Software Settings for a Schiit DAC w/ EQ

[PC Audio Lover]

Just picked up my first DAC ever... the Schiit Mimir. But I'm a bit confused about some of the driver/OS/software settings to use if I'm trying to get the max sound quality and leverage the Mimir's processing. Here's what I'm using in my digital chain:

• Desktop PC w/ Windows 11 Pro
• Equalizer APO w/ Peace (I'm using equalization after the assistance of a UMIK-1 due to my small office room with a bunch of modes. I'll try to do room treatment later but for now this does help tame some issues.)
• Tidal
• Schiit Mimir (connected via Unison 384 which supports up to 32 bits/384kHz)

(It should be noted that on top of a lot of music listening, I will sometimes game on this PC including some older games, and when I do switch over to my Sound Blaster ZxR card to try and get all the effects in those engines. Not even sure if that matters anymore to be honest, but I digress...)

The specific settings I'm a little confused about are the following:

• Windows > System > Sound > Properties > Output Settings > Format > bit depth and the sample rate. Should I max those out to 32/384 since the Mimir can handle it? Or by doing so is it relying on Windows to upsample everything before it sends it out? (Which I hear isn't desired.) I keep finding conflicting information about this setting and don't understand if the setting dictates the max output ALLOWED, or if Windows just takes everything and upsamples it then outputs it in that setting. (The ideal, I would surmise, would be to use whatever setting allows the sound to be output via whatever the format of the file dictates.)
• Windows > System > Sound > Properties > Advanced Settings > Audio enhancements. It defaults to "Device Default Effects" but I can change that to "Off". I've seen people say you shouldn't turn it off if you're using software EQ programs, but I've tested it and my Equalizer APO settings still seem to work even with this off. I'm more concerned about turning this off if it effects game engine processing, otherwise I'd probably turn it off for music listening "just in case". Can't find an informed, reliable source on this setting.
• Tidal > Settings > Playback > Sound Output > Settings> Use Exclusive Mode. I hear a lot of people will use this to ensure "bit perfect" signal to the DAC, which I take to mean something like "completely untouched by software". Unfortunately, I've noticed when I turned this on that Equalizer APO no longer works... at least with whatever settings I have in Equalizer APO. (It also has the practical negative effect of not allowing other applications to output sound, which I guess I could live with every time I pull up Tidal knowing I'm just trying to listen to music. But it's the lack of equalization that's the bigger issue.)
• Equalize APO w/ Peace ... maybe there are some settings I should be looking at here and have not?

If there's an informed user out there that could provide some insight, it would be much appreciated.

NOTE: I know the Mimir does have the option to do a little bit of EQ via Forkbeard and I do have the module. Sad it's only 3 bands... I know Schiit says "it's enough" but maybe not for my needs. Either way, it's a moot point for me at the moment, as I discovered after purchasing it they have no Android app.

In case you're curious, the rest of the chain in my office's nearfield 2.1 setup post-Mimir is:

• Fosi ZP3 Pre-Amp (I have it for the HPF.)
• Emotiva A2M Flex Amp
• Ascend Acoustics Sierra-1 V2 Bookshelves
• SVS SB-1000 Pro Subwoofer

 

[staticV3]

Should I max those out to 32/384 since the Mimir can handle it? Or by doing so is it relying on Windows to upsample everything before it sends it out? (Which I hear isn't desired.) I keep finding conflicting information about this setting and don't understand if the setting dictates the max output ALLOWED, or if Windows just takes everything and upsamples it then outputs it in that setting.

Windows takes everything and resamples it to the sample rate that you set in the Sound Control Panel.

I'm more concerned about turning this off if it effects game engine processing, otherwise I'd probably turn it off for music listening "just in case". Can't find an informed, reliable source on this setting.

Game engine processing does not use APOs, so you can safely turn that setting off.

You can also use the EQApo Configurator to force any external APOs off, which depending on the output device can improve fidelity. See here for more:

Ending the Windows Audio Quality Debate

A debate has been long raging in the audio community that Windows isn't fit for use in a high resolution audio system. Sceptics counterargued that bits are bits and Windows audio degradations are a figment of an over-active audiophile imagination. I decided to investigate this as I kept...

www.audiosciencereview.com

Tidal > Settings > Playback > Sound Output > Settings> Use Exclusive Mode. I hear a lot of people will use this to ensure "bit perfect" signal to the DAC, which I take to mean something like "completely untouched by software". Unfortunately, I've noticed when I turned this on that Equalizer APO no longer works... at least with whatever settings I have in Equalizer APO. (It also has the practical negative effect of not allowing other applications to output sound, which I guess I could live with every time I pull up Tidal knowing I'm just trying to listen to music. But it's the lack of equalization that's the bigger issue.)

