DAC blind tests? EVER?

[Dunring]

This describes my experience as well. Bought a lot of dacs over the last three months. Doing quick A/B tests most of them sound slightly different, like changing filters. But the sound never seems better or worse. In actual times of listening and enjoying music there is no way I could pick one from another.

These days I just look for a DAC with 3 prong ground and balanced XLR just to weed out as much possible interference or ground issues as possible. Just got in a Topping E70 velvet and it's got it all for connectivity and nice display. That and the D30Pro are my favorites.

 

[jooc]

This is a famous test from Tom's Hardware I think everyone should reference from time to time:

What Does It Take To Turn The PC Into A Hi-Fi Audio Platform?

Most hi-fi audio is stored in digital form. With advancements in lossless compression, bit-perfect ripping/streaming, HD audio formats, multi-terabyte storage, and PC-friendly DACs, has the PC earned a place among high-end audio gear? At what price point?

www.tomshardware.com

I don't even want to summarize because it's fun to just read through, but very basically a test subject who owned a 70K stereo system couldn't discern between on-board PC DAC and a super expensive DAC if volume was precisely equalized.

Here are the test subjects' usual stuff if interested (I found it interesting) :

View attachment 340397

Here's a great quick reference on why it's necessary to volume match sources (page 8 of same article) :

What Does It Take To Turn The PC Into A Hi-Fi Audio Platform?

Most hi-fi audio is stored in digital form. With advancements in lossless compression, bit-perfect ripping/streaming, HD audio formats, multi-terabyte storage, and PC-friendly DACs, has the PC earned a place among high-end audio gear? At what price point?

www.tomshardware.com

.

"...We've never seen a properly-conducted blind test where individuals could reliably tell between 24- and 16-bit audio. "

I wonder how this statement plays in this forum, do most people agree or no?

 

[BobbyTimmons]

The issue in that is that any decent DAC uncolored from maybe $100 and up sounds the same as any other decent uncolored DAC.

DACs should all sound the same from $20 up these days.

 

[jooc]

DACs should all sound the same from $20 up these days.

This may very well be the case, but for whatever reason in most Windows PCs still to this day, the onboard DAC --> aux out --> any powered speakers or amp sounds horrific.

If you were to pull the DAC chipset in the PC out and put some modestly nice components around it it might perform OK, but whatever these PC builders are doing it destroys the sound in any non-Apple computer I've ever tested. I'm not talking about measurements, I mean it's so bad that after volume matching you can audibly hear the difference between onboard DAC and a very cheap ($20) external DAC. In that Tom's guide test I note that they were using a different DAC daughter board than I've ever listened to, but they may have got lucky on what they happened to choose to test.

I noticed this back in 2014 and it doesn't seem as though much progress has been made, it's like they have an attitude of 'don't bother, if the laptop buyer cares about using anything but built-in speakers on this thing they'll buy external components anyway," which is probably true.

 

[BobbyTimmons]

This may very well be the case, but for whatever reason in most Windows PCs still to this day, the onboard DAC --> aux out --> any powered speakers or amp sounds horrific.

If you were to pull the DAC chipset in the PC out and put some modestly nice components around it it might perform OK, but whatever these PC builders are doing it destroys the sound in any non-Apple computer I've ever tested. I'm not talking about measurements, I mean it's so bad that after volume matching you can audibly hear the difference between onboard DAC and a very cheap ($20) external DAC. In that Tom's guide test I note that they were using a different DAC daughter board than I've ever listened to, but they may have got lucky on what they happened to choose to test.

I noticed this back in 2014 and it doesn't seem as though much progress has been made, it's like they have an attitude of 'don't bother, if the laptop buyer cares about using anything but built-in speakers on this thing they'll buy external components anyway," which is probably true.

Ok external DACs will all sound the same from around $20 up.

 

[Killingbeans]

"...We've never seen a properly-conducted blind test where individuals could reliably tell between 24- and 16-bit audio. "

I wonder how this statement plays in this forum, do most people agree or no?

Yes, under normal listening conditions. If you crank the gain like crazy and do instantaneous switching with short samples you'd might be able to detect a tell.

Doesn't really count as a "difference" in my book.

 

[BDWoody]

"...We've never seen a properly-conducted blind test where individuals could reliably tell between 24- and 16-bit audio. "

I wonder how this statement plays in this forum, do most people agree or no?

Listening to normal music in a normal room, yes.
It might be possible with extreme gain riding to bring the otherwise inaudible noise up to where a difference could be heard, but effectively no one is going to hear a difference otherwise.

