RME ADI-2 DAC fs - Official Thread

Official AMP/DAC/DAP Model Discussion
182

I wasn’t sure where to put this since there isn’t a DAW/Audio Interface type area. Psalmanazar on SBAF mentioned Steinberg audio interfaces when I was all woe is me and my Focusrite Scarlett died.

I bought a Clarett and thought it was great except the random dropouts I kept getting (worse than the Scarlett). I bought the Axe I/O because it was a new guitar-focused interface and sounded good but the latency was bad.

Got annoyed. Found a used Steinberg UR28M and bought it. Latency is mediocre on it but it works really well, vocals and guitar recording are good on it, etc. Since it was used, I couldn’t return it and decided to find a way to live with it.

Now on to why we’re in the ADI-2 DAC thread.

I remembered that you could use it in full duplex mode. I’m now outputting the digital signal from the UR28M interface via coax to the RME ADI-2 DAC. That’s what goes into my DAW, and the latency on the RME drivers is nucking futs. It’s really fast.

As measured by RTL Utility, I’m getting less than 2.5ms round trip latency. That’s bonkers! As an added bonus, I have the RME outputting to a tube headphone amp (ECP Audio L-2) for extra sauce.

The key was that I had to go into the ADI-2 DAC’s I/O menu and choose USB (Rec coax) so it could record the digital input but play back USB->analog. Initially I was a little dismayed because it was just giving me the raw signal to my headphones, then I decided to look at the manual and the solution was right there.

5 Likes
183

A close-up of this beauty:

Imgur

Love how clean and futuristic it looks.

6 Likes
184

I’ve been using the RME ADI 2 DAC with a Sbooster 12v power supply in the recent weeks and after switching back to the RME psu a few times I find the Sbooster to make a positive difference on my system (pi2designs 502dac - RME ADI 2 - Phonitor e - 2x HD650 with Forza Audioworks Claire Hybrid (one cable balanced and one cable unbalanced)). The RME has more slam, soundstage improves and micodetails/dynamics are enhanced. I’ve blindtested DAC filters and find it easier with the sbooster compared to the stock PSU. Unfortunately it is uncomfortable to blindtest the powersupplies because changing them is too much work.

The build in headamp also scales with the bigger psu, on “hipower” this sounds very good with the HD650 now and I found myself a few times turning the Phonitor volume knob just to notice the HD650 had been plugged into the RME. The Phonitor is still better, especially at very high volumes. It has better control and less fatigue but the RME as AIO slams very nicely now. I never liked it much in the past as AIO for the Sennheisers.
I’m using the RME as AIO at -40 to -50db with hipower and HD650.

I’ve also taken some time to compare filters. My favourite has always been “sharp”. It has better macrodynamics and better detail than “sd sharp” and is also a hair warmer. “sd sharp” has some slight haze and is more polite/compressed than “sharp”. I think “sharp” decays faster. I find myself raising the volume by 1-2 db after switching from “sharp” to “sd sharp” to make up for the reduced slam.
“slow” is also nice but I haven’t made up my mind about it yet. It’s probably a bit coloured for may tastes. “NOS” lacks a bit compared to the Holo Cyan which I also own.

Would be interested to read more impressions on the filters!

3 Likes
185

I’d be willing to bet that the effect of the Sbooster, if there is one, has nothing to do with the performance of the RME unit itself and is, instead, due to there being less noise fed back onto the AC supply where it might affect other components.

I say this for several reasons:

  • The first is that the very first thing that happens to your power after the power-input jack on the RME is that it hits a very-high-frequently SMPS inside the unit, followed by a series of regulators, level shifters and so on. While SMPS aren’t good for what happens to the upstream AC supply, done properly (which RMEs are), they’re almost completely immune to upstream power quality issues.

  • Having more power available to the RME unit (i.e. the higher current rating of the Sbooster vs. the included PSU) does nothing to its performance … the RME draws what it draws and it won’t draw the full 2A of the stock supply, let alone more.

  • I’ve measured the RME’s performance using a number of PSU options, ranging from a car battery with, and without ULDO regulation, a SotA lab-supply (which achieves better noise, rejection, stability and precision in practice than any of the “audiophile” supplies I’ve seen even claim) and various after-market supplies. Performance changed by less than 0.4 dB between best and worst … that was down around -130 dB, which is a) below any possible audibility and b) essentially down to variation in tests (a difference in how tightly the cables connected, were run, or how oxidized one part was, can easily yield such an error at such tiny levels).

Which is not to say you’re not hearing a difference, but it’s almost certainly not because the Sbooster is making any measurable or audible difference to the performance of the RME unit itself.

If you’re using AutoRef mode, to maximize the dynamic range performance of the RME, then this means you’re listening at about 50dB - 60 dB average level with the HD650. That’s pretty quiet … so much so that I’d be surprised if there was any slam to their delivery, and well within the region where using the “loudness” function would be useful to address equal-loudness-countour issues.

