Ending the Windows Audio Quality Debate

[markanini]

Combined with studio speakers with extended high frequencies and ruthless honesty about bad mixing-mastering and any type of distortion, be sure it doesn't cognitive bias. Feel free to test yourself anyway. Just try to play the same song on -0.1 and -8db preamp gain on a matched listening level. Probably will not be that significant with hifi gear with 'smoother' treble though.

We don't know your full signal chain, but it's unlikely that Windows Audio is affecting the quality, and likely that you are misattributing something. Even trained listeners can be biased by volume, I know as a small-scale mixer, and the developer of plugins like ABLM know this.

 

[Gitfiddle]

The truth hurts? What he suggested has a scientific backing. Simple -- and not even necessarily large -- level difference has interesting psychoacoustic effects.

Not even just psychoacoustic effects, check out the distortion between 86 DB and 90 DB here on these budget monitors. 4 DB and these things go crazy!
https://www.audiosciencereview.com/forum/index.php?threads/edifier-mr4-review-budget-monitor.29068/

 

[mikro makro]

We don't know your full signal chain, but it's unlikely that Windows Audio is affecting the quality, and likely that you are misattributing something. Even trained listeners can be biased by volume, I know as a small-scale mixer, and the developer of plugins like ABLM know this.

Setup: atc scm12 pro speakers, rotel rb-1552 mkII power amp, ssl2 audio interface, mogami interconnects between interface->amp and van damme speaker cables. The thing i don't understand, why the signal acts differently (on ear matched listening level !) as i change media player/preamp/system volume level despite i'm using wasapi driver in exclusive mode ? Around 0db it sounds a bit compressed, bass is more prominent with less sub frequencies, highs are more forward and the overall sound is more like windows limiter is still in the signal chain. Below -4db it starts calming down. Maybe some driver issue ?

 

[mikro makro]

Of course i'm compensating the listening level on the audio interface as i turn volume up or down in the computer and it acts like just i described it even at low listening level. It's just me or anyone else hears the same ?

 

[Jon L]

Summary Recommendations:

• Install Equalizer APO and use it to disable original APOs and set EAPO's preamp gain at ~ -4 dB to avoid upsample overs, filtering induced peaks, and the Windows CAudioLimiter

If I recall correctly, I didn't have to purposely go into Test Mode to uncheck "Original APO." I just looked, and original APO's were already unchecked without my intervention.
Thanks for doing the measuring work. I am finicky, too, but I have been using Windows audio with EAPO with great success sound-wise, at least as good compared to any ASIO, WASAPI implementations..

 

[Gitfiddle]

Of course i'm compensating the listening level on the audio interface as i turn volume up or down in the computer and it acts like just i described it even at low listening level. It's just me or anyone else hears the same ?

At low listening levels one way I can Introduce audible changes at level matched playback is when I have a moderate digital EQ running and didn't have enough -gain in the EQ software itself. This introduces audible pops and clicks and other oddities. (Assuming levels under 65DBA is low.)

I use voicemeeter banana in WASAPI mode and perhaps it behaves differently than EQ apo.

Since you have an audio interface you should be able to measure the difference using REW software. You could do measurements something similar this: PC out-> DAC -> preamp-> Audio interface -> PC in -> REW software. No need to measure speakers or main amp since the change in gain and measurable performance shouldn't require them if this is windows related. This is your best option to validate what you are hearing and get actual help here.

 

[mikro makro]

Just installed REW, i must find out how to do a loopback record measurement with it and if it's done, will post the results. In the meantime i asked two of my friends to do a quick listening test (using wasapi excl driver on their media player and level matched listening) at their home without suggesting them what they should hear. Both of them told me exactly the same what i've heard. Around 0db compressed, unpleasant highs, too 'forward' sound, at -8db, much more laid back, cleaner. I know it's not a scientific test, especially with only 2 people involved but if there's no distortion around 0db, it should sound exactly the same as at -8db. Feel free to test it yourself on your own equipment.

 

[MaxwellsEq]

Just installed REW, i must find out how to do a loopback record measurement with it and if it's done, will post the results. In the meantime i asked two of my friends to do a quick listening test (using wasapi excl driver on their media player and level matched listening) at their home without suggesting them what they should hear. Both of them told me exactly the same what i've heard. Around 0db compressed, unpleasant highs, too 'forward' sound, at -8db, much more laid back, cleaner. I know it's not a scientific test, especially with only 2 people involved but if there's no distortion around 0db, it should sound exactly the same as at -8db. Feel free to test it yourself on your own equipment.

In your opinion, what is the electronic mechanism causing this?

 

[mikro makro]

There's a good chance that the problem can be tracked back to dac, not windows audio.

