BLIND TEST Results: "Do digital audio players sound different playing 16/44.1 music?"

[davewantsmoore]

1 hour ago, DJGopal said:

peak normalising algorithms deliver.

Ahhhhh....

 

Quote

Of course, blind testing requires that we normalize the average sample volume across the 4 devices. This was performed with Adobe Audition CS6 in 32-bits resolution then saved back out as 24-bits. At most I had to boost the softest recording by +1.16dB and the loudest recording attenuated by -1.25dB. This resulted in all the recordings being only +/-0.05dB difference which should be well below threshold of human ability to consciously or subconsciously discern. Remember, listeners tend to show preference towards louder recordings. Exactly what the threshold of "just noticeable difference" varies depending on the research and can be debated (like here) depending on signal type. Typically, we won't be able to detect volume differences as easily with actual music than simple test signals like a pink noise tone. The lowest I've seen people suggest for controlled tests like performing an ABX is 0.1dB. Thus my decision to accept the slight variation of up to +/-0.05dB between the samples I believe is reasonable.

 

Pure "peak normalisation" is certainly an issue when mixing very disparate tracks together (which is what the article is discussing)

 

But in this case.... When the audio tracks are all very close to the same....   Then the assumption that peak normalisation would be incorrect, relies on the assumption that the peaks are significantly differing in levels than other parts of the signal.

 

Once you have "normalised to the peaks"  (or whichever slice of the waveform you might pick) .... then you can sanity check this assumption by comparing other parts of the waveforms with each other.

 

"Normalisation" is likely a confusing word for everybody .... as like the article says it can have blurry meanings.    Perhaps he should have used "volume match".

 

 

 

The simple fact is that given nearly 3dB between loud and soft.

Quote

At most I had to boost the softest recording by +1.16dB and the loudest recording attenuated by -1.25dB

That is was essential for Archmagio to match the levels of the tracks.

 

Aside from "fancy algorithms".... there's no true way to do this without picking a part of the waveform, adjusting to match levels.... and then sanity checking elsewhere.

 

 

 

 

 

Edited May 20, 2019 by davewantsmoore

[davewantsmoore]

2 hours ago, rocky500 said:

Good luck with any normal person ever been able to do a reliable DBT at home.

PS - You're zeroing in on the answers to the big questions....

 

Why is it that people say "oh each to their own" ... and "we all hear a little bit different" ......  when studies consistently show this is much more false than true

Why is it that people hear differences between devices ... when those devices can be shown to have true differences somewhere between zero and tiny.

Why is it that certain "tweaks" which can be shown not to change the signal .... have people that test them and conclude they're audible

Why is it that certain "tweaks" which are shown to change the signal .... have people who test and conclude they're inaudibe

 

The answer is all contained in how they are tested (auditioned)... and how we hear....  and how tests that people typically do, are full of things which compromise them to the point where it is just guessing with a side helping of confirmation bias.

 

[Dean Gale]

Apart from matching the volumes levels the playback shouldn't be altered , that's like testing a car without driving it , earmuffs and blindfolded.  Thank god I didn't have to waste my time reading it. Sorry just having a laugh . 

[davewantsmoore]

1 hour ago, DEANO23 said:

Apart from matching the volumes levels the playback shouldn't be altered

Indeed.  That's all he did.   Just matched the volumes.

 

Yes - there is an ADC/DAC in the loop .....  but the world is very inconvenient without that (ie. how to you replay your different samples for the testing)  ..... and you can also run tests to demonstrate that it is inaudible.

 

The thing is, that is very well know that how you play the sames (eg. 1 second of sample A, then B, then C, or whatever) is very very very important to the test.   Much more important than (for example) whether you have a (known transparent) ADC DAC in the loop.

 

 

 

[rantan]

1 hour ago, davewantsmoore said:

PS - You're zeroing in on the answers to the big questions....

 

Why is it that people say "oh each to their own" ... and "we all hear a little bit different" ......  when studies consistently show this is much more false than true

Why is it that people hear differences between devices ... when those devices can be shown to have true differences somewhere between zero and tiny.

Why is it that certain "tweaks" which can be shown not to change the signal .... have people that test them and conclude they're audible

Why is it that certain "tweaks" which are shown to change the signal .... have people who test and conclude they're inaudibe

 

The answer is all contained in how they are tested (auditioned)... and how we hear....  and how tests that people typically do, are full of things which compromise them to the point where it is just guessing with a side helping of confirmation bias.

 

Ok, so when you personally go to buy a new component or a pair of speakers or a cable of some sort, you always conduct formal DB testing prior to purchase?

 

If so, I admire your committment and also if so, why doesn't everybody buy a cheap 50 watt amplifier?

 

There is either a difference or not so you tell me.