No confusion there, that's exactly how it works.

 

[PC Audio Lover]

Windows takes everything and resamples it to the sample rate that you set in the Sound Control Panel.

1. Should I assume it's not desirable to have Windows do any upsampling/downsampling, and instead have my DAC on NOS or allow it to use it's own filters/processing for upsampling?
2. Should I take this to mean the only way to avoid the Windows resampling is to use Exclusive Mode in Tidal? Or are there other settings, such as in Equalizer APO, that could help with this? Because I can't figure out how to avoid Windows messing with it and still have my current EQ solution via Equalizer APO.
3. SIDE QUESTION: I'm curious if anyone knows, if you choose a bit depth and sample rate setting in Windows that matches a Tidal track exactly, if Windows is still somehow resampling the signal through some sort of DSP?

Game engine processing does not use APOs, so you can safely turn that setting off.

Thank you! I was actually wondering about older systems like EAX that I believe SoundBlaster supports. But this is useful info.

 

[DVDdoug]

Just picked up my first DAC ever... the Schiit Mimir. But I'm a bit confused about some of the driver/OS/software settings to use

Generally, the "digital side" and electronics are better than human hearing. Sometimes you can get noise from analog electronics (hum, hiss, or whine in the background). And, you need enough amplifier to go as loud as you want. With electronics, frequency response is usually flat over he audio range (assuming no EQ/tone controls) and distortion is almost always below audibility unless you over-drive an amplifier (or something) into clipping (distortion).

You can also push the digital data into clipping if you boost with EQ. But your digital EQ should have a "preamp" setting to bring the overall levels down to a safe level. (It's called a "preamp" function but it's normally used to attenuate rather than amplify.)

Of course, you can "damage" the sound digitally by using a low-resolution format (8-bits / 8KHz is pretty bad for music ) or low-bitrate MP3 can sound pretty bad. High bitrate can often sound identical to the uncompressed original or it can be very hard to hear the difference. It IS lossy compression and data is thrown-away to make a smaller file, but it's not nearly as bad as its reputation.

SPEAKERS make a BIG difference. All speakers sound different (better or worse), the room makes a difference, and of course EQ/tone controls can make a BIG difference.

I'll try to do room treatment later but for now this does help tame some issues

Unless you have a "bad sounding" room with mostly reflective surfaces it may not be worth it. And most listeners don't want a "dead" room.

...Except for the bass range which can almost always benefit from treatment. But that requires bass traps. Bass traps reduce the bass that would otherwise be reflected so they smooth the bumps AND the dips. Regular acoustic treatment doesn't help with low frequencies. "Traditional" bass traps are very large and bulky so they make bedroom of living room much smaller. There are thinner "membrane" traps but you still need to cover a significant percentage of wall area. And they are large and you need multiple bass traps to be effective so it gets expensive.

Windows > System > Sound > Properties > Output Settings > Format > bit depth and the sample rate. Should I max those out to 32/384 since the Mimir can handle it? Or by doing so is it relying on Windows to upsample everything before it sends it out? (Which I hear isn't desired.)

It's going to re-sample, which isn't going to make any difference if you up-sample. And the bit depth is always re-sampled (if necessary) to match the hardware. (Although no DAC is accurate to 32-bits... The practical limit seems to be around 18-20 bits.)

Generally "CD quality" 16/44.1) is better than human hearing.*

Windows > System > Sound > Properties > Advanced Settings > Audio enhancements. It defaults to "Device Default Effects" but I can change that to "Off".

That's up to you. But if you want "accurate" sound, leave them off.

Tidal > Settings > Playback > Sound Output > Settings> Use Exclusive Mode. I hear a lot of people will use this to ensure "bit perfect" signal to the DAC, which I take to mean something like "completely untouched by software". Unfortunately, I've noticed when I turned this on that Equalizer APO no longer works..

Correct. If you alter the "bits" with EQ, it's no longer "bit perfect". But for room correction (or other "correction") you WANT to alter the data/sound.

"Bit perfect" settings are more like "insurance" to make sure the digital side isn't altering the sound in any way.

Equalize APO w/ Peace ... maybe there are some settings I should be looking at here and have not?

Don't try to fix ever little "wiggle", especially at the higher frequencies. Moving your head (or moving the microphone) will change everything.

And don't use too much boost in the low range where you're getting room mode cancelations. A little boost (maybe +6dB which is 4 times the amplifier power) is usually OK but it takes "infinite" power and "infinitely large" subwoofers to overcome cancelation. Big cuts where you have bumps-up in frequency response are OK. The good news is that the bumps are more annoying than the dips. The bumps and dips are also different at different places in the room but moving your head slightly doesn't make any difference.

