• Registered: 2019-03-22
  • Posts: 24

Topic: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

I have performed extensive testing of my hybrid setup. And I have determined a -0.8dB inherent signal loss on the BF Pro FS with a -20db sinewave signal being run through the line level input and line level output at unity gain.

There is also a -0.6dB signal loss when using the DIGITAL input and line output with the same test.

Is this common for gear to have inherent signal loss?

I tested a physical loopback connection on my AIO Pro and determined there is a -0.3dB inherent loss there. So when I put the BF Pro in the chain to test the specific gear, I subtracted that -0.3dB from the overall loss.

Just to conclude, the combination of inherent loss between the AIO Pro and the BF Pro FS together is between -0.8 and -1.1 depending on whether you're using Digital inputs or line level inputs on the BF Pro FS.

I don't really understand what the tolerances are for professional gear in terms of inherent signal loss. But is -1dB considered acceptable in a signal chain between two pieces of gear?

Thanks

  • mkok
  • DAW Mastermind
  • Offline
  • Registered: 2020-11-04
  • Posts: 632

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

Wow. I just use mine.

Babyface Pro Fs, Behringer ADA8200, win 10/11 PCs, Cubase/Wavelab, Adam A7X monitors.
  • waedi
  • DAW Mastermind
  • Offline
  • From: Swiss Zürich
  • Registered: 2020-12-08
  • Posts: 4,539

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

...at unity gain. Why ?
Pull up the fader a little bit and you can have signal-win instead of signal-loss.

M1-Sequoia, Madiface Pro, Digiface USB, Babyface silver and blue

4 Reply by unpluggged (edited by unpluggged 2024-11-15 22:05:13)

  • Registered: 2023-06-30
  • Posts: 161

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

Waltong wrote:

I have performed extensive testing of my hybrid setup. And I have determined a -0.8dB inherent signal loss on the BF Pro FS with a -20db sinewave signal being run through the line level input and line level output at unity gain.

How did you determine that? And what is "-20 dB sinewave"?

Waltong wrote:

There is also a -0.6dB signal loss when using the DIGITAL input and line output with the same test.

And what does this mean?

Waltong wrote:

I don't really understand what the tolerances are for professional gear in terms of inherent signal loss. But is -1dB considered acceptable in a signal chain between two pieces of gear?

How did you perform the test specifically?

Fireface UCX II + ARC USB > ADI-2 Pro FS R BE > Neumann KH 750 DSP + MA 1 > KH 120 A
  • From: Germany
  • Registered: 2007-04-10
  • Posts: 9,774

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

This is not a "loss", but a slight discrepancy between analog levels relative to 0dBFS, which is by design. I believe it's been explained here somewhere earlier.
I'm not quite sure how you are using digital inputs and line outputs together here. There's certainly no change in level across a digital loopback.

Regards
Daniel Fuchs
RME

RME Support's Website

6 Reply by Waltong (edited by Waltong 2024-11-15 23:06:42)

  • Registered: 2019-03-22
  • Posts: 24

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

RME Support wrote:

This is not a "loss", but a slight discrepancy between analog levels relative to 0dBFS, which is by design. I believe it's been explained here somewhere earlier.
I'm not quite sure how you are using digital inputs and line outputs together here. There's certainly no change in level across a digital loopback.

It is quite a simple test, connect the physical output to the physical input of the interface and measure the difference of the signal output vs the signal input.

You are saying the discrepancy is inherent due to dBfs conversion on the D/A and A/D?

Using the BF Pro FS in standalone mode allows a digital input signal coming from the AIO Pro to be internally routed within the BF to the analog outs and back into the AIO Pro using the same test method.

If the AIO Pro has a -0.3db discrepancy and the BF Pro FS has at least 2x that loss, the gain needs to be made up on the signal path to reach unity gain and have proper gain staging across components of gear.

You are saying these are expected tolerances with RME gear? They become cumulative signal losses as gear gets interconnected and daisy chained.

7 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,599

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

What Daniel means is that analog has tolerances. Therefore all our interfaces are intentionally designed to have their output level a tiny bit (like -0.2 dB at 0 dBFS) lower than what the reference level says, and all inputs are a tiny bit less sensitive (+0.2 dB needed for 0 dBFS). As with analog these deviations can vary a bit.

Reason for this is simple: if you use an analog loopback at identical ref levels the AD side should not be overloaded. The method decribed above is an easy way to prevent this despite the analog's discrete world deviations. Overs can still happen with ISPs (inter sample peaks), though.

Regards
Matthias Carstens
RME

MC's Website

  • Registered: 2019-03-22
  • Posts: 24

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

MC wrote:

What Daniel means is that analog has tolerances. Therefore all our interfaces are intentionally designed to have their output level a tiny bit (like -0.2 dB at 0 dBFS) lower than what the reference level says, and all inputs are a tiny bit less sensitive (+0.2 dB needed for 0 dBFS). As with analog these deviations can vary a bit.

