RME Babyface Pro FS Portable Interface Review

[Zensō]

Is this still a problem with the new FS version?

Thanks.

CC @MC_RME

I have the older non-FS version and have not noticed this as being an issue.

 

[MC_RME]

Neither have I nor anyone else. Strange. And the 'plastic' is not ours. These are professional Amphenol jacks where we just not use the 'housing' but the aluminum chassis instead.

 

[JBX1666]

Neither have I nor anyone else. Strange. And the 'plastic' is not ours. These are professional Amphenol jacks where we just not use the 'housing' but the aluminum chassis instead.

Thanks for taking the time to reply so promptly.

Good to hear as I switch cables multiple times every day, and the Babyface Pro FS is on my shortlist. I was actually waiting for a UCX update, but not sure if we will ever see that

Best regards,

JB.

PS this review is a great breakdown of the BFP FS, well done @amirm

 

[soundoffhear]

+4 dbu nominal sensitivity is sort of a pro standard. It expects 20 db of headroom above that or a max output of +24 dbu or about 12.28 volts.

One feature I would like to point out, related to the SNR measurement: The XLR output max level is +19 dBu balanced. This new FS version has a switch on the bottom that reduces the output level to +4 dBu max. This is very useful as active monitors these days are very sensitive. At +19 one would have to reduce the level in the digital domain a lot, loosing about 15 dB of SNR. Thanks to the switch the full SNR is moved down to that lower level.

I have the original Babyface Pro.

Although what I'm about to layout is pretty complex... I have an essential question, that I have yet to get a straight answer on.

Does pro audio gear operate at it's best performance/efficiency when hovering around +4dBu?

The following factors are causing me to ask and want to account for that question:
1) I have Yamaha HS8 active monitors and I'm about to calibrate them to a reference listening SPL of 73dB.
2) I also use K-20 Metering System in my audio production workflow.
3) The only remaining part in my signal chain that I need to account for is where my Babyface Pro output level will be 'set and forget'.

Given the FS revision includes a +4/+10dBu switch and the comments about this feature; is there an overall advantage/disadvantage to adjust the output of the Babyface Pro to achieve +4dBu for my reference level?

If it's confirmed as an advantage to target +4dBu as a "zero reference level":
Wouldn't the output fader on the RME Babyface Pro need to be at +5dBfs in order to achieve a +4dBu result in my K-20 metering situation?

Math would go like this:
Starting With Babyface Master Output Fader at 0dBfs
+19dBu (audio interface output spec) minus 20dBfs (K-20 Metering System reference volume reduction from input source) = -1dBu (actual output from audio interface)
-1dBu plus 5dBfs (output fader compensation) = +4dBu

 

[MC_RME]

As soon as you set this up you will notice that the HS8 are much too loud, so the BF switch in the +4 dBu setting is the only choice that makes sense. No need for complicated and confusing mathematics...

 

[L5730]

@soundoffhear
I hooked up the BFProFS to active monitors. The underside switch was in the +19 dBu position as it came.
Then had to turn the Main fader down something like 15 dB or whatever for reasonable listening levels (didn't touch the gain on the monitors/amps).
I read the manual, also read the limited literature that came with the monitors. I switched the switch to +4 dBu, I can now use the Main fader more sensibly.

I think the switch is calibrated the same as the gain on the totalMix faders, so it's pretty accurate.

 

[soundoffhear]

As soon as you set this up you will notice that the HS8 are much too loud, so the BF switch in the +4 dBu setting is the only choice that makes sense. No need for complicated and confusing mathematics...

Did you realize I'm calibrating my reference level (measured by SPL meter at listening position)? And also my main question is whether +4dBu is an ideal "target output" level for pro audio gear?

I think you missed two factors that I'm working with:
1) I'm using K-20 metering system, so the working/listening level will be kept turned up +20dBfs on the RME fader no mater what. So, even if I had the option to toggle a switch to be +4dBu for the nominal level (0 on fader), I wouldn't use it because the RME output fader doesn't reach +20dBfS from the 0 mark. And even if Total Mix FX was capable of turning up +20dBfs from the 0 mark, I would still much rather have the fader sit at +5dBfs.
2) It will be this RME output level that I will use to then calibrate my monitors to 73dB SPL average output

This would only result in max peak volumes of 93dB SPL (not too loud)

 

[L5730]

@soundoffhear
The manual for my monitors said "This input is an electronically balanced input and is compatible with both -10 dBV and +4 dBu signals."

This to me suggests that if I were to set the BFProfs switch to +19 dBu there is a risk of clipping the input of the amp, if I raised the main fader in totalMix.