 

I am listening.

[sir sanders zingmore]

11 minutes ago, rantan said:

f so, I admire your committment and also if so, why doesn't everybody buy a cheap 50 watt amplifier?

Not enough power 

[rantan]

26 minutes ago, Sir Sanders Zingmore said:

Not enough power 

Ok then double of nuffink

[davewantsmoore]

38 minutes ago, rantan said:

you always conduct formal DB testing prior to purchase?

No.  Why would you think that?

 

I figured with what I had said that you would (if anything) think the opposite.

 

38 minutes ago, rantan said:

why doesn't everybody buy a cheap 50 watt amplifier?

Because it might not have enough power, or might have too much distortion  (?!)

 

38 minutes ago, rantan said:

There is either a difference or not so you tell me.

I'm not sure what it is that you want to know.

[rantan]

1 minute ago, davewantsmoore said:

 

I'm not sure what it is that you want to know.

HI Dave, with respect what I am asking is why some people think there is no audible or measurable difference between amplifiers...for example. If this is true why does a 100 watt Luxman/Accuphase /etc sound better than a 100 watt Onkyo if neither amp is clipping.

[davewantsmoore]

7 minutes ago, rantan said:

If this is true why does a 100 watt Luxman/Accuphase /etc sound better than a 100 watt Onkyo if neither amp is clipping.

If they sound different, then the signal exiting one amp must be different from the other.

 

By comparing those signals, we can see how they are different.     We can observe differences in the signals which are orders of magnitudes smaller than signals which actually cause the speaker cones to move (which you can confirm by observing speaker cones).

 

Once you've checked the differences between the signals, then you have the answer.

 

Common / major examples:

• Levels are not matched
• Non-linear distortion
• Non-zero output impedance (issue only shows once connected to a speaker)

 

Once you've zero'ed out these differences.... then it really IS surprising how similar competently designed amplifiers sound, when auditioned properly (ie. very fast switching, not moving from your seat, etc.) and when driving sane speaker loads.

Edited May 20, 2019 by davewantsmoore

[rocky500]

Just came across an interesting video

 

[rantan]

58 minutes ago, davewantsmoore said:

 

Common / major examples:

 

• Non-linear distortion
• Non-zero output impedance (issue only shows once connected to a speaker)

 

Once you've zero'ed out these differences.... then it really IS surprising how similar competently designed amplifiers sound, when auditioned properly (ie. very fast switching, not moving from your seat, etc.) and when driving sane speaker loads.

 In both cited examples the premium amps would be superior.

 

I agree that level matching is important and necessary.

I have actually participated in several blind tests over many years and apart from useless snake oil gadgets,I could always identify a difference between ordinary components when compared (level matched ) to premium quality components, particularly amplfiers and cd players.

Not you Dave ( your remarks are always informative ,well reasoned and civilised) but some *other* testosterone riddled acolytes of Ghengis Khan need not bother disparaging or attempting to discredit my remarks to make themselves feel better about the world, so they can save their energy for other occasions.

 

[sir sanders zingmore]

7 minutes ago, rantan said:

your remarks are always informative ,well reasoned and civilised)

+1

[Dean Gale]

If they sound different, then the signal exiting one amp must be different from the other.
 
By comparing those signals, we can see how they are different.     We can observe differences in the signals which are orders of magnitudes smaller than signals which actually cause the speaker cones to move (which you can confirm by observing speaker cones).
 
Once you've checked the differences between the signals, then you have the answer.
 
Common / major examples:

• Levels are not matched
• Non-linear distortion
• Non-zero output impedance (issue only shows once connected to a speaker)

 
Once you've zero'ed out these differences.... then it really IS surprising how similar competently designed amplifiers sound, when auditioned properly (ie. very fast switching, not moving from your seat, etc.) and when driving sane speaker loads.

Isn't the way things get distorted that we perceive when the volume differential is taken out . I.e. 2nd order harmonic , etc , etc.

[davewantsmoore]

12 hours ago, DEANO23 said:

Isn't the way things get distorted that we perceive when the volume differential is taken out . I.e. 2nd order harmonic , etc , etc.

Not sure of the question, sorry.

[Dean Gale]

Not sure of the question, sorry.

Okay

[MLXXX]

On 20/05/2019 at 7:42 AM, Rickus said:

A long read but some interesting conclusions.

 

http://archimago.blogspot.com/2019/05/blind-test-results-part-1-do-digital.html
http://archimago.blogspot.com/2019/05/blind-test-results-part-2-do-digital.html
http://archimago.blogspot.com/2019/05/blind-test-results-part-3-do-digital.html

Thanks Rick for starting this thread. The author has gone to a lot of trouble in analysing and presenting the results.