DAC burn-in/break-in? My Gustard X16 measurements

[Pdxwayne]

As expected, blind tests are very tough and very time consuming, especially when listening for small changes in volume.....

I managed to complete only 6 blind listening comparisons this morning and I decided to stop as it already took me 30 minutes just to do 6 comparisons. I started to get impatient and seem to loss the ability to detect slight change in volume near the end.

The number of tries not yet reached statistical significant.

First 5 correct, last one failed. 5 out of 6 correctly identify which DAC is which.

Test song is Pocket Piano from this album:
https://tidal.com/playlist/e368f41b-c5d9-4121-aa2a-e1c49ce1e17a,

The x16 started cold, was turned off for 12 hours. Tests started about 10 minutes after powered on. E30 been on for more than 3 days.

Wife is the one who helped me with the blind tests.

She flipped a coin each time to decide which button in RCA quick switch to press before I start my listening. She would make sure both buttons were unselected before she select the button.

I am going to call it quits while I am ahead.

Really, not worth spending more time to investigate as the differences are now too small to bother me. Time to enjoy music!

: p

Again, I thank everyone here for being patient with me and many for providing me tips and advises!

 

[Jimbob54]

As expected, blind tests are very tough and very time consuming, especially when listening for small changes in volume.....

I managed to complete only 6 blind listening comparisons this morning and I decided to stop as it already took me 30 minutes just to do 6 comparisons. I started to get impatient and seem to loss the ability to detect slight change in volume near the end.

The number of tries not yet reached statistical significant.

First 5 correct, last one failed. 5 out of 6 correctly identify which DAC is which.

Test song is Pocket Piano from this album:
https://tidal.com/playlist/e368f41b-c5d9-4121-aa2a-e1c49ce1e17a,

The x16 started cold, was turned off for 12 hours. Tests started about 10 minutes after powered on. E30 been on for more than 3 days.

Wife is the one who helped me with the blind tests.

She flipped a coin each time to decide which button in RCA quick switch to press before I start my listening. She would make sure both buttons were unselected before she select the button.

I am going to call it quits while I am ahead.

Really, not worth spending more time to investigate as the differences are now too small to bother me. Time to enjoy music!

: p

Again, I thank everyone here for being patient with me and many for providing me tips and advises!

I commend your dedication.

Now please, don't buy anymore hifi kit for a while ;-)

 

[Pdxwayne]

I commend your dedication.

Now please, don't buy anymore hifi kit for a while ;-)

Haha, I agree. Have a great rest of weekend!

 

[Pdxwayne]

This week I tried foobar2000 ABX for the first time.

I did pretty good in one trial comparing original song vs a song with gentle slope of 0.2db from ~1.3Khz to ~9Khz. I did pretty good and got 12 out of 16. See the post in https://www.audiosciencereview.com/...no-v-stereo-what-is-audible.21864/post-821300.

So I thought this will be a good one to use to further check my ability to hear a 0.2db imbalance of right channel with actual stereo music.

I used Audicity to adjust right channel of original clip down 0.2db.

Then use Deltawave to confirm channel db different between original vs adjusted clip.

As you can see, left channel is the same:

Right channel has a different of 0.2db:

Then I tried ABX the two files. First time I failed (40% p-value). I was focusing on sound stage and that didn't help. Second time I did better. I actually focusing on loudness difference and got 1% p-value.

So, I guess indeed I can sense a difference of 0.2db using music?

 

[charleski]

Wife is the one who helped me with the blind tests.

She flipped a coin each time to decide which button in RCA quick switch to press before I start my listening. She would make sure both buttons were unselected before she select the button.

I suppose it should be pointed out that you're describing a single-blind testing regime. Double-blind designs are preferred as they're less likely to produce inadvertent unblinding of the test subject through unconscious cues from the person doing the switching. Unfortunately these generally require extra personnel or equipment. In this situation one solution would be to have your wife perform the switching in a separate room, and have no communication with you other than, say, flashing a light to indicate that a new trial was commencing.

Details of experimental design are quite important in psychological testing.

 

[Pdxwayne]

I suppose it should be pointed out that you're describing a single-blind testing regime. Double-blind designs are preferred as they're less likely to produce inadvertent unblinding of the test subject through unconscious cues from the person doing the switching. Unfortunately these generally require extra personnel or equipment. In this situation one solution would be to have your wife perform the switching in a separate room, and have no communication with you other than, say, flashing a light to indicate that a new trial was commencing.

Details of experimental design are quite important in psychological testing.

Thanks. I can't see nor hear her actions. She sat on the floor with the switch and hand movement outside of my view.

Also, all she did was pressing the switch button, without knowing anything about which DAC is which. I thought if she doesn't know which is which, she is also blind. : )

Thus I called it double blind. If this is not both blinds, I don't mind call it single blinds.
: )

 

[charleski]

Thanks. I can't see nor hear her actions. She sat on the floor with the switch and hand movement outside of my view.

Also, all she did was pressing the switch button, without knowing anything about which DAC is which. I thought if she doesn't know which is which, she is also blind. : )

Thus I called it double blind. If this is not both blinds, I don't mind call it single blinds.
: )

Single-blind is still infinitely better than no blinding at all. If you’re doing an ABX paradigm, then your wife would presumably know whether X was A or B, which is what matters here.

 

[Pdxwayne]

Single-blind is still infinitely better than no blinding at all. If you’re doing an ABX paradigm, then your wife would presumably know whether X was A or B, which is what matters here.

She doesn't know which DAC is which. All she sees is the RCA switch. She doesn't care how the switch was connected to the DAC anyway. I will listen, then tell her I think switch A or B was pressed. She will confirm if I selected switch A or B correctly or not. That is all she knows.