HydrogenAudio

Hydrogenaudio Forum => Listening Tests => Topic started by: apastuszak on 2019-08-17 18:16:00

Title: Help me make sense of a listening test on Reddit
Post by: apastuszak on 2019-08-17 18:16:00
I'm reading this post on reddit out of curiosity;

https://old.reddit.com/r/headphones/comments/c4ze3g/testing_to_see_if_tidal_uses_dspalters_their/

The quick summary goes as follows:

1. Posted claims he can hear a different with Tidal streaming through an app called roon, vs Tidal streaming through it's native app.
2. When he captures the output of both apps using Audacity and then uses an audio DIFF tool to do a null test, the captures show as identical.
3. When he runs the same two streams through an audio interface and then captures that and does same DIFF, they are not identical.

I'm not trying to put the person on the spot, or prove them right or wrong.  I'm just trying to wrap my head around what they did in this test, and if their test method is sound or if there is something more that needs to be done or something that needs to be modified in the testing.

This is an exercise in expanding my knowledge of listening tests.
Title: Re: Help me make sense of a listening test on Reddit
Post by: saratoga on 2019-08-17 21:06:08
I only skimmed it, but having different output from the OS's mixer isn't too uncommon.   Things like volume are often adjusted during mixing. 

It'd be a more interesting test if he'd look at what exactly is different.  If it's just volume adjustment, the signal won't null, but the different is basically irrelevant.   
Title: Re: Help me make sense of a listening test on Reddit
Post by: Rollin on 2019-08-17 21:21:55
When he runs the same two streams through an audio interface and then captures that and does same DIFF, they are not identical.
So he records from DAC's analog out? And how he align start of both recordings in this case? Without perfect alignment null test is meaningless.
Title: Re: Help me make sense of a listening test on Reddit
Post by: apastuszak on 2019-08-17 21:24:25
Well, from what I read, he captures from the DAC directly into Audacity.  I wonder if that even possible, or if that's just going digital->digital.
Title: Re: Help me make sense of a listening test on Reddit
Post by: Rollin on 2019-08-17 21:36:12
I wonder if that even possible, or if that's just going digital->digital
RME ADI-2 DAC doesn't have digital output - https://www.rme-audio.de/en/products/adi_2-dac.php (https://www.rme-audio.de/en/products/adi_2-dac.php)

So what otput to what input he hooked exactly?
Quote
I hooked up the output of my RME ADI-2 DAC to my audio interface input

Oh, i downloaded difference file. Significant differences are only in left channel. So this can be channels disbalance of analog output/input (if he used analog).
Title: Re: Help me make sense of a listening test on Reddit
Post by: apastuszak on 2019-08-17 22:39:02
I wonder if that even possible, or if that's just going digital->digital
RME ADI-2 DAC doesn't have digital output - https://www.rme-audio.de/en/products/adi_2-dac.php (https://www.rme-audio.de/en/products/adi_2-dac.php)

So what otput to what input he hooked exactly?
Quote
I hooked up the output of my RME ADI-2 DAC to my audio interface input

Oh, i downloaded difference file. Significant differences are only in left channel. So this can be channels disbalance of analog output/input (if he used analog).

Wouldn't that disbalance always exist regardless of software used?
Title: Re: Help me make sense of a listening test on Reddit
Post by: Rollin on 2019-08-17 23:08:02
Wouldn't that disbalance always exist regardless of software used?
Yes, it would existed, of course.
Anyway question remains about what output/input were used and how alignment was done.
And don't forget that theoretically this "test" can be just fake and BS. Technically, it is easy to do such test to verify, but it is not easy to waste money on Tidal and Roon.
Title: Re: Help me make sense of a listening test on Reddit
Post by: apastuszak on 2019-08-18 02:23:55
Wouldn't that disbalance always exist regardless of software used?
Yes, it would existed, of course.
Anyway question remains about what output/input were used and how alignment was done.
And don't forget that theoretically this "test" can be just fake and BS. Technically, it is easy to do such test to verify, but it is not easy to waste money on Tidal and Roon.

Tidal is a streaming service.  So, if you're going to pay for streaming anyway, Tidal does have some convenience options that might make it worth it for someone.

Roon, on the other hand, I don't get $120/YEAR for a music player, when options like foobar2000 exist for local play, and airsonic and ampache exist for personal streaming.
Title: Re: Help me make sense of a listening test on Reddit
Post by: ani_Jackal3 on 2019-09-16 12:23:53
That sub is guilty of people being clueless. Could very well being making stuff up to gain points on Tidal bad spotify good attitude there. Then again it's Reddit where few think the THD measurements done at 90 - 100db are audible under 90db.