Frugal DACs...

[Occam]

Hmmm... bargain DACs.
......
DO try for a double blind test before spending the money, though....

DACMan

DACMan - I assume you're being facetious in suggesting DBT when a hobbyist is choosing a DAC, because frankly, its simply too daunting a task. Lets assume the following -
You've actually got a computer controlled switching unit/preamp which can switch (or not) with the same noise (or lack thereof) between those 2 DACs, and make a record of the listener's switch initiation and the resulting state, A or B DAC in circuit. We also need some mechanism to record the person under test response to state changes.

But Wait, there's more!!!

1. Unless the analog output levels are the same we need level matching between on a per input, as different volumes invalidate comparisons. Yes, there are preamps with this facility, I just don't happen to have one.

2. We must isolate the digital input signals from each other so that reflections don't cross contaminate the results. One could use separtate digital transceivers/logic gates (multiple outputs of the same signal from a computer card) or transformers, An AES to dual 75 ohm winding transformer, with Micro-Circuits(tm) attenuators at the spdif terminus would adjust levels and minimize reflections/cross contamination.
Same goes for a USB source.

3. Unless at least one of those Dac DUTs are battery powered, we need to isolate the AC mains power between the units, to minimize 'backwash' cross contamination.

I've previously done single blind testing to evaluate power cords, analog and digital ICs connected to 2 exactly the same DACs (and then swapped DACs to re-verify conclusions). It is a non-trivial task. It is absurd for you to seriously suggest a DBT. Whenever I hear someone suggest a DBT, I can be reasonably certain, (witha few rare exceptions) they've not done a valid DBT themselves. YMMV

FWIW,
Paul

 

[opus111]

DBT is pointless because all normal listening is done sighted - i.e. we all know what we are listening to. If you do try a DBT, which if done properly will eliminate the effect of placebo then what you hear when listening sighted will differ because under DB conditions placebo contributes nothing. But in day to day listening, placebo is always present hence its meaningless to exclude it when testing.

 

[jkeny]

Totally agree Paul & furthermore the additional devices needed to achieve level matching, A/B instant switching (because audio memory is considered so poor) have to be tested to be NOT introducing any additional variables. Not to mention controlling for other variables that are part of any rigorous perceptual blind testing defined in Mushra "MUltiple Stimuli with Hidden Reference and Anchor" (if the differences are considered obvious) or probably even more appropriate - BS.1116 : "Methods for the subjective assessment of small impairments in audio systems including multichannel sound systems".

So yes agree with Opus too, listen to the DACs in your system over an extended period of time i.e live with them for a while!

 

[Gregadd]

DBT is pointless because all normal listening is done sighted - i.e. we all know what we are listening to. If you do try a DBT, which if done properly will eliminate the effect of placebo then what you hear when listening sighted will differ because under DB conditions placebo contributes nothing. But in day to day listening, placebo is always present hence its meaningless to exclude it when testing.

...and for what we are listening. Most important we release all the variables we were trying to hold constant

 

[Occam]

Actually, I find blind testing quite useful, and within my very narrow needs (evaluating cables) not all that difficult. But my use is quite specific, I'm not evaluating different dacs, but rather the same DACs where the DUT is a single (or pair of) cable(s). Nor is the goal to prove anything, but rather to guide my own development process. And some of those metrics while the outcome of a purely subjective experience are quite numeric, and not subject aural memory constraints, specifically, pointing.
The late Bob Crump, posited that the optimal cable direction, giving the most accurate tonality, dynamics,etc... is the orientation that gives the most accurate height information. The engineering behind not only portraying lateral information, points on an arbitrary 'compass', but also height and depth information via manipulation of phase information and delays is reasonably well known. The 'up and over' track on the 2nd Stereophile Test CD is a good example. One can analyze this sound staging, reported numerically, by simply pointing and noting distance. [inclination and distance are 'duals' of height and depth]
While I'm an agnostic on most DBT [not that I don't believe in them, but rather that I find them well nigh impossible to do correctly.] there are subjective measures that are numeric, and repeatable, nor subject to rapidly decaying aural memory and that is the ability of a component to subjectively allow the objective reporting of polar coordinates as engineered into recordings. YMMV. Sorry for the thread drift.

Regards,
Paul

 

[Mortsnets]

-- Mortsnets, Welcome! :b ...And Happy New Year!

