DAC/Pre's Digital attenuation vs. analog


I'm trying to decide whether to buy a W4S DAC2 and use the digital preamp, or to get the DAC1 with a separate component preamp (such as a Bryston BP25). I don't care all that much about the cost; just performance.

What are the advantages/disadvantages of going either way?

(I'm also considering a Benchmark HDR or USB).
Hi Robert, whether going digital or analog with attenuation it's all really about the impedance matching. Just like having to impedance match your speakers with your amps, you also have to do the same with your source and your amp, if you haven't already. Digital control can be attractive because you don't have anything else to buy, it's already remote controlled and there is nothing, no circuitry or components, being added to the signal path that can to any degree degrade audio performance. The downside is that the digital chip used in in your DAC or player should use at least 32-bit proccessing or, better yet, the newer 64-bit processing now becoming available in order to reliably avoid the sonic pitfall of beginning to lose digital resolution as attenuation is reduced below the standard (for Redbook CD) of 16 bits. So, a 32-bit player can offer 16 db of digital attenuation before reaching the 16-bit level, after which, with further reduction, you begin losing digital resolution. That, at some point, begins to darken the sound and small musical details increasingly begin to fall away from the presentation. This effect becomes harder to perceive as the overall volume level drops however and some have even claimed to have not heard a significant adverse impact on the sound as a result (although, I don't find myself among them). But, 16 db should be enough attenuation for most people and purposes and the 48 db offered by 64-bit chips would be enough for anyone. Analog controls have their own advantages and disadvantages. A big advantage is that you are freed from the restriction of having to pay for high processing power in your source. That fact may also give you a little more leeway when it comes to impedance matching your source and amp before you buy. For an example, I have in mind a product like Scott Endler's Stepped Shunt Attenuators (which, as his site will inform you for performance reasons, you can ONLY use by installing them physically at the inputs of your amp(s), without any cabling between them and the amp). These things use surface mount resistors which I can tell you are very transparent indeed (I've been using a pair for years) and there are only 2 resistors in the signal path per channel at a time - one for the ground and one for the value of attenuation that's being selected. There's even a possible way of adding remote control to the 24-step versions available in a DIY kit from Bent Audio, though it's not particularly cheap, itself. The Endler attenuators are about $200 a pair, still about 1/10th the price of a good and sonically equivalent active preamp. In my rig (a relatively well impedance-matched system from source to speakers) I heard little difference in sound quality going back and forth from the Endlers at -4db attenuation (full output) to removing them and listening to the DAC fed into the amp with no attenuation whatsoever (on the same quietly recorded CD's). Only a faint reduction in the silkiness of the highs was the only difference that test has revealed to me, but, again, and I can't stress this enough - impedance matching when going preampless is everything, so take your time with that part and it will reward you well with the rest of it, either way.
Doh!!, just reread your OP and realised you're looking for specific advice, not general, sorry. My vote (of the 3 you mention) might be for the DAC2, provided you feel 16 db of attenuation is ok for you. This would inherently be a bit better (or at leat more direct in terms of transparency) than either the DAC One or the Bryston. However a quality active preamp usually works to 'improve' for lack of a better word the soundstaging of a well suited amp. An amp fed directly may offer a more live sound and an amp with an active pre in the price range of the Bry may have a wider or deeper stage (or both), but the degree to which that's true may vary depending on both the quality of that preamp (again, shouldn't be a real issue for the Bry) and how well it was impedance matched to its amp - like I say it all tends to come back to matching, but amps driven straight can (though not always) have slightly smaller stages (though no less coherent or 3D). If there's any way you can audition your amp temporarily, before you decide on anything, being fed directly from a CD source with 32-bit attenuation, then you can compare to your current pre if your using one, both in and out of the signal path, and get at least some impression of the general differences involved. The Bryston may end up being somewhat different in this regard than with your current pre, but this audition should show you the basics. I have actually valued transparancy more than soundstage in this circumstance with my own gear, but of course it may be the other way around for you.
I have done a few shootouts between setups using volume controls on the DAC, and uning preamps, and finally ended up settling for "DAC direct" (PS audio perfectwave MKII). Prior to the arrival of the MKII, I found a slight advantage in using an very expensive (tube) preamp. Based on my experience, I am quite confident concluding that for a given budget, and with no need for input switching, DAC direct will give you the best results - considering you can buy a better DAC with the money you save on the preamp. If you don't have the time/appetite to circulate different pieces of gear through your system and make the decision based on your own observations, I would have no problem recommending going the DAC2 route. You will also save cabling funds, which can all be applied towards better hardware elsewhere in the chain.
I'm having a hard time agreeing with "32bit offers only 16dB of attenuation before eating into the 16bit resolution of Redbook".
As far as I know, every bit accounts for 6dB, thus 16x6 would mean that a DAC chip working internally on 32 bit will offer 96dB of digital attenuation of a redbook stream before loss of information occurs. And that's pretty much the entire dynamic range of redbook.
Attenuation in the digital domain is no longer the bad thing it used to be, with the advent of these 24-32 bit DAC chips. The myth, though, will continue, I'm sure.
Hi guys. I should have mentioned that my main concern is at low volumes. I was thinking that maybe digital attenuation would be to my dis-advantage. I'll be playing .flac files through foobar2000 exclusively (which has a digital volume control itself). Would it be best to leave the foobar volume control all the way up?
Using digital volume exclusively for volume control will result in a significant reduction in detail and imaging. It's okay to do about -9dB, but not more. This requires that the DAC have some kind of gain control or a good volume of its own.

