Discussion on DACs with a blind test at end

[muon*]

Using recorded files for comparisons add confounding variables, IMO.

 

Anyway, I've previously posted my view of this type of "science".

[POV]

9 hours ago, muon* said:

Using recorded files for comparisons add confounding variables, IMO.

 

Anyway, I've previously posted my view of this type of "science".

 

I really think you should look into what a confounding variable actually is because you apply the term incorrectly.  The recording itself is not a confounding variable.  The process could introduce confounding variables if say for example the recording process is not consistent, or if different recording settings, equipment, or resolution are used for each DAC.  This would then introduce systematic differences that are unrelated to the DACs themselves making it impossible to isolate the effect of the DACs and consequently would be confounding.

 

But the use of recordings for comparisons does not necessarily add confounding variables, and in actuality it's a very simple process to ensure that confounding variables are not introduced.  In the example video, there does not appear to be any confounding variables introduced.

 

Edited March 14 by POV

[Satanica]

On 13/03/2025 at 7:16 PM, brodricj said:

Blind A/B comparison testing of anything in audio is a pointless exercise.

I guess it's pointless to those who choose to ignore them and the results. But that's a choice, not a fact, not information.

[brodricj]

The result of a blind test is always the same. Have 10 people in a room with 10 different opinions and they are all correct.

[aussievintage]

1 hour ago, brodricj said:

The result of a blind test is always the same. Have 10 people in a room with 10 different opinions and they are all correct.

 

The target outcome of a test is NOT opinion.  That's the point.

[aussievintage]

Why do I get the feeling people are jumping in for the same old argument, perhaps without even bothering to understand, or even watch, the video? 

[brodricj]

1 minute ago, aussievintage said:

The target outcome of a test is NOT opinion.  That's the point.

Opinion is the only thing that matters.

[aussievintage]

NM

Edited March 14 by aussievintage

[Volunteer]

 

 

One or two posters are attempting to derail discussion. Think carefully before continuing.

 

As a helpful indicator, look at the hierarchy below. If you are dragging the conversation into the contradiction layer, then you should probably disengage and move on as per website guideline advice (hint: you're not adding anything).

 

[bob_m_54]

So, did anyone actually try to pick how many times the audio chain was switched. I did, and didn't have a hope of even guessing it.

[aussievintage]

22 minutes ago, bob_m_54 said:

So, did anyone actually try to pick how many times the audio chain was switched. I did, and didn't have a hope of even guessing it.

 

I did of course.   I listened very carefully and could not pick anything for certain.  

[aussievintage]

 

When the results are released,   if his assertions are correct, the test should show that, for anyone who actually counted a number of switches, then the numbers will be fairly random.  

[sir sanders zingmore]

1 hour ago, aussievintage said:

 

I did of course.   I listened very carefully and could not pick anything for certain.  

Same here. 
 

As an aside I do wish the presenter didn’t feel the need to shout quite so much 

[Martykt]

10 hours ago, aussievintage said:

Why do I get the feeling people are jumping in for the same old argument, perhaps without even bothering to understand, or even watch, the video? 

His click-baity title for a start... 

 

A minute in to the video and it's already very clear what the findings are going to be...

 

I like how his first £40,000 example of a "snake-oil" dac doesn't even use off the shelf dac chips, instead using field programmable gate arrays (dcs ring technology) !!

 

He also ignores that the output stage in a dac and also the particular type of filter used will contribute to how a particular DAC sounds.

Not to mention some DACs that use off the shelf chips also use multiple dac chips and do things like upscaling etc.

 

The more expensive DACs quite often use other technology like R2R or ladder technology too which don't use an off the shelf chip either...

[Martykt]

8:30 in and he does talk about R2R and DACs that use industrial chips saying that they will sound inferior.

 

Is it me or is he then contradicting himself by saying you won't be able to hear a difference between two different DACs that have been volume matched?

 

back to video.

[Wagon74]

2 hours ago, Martykt said:

His click-baity title for a start... 

 

A minute in to the video and it's already very clear what the findings are going to be...

 

I like how his first £40,000 example of a "snake-oil" dac doesn't even use off the shelf dac chips, instead using field programmable gate arrays (dcs ring technology) !!

 

He also ignores that the output stage in a dac and also the particular type of filter used will contribute to how a particular DAC sounds.

Not to mention some DACs that use off the shelf chips also use multiple dac chips and do things like upscaling etc.

 

The more expensive DACs quite often use other technology like R2R or ladder technology too which don't use an off the shelf chip either...

Completely agree. Watching the video, the presenter engages in a circular argument. It asserts all DACs essentially at their core is just the same DA chip and any contrary argument is easily explained by hype, marketing, bias, or subjective psychological factors. I do think it is a video to justify the click-bait title. I disagree with fundamentally nearly every assertion made, to the point I don't know where to begin?

