Skip to main content
Topic: Multi-bit DACs (Read 18733 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: Multi-bit DACs

Reply #50
What's wrong with an NOS DAC? 

They cost more than better-performing current technology.
They confuse people who don't know better.
They are physically larger.
They take more power.
They have limited availability.
They have limited flexibility.

Basically, they are toys.

If you like to toy around with your audio system and don't care about costs or performance, then they are just what you want.

Some of us march to a different drummer and want low cost, high performance, etc.

Most people want to listen to music, spoken word,  and drama as opposed to play with toy audio systems.

Re: Multi-bit DACs

Reply #51
It is said that NwAvGuy designed the Odac to prove NOS DAC's (and a list of other design features like asynchronous USB claimed to be essential for superior DAC performance) were not needed.  Tom's Hardware did their now famous blind test where a motherboard Realtek 889 sounded as good as the Odac and a $2k Benchmark DAC.  Makes me wonder how much difference there is between my ALC283 and the ALC889 with it's superior published SN ratio.

http://www.tomshardware.com/reviews/high-end-pc-audio,3733.html

There is my favorite test where a consumer DVD player hooked up to a $200 studio amp with a $2 cable was indistinguishable from an $11k stack of gear:

http://matrixhifi.com/ENG_contenedor_ppec.htm

What really blows me away is the A500 input volume controls were shown by some guy to increase distortion from .1% to .4% when turned down, yet these had to be used in the test as DVD players ordinarily do not have variable output.

The moral of the story is electronics usually don't make a difference.  If they do it's something like trying to drive 600 ohm cans to high volume with a mobile phone.  The playback experience is mostly dependent on the recording, loudspeakers/headphones, and room acoustics, in that order.

What I can guaranty you is high end audio gear will look pretty in your listening area.

Re: Multi-bit DACs

Reply #52
There is my favorite test where a consumer DVD player hooked up to a $200 studio amp with a $2 cable was indistinguishable from an $11k stack of gear:

http://matrixhifi.com/ENG_contenedor_ppec.htm
I remember this and I also remembering others attempting to draw the same conclusion.

Suppose I constructed a test in order to determine whether people prefer milk, juice or water.  Let's say the results came back using a very limited number of participants showing a preference of roughly a third for each.  Would that mean that milk, juice and water are indistinguishable from one another?  There is very little daylight between this the above quote.

EDIT:  Let me get a bit closer.  Let's instead say the test was designed to measure a preference just between beer and wine and roughly equal numbers of people chose one, the other or abstained.  Does this show that no difference could be discerned between beer and wine?  No, the numbers demonstrated that the preference was split roughly equally between the three categories, no more, no less.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Multi-bit DACs

Reply #53
It is said that NwAvGuy designed the Odac to prove NOS DAC's (and a list of other design features like asynchronous USB claimed to be essential for superior DAC performance) were not needed.  Tom's Hardware did their now famous blind test where a motherboard Realtek 889 sounded as good as the Odac and a $2k Benchmark DAC.  Makes me wonder how much difference there is between my ALC283 and the ALC889 with it's superior published SN ratio.

The ODAC and Objective2 headphone amp were designed to show you that audiophile gear was too expensive.  Objective was put in the name, because you were supposed to be able to objectively tell that his product was as good as offerings that were significantly better.  The CMOY amp was designed for the very same reason.  Chu Moy was trying to show up the big "audiophile" companies at the time, and released the plans for CMOY for free.  Before you know it, there were CMOY amps popping up all over eBay for around US$50.

But Chu Moy passed away recently and NwAvGuy seems to have fallen off the face of the planet.  And I've seen posts by people that claim were in direct communication with him that said he claimed to have received death threats.  His domain name keeps getting renews, so he's probably still around somewhere.

Re: Multi-bit DACs

Reply #54
Let's instead say the test was designed to measure a preference just between beer and wine and roughly equal numbers of people chose one, the other or abstained.  Does this show that no difference could be discerned between beer and wine?  No, the numbers demonstrated that the preference was split roughly equally between the three categories, no more, no less.

The point being that testing for audible differences and testing for preferences are two different things.

They are related in that you must be able to reliably hear a difference in order for the preference testing to be meaningful.  "I hear no difference but I prefer one or the other..." is illogical.