Sad it's only 3 bands... I know Schiit says "it's enough" but maybe not for my needs.

It's enough to adjust for your taste or for recordings that need a little tweak, but it's not precise enough to fix-up room modes.


* In proper-controlled blind listening tests, almost nobody can hear the difference between a high-resolution original and a copy down-sampled to 16/44.1.

 

[staticV3]

Should I assume it's not desirable to have Windows do any upsampling/downsampling, and instead have my DAC on NOS or allow it to use it's own filters/processing for upsampling?

Windows resampling ain't that bad.

I'd choose EQ but resampling over bit-exact but no EQ every time.

If you're curious, you can download Audacity and capture all post-APO Windows audio by setting Audacity's input to Wasapi Loopback.

Make sure you match Audacity's Project sample rate to the one set in the Sound Control Panel before you hit Record and you can compare what different Windows sample rates do to the audio.

Programs like Room EQ Wizard and Pkane Deltawave are very handy for analysis.

Should I take this to mean the only way to avoid the Windows resampling is to use Exclusive Mode in Tidal? Or are there other settings, such as in Equalizer APO, that could help with this?

To avoid resampling, your playback software needs to use one of three APIs:
Wasapi Exclusive
ASIO
Kernel Streaming

That's what Exclusive Mode in Tidal does. It switches the playback API, likely from Wasapi Shared to Exclusive.

No setting inside EQApo or Peace can enable bit-exact playback.

SIDE QUESTION: I'm curious if anyone knows, if you choose a bit depth and sample rate setting in Windows that matches a Tidal track exactly, if Windows is still somehow resampling the signal through some sort of DSP?

You can check with Audacity and Deltawave.

IIRC, though the output is technically not bit-exact, it's still audibly flawless.

 

[PC Audio Lover]

You can also push the digital data into clipping if you boost with EQ. But your digital EQ should have a "preamp" setting to bring the overall levels down to a safe level. (It's called a "preamp" function but it's normally used to attenuate rather than amplify.)

I indeed use my digital EQ pre-amp to bring it down and avoid clipping due to some of what I'm trying with EQ right now. More on that below.

Unless you have a "bad sounding" room with mostly reflective surfaces it may not be worth it. And most listeners don't want a "dead" room.

Unfortunately right now I do! Small... kind of empty... wood floor. I have to do a little something soon.

...Except for the bass range which can almost always benefit from treatment. But that requires bass traps. Bass traps reduce the bass that would otherwise be reflected so they smooth the bumps AND the dips. Regular acoustic treatment doesn't help with low frequencies. "Traditional" bass traps are very large and bulky so they make bedroom of living room much smaller. There are thinner "membrane" traps but you still need to cover a significant percentage of wall area. And they are large and you need multiple bass traps to be effective so it gets expensive.

Yeah, I'm aware. I'm thinking maybe just get around eight 6" GIK bass traps and hang them all over the bare walls to help with the 100Hz+ issues, early reflections, etc. Only so much I can do. But I digress...

Don't try to fix ever little "wiggle", especially at the higher frequencies. Moving your head (or moving the microphone) will change everything.

Oh yes, I've noticed. When I take measurements the slightest of movement in mic positioning changes that high end measurement drastically. I typically take a few measurements in various positions and average them to get some idea of what it's like at listening position. (What's amazing to me is how much my BRAIN seems to average things out perceptually as my head bobs around.)

And don't use too much boost in the low range where you're getting room mode cancelations. A little boost (maybe +6dB which is 4 times the amplifier power) is usually OK but it takes "infinite" power and "infinitely large" subwoofers to overcome cancelation. Big cuts where you have bumps-up in frequency response are OK. The good news is that the bumps are more annoying than the dips. The bumps and dips are also different at different places in the room but moving your head slightly doesn't make any difference.

Ya know... I was told not to bother tackling dips with EQ because if you have a room mode caused by deconstructive interference then adding more power to that frequency will likely only make it worse. But when toying with the auto EQ in REW and allowing up to almost 10dB in boost (which it basically gave me at around 4 different points in the bands), and then re-measuring the result, it sure did SEEM like it brought those dips way up to the desired room curve. I don't exactly understand the result myself!

It's enough to adjust for your taste or for recordings that need a little tweak, but it's not precise enough to fix-up room modes.

Yeah. Since Forkbeard can be updated, would be so nice if down the road they figured out a way to leverage their processing power on this thing to give us more bands.

 

[staticV3]

Oh yes, I've noticed. When I take measurements the slightest of movement in mic positioning changes that high end measurement drastically. I typically take a few measurements in various positions and average them to get some idea of what it's like at listening position.

Moving Microphone Method is the answer

Also using REW's Uncorrelated Noise in case you're measuring both speakers at the same time.