There are five types of volume control:

1) Digital volume control - reduces resolution as the volume is decreased

2) resistive attenuation - delivers detail, but kills dynamics with most amps

3) active gain control - adds distortion, noise and compression, but has good drive, so dynamics are good.

4) Transformer scaling attenuation - with strong DAC outputs and good transformer linestages, like Music First TVC, this is more transparent and preserves the dynamics. Easy to match to amps

5) D/A reference voltage control - This volume adds no noise, distortion or compression, in fact it actually improves S/N ratio as the volume is decreased. No loss of resolution. The best of all possible options.

Steve N.
Empirical Audio
"5) D/A reference voltage control - This volume adds no noise, distortion or compression, in fact it actually improves S/N ratio as the volume is decreased. No loss of resolution. The best of all possible options."

Thanks Steve. How many companies make something like this and @ what price point?
Robertsong - there are two I believe, and Empirical Audio is one of them. $6K pricepoint for the DAC. Includes preamp, DAC and USB interface.

Steve N.
Empirical Audio
Isn't it fascinating how different well respected digital audio design engineers can express diametrically opposing views on something as basic as loss of resolution in digital volume controls or lack thereof. Engineering is supposed to be an exact science! In the absence of consensus I trust my ears, and ended up ditching a 10k preamp in favor of digital volume control.
04-26-12: Ivan_nosnibor
Hi Robert, whether going digital or analog with attenuation it's all really about the impedance matching.
please explain this statement as I do not understand how digital attenuation & impedance matching of source & power amp are related. Sources are usually very low output impedances digital attenuation or not.

So, a 32-bit player can offer 16 db of digital attenuation before reaching the 16-bit level,....
how did you come up with these numbers of 16dB of attenuation using a 32-b D/A?? Can you share your calculations? Thanks.
I think it's very wrong but it would be good to see your calculations before saying more.
Right now my CDP has a 24-b D/A that provides 50dB of total attenuation where at 50dB of attenuation you get zero output volume. So, how did your 32-b D/A provide only 16dB of attenuation?
Edorr - what kind of digital volume control do you use?

Is it a slider in itunes or Pure Music etc? Do you do any gain control in the DAC?

The devil is in the details.

Steve N.
Empirical Audio
Lots of technical talk (again), but my two cents.... buy the W4S DAC-2 (I owned one), it is a really nice piece for the money.

STOP WORRYING ABOUT ALL THIS DATA and SPECS and tech. discussions about digital volume!!!

And when I sold my DAC-2, I sold it for about $1,200. Not bad spending $1,500 for it, using it for a year or so, then getting back $1,200.
i have the ps audio separates (pwt and pwd).

the pwd has a digital volume control.

i prefer the sound using the digital vol control at 61 percent of max and using more gain on the preamp.

i think the digital volume control is inferior to the analog vol control on my passive and tibe preamps.