However, just because the listeners have no preferences, doesn't mean that they hear no differences.

Also, if differences are so small that they are hard to hear reliably, listeners often have no preferences. The alternatives sound almost the same, so why prefer one over the other? YMMV.

Re: Multi-bit DACs

Reply #55
IIRC, there was nothing to randomize in that test.  There was just an A and a B.  The users knew at all times whether they were sampling A or B, they just didn't know which system was which throughout the test.  Then again, if they allowed a sighted audition and there was some audible tell (due to the actual equipment being evaluated, or due to some flaw in the setup or procedure, it doesn't matter), they did know which was which throughout the test.  Even if there absolutely wasn't any audible tell (including the actual equipment), a sighted audition gives the participants an opportunity to get impressions and ascribe differences which they are free to apply to either A or B at any time during the test before they ultimately give their answer.  These impressions are truly imaginary, of course, but it doesn't matter.

"I hear no difference but I prefer one or the other..." is nothing I would expect anyone to say, but if someone can't actually hear a difference but thinks he can then he can also have a preference.  When there is just an A or a B, but never an X (unless you count the one and only time the veil is pulled over the equipment) he can then freely and consistently apply this imagined preference.

There can be differences with preferences, there can be differences without preferences.  Differences can be ascribed whether they are due to something actually audible or simply due to placebo.

"why prefer one over the other?"
Humans are silly?  I don't know.

"YMMV."
That's the safe bet.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Multi-bit DACs

Reply #56
The Tom's Hardware test is problematic as it had only 2 participants.

Re: Multi-bit DACs

Reply #57
3, Getting analogue audio out of a device having only a digital output, like many TV's.  The cheap and popular FiiO Taishan is frequently used for this purpose.

Yup - my stereo is two channel analog. I don't like surround sound, it makes me feel boxed in like I'm in a closet. Well done surround sound not so much, but movies likes to make you feel like everything is happening around you. Something about me psychologically does not like that.

So I don't need an expensive stereo with digital in, but my TV only has digital out. That device was cheap and solved the problem. Never seems to have a problem, I forget it is there, everything sounds good.


Re: Multi-bit DACs

Reply #59
The Tom's Hardware test is problematic as it had only 2 participants.

A test with only two participants and no other evidence can be problematical when the hypothesis that is supported is negative. OTOH, a test with two participants that supports a positive hypothesis can be the start of something big.


http://www.tomshardware.com/reviews/high-end-pc-audio,3733-18.html   which is titled "Page 18:Why Audio Formats Above 16-Bit/44.1 kHz Don't Matter"  looks like a negative hypothesis to me.

Re: Multi-bit DACs

Reply #60
And none of them (a bit guilty here myself) ever give the onboard DAC even a brief chance.

I did try it, but there was really bad load-dependent noise, I could hear high-pitched squealing when I moved windows around. I've used the onboard sound on other systems that have been just fine, but mine is just a hunk-a-junk, unfortunately.

I was perfectly fine with the onboard DAC.  Then I bought a pair of Klipsch Promedia 2.1 speakers for my PC and plugged them in my docking station and got a buzzing noise sometimes.  So I decided to try an external DAC.  I bought a used Schiit Modi for $90 from someone.  I figure I can try and see if it eliminates the buzzing.  And it will also shut up all the placebophiles that keep telling me I don't have a leg to stand on because I am using "on-board."

Am I going to hear a difference with the on-board DAC?  Other than maybe eliminating the occasional buzzing, I would say no.

These Pro Media speakers are disgustingly loud.  I usually keep my PC volume at 100% and adjust the volume on my speakers.  But when I do that, the lowest volume on the speakers is way too loud.  I'm thinking the speaker out jack on my laptop must be amped.

The buzzing in my onboard would occur every time I moved the damn mouse.  It was so bad and irritating to me, I felt the urge to strangle someone over it.  Getting an internal sound card with a built-in headphone amp with a break out box to sit on my desk (although it actually sits on top of my computer) saved me a stay at the local insane asylum.  I'm using a desktop PC by the way.

Instead of paying money to fix someone else's problem, you should just returned the PC and got a working one.

 
SimplePortal 1.0.0 RC1 © 2008-2018