Beyond which, that’s a lot of digital attenuation, even for a software-volume control as good as the (42-bit precision) implementation on the RME-ADI 2 series.

…

On the filters …

My preference is also for the “Sharp” filter (which is a linear-phase filter). The “SD Sharp” filter (minimum-phase) messes with my perception of timbre, which I seem to be trying to subconsciously adjust to, and the end result is that listening gets fatiguing in ways that it doesn’t with the “Sharp” setting.

Curiously the “NOS” and “Slow” filters I don’t like at all. They don’t do much to address aliasing, which may or may not be the reason I don’t really like them … but perhaps not because the CYAN in NOS mode has aliasing issues that are the same, or larger, and I don’t have issues with that on the Holo Audio unit.

However, when fed from the Hugo M-Scaler, with the RME unit in “NOS” mode, well, now that’s an interesting proposition. Not a remotely worthwhile one, but there are audible (and measurable) differences that I did not expect to find.

7 Likes
186

I’ve had the RME psu plugged into a seperate outlet and I still have the wallwart that powers the Raspberry Pi computer (the 502DAC transport is powered by a seperate LPS) plugged into the powerstrip that holds the audio equipment. The sbooster includes some power filtering and I don’t use any power conditioners etc elsewhere. This might be another reason for the perceived difference. I guess your using some kind of power conditioners.
Do you know how finicky the Phonitor is with power related issues?

-40 to -50db is using the headphone output of the RME in hipower mode with the HD650.
Listening to rock music with DR8 average I find -45db pretty loud, with classic at -40db or -35db that’s also quite loud. I can’t measure though.

As a DAC I use “Rev level 7 and volume 0” or “Ref level 13 and volume -6”. Balanced in to the Phonitor that’s 3,46 Vrms I think.
I use it with the Phonitor mostly.

1 Like
187

Not when I did my measurements. All done off a common, shared, AC supply (probably noisier than most home environments).

It doesn’t seem particularly picky. No notable audible or measurable artifacts between raw house-AC and my dedicated trunk + conditioners. Seems pretty well filtered.

SMPS are almost immune to mains-borne noise by their nature. They’re more likely to add their own issues than to pass a problem from the source supply. Putting an LPS in front of any competent SMPS is largely redundant … unless the SMPS is really awful - and most professional one’s aren’t.

I’d still wager any difference in what you hear has nothing to do with the power reaching the RME unit, and is, instead, down to the Sbooster not contaminating it’s AC supply. Though the RME PSU is very well filtered in that regard also.

If you find it loud, then you don’t need to measure it. Everyone has different sensitivities to volume (my wife things my normal level is painful). But, just on the math, those will be 50-60 dB listening levels, which is pretty quiet. Check out the “loudness” functions on the RME, you may find them revelatory in this case.

1 Like
188

“LPS Nervosa” has become a real thing of late.

Cheap SMPS, and those operating too close to audible frequencies, or with poor filtering and post-conversion regulation are problematic. Most decent gear doesn’t ship with crappy SMPS, though. RME are a case in point … they’re both excellent in their output and well controlled in what they allow back to the AC supply.

Replacing an outboard SMPS with a GOOD LPS, where the device receiving power doesn’t then feed everything through another, on-board, SMPS can have real (if typically rather small) benefits. Using an LPS to feed an SMPS will almost always only yield benefits to OTHER components.

In almost all such cases (i.e. LPS->SMSPS) I’ve heard, or measured, there’s a bigger benefit in spending the money you would have spent on an LPS in just buying a better component in the first place.

5 Likes
189

Thanks for your expertise, it’s always so helpful to get this kind of input!
I’m used to the HD650 sound so I don’t use any DSP settings. I’ve tried the loudness but would prefer to use the treble and bass balance instead because it’s faster to adjust and it allows me to actually get what I want.

I actually bought the RME (used) for the sbooster, which I already owned, after the first listening tests, so it is a free ride. Some people buy a new and “better” 1k+ amp, dac and/or headphone every couple of months, that’s at least as much nervosa.
The only way to make sure would be a blind test based on two complete setups to allow easy switching between both systems.

2 Likes
190

I tend to think of “nervosa” in audio being mostly about people buying tweaks to fix “issues” that really aren’t issues in the first place, usually based on not properly understanding a) what the tweak will (or can) do and b) how the equipment they are applying it to is designed (which comes into the relevance of the tweak in the first place) - and then hearing major differences that aren’t actually there.

Repeatedly.

There are cases where using an external LPS can offer a significant advantage over the manufacturer’s supplied wall-wart or power brick. They’re just not typically found when the internal power handling on the device itself begins with an SMPS.