Reason for this is simple: if you use an analog loopback at identical ref levels the AD side should not be overloaded. The method decribed above is an easy way to prevent this despite the analog's discrete world deviations. Overs can still happen with ISPs (inter sample peaks), though.

Ok, this should be down to the user not running the signal too hot in/out of the digital domain. And the rest of us looking to operate equipment properly don't need our signal modified on behalf of these folks that need intersample peaks tamed because they are trying to run their true peak signal hot near 0dbfs. With modern converters and 24bit+ noise floor/resolution, it is totally unnecessary even in mastering to come close to 0dbfs. We aren't in the 90's where 16bit recording benefited from working with less headroom and getting near enough 0dbfs where ISP would have an impact on overages. Therefore, I still don't understand the design choice.

RME Support wrote:

There's certainly no change in level across a digital loopback.

I ran that digital loopback test and there is indeed a -0.5dB difference. This is using spdif in/out of the AIO Pro to the spdif in/out of the BF Pro fs in standalone mode. To your point, there is no A/D or D/A at play here, and I'm still getting signal loss.

  • mkok
  • DAW Mastermind
  • Offline
  • Registered: 2020-11-04
  • Posts: 632

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

What is your use case? Does this matter for what you are doing? If not I don’t see the point.

Babyface Pro Fs, Behringer ADA8200, win 10/11 PCs, Cubase/Wavelab, Adam A7X monitors.
  • From: Germany
  • Registered: 2007-04-10
  • Posts: 9,774

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

Waltong wrote:

Ok, this should be down to the user not running the signal too hot in/out of the digital domain. And the rest of us looking to operate equipment properly don't need our signal modified on behalf of these folks that need intersample peaks tamed because they are trying to run their true peak signal hot near 0dbfs. With modern converters and 24bit+ noise floor/resolution, it is totally unnecessary even in mastering to come close to 0dbfs. We aren't in the 90's where 16bit recording benefited from working with less headroom and getting near enough 0dbfs where ISP would have an impact on overages. Therefore, I still don't understand the design choice..

There is no "modification" and there is no "loss". How important is it in your individual workflow to pass a signal from the analog output of an interface to the analog inputs of the same device straight, without passing through some external signal processor or the like, just running it through DA and AD? Why are you afraid of a signal level change of less than one dB in a use case you will hardly ever need? You can easily compensate for it one way or another without losing any signal quality. You are making this sound like a conspiracy where evil forces are trying to adversely affect your audio signals...

RME Support wrote:

There's certainly no change in level across a digital loopback.

I ran that digital loopback test and there is indeed a -0.5dB difference. This is using spdif in/out of the AIO Pro to the spdif in/out of the BF Pro fs in standalone mode. To your point, there is no A/D or D/A at play here, and I'm still getting signal loss.

How exactly does this setup work? How is the signal going through the BF here? Feel free to provide screenshots and images... Also, why didn't you simply try a physical SPDIF loop on the Babyface?
You said "There is also a -0.6dB signal loss when using the DIGITAL input and line output with the same test", so there is conversion involved...

Regards
Daniel Fuchs
RME

RME Support's Website

  • Registered: 2019-03-22
  • Posts: 24

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

RME Support wrote:

You are making this sound like a conspiracy where evil forces are trying to adversely affect your audio signals...[

Wow, a lot of hyperbolic language coming in here "evil forces" etc.

It's really straightforward. Whether it's .5db or 5db difference, accuracy matters when tracking down signal integrity across an audio chain. This is so you can isolate equipment influence in a troubleshooting scenario. This is basic audio engineering 101.

So I'll match your hyperbolic language and say you might be in over your head on this one. I just asked for an accounting of signal integrity across RME gear. Kind of sad how you respond so defensively/unprofessionally. You act like the sky is falling because someone wants to be certain.

12 Reply by vinark (edited by vinark 2024-11-19 18:26:46)

  • vinark
  • DAW Mastermind
  • Offline
  • Registered: 2009-06-05
  • Posts: 2,779

Re: Signal Loss on RME gear? BF Pro has inherent loss up to -0.8dB

In any case you have your reply, from the horses mouth (pun intended, no offence ment @MC and Daniel). The discrepancy is intentional and not a dysfunction.
I understand why you were worried, but is it possible to work with now you know?
This has always been a problem between devices and brands, but with pure analog gear where 0dbfs and digital clipping were not an issue (just a little more THD), most devices offered a way to fine tune to your (house) standerd internally. This is obviously not possible with the babyface (which I use and love myself) and would only have caused many problems. Which RME tried to prevent with their pretty sensible compromise on a portable device that does not have the pro market as a first target.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
BFpro fs, 2X HDSP9652 ADI-8AE, 2X HDSP9632