RP!  Welcome back!   Nice to see you here again.   

My thoughts are to set your Amp input sensitivity per the the Mfr's guidelines, and let Auto-Ref work it's Magic!    Max Out your SNR!    Off the Cuff, I see no need for a fixed level. We would normally reserve that for an instance where one uses a Preamp, or Integrated Amp.   In your case.  No.

Push the Button, and Enjoy.

All the Best for Happy Listening!

Curt

PS:  How's the Shoulder?

(for those wondering, RPGroman was our only known Casualty from the Speaker/Furniture moving exercise we did some months ago)  There may have been others...but?  Dead Men tell no tales.

RP,

Auto-Ref is it's own deal.  IT decides the correct output level, and it works a treat!

Fly Fishing?  When might we anticipate the Trout (Forelle) Feast??    I'll bring the Cocktail Sauce!!

Curt

"Catch and Release" hey?  Does your Wife know about this?  Sounds to me like the Group intends to do a bit of Nightclubbing.

Were that the case, get that Arm in a Sling, and play the Sympathy Card!!   Highly effective.  Ask about their Home Demo Policy! 

I digress.  Perhaps you can tell us more about your new Amp.  Model, etc.   From there, we can make more informed suggestions, and have your entire system operating at it's maximum potential.

Best to You!

Curt

That Benchmark Amp just recently came up in discussion on the Forum.  Do a bit of Poking around.  You may find some helpful insights.

Thanks for that suggestion.  I'm usually sitting too far away (and using the remote) to notice that detail on the screen. But, now that you've pointed it out, I do see it!  Much appreciated, Matias. 

And, thanks to you, too Curt.

I've long suspected that if RME did an amp it would likely perform very similarly to the ABH2!

An update after living with the AHB2 for a little while.  First of all, it is all that's advertised/written/measured.  Really any outstanding piece of kit....and so tiny!  Plus, it just looks so cool sitting next to the ADI-2 and makes my Revels sing!

Re. levels, the mid-gain setting on the AHB2 seems just too loud.  I can get the ADI2 up to about -14/15 db after which it's deafening.  I played some of my more majestic (as in super loud!) symphonic tracks and, at my listening position, the peaks/crescendos were hitting 85-90 db on my hand-held meter at -15 on the ADI.  That's borderline too much for my ears.   So, I've set the AHB2 to low gain and those same 85-90 db meter readings are achieved around -6 or so on the ADI.

My rudimentary understanding is that it is theoretically better performance on the ADI the closer it can get to-0.    That would appear to point towards keeping the AHB2 at its low gain setting.  Am I correct in my understanding???

I am using auto ref level "ON" and, unless somebody contradicts, that's where it'll stay.

Finally, for general "polite"/dinnertime listening, somewhere around the -25 to -30 range on the ADI is about where we've settled. The low volume ref is still set at it's default -30 setting.  Given the above, do I need to adjust the low volume reference a bit?  If so, what would be the recommendation on that??

I wrote that several times before: there is no loss of whatsoever. The limit is not the ADI or the AHB-2, it is the ambient room noise and your ears. There is no distortion at lower levels with your system, so the only lower limit is the self-noise of both units - imagine the low level parts of the music would vanish within the noise - uh oh! But that is not happening. Are you able to hear any noise when not playing music? No, because it is so low it stays below your hearing threshold.

So what happens when you play music at -50 dB on the ADI? Parts of the music will be below your hearing threshold, that's all. No need to speculate about bits lost or reduced dynamic range or low level distortion and all this outdated crap.

And when you crank up the volume until it hurts so that you are able to easily hear even the lowest volume parts of the music, fun is gone.

Your personal sweet spot is somewhere in-between and that is fine and nothing to worry about. And you don't loose anything. Enjoy! (copyright Curt962

Thanks to you both.  No noise...in fact the background is dead quiet.  Absolutely zero  hiss or buzz from the speakers w/no music playing, either (that's a first for me!).

Consider me reassured that I've managed to bumble along and have got it more or less right ;>).

Again....if anybody's considering the AHB2 and ADI2-DAC combo, it is a definite audio winner!!!

I'm aware that I'm reviving an old thread, but this topic is somewhat timeless and frequently comes up.

