Configuration recommendations for RME ADI-2 FS DAC v2, Geshelli Labs Erish, Sennheiser HD 660 S

[muss3d]

Hi,

I'd like to configure my DAC/AMP and music software to the best possible settings.
Ideally I'd like to find a "configure-it-once-and-forget" solution, so I can simply enjoy music rather than fiddle with settings every 5 minutes
And I was wondering if anyone would be interested in sharing their configuration for RME ADI-2 DAC and music software they use or any configuration suggestions.

So, ATM I have:

• RME ADI-2 FS DAC v2
• Geshelli Labs Erish
• Sennheiser HD 660 S
• Shure 846
• Stock TRS cable
• and one of the cheapest 4-pin XLR cables for Sennheiser HD600 series https://www.amazon.co.uk/gp/product/B07M9VVQX2/

My current RME ADI-2 FS DAC v2 configuration looks like this:

• Analogue outputs
  • Are set to Fixed Output Level (just like suggested in the video at the end of this post, Apparently this sets the output level to 2 volts, equalling to +8dBu)
  • Auto Ref Level: off
  • Ref Level: +13dBu
  • Lock volume: on
  • Volume: -5dB
  • Parametric EQ: on (see Oratory1990 preset for Sennheiser HD 660 S. I used the first 5 filters from the "Parametric EQs" table.)
  • Bass/Treble: off
  • Loudness: off
  • All other settings are left at default values
• Phones output
  • Hi-Power: off
  • Auto Ref Level: on
  • Lock volume: off
  • Parametric EQ: on (see Oratory1990 preset for Sennheiser HD 660 S. I used the first 5 filters from the "Parametric EQs" table.)
  • Bass/Treble: off
  • Loudness: off
  • All other settings are set to default values
• IEM output
  • Parametric EQ: on (see crinacle preset for Shure 846 (blue filter)). I used the first 5 filters from the "Parametric EQs" table.)
  • Bass/Treble: off
  • Loudness: off
  • All other settings are set to default values
• General Options
  • Phones / DSD:
    • Mute Line: Toggle Ph/Line
  • Display
    • Hor. Meter: Dual

Erish is on low gain.


On Linux is use mpd as my default player and ncmpcpp as CLI GUI:

Output is configured like so:

audio_output {
    type                "alsa"
    name                "Alsa RME"
    auto_format         "no"
    auto_resample       "no"
    auto_channels       "no"
    dop                 "yes"
    dsd_usb             "no"
    replay_gain_handler "none"
    mixer_type          "none"
}

ALSA configuration /etc/alsa/alsactl.conf:

ctl.hw {
    @args [ CARD ]
    @args.CARD {
        type string
        default "0"
    }
    type hw
    card $CARD
}

The other and /etc/asound.conf is empty.


PulseAudio config /etc/pulse/daemon.conf is not really relevant in this context, but I'll paste just in case someone's interested:

default-sample-format = float32le
default-sample-rate = 44100
alternate-sample-rate = 384000
avoid-resampling = yes
high-priority = yes
resample-method = soxr-vhq

On Windows 10 I use:

• Latest madiface_win_09722.zip
• Foobar2000 with ASIO support 2.1.2

Thanks,
muss3d

Fixed Output Level configuration for RME-ADI 2 DAC (starts @ 4:14)

 

[dfuller]

If it was me... I'd just use the built in headamp on the ADI-2. It's very good and has more than plenty of power for HD660s, which to be entirely honest don't need a ton of power (150 ohm nominal impedance and 104dB/1V sensitivity = not much power needed to reach painfully loud volumes)

 

[muss3d]

@dfuller I agree that the built-in headamp is very good. I got Erish just to have a balanced output.
To my mind, Erish in some (mostly acoustic) tracks has a bit better channel separation and that's about it.

My question here, was more about the recommended "Line Out" volume settings on ADI.

 

[AnalogSteph]

Play a 23 Hz, 0 dBFS sine with the HD660S PEQ on. Repeat with same @ -1.5 dBFS for reference. I am not convinced that the 0 dBFS won't clip. (The reason being the +6.4 dB PEQ boost @ 23 Hz @ -5 dB digital attenuation.) Yes, it's an edge case, but I mean, it's not like you're short of dynamic range, right?

To my mind, Erish in some (mostly acoustic) tracks has a bit better channel separation and that's about it.

I'll let you in on a secret - but don't tell anyone, 'K?

You know what reduced channel separation as a result of common ground return conductor resistance sounds like?
Increased stereo width.
Also, even a 20 dB channel imbalance is quite uncomfortable in headphones. BTDT.

In sum, I'm afraid there's about a 95% chance that either (a) level matching wasn't perfect or (b) you were fooling yourself, perhaps because changing the cable takes a while and as such a direct A/B is not possible. Nobody said listening tests were easy.

BTW, looking at the website for the Erish, I noticed that their "vision of what fully balanced should mean" would seem to have one problem: Input common-mode rejection is dependent on volume potentiometer channel tracking. A mere 0.5 dB of mistracking lets CMRR plummet to a measly 26 dB. I guess nobody noticed because the unit is using a floating power supply (so between that and the audio cable's shield connection, there isn't a great deal of common-mode to contend with anyway), and they were only measuring with the volume at full tilt.
If they ever wanted to put this circuitry into an IEC Class I box with integrated power supply, the input would have to get a balanced/balanced (fully differential) receiver like fig. 14 or fig. 13 (perhaps with an OPA1637).

 

[muss3d]

Will test it soon. Thanks @AnalogSteph

 

[muss3d]

hahahaha I like that lil trap that you set for me @AnalogSteph

I played a 23 Hz tone with:

• PEQ on
• Auto Ref Level on
• low-power mode on

I managed to get to -15 dBr. It got simply too loud and uncomfortable.
Moreover whenever ADI-2 was changing the reference level, the "switch" was really loud and harsh . Thus I gave up with further testing.

In sum, I'm afraid there's about a 95% chance that either (a) level matching wasn't perfect or (b) you were fooling yourself, perhaps because changing the cable takes a while and as such a direct A/B is not possible. Nobody said listening tests were easy.

Hard to tell. I'm leaning towards (b).