Similarly, putting USB “re-clockers” on USB 2.0 Async DACs is redundant. All Async DACs buffer the input and use their internal sample clock to pull data from that buffer and to drive the DAC IC. They’re re-clocking already - and at the point they should … closest the to the DAC and it’s sample clock. USB Audio doesn’t have a sample clock in the first place, it’s just the USB packet clock that such devices re-clock and that has no relation to sample jitter whatsoever.

There are plenty of other examples.

Actually buying different, similarly-leveled, components, repeatedly, while not necessarily a good practice does, at least, offer a MUCH better chance that there will be both audible, and measurable, differences in the end result … which is rarely the case with the “nervosa” type activities.

It’s really more of a problem on the lower-end of the price spectrum, where the individual “tweaks” frequently cost more than the component being “improved” and, as often as not, aren’t really helping.

Anyway, the point isn’t to be judgmental here - and in real terms the “nervosa” issues come down to an ill-informed belief that, say, LPS are always better than SMPS that is then widely applied on the basis of “FOMO” even when it is doing no actual good and might actually be making things worse.

1 Like
191

I should add that, blind-tests are (potentially) useful for determining if there are detectable differences.

They don’t help with determining “improvements” though, as that part is subjective.

When I do my comparisons, I like to do as blind-a-test as possible (trivial with DACs for example, but near impossible with headphones). I also like to employ measurements (though I tend not to share those). This is helped with some special, and sadly not very readily available/affordable tools, such as a hardware-based level-matching real-time A/B/X comparator system, and made easier by having a wife that is a scientist by training and able to assist with such things.

This helps keeps things honest, as it can be the case that I do not necessarily know if I’m listening to what I think I’m listening too. She’s been known to do things like put two identical DACs into the setup. (sometimes slightly increasing the volume of one vs. the other). Or deliberately put one out of phase with the other.

A couple of years ago I did back-to-back comparisons between the Modi MB and the Bifrost MB. It took me a lot of sighted listening, and failed A/B/X tests, to eventually learn to tell the two apart reliably. Even then it required huge concentration and listening in a fashion entirely at odds with being able to enjoy the music itself.

When you actually have people go through this sort of thing, it’s often surprising how their opinions on the scale of changes, or what they can really hear/differentiate, change.

2 Likes
192

Do you prefer the same filters for speakers too?
I find it interesting that I prefer the SD slow for speakers and the Sharp for headphones. Go figure…

1 Like
193

Can’t say, as I haven’t used the RME ADI-2 DAC in my speaker rig.

It’s not a very useful test for me … as that speaker rig is a full Linn Exakt setup, so the output of the RME would have to through the Akurate DSM’s ADC and then the 4x AKM 4497 DACs (and supporting Chakra amps) in each speaker.

Wouldn’t surprise me if different filters were preferable between headphones/speakers though … lots of potential variation there!

3 Likes
194

Step 1 of my expensive system versatility increase complete.

RME ADI-2 DAC sold.

“But wtf I thought you loved it?!”

Yep.

I wanted even more versatility, though. I can use the ADC and USB -> S/PDIF capabilities of the RME ADI-2 Pro FS, so that’s on order now.

4 Likes
195

While I wait on the ADI-2 Pro FS (maybe we should combine this thread into one for both units since they’re so similar?), I broke out my modded MHDT Labs Orchid.

I’d forgotten how “analog” these things can sound. I’m missing the crisp edges and cleanliness of the RME something awful, but the Orchid really is a romantic piece of kit.

Coupled with the SPL Phonitor XE and you can hear exactly what the DAC sounds like. It’s certainly softer, more flowing, and more of a relaxed sound compared to the RME.

Looking forward to the ADI-2 Pro FS, though.

3 Likes
196

This is a good point… I should try my other DACs and my incoming DAC with the Phonitor XE to see how the DACs sound in comparison, I’ll probably just do my El DAC, RME ADI-2 DAC and my incoming (Mass)Drop X Airist R-2R DAC…

2 Likes
197

New firmware day!!! This one is a doozy! New color “Amber” option for level meter! Plus a bunch of new features and fixes!

https://www.forum.rme-audio.de/viewtopic.php?id=28859

Have fun and practice safe updating :wink: you filthy animals! :crazy_face:

5 Likes
198

We need purple.

3 Likes
199

Concur!!! I should suggest it! Or we should start a petition lol

2 Likes
200

Excellent!

“Push Vol to Mute” was long overdue!

3 Likes
201

I switched from Cyan to Amber. Amber’s probably a better color at night, but I like the Cyan a bit more.

Cyan brings more brightness to the sound, while the amber adds a little more warmth to the music. The neutral green default setting isn’t too bad either. I hope they add purple for additional clarity, or red for when I want additional bleeding edge brightness.

4 Likes