 

If we take a step back and applied the same review style to say a motor journalist. Only in this case the assertion is all high priced cars are essentially the same and the reviewer has never driven any of the high end cars in question, since doing so will not be objective (remember - hype, marketing, bias, and psychological factors). So instead we will all use his simulation driving program and you tell him how many times he swapped cars during the simulation?

 

 

 

 

 

 

[aussievintage]

8 hours ago, Wagon74 said:

Completely agree. Watching the video, the presenter engages in a circular argument. It asserts all DACs essentially at their core is just the same DA chip and any contrary argument is easily explained by hype, marketing, bias, or subjective psychological factors. I do think it is a video to justify the click-bait title. I disagree with fundamentally nearly every assertion made, to the point I don't know where to begin?

 

If we take a step back and applied the same review style to say a motor journalist. Only in this case the assertion is all high priced cars are essentially the same and the reviewer has never driven any of the high end cars in question, since doing so will not be objective (remember - hype, marketing, bias, and psychological factors). So instead we will all use his simulation driving program and you tell him how many times he swapped cars during the simulation?

 

 

I also agree with the criticisms of his style and the title, but to use an old saying, do not throw the baby out with the bathwater.   

 

Analogies are always difficult, and I am sorry, but that one is way off the mark.  He is not using a simulation and is talking about real DAC chips that are indeed used in many devices.  His main point is that they all strive to be transparent (or should), so by definition, they should sound the same, and if they don't, in any way, that is an error in their performance.  If they are transparent, you should not hear any difference, and he gives you a test to prove him wrong if you can.  Hopefully he is being honest about it, and will reveal the results soon.

[Addicted to music]

12 hours ago, Martykt said:

His click-baity title for a start... 

 

A minute in to the video and it's already very clear what the findings are going to be...

 

I like how his first £40,000 example of a "snake-oil" dac doesn't even use off the shelf dac chips, instead using field programmable gate arrays (dcs ring technology) !!

 

He also ignores that the output stage in a dac and also the particular type of filter used will contribute to how a particular DAC sounds.

Not to mention some DACs that use off the shelf chips also use multiple dac chips and do things like upscaling etc.

 

The more expensive DACs quite often use other technology like R2R or ladder technology too which don't use an off the shelf chip either...

 

There are lots of FPGA on the market,  regardless of price they dont performed on the test bench as the off the shelf standards.  But that doesnt say its not worth the pennies.   I heard that DCS 4 stack on D'Agonstino Electronics with Wilson Reference speakers from a Standard Apple Airbook as a source......absolute divine,  one of the audiance in the room wanted the files i had so i gave it to him.... I think there was approx $450K in that room alone.   You had to be there to hear it,  and then there was the TAD room with Andrew Jones...I  had to evaluate what i had.  One of the best highlights of going to a HIFI show that i wont forget! especially when i got back into the hobby.  But obviously it wont please everyone as i heard comments from the audiance being critical, that said go back and listen to your own rig and do a compare.

 

12 hours ago, Martykt said:

8:30 in and he does talk about R2R and DACs that use industrial chips saying that they will sound inferior.

 

 

 

Id be interested in whch ones he calls out as inferior!   lots of them out there and i can tell you hes right,  some manufacturers and i wont call them out have no idea how to do SQ.  Youd better off looking at something measured by Amir at a fraction of the cost. 

 

 

43 minutes ago, aussievintage said:

 

 

If they are transparent, you should not hear any difference, and he gives you a test to prove him wrong if you can.  Hopefully he is being honest about it, and will reveal the results soon.

 

This,  in my opinion in audio, is that well engineered products with valves being an acception,  should not take 100s of hours to get there, if the manufacturer ever saids this, look else where.  they sould be transparent and measure the same whether its switched on or been on for 100s of hours, the performance should not change.

[Martykt]

21 minutes ago, Addicted to music said:

There are lots of FPGA on the market,  regardless of price they dont performed on the test bench as the off the shelf standards.  But that doesnt say its not worth the pennies.   I heard that DCS 4 stack on D'Agonstino Electronics with Wilson Reference speakers from a Standard Apple Airbook as a source......absolute divine,  one of the audiance in the room wanted the files i had so i gave it to him.... I think there was approx $450K in that room alone.   You had to be there to hear it,  and then there was the TAD room with Andrew Jones...I  had to evaluate what i had. 

FPGA are not special in themselves, it's more about how you utilise them.

Absolutely divine correlates with my experience with DCS DACs, extremely well engineered and sound exceptional.