TL;DR: What I believe so far:

1) analog gain/volume controls are outdated for good reasons and should be avoided

2) there is no need to be afraid of DSP-based gain/volume controls given today's extreme performances (32 bit, floating point, dither and whatnot)

3) the SNR of a DAC decreases at lower volumes -> just like the ear's

Since I myself am still confused about the possible benefits of keeping up the SNR in practise, also due to marketing statements from Benchmark, there it goes (again):

That is referring to the practical (dis)advantage and I entirely agree.

Exacly, and that loss of SNR at lower volumes given that fixed noise floor (the very definition of it, actually) leads to unavoidable "loss" of SNR and dynamic range altogether when the SPL drops.

Without any intent to bash them, but I remember very well that part of Benchmark's argument for using an potentiometer in the analog path of the DAC1 back then (with all its disadvantages) went exactly in that direction.

So what I still don't get - yes, some solutions might (theoretically) keep the SNR constant and high at even low playback levels (pushing the SPL levels way into the < 0dB range), but since the ambient noise and hearing flaws will entirely mask it anyway, what is to be profited?

It would be great if you could elaborate what the core purpose of the "auto ref" feature in the RME ADIs actually is. I think in this older posting of yours, you also were referring to then auto fef - feature, praising the "better real-world dynamic and signal to noise and distortion values" of RME products (after having had the Glühwein of course):

https://forum.rme-audio.de/viewtopic.ph … 20#p136920

I understand that it can be beneficial and desirable to keep the SNR high for the analog interconnects like Cinch or XLR, and don't doubt that RME devices are better in that regard than many other products.
But for instance a RME ADI and a pair of headphones directly connected to its integrated amplifier - does auto ref featue have any benefit? No, right? Because the DAC and the amplifier are already an ideally matched pair.

Thanks, Ramses, for this Excel spreadsheet.
I've been using it for quite some time now and it's absolutely spot on.
Unfortunately, using Parametric EQ and Loudness features pushes this chart into unknown territory.

Device: ADI-2/4 Pro SE

The logic behind AutoRef is unclear in this case. It always wants to move me to the next Ref level when I haven't asked it to.
For example, my target output for the XLR is +8.5 dBU. I also turned on Parametric EQ (max 6 dB) and Loudness (max 6 dB).
AutoRef in this case keeps the Ref level at +19 dB, essentially attenuating -10.5 dB and reducing the SNR to 109.5 dB. If the Ref level was +13 dB, the SNR would be 115.5 dB.

I have tried to understand the calculation. I assume the question was of a theoretical nature, because there's nothing to worry about here.

If 8.5 dBu is the target level (see (1) in the picture), then I assume you mean the maximum. Then that means a volume setting of -15.5 dBr at Reflevel +13 dBu and SNR of 115.5 dB.

If you now increase the level by 6+6=12 dB using PEQ and dynamic loudness, then you have to lower the output volume by 12 dB to stay at a maximum output level of 8.5 dBu (see (2) in the picture).

With this volume setting of -27.5 dBr at a ref level of +1 dBu, you will achieve 114.5 dB SNR.
This is still an excellent value at which no noise will be audible.

If the device had no Auto Ref level, you would normally have an SNR of 120 - 27.5 = 92.5 at the volume setting of -27.5 dB. So Auto Ref Level increases the SNR here by 22 dB (114.5 - 92.5). These are excellent values that are far above the dynamic range of a CD recording (96 dB).

Good question. Let me think about it. I could be wrong, but I think you can see it like this:

The device is as quiet as SNR can be (115.5), the dynamic is as loud as the audio signal actually is (-30 dBFS).

Thanks KaiS - you answered my question.

As far as full scale is concerned: in relation to a complete track, this will certainly be reached temporarily at some point during a peak. But depending on the work (mainly classical music), there are often long passages that are quite (very) quiet. Only sound engineers who want to destroy the character of a work then use compressors or similar means. That's why some music just sounds boring and stressful to me and I put it on the shelf after a short time. That's why I very rarely buy music, or only if I've heard it before.

hm what is direct metal mastering ?
The guy doesn't look like Metallica, more like Flippers.