 

[Bernard23]

Haven't yet read through all of th 43 pages, and am cpompletely new to this, mostly because I am deeply mistrusting of Windows and what it might or might not be doing. I thought I'd try it out as I really need to EQ my phones and it's the only way i can use my venerable JDS Element 2 with DSP as it only has a USB-B input! I've installed it, saved a few profiles and starting comparing it using Qobuz and a pair of daybreaks that don't need any EQ for me. Although it's not hugley different, I'd reliably distinguish between Qobuz in exclusive mode being better than Qobux non-exclusive and EAPO switched in (flat EQ settings). I can't tell any difference between toggling EAPO on or off. I also notice that my device config screen doesn't look the same, i cannot switch off old profiles, I can only add output devices that I'd like to apply. If I've understood the OP, there should be a benefit to toggling EAPO ON vs OFF?
What am I missing?

 

[DEALUX]

Haven't yet read through all of th 43 pages, and am cpompletely new to this, mostly because I am deeply mistrusting of Windows and what it might or might not be doing. I thought I'd try it out as I really need to EQ my phones and it's the only way i can use my venerable JDS Element 2 with DSP as it only has a USB-B input! I've installed it, saved a few profiles and starting comparing it using Qobuz and a pair of daybreaks that don't need any EQ for me. Although it's not hugley different, I'd reliably distinguish between Qobuz in exclusive mode being better than Qobux non-exclusive and EAPO switched in (flat EQ settings). I can't tell any difference between toggling EAPO on or off. I also notice that my device config screen doesn't look the same, i cannot switch off old profiles, I can only add output devices that I'd like to apply. If I've understood the OP, there should be a benefit to toggling EAPO ON vs OFF?
What am I missing?


View attachment 491658

Equalizer APO doesn't work with exclusive mode. Exclusive mode disables "enhancements" which includes EQ.

 

[Bernard23]

Equalizer APO doesn't work with exclusive mode. Exclusive mode disables "enhancements" which includes EQ.

I know that. My question is still valid! I amusing qobuz in non exclusive mode to check EAPO. I then switch it on as a reference, so I'm comparing

1. streaming without EAPO in exclusive mode = best sounding
2. streaming in non exclusive mode with EAPO <ON>
3. streaming in non exclusive mode with EAPO <OFF>

What surprise ls me is that 2 and 3 sound identical, I was expecting EAPO to make an improvement according to the OP, unless 9f course the improvements in distortion are inaudible?

I'd hoped that 1 and 2 would sound the same and better than 3

 

[markanini]

I am deeply mistrusting of Windows and what it might or might not be doing.

If you trust your audio interface, EqualizerAPO, and an automatically generated EQ profile, you can trust Windows audio, especially since it’s now well-documented how and when Windows processes sound. One instance is sample rate mixing using a windowed sinc function, which performs excellently according to objective measurements. There’s also a soft clipper processor at the end of the audio chain. Skipping resampling and soft clipping might seem like a good idea, but issues arise if your audio source has a different sample rate, the audio is highly dynamic, or your playback software is set to play with positive gain.

 

[Bernard23]

If you trust your audio interface, EqualizerAPO, and an automatically generated EQ profile, you can trust Windows audio, especially since it’s now well-documented how and when Windows processes sound. One instance is sample rate mixing using a windowed sinc function, which performs excellently according to objective measurements. There’s also a soft clipper processor at the end of the audio chain. Skipping resampling and soft clipping might seem like a good idea, but issues arise if your audio source has a different sample rate, the audio is highly dynamic, or your playback software is set to play with positive gain.

Thank you, what changes / improvements are there between W10 and W11?

Of course the caveat is I'm making these judgements purely subjectively with all of the uncertainty that's associated with it.

 

[DEALUX]

What surprise ls me is that 2 and 3 sound identical, I was expecting EAPO to make an improvement according to the OP, unless 9f course the improvements in distortion are inaudible?

You probably misread what he said. If you install EQ APO, I think its implementation of the APO driver is always on. You're just toggling the EQ on and off.

 

[Bernard23]

You probably misread what he said. If you install EQ APO, I think its implementation of the APO driver is always on. You're just toggling the EQ on and off.

I certainly did! Makes sense then. Of course, with EQ most phones are going to sound much better than any minute differences between steaming in excl or non-excl mode.
I should add I really like the graph window feature in Peace.

 

[magreen]

I've made great use of EAPO for PEQ/room correction in Windows 10 based on the advice in this thread. Should I be worried about updating to Windows 11, and it affecting the Windows mixer audio quality?

 

[Robbo99999]

I've made great use of EAPO for PEQ/room correction in Windows 10 based on the advice in this thread. Should I be worried about updating to Windows 11, and it affecting the Windows mixer audio quality?

As far as I know Windows 11 uses same audio system as Windows 10, I haven't heard of any distinctions.

 

[magreen]

As far as I know Windows 11 uses same audio system as Windows 10, I haven't heard of any distinctions.

Great to hear, thanks. Sounds like I don't need to fear updating when the time is right.