Bob

Thanks, looks like a great board with a lot of expertise. Bob Walters of the Bay Area Audiophile Society recommended it. I really like that you guys have a place for the frugal end of things.
Craig

 

[NorthStar]

Thanks, looks like a great board with a lot of expertise. Bob Walters of the Bay Area Audiophile Society recommended it. I really like that you guys have a place for the frugal end of things.
Craig

---- Hey Craig, frugality has certainly its very important place among the Best. :b

 

[opus111]

Thanks, looks like a great board with a lot of expertise. Bob Walters of the Bay Area Audiophile Society recommended it. I really like that you guys have a place for the frugal end of things.

That's cool - warm welcome from me - when my dream comes to fruition, the 'frugal' end will be mainstream and newbies will be more likely to say 'I love that you have a place for the esoterica'

 

[Mortsnets]

The Bay Area Audiophile Society recently had a very well attended event comparing budget USB DACs. I wasn't able to attend but via email Jason reported that the Halide HD was very impressive especially considering the price includes built-in cables.

The DACs were:
Audioquest Dragonfly ($250)
iFi iDAC ($300) with iUSB ($200) power supply
Schiit Bifrost ($450 with USB)
Halide DAC HD ($495)

 

[mep]

The Bay Area Audiophile Society recently had a very well attended event comparing budget USB DACs. I wasn't able to attend but via email Jason reported that the Halide HD was very impressive especially considering the price includes built-in cables.

The DACs were:
Audioquest Dragonfly ($250)
iFi iDAC ($300) with iUSB ($200) power supply
Schiit Bifrost ($450 with USB)
Halide DAC HD ($495)

So the most expensive cheap DAC won?

 

[Mortsnets]

Well the Halide includes a Wireworld Starlight ($90) usb cable built-in. Also they didn't seem to like the iFi iDAC without the optional power supply so you could say the second cheapest one won.

 

[opus111]

Audioquest Dragonfly ($250)
iFi iDAC ($300) with iUSB ($200) power supply
Schiit Bifrost ($450 with USB)
Halide DAC HD ($495)

All those DACs are using CMOS S-D single chips with voltage output. Which means on-board CMOS opamps - oodles of noise modulation from both the S-D processing and the opamps. I think the first two use ESS9023, the Schitt uses AKM4399 and the Halide, WM8716.

 

[opus111]

To add a bit more detail on noise modulation, here's a snippet from the AKM4399 datasheet to illustrate :



I'm able to decode these figures to show noise modulation because they also include a helpful FFT plot right at the back of the datasheet :



The FFT is showing the THD for a full scale 1kHz sine stimulus. The highest harmonic shown is -109dB for the 3rd. Armed with this information go back to the first snippet and note the THD+N for full-scale 1kHz sine stimulus is shown as -105dB. This tells me that the THD+N figure is not determined by the THD component, rather the 'N' component. With full-scale sine stimulus, the DAC is putting out -105dB white noise.

The noise is not constant notice because the -60dBFS stimulus figure shows a THD+N of -60dB. Therefore the noise has fallen 15dB in level to -120dB.

Note 5 tells us the figures were obtained in averaging mode on the AP2. Thus its safe to assume the peak noise modulation is considerably worse than 15dB.

 

[NorthStar]

-- I really like the AKM DACs, in general; very articulate, detailed, and with tight deep bass.
Anthem uses them in their high-end pre/pros. ...So are other high-end audio manufacturers.

Wolfson DACs are generally smooth. ...Good match for bright speakers, I think.

 

[opus111]

I guess Bob you've never heard a NOS DAC with low noise modulation? When you do, there's no going back as these parts all sound desperately dynamics-challenged

 

[Mortsnets]

I guess Bob you've never heard a NOS DAC with low noise modulation? When you do, there's no going back as these parts all sound desperately dynamics-challenged

So you prefer the TDA1543 NOS for inexpensive DACs to all of these? How about for frugal USB conversion?

 

[NorthStar]

I guess Bob you've never heard a NOS DAC with low noise modulation? When you do, there's no going back as these parts all sound desperately dynamics-challenged

Hey, I am very restricted in my 'physical' audio evolution. ...My analog r.a.d.i.o. tuners use no DACs at all. :b ...Neither my turntables.

 

[opus111]

Yes, when the comparison is made on SQ rather than THD and SNR, the TDA1543 easily wins over these.

For frugal USB-SPDIF, I use one of these http://www.ebay.com/itm/HA-INFO-U2-..._Audio_Amplifiers_Preamps&hash=item3a7d0d4f7c

There are plenty of other options though - best go for a device using the Tenor chip I reckon.

 

[opus111]

...Neither my turntables.

Curious here - you don't notice your turntables sounding much more dynamic than Wolfson or AKM DAC chips?

 

[NorthStar]

-- Richard, are you familiar? ....

​

=> http://www.positive-feedback.com/Issue38/mhdt_paradisea.htm