I'd be inclined to set to +4 dBu and then just remeasure your monitors, then adjust the level control on them for your desired SPL.
I created my own master fader graphic for Reaper DAW and have it to display -18 dB RMS as the zero point.

HS series manual:
"LEVEL control:
Adjusts the output level of the speaker. When set to the 12-o’clock position the speaker is optimized for a nominal input level of +4 dB."

Are you wanting the LED meters on the RME interface to follow K-20 metering, or are you not so fussy about that?

 

[soundoffhear]

@soundoffhear
The manual for my monitors said "This input is an electronically balanced input and is compatible with both -10 dBV and +4 dBu signals."

This to me suggests that if I were to set the BFProfs switch to +19 dBu there is a risk of clipping the input of the amp, if I raised the main fader in totalMix.

I'd be inclined to set to +4 dBu and then just remeasure your monitors, then adjust the level control on them for your desired SPL.
I created my own master fader graphic for Reaper DAW and have it to display -18 dB RMS as the zero point.

HS series manual:
"LEVEL control:
Adjusts the output level of the speaker. When set to the 12-o’clock position the speaker is optimized for a nominal input level of +4 dB."

Are you wanting the LED meters on the RME interface to follow K-20 metering, or are you not so fussy about that?

The LED meters on the RME don't need to follow K-20 metering. I just need to know if it is optimal to adjust the output level of the RME to hit +4dBu. Where the output fader exists on the RME meter, I'm agnostic about.

One thing to note: nominal value is +19dBu for the Babyface Pro. This means when the output fader is at 0 (nominal) the physical output is +19dBu. Requiring a person to turn down the fader to -15dBfs to reach an output of +4dBu.

However, with the K-20 metering system already subtracting -20dB as an input into the RME from the DAW for headroom, it puts the RME nominal (0 mark on the fader) output at -1dBu.

All of my amplitude will be controlled in my DAW environment. I will only be touching the RME output fader if I'm listening to a commercially mastered source input. At which time, I will use the DIM function (mine is set to -20dB) to normalize the output volume.

 

[L5730]

I don't know if we are on the same page.
K-## metering doesn't mean 'subtracting' anything, to me it means that peaks can go to 0 dBFS but the Loudness (or weighted RMS) is at whatever the K- number is.
Grab a current modern hyper loud track coming in at -6 LUFS, and it'd gotta be dropped by a whole 14 dB! to measure 0 on the K-20 meter.
Wasted headroom, but that's because it's hyper compressed/limited.
Something done at K-20 could still have 0 dBFS peaks, and so maximum peak-peak output voltage.

From the BFPfs manual:
"On the underside of the Babyface Pro FS there is a 3/8" thread for mounting
the unit on tripods, as well as a recessed slide switch that sets the maximum
XLR output level to +19 dBu (default) or +4 dBu. +4 dBu is recommended
when connecting sensitive amplifiers or active monitors. It helps to reduce
noise and to less lower the fader in TotalMix FX because the output level is
then 15 dB lower"

Better to use +4 dBu rather than +19 dBu output as it means less digital attenuation in the interface, if that matters at 32bit float (?).

It makes sense to just set to +4 dBu on output, IMO, as that is the voltage that the amp in the speaker is expecting. Then adjust the volume controls on the speaker/amp to meet the desired SPL with pink noise - as you would have done already.

Suggests that the amp is expecting -10 dBu at Max. volume (or minimum attenuation). Halfway on the volume knob, showed it was expecting +4 dBu for normal use.

 

[AnalogSteph]

Yamaha's specs unfortunately are rather vague. "A nominal +4 dBu" does not necessarily mean that this is input sensitivity for maximum output (but maybe 85 dB SPL @ 1 m or so). One would wish they had a sensitivity spec similar to Genelec or Neumann (acoustic output level at given input level and monitor settings). Oh well.

With these monitors, you cannot clip the input stage before the power amps / drivers I think. They would go VERY LOUD first.

Going through the docs for the "old" Babyface Pro I don't think this had the +4 dBu output level option yet at all... this seems to have been new in the FS model. So setting the Yamahas to lowest sensitivity (+4 dBu) would pretty much be a must - and at this point I still wouldn't be sure that BF noise would be completely negligible... +19 dBu - 118 dB(A) = -101 dBu(A). With some monitors that will translate, say, 0 dBu in to 108 dB SPL @ 1 m anechoic, that would be +7 dB SPL(A) @ 1 m... this is still audible. Hence why this feature was introduced in the first place.

If you insist on level LEDs following the metering, in all likelihood a monitor controller or mixer is going to be required to take up the level difference between 0 dBu out and monitor input required for 73 dB SPL @ listening position when set to +4 dBu in. Determining said difference would more or less amount to measuring Neumann-style input sensitivity.

