate_Crop.png


This is Mic input.
Line input is a copy/paste.


THD vs Frequency (Inputs)

1718558791005.png


We mostly see high frequency noise at play here.


(Continued in next post)
 

Attachments

  • 1719503265221.png
    1719503265221.png
    7.6 KB · Views: 40
Last edited:
RME UFX III - Outputs (DAC) measurements

Output ranges



1719503494234.png


The main output (XLR) may be set to 4 discrete levels:
for +24dBu, +19dBu, +13dBu or +4dBu at 0dBFS

The other line outputs don't include the +24dBu range
Headphones are proposing "High" (19dBu) and "Low" (4.2dBu) ranges.


SINAD (Outputs)

2024-04-11 16_19_17-Greenshot SINAD 4V_crop.png


Again, 4V SINAD is not the most relevant level for a studio interface, since that level is not a reference in a studio.
Still, we have it, and it reaches the "Excellent" category, as per Amir's rating.

All 8 line outputs reach a similar score.

If we push things a bit, we see better result still.

UFX III Main L - 19dBu - SINAD 6.9V_crop.png



Here is the SINAD evolution vs Level, for the different ranges

1718561925097.png


We see a few interesting results:
  • There is no increased distortion at max level. So the best result is at full output scale.
  • Of course, we see the noise difference with each range: 13dBu has a better SINAD than 19dBu at the same output level, but will have a lower max output level.
  • 4dBu and 13dBu plots are superposed.
    That means that the 4dBu output range is just the 13dbU range with a digital attenuation, added for convenience. There is no noise benefit.


Dynamic range (Outputs)

1718562009126.png


Here is the Dynamic range for the 19dBu output
This is CCIR-2k weighted, to enable comparison to Amir's results.
Here are the different values

1718562034231.png


1718562052658.png


1718562068137.png


The +4dBu is behind, which is expected since we know it's just the 13dBu output digitally attenuated by approx. 9dB.
The +13dBu is a few tenths of a dB below the spec. Which is no big deal.
All results (except the 4dBu) remain in the "excellent" category.


Multitone (Outputs)

2024-04-11 16_10_48-32 Tone Input Signal _ 192 kHz Sampling Rate_crop.png

(This measurement seems a bit overloaded. I'll try to re-run it when time allows.)


Linearity (Outputs)

1718562465800.png


What else ?


Sine 50Hz under 600 Ohm load measurement
(Torture test)

1718562625058.png


Torture ?
You must be kidding.


Output impedance

1718562777013.png


Specs says 150, 75 and 2 ohm.


IMD+N (Outputs)

1718563246777.png


Of course, +4dBu has higher noise. We now know why.
Other than that, it's pretty good, isn't it ?


Jitter (Outputs)

1718563023058.png


What exactly is Jitter, remind me ?


Filter (Outputs)

@44.1kHz

1718563347601.png


1718563394223.png


(I have the plots at 48kHz and they are very similar... but shifted in frequency, of course)


Filters Impulse responses (Outputs)

2024-05-25 12_10_46-Multi-Instrument Pro 3.9.9.6   -   [+3DP+DLG+LCR+UDP+VBM+DHS]   -   _ASIO ...png



THD+N ratio vs Frequency (90kHz BW) (Outputs)

1718563772578.png


This measurement is worth a comment:
As may be seen by looking at the harmonics, the 2 peaks in THD at approx. 100Hz and 640Hz are not related to audible distortion.
They seem to be artifacts of the sampling frequency.
So, to make it short, they are coming from my measurement methodology.


So, with all this, what to say ?
This machine is pretty good on all fronts, so far.
It combines an excellent, almost state-of-the-art, mic preamp stage with a pretty decent DAC/Output stage.

(Continued in next post)
 
Last edited:
RME UFX III - Headphones measurements

1718564169787.png


You just can NOT saturate the low power output.
In high power, distortion raises a bit early, if we compare to the best specialized headphones amps, especially for low impedance.



1718564204407.png


Power is there.
Not the power of a dedicated Headphones amp, nor the power of an ADI-2 range device, but plenty enough for normal studio use.
Few people are using very insensitive planar headphones in a studio, to my knowledge.


1718564228633.png


Maximum power is reached around 60 Ohm.

I measured 85dB SNR ratio in low power mode at 50mV output under 300 Ohm.
(This is better than a Chord Dave, as an example.)


But I kept the best for the end:

All last generation RME Interfaces (except the Babyface) now allow a "Room EQ" and some "Crossfeed" on the outputs.

1718567206402.png


Look at this beauty:
We have 9 bands fully parametric EQ to correct for room or headphones non linearity.
And you may, of course, save different presets.

This, together with the 2 headphones outputs, is pushing the bar to the next level, IMO.

I've been using it with my UCX II since it was released, and that's really a serious improvement.

(Continued in next post)
 
Last edited:
RME UFX III - Other measurements - Conclusion

Round Trip Latency measurement


RTL measured in ms, as a function of sampling frequency and buffer size

1718567837842.png


Totalmix


1718569000103.png


Every RME interface comes with TotalMix
This is a virtual mixer running in hardware within the interface hardware itself.
It has, of course, a control software that may run on a PC or on an iPad.

I won't go into the details.
This is far beyond the scope of this review.