 

30 minutes ago, Addicted to music said:

Id be interested in whch ones he calls out as inferior!   lots of them out there and i can tell you hes right,  some manufacturers and i wont call them out have no idea how to do SQ. 

R2R is quite a fiddly technology where it requires lots of very precise resistors.

This makes it expensive to get right.

Some of the best DACs in the world use R2R or ladder technology.

Done poorly however...

 

43 minutes ago, Addicted to music said:

Youd better off looking at something measured by Amir at a fraction of the cost. 

Now now don't bring Amir into this...

[Addicted to music]

9 hours ago, Martykt said:

 

R2R is quite a fiddly technology where it requires lots of very precise resistors.

This makes it expensive to get right.

Some of the best DACs in the world use R2R or ladder technology.

 

 

Outside the scope of this thread but: the marketing bangs on about resistor accuracy where today we are able to get precision resistors as accurate as 0.02%,  you could not get that accuracy back in the 80s or 90s.    But what they don’t tell you is as temperature deviates from room temperature the value can varied to almost 2%.   Different types (materials) of resistors will behave differently to there values,  this doesn’t even bring into thermal noise into the equation where the resistance value contributes to the noise that’s inherent to all resistors, metal film being the lowest of them all.   

[andyr]

52 minutes ago, Addicted to music said:

 

Outside the scope of this thread but: the marketing bangs on about resistor accuracy where today we are able to get precision resistors as accurate as 0.02%,  you could not get that accuracy back in the 80s or 90s.    But what they don’t tell you is as temperature deviates from room temperature the value can varied to almost 2%.   Different types (materials) of resistors will behave differently to there values,  this doesn’t even bring into thermal noise into the equation where the resistance value contributes to the noise that’s inherent to all resistors, metal film being the lowest of them all.   

 

So your view, Peter, is that R2R DACs... are a piece of Schiit?  

 

[Almaz]

I agree with the presenter in as far as his point that if two DACs are made using the same chip then theoretically they should sound the same regardless of how much they are worth. The difference is in the circuitry used around the chip. It is my opinion that like amplifiers, CD players, and DACs, each have their own sound signature. Which is to do with the circuitry and the quality of the components used to make that piece of equipment. So wilst I struggle to get my head around that a $40,000 DAC is gonna to give you great value for money. I do think that a well constructed DAC using quality components and good circuitry will always sound better than something built on the cheap. Not to mention valves V solid state construction.

I did however enjoy the video.

Edited March 16 by Almaz

[aussievintage]

9 hours ago, Almaz said:

I do think that a well constructed DAC using quality components and good circuitry will always sound better than something built on the cheap.

 

The thing that seems to cause people to struggle the most is the quality of some of the Asian products.  Although being sold cheaply, a lot of them are turning out to be of excellent quality, and hence can challenge many of the much more expensive options.

 

9 hours ago, Almaz said:

if two DACs are made using the same chip then theoretically they should sound the same

 

I think he is even going further and proposing that even with a different chip, if a DAC is made with the aim of sounding transparent, then it should also sound the same as all others with the same target.

 

 

[aussievintage]

11 hours ago, Addicted to music said:

 But what they don’t tell you is as temperature deviates from room temperature the value can varied to almost 2%.  

 

One solution could be what we used to call an "oven" and used for crystal oscillators.  Put the resistor network in an enclosed space that is heated above ambient sufficiently, and/or controlled so that ambient temperature changes have negligible affect.

 

First DAC I ever built used a resistor network.  Kinda nice to see old designs persisting  

[Keith_W]

The DAC market (and also amplifiers and other electronics) is split into two. 

 

The first market attempts to reproduce the original signal with no added sound or character of its own - "high fidelity" means it is true to the original signal. Modern DAC's are able to do this to below audible limits and without using exotic technology. 

 

The second market is to create "interesting" sound. These DAC's typically measure poorly because of all the extra noise and distortion. However, some people argue that they sound better. 

 

I don't disagree that a bit of noise and distortion can sound better. It is well known in photography that adding noise can make images look sharper and more appealing. Compare these two images: 

 

 

One of them has noise added, the other is the original image. I won't tell you which is which - study the images (in particular, look at the hair), decide which one looks sharper, then read the article to find out. 

 

To my knowledge a similar study has not been done in audio, but it would go a long way towards explaining why people prefer the sound of turntables for example. That "analog sound" from turntables, R2R has ample noise and distortion and I suspect that is why it sounds more appealing. People say that DAC's with copious distortion (e.g. any DAC with output transformers) sound more "analog", even though they tend not acknowledge the "distortion" part. 

 

IMO there is nothing wrong with a bit of noise and distortion - if it sounds better to you, then it IS better.