 

[soundoffhear]

I think we are missing the essential points:

1) With K-20 Metering, the nominal 0 level on the fader in the DAW is -20dB lower than typical 0dBfs. Meaning you're working at an average amplitude that is -20dBfs lower than digital clipping levels at 0dBfs
2) With the original Babyface Pro this would reduce the +19dBu default output down to a -1dBu output when the RME fader is at nominal 0 dBfs
3) So I see no advantage to the +4dBu attenuation switch within a proper K-20 metering workflow, as the digital signal is already being targeted -20dB lower for the average volume

 

[L5730]

^ but couldn't peaks/transients be +20 dB (0 dBFS), depending upon the program material?
Surely one would want to avoid clipping - hence the ability to turn the output voltage down on the interface with a switch.

RMS/loudness/VU would be -20 dBFS, because that's what the K-system is for. Target loudness really, or that's how I see it. K-20 or K-14 just allows for more headroom. The zero point in the scale is still 85 dB SPL. In each case, I assume one adjusts the monitoring level to reach the desired SPL at the desired digital RMS/loudness/VU level.

As far as I know, no one makes songs play at K-0 to K-4 (theoretical, these don't exists for good reason). Maybe K-6, if it was a thing :s

 

[AnalogSteph]

3) So I see no advantage to the +4dBu attenuation switch within a proper K-20 metering workflow, as the digital signal is already being targeted -20dB lower for the average volume

I don't get that logic. All the metering system does is specify the amount of headroom, i.e. the difference between 0 dBFS and average level. It does not magically reduce any levels. You still have to translate 0 dBFS to monitor input sensitivity and then acoustic output level in a sensible manner.

Let's look at the Neumann KH120A for example. Its output level can be set to either 94, 100, 108 or 114 dB SPL @ 1 m anechoic for 0 dBu input, with a pot for up to an extra 15 dB of attenuation. Meanwhile, maximum output @ 1 m is about 111 dB SPL with a following wind. At highest sensitivity, this is reached with just -3 dBu of input, so this is as much as you need - for 0 dBFS level!
For best match to a 0 dBFS = +4 dBu output, you'd pick the 108 dB setting and turn down the pot by a dB.
For best match to a 0 dBFS = +19 dBu output, you'd pick the 94 dB setting and turn down the pot by 2 dB.
And this I remind you would result in 111 dB SPL @ 0 dBFS or 91 dB SPL average in K-20, which may still be substantially louder than you want.

A number of monitors out there basically just have the equivalent of the 108 dB setting and no level pot, or their adjustment range is from like 114 dB to 102 dB SPL @ 0 dBu. Required 0 dBFS levels: Anywhere from -3 to +9 dBu, potentially lower if you don't want it as loud...

 

[soundoffhear]

I don't get that logic. All the metering system does is specify the amount of headroom, i.e. the difference between 0 dBFS and average level. It does not magically reduce any levels. You still have to translate 0 dBFS to monitor input sensitivity and then acoustic output level in a sensible manner.

As I said, my use case is for Music Production where the volume will average around -20dBfs RMS, with peaks here and there going up to -10dBfs maybe. That is 90% of the audio I will be listening to, at a SPL of 73dB. For the other 10% of audio (commercial masters) I will check for reference, I will use the DIM feature on the Babyface to attenuate the volume -20dB.

All and all, this puts my digital clipping level at a maximum of 93dB SPL coming from my Monitors at any given time.

With +19dBu being the default output of RME Babyface Pro, I'm already operating at a reduced input level by coming from my production sotware target level of -20dBfs. My monitors won't really ever pass a 93dB SPL threshold with peaks in the worst case scenario. I'm not seeing the point of the +4dBu pad in my use case.

In fact, the RME Babyface Pro is mainly purpose built as an audio production interface. And given the modern consensus international standards from EBU R 128 to target a -23 LUFS; working at a reduced -20dBfs average RMS target is the responsible way to use the interface. RME already built in a +19dBu output which roughly corresponds with this -23LUFS international target standard.

My question is if I should kick up the dBfs +5 on the RME fader to equal an average +4dBu output level. Is there is an efficiency or performance benefit to target +4dBu output? Otherwise, I'll leave the RME output fader at 0dBfs, have an average output of -1dBu and call it a day.

 

[L5730]

http://www.sengpielaudio.com/calcul...recording level (pro,a voltage of 1.228 volts.

As far as I understand it, ignore K-metering for a moment, as all that is important is Full Scale digital signals (transients/peak) and how that translates into voltage output from the interface, and subsequently whether it will clip the amplifier input or not. A 0 dBFS peak signal is going to be 0 dBFS if you use K-20, K-14 or anything else, it's digital maximum.