 

[board]

The Naim definitely doesn't have any components on the PCB indicating that reverb has been added. It would be a PITA to implement without digitizing the signal first and converting it back afterwards. You could do it with a small spring tank, but that wouldn't exactly be a subtle solution

It was of course guesswork on what it might be based on just listening, but reading your comments and those of others makes me thing I simply guessed wrong.
Although this goes even more into anecdotes, then I did audition another amp once where it seemed even more like a bit of reverb was added.
But again, it's guess-work, and if adding reverb is as difficult as you describe then that most likely wasn't what I heard.
It sounds like adding reverb to a DAC is easier then.

 

[Frank2]

Good discussion. What can create an illusion of wider and deeper soundstage?

A small delay of a few ms between the left and right channel.
It is called the Haas Effect:

What Is the Haas Effect and How to Use It

The Haas effect can give your mix a perceived increase in level and a sense of spaciousness. Learn more about what the Haas effect is and how to use it in your productions, as well as why you might not want to.

www.izotope.com

 

[ahofer]

"...We've never seen a properly-conducted blind test where individuals could reliably tell between 24- and 16-bit audio. "

I wonder how this statement plays in this forum, do most people agree or no?

My understanding is that this is true, but just one or two might have shown some trained listeners doing a creditable job by looking for very specific tells. May have been gain-riding accounting for the tests not being “properly done”.

 

[board]

Easy way to find out. Record the electrical signal at the speaker terminals, then compare the recordings to see if you can hear the same differences in an ears-only comparison. @pkane's software or foobar will let you do an ABX comparison.

I won't be able to record the electrical signals at the speaker terminals, as I'm simply not set up for that.

 

[SIY]

I won't be able to record the electrical signals at the speaker terminals, as I'm simply not set up for that.

You don't have a sound card or audio interface and four resistors?

 

[board]

A small delay of a few ms between the left and right channel.
It is called the Haas Effect:

What Is the Haas Effect and How to Use It

The Haas effect can give your mix a perceived increase in level and a sense of spaciousness. Learn more about what the Haas effect is and how to use it in your productions, as well as why you might not want to.

www.izotope.com

Do you, or anyone else here, by any chance know if this delay could be part of the explanation of the appeal of vinyl?
Often it's been claimed that vinyl has a deeper and wider soundstage, etc.
I've usually put it down to frequency response (or EQ), which actually can affect the soundstage, perception of depth, etc., especially if you lower the harshness region.

 

[board]

You don't have a sound card or audio interface and four resistors?

I have a Focusrite 2i2, but no resistors.
In any case, I only have one amp now - the Nord I sent back; the Naim I sold.

 

[SIY]

Often it's been claimed that vinyl has a deeper and wider soundstage, etc.

People claim all sorts of things when they don't use basic controls in listening tests.

Re: vinyl, Mike Uwins pretty much answered that question with well-designed tests.

 

[board]

People claim all sorts of things when they don't use basic controls in listening tests.

Re: vinyl, Mike Uwins pretty much answered that question with well-designed tests.

I will print and read the paper you linked to, as it seems very interesting, but I won't be able to do it for a while (and it's also 26 pages). I did a search (for "phase", "delay", and "timing") and couldn't find what I was asking about - specifically if a bit of delay in one channel, causing a change in how we perceive soundstage, is something that can be found on vinyl.
Can you point me to the section where he mentions this?

 

[SIY]

I will print and read the paper you linked to, as it seems very interesting, but I won't be able to do it for a while (and it's also 26 pages). I did a search (for "phase", "delay", and "timing") and couldn't find what I was asking about - specifically if a bit of delay in one channel, causing a change in how we perceive soundstage, is something that can be found on vinyl.
Can you point me to the section where he mentions this?

I think you may want to read through the paper before starting to wave hands about a mechanism that explains imaginary differences. His work pretty much demolished the idea that preference for vinyl is based on actual sonics.

 

[oleg87]

"...We've never seen a properly-conducted blind test where individuals could reliably tell between 24- and 16-bit audio. "

I wonder how this statement plays in this forum, do most people agree or no?

There is no mystery about the audible effect of bit depth. You could easily contrive a test that would make 16bit quantization noise audible, but with real-world content anyone would willingly subject themselves to? I haven't personally encountered such material.

 

[board]

I think you may want to read through the paper before starting to wave hands about a mechanism that explains imaginary differences. His work pretty much demolished the idea that preference for vinyl is based on actual sonics.

It seems to me like you're making a lot of assumptions about what my opinion/position is.
I asked a simple question. That was all.

 

[SIY]

I asked a simple question.

You asked a question with a... questionable premise. Examine that premise first.