But I just wanted to measure 2 things about it:

Internal headroom
I've done a simple test:
How many tracks with a 1kHz sine at 0dBFS may I sum before saturation ?
The answer is
16.
That's 20 * LOG10(16) = 24 dB headroom
(I found after this that this is exactly the value specified in the manual.)

As a comparison, my venerable Yamaha DM1000 40 channels mixer allows 8 channels, or 18dB headroom.

There is little chance the internal headroom will limit you.


EQ quality

I've performed 3 tests:

Conformance test
The goal is to check which standard the UFX III follows for parametric EQ.
This may be important if you want to apply some EQ settings from other people, like Amir, or from tools like REW or AutoEQ, amongst others.

Our fellow member @JohnPM, REW's author, details some possible EQ formula :
(More details may be found here)


1719500235790.png


A simple test shows RME is following RBJ formula - which is the most common nowadays.

("RBJ" stands for "Robert Bristow-Johnson")


So, for this EQ

EQ 2.png


In theory, we should get this

image.png

(As may be seen, the response close to 20kHz depends of the sampling rate.)

and in practice, I measured this

1719553286687.png


Except for the change in Y scales, the plots are identical.

EQ impact​

Measured on main output, 19dBu range

Here is Multitone 32 with EQ off

1719501213290.png


and here with EQ ON

1719501277098.png


We see a few spikes in high frequencies, but we also raised the level by 12dB there, and their level is more than 115dB below signal.
So nothing to worry about.


As a second check, I used Amir's EQ test

EQ 1.png


It's simpler and just cutting, not adding, so easier to check with SINAD

Without EQ

1719501424631.png


With EQ

1719501490630.png


I think that's enough... ;)


Conclusion

I think it's pretty clear at that stage that this device is really excellent.

It may not be the best for all aspects.
But what I think is the most impressive is that it's pretty close to the best in ALL area.
I couldn't even find one weak point left.
(And, believe me, I tried hard)
And for such a complex device, that's is quite an achievement !

What is even more impressive, maybe, is that it seems the very same technology, in term of basic hardware (Mic inputs, Totalmix), has been included in the lower end RME interfaces, the UFX II and 802fs.
I didn't check them, but if that's true, that makes them very interesting devices.

With RME, the good thing is that you have some kind of fool-proof check when you do your own measurements.
If your results don't match RME specs, you'd better check them twice. Because, on their side, they check everything.
My measurements are mostly better than their specs. As usual.
I found one or two exceptions, but we speak 0.5dB or so, so they are mostly irrelevant.


So, in short, for me,
This is just excellent !



Closing words


This review was quite a journey.
I started 4 months ago !
One of the most difficult part was to try to summarize relevant infiormation while not flooding the reader with just too many details.
Let me know what you think.

Thanks to everyone who made this review possible !
 
Last edited:
Phenomenal bit of kit and a very thorough and interesting review! Thanks!
 
Extensive and exhaustive testing work @Rja4000. Well done and congratulations on Front Page promotion. Sweet. ;)
 
THANK YOU SO MUCH FOR THIS AWESOME TROVE OF INFORMATION!

AND TO AMIRIM FOR MAKING IT ALL POSSIBLE!!

WE LIVE IN THE BEST TIMES EVER FOR AUDIO!!!

I recently spent $10k on an alpine 17" screen system for my truck and am looking forward to having all 1.6G of my music downloaded on a disc soon.

Turns out when you let a subwoofer do all the heavy lifting < 500Hz and just let the mains play above that, with zero distortion cause they're not being driven past their comfort zone.

Between that and my home 120dB Wall Of Sound I'm gonna go deaf!

HEY: Currently stuck in a groove on the Scorpions, EVERY TUNE THEY MADE ROCKS IN IT'S OWN WAY, and well recorded.

Highly recommended.
 

Attachments

  • IMG20221121193457.jpg
    IMG20221121193457.jpg
    126.7 KB · Views: 177
Incredible review, excellent interface.
 
Fantastic review... the worst things you can say about this device are the instrument input is 1dB down at 30khz and the headphone amp isn't TOTL. Otherwise they seem to have knocked it out of the park. If I was putting together a serious studio, I think this would be at the top of the list. As far as fidelity goes it looks like you can make world-class recordings with this thing.
 
Measurements were performed mainly with my usual RME ADI-2/4 Pro SE

Excellent work!

Can you show a system loopback of the RME ADI-2/4 Pro SE on the AP's bench mode screen at the measurement levels you picked?

As it stands, we have no idea how much is contributed from the A/D-D/A (instrument) and how much is contributed by the A/D-D/A (DUT).
 
Superb ! Very in-depth and dense review. i'm gonna read it again and dive in it more carefully.

Of course, as most of us could expect, performances from a top line RME machine are beyond reproach. This brand is definitely one of the top pro audio electronics makers.

A question (sorry if I missed a point) : what means exactly EIN and why is this criteria so important ? I'm not used to read this acronym often on ASR or elsewhere.
 
@Rja4000: Thank you for your fantastic review!

I am very glad to see this top performance, as I sold my miniDSP SHD to replace it with a UFX II which - in its latest hardware rev.7 - has identical AD/DA boards as UFX III. Only difference being the missing MADI interface.

So far I am very happy with my UFX II. The flexibility of TotalMix is fantastic (a little bit overwhelming at first), and communication with RME in case support is required is very straightforward.

BR, Mark
 
Back
Top Bottom