I'd rather provide the amplifier a maximum of 3.4 v peak-to-peak (+4 dBu setting) than potentially 19.5 v peak-to-peak (+19 dBu setting). Allows the volume pot. on the amplifier(s) to be more middle of travel, if that matters.

The digital headroom is already defined by yourself in choosing the K-system. You have chosen to give yourself 20 dB of (digital) headroom by using K-20.

My question is if I should kick up the dBfs +5 on the RME fader to equal an average +4dBu output level. Is there is an efficiency or performance benefit to target +4dBu output? Otherwise, I'll leave the RME output fader at 0dBfs, have an average output of -1dBu and call it a day.

At +4 dBu setting a digital Full Scale sine wave will generate 1.22 v RMS.
At +19 dBu setting a digital Full Scale sine wave will generate 6.9 v RMS.

Your question, if I take the liberty to rephrase, is it better to set to +19 dBu output and turn down the amplifier input trim, or +4dBu and turn it up?
If I read the manual correctly, the BFPfs will generate a little less noise if you flick the switch to +4 dBU.


Your DAW metering system defines the digital headroom.
Your amp/speaker volume control defines the listening level (SPL).
Your interface/DAC controls the maximum peak voltage into said amp/speaker - set to match expected voltage range.


But you said you had the old BFPro, without the switch that the FS has. So you will always be a +19 dBu, unless you dial down the output in the interface (digitally), or from the DAW (digitally, and possibly high bit depth precision). You probably aren't clipping the input on the amp/monitors else you could hear it.

 

[Mesh]

Thanks for this review, Amir!

This unit is USB bus powered. As such we had to make sure it will not cease operation the moment you record the take of your life. The whole unit is carefully balanced between power consumption and performance. The reason for the two headphone outputs is part of that - a high power headphone output can easily shut down USB bus power with higher levels on low impedance phones. So we added two outputs with separated driver stages. One gives a high output level, but its output impedance (a 10 Ohm resistor) limits the available current. Note that changing this to 0 Ohm would not have resulted in higher output power as measured on a 300 Ohm load! Only higher levels could do that, and these are not possible without upping the power consumption. Also the effects on frequency response and damping are no problem with phones of 80 Ohms and up.

For low impedance headphones 10 Ohms is too high, so we added a second output (3.5 mm) that is limited to a 6 dB lower output level, but near zero Ohms output impedance. From my practical experience with a lot of different headphones, you will not have a problem with either low or high impedance ones, and you could even connect two at the same time (thanks to the separated driver stages).

One feature I would like to point out, related to the SNR measurement: The XLR output max level is +19 dBu balanced. This new FS version has a switch on the bottom that reduces the output level to +4 dBu max. This is very useful as active monitors these days are very sensitive. At +19 one would have to reduce the level in the digital domain a lot, loosing about 15 dB of SNR. Thanks to the switch the full SNR is moved down to that lower level.

If you compare the ADI-2 DAC FS with the 3.5 mm of the BF Pro FS, what are the differences between them (or are they effectively equal for practical purposes)? I have tried to compare the specs from your website, but the metrics used are different for the IEM 3.5 mm part. It would be brilliant to get comparable numbers or at least an overall take on it in practice terms.

EDIT: Nevermind! I had missed the https://www.audiosciencereview.com/...ble-interface-review.12313/page-4#post-390969 post where you showed this, and I can compare with the measurements Amirm has done on the ADI-2 DACs also. All sorted now, and looks to be essentially the same for practical purposes.

 

[formdissolve]

FYI Benchmark made excellent XLR to RCA cables with Pin 3 floating as per spec (electronically balanced) that can be used with this Babyface, but are currently out of stock. Not sure if they will make them again.

https://benchmarkmedia.com/collecti...ark-xlrfm-to-rca-adapter-cable-pin-3-floating

I may have purchased the last pair of 3 feet cables... They are made from Canare L-2T2S which has a respectable capacitance rating of 33 pf per foot (between conductor to shield - PDF here).

 

[jerryfreak]

id solder those up for cost of parts+ shipping if anybody wants. its like 15 mins of work

 

[Monstieur]

FYI Benchmark made excellent XLR to RCA cables with Pin 3 floating as per spec (electronically balanced) that can be used with this Babyface, but are currently out of stock. Not sure if they will make them again.

https://benchmarkmedia.com/collecti...ark-xlrfm-to-rca-adapter-cable-pin-3-floating

I may have purchased the last pair of 3 feet cables... They are made from Canare L-2T2S which has a respectable capacitance rating of 33 pf per foot (between conductor to shield - PDF here).

I bought these, unscrewed the shield, and disconnected the pin.
https://www.amazon.com/gp/product/B07VBV78YT/