I agree, you may get a slightly better results throwing an expensive tube preamp in the mix, but why? Why add that expense, another power cord and IC - why add the additional box?

The volume control on the DAC-2 is good enough for anyone EXCEPT a tech engineer who likes to read books and specs instead of actually sitting down and listening to music.

I am so tired of seeing all this tech talk about what will sound better... what... on paper it looks like it will sound better? How do you know it will sound better to person A, person B or person C, D, E, F..... ? Answer: YOU DON'T. Talk all the tech you want about volume control on DACs... I prefer to try it and determine MYSELF what sounds better.
I use a PC with JRiver 17 (tried to get the JPlay plugin to work - no joy). From what I gather, playing in any WASAPI / ASIO mode the slider does not work.

I control volume digitally in the Trinnov 4 channel processor


This processor sits before my PS audio DAC. I run the dac at 100. I am actually thinking of recalibrating my system, running the PS audio at 80, and doing less digital attenuation in the Trinnov.

By the way, does your DAC have remote volume control? I may decide to give it a whirl, but no remote control would be a dealbreaker.
Reading all the specs and tests in the world (on paper) about volume control will not tell you ANYTHING about how good a DAC will sound. Please stop this nonsense.
Edorr - The Overdive DAC has manual volume. It is more convenient to use an iPad anyway because you will need 2 remotes if the DAC has remove volume control, one for volume and one for music selection.

Does your Trinnov have digital in and digital out?

Steve N.
Empirical Audio
Steve, the Trinnov has digital in/out. It also operates as a crossover. I use the Trinnov DAC for my subs, and the Trinnov digital out into my PS audio DAC.

I do lots of late night listening to music that don't need the subs, so I could play this music in 2.0 mode and use an analog volume control on the DAC if this is indeed better.

If I get your DAC but then have to control volume digitally on the Trinnov on the iPad (which is what I am doing now), I would not have the analog volume control benefit.
Audiofreak32: Why the W4S DAC-2 over the Benchmark HDR? I assume your choice is not based on the volume control. Is it just a personal preference? I've read several threads and it seems to be evenly split over which DAC people here prefer.

Which DAC did you upgrade to?
Not sure what the Benchmark unit costs, from who I have talked to, the W4S DAC-2 beats any of the Benchmark units. I really liked my DAC-2... was connected XLR direct into a BAT VK-55, sounded great. One small box does so much. They can price it that way b/c they do not sell at dealers. And like I mentioned... the resale is great if you decide to try something else. I also tried their big class D mono amps - not the same experience AT ALL. I guess I just prefer the tube sound, but I did not like any of those new amps with the B&O ICE modules... and I tried several.

I was using the DAC-2 off a netbook PC with mainly FLAC and WAX at 24/96 and some 24/192.

I got a great deal on a demo LINN Akurate DS (latest version) and it was quite an upgrade, so I sold my W4S unit.
Yes, just personal preference (although I have not personally owned the Benchmark, I am going off what others have said). I think the DAC-2 is a really great piece of gear. And if it were me, and I was only doing digital files, I would go without a preamp. If you have other sources (CD player, LP), then I suppose you would have to have at least a phono preamp. You could hook a CD player into the DAC-2 but this would make no sense... it would never sound as good as a hi-res file.
"If I get your DAC but then have to control volume digitally on the Trinnov on the iPad (which is what I am doing now), I would not have the analog volume control benefit."

Yes, you could still get the benefit if you adjust the digital volume on the Trinnov to max and then set the manual volume on the Overdrive to a good listening level with your quietest track. Then, when a louder track comes on, just reduce the digital volume on the Trinnov 2-9dB. This will not impact SQ at all. You still have remote control.

Steve N.
hi steve:

the problem is the digital volume control. when you set it to max, the analog vol control on the preamp is providing less gain.

my experience, as stated, is that is preferable to minimize the contribution of "digital" gain in favor of analog gain.
Steve, that sounds like a reasonable approach. One 3 feet trip to the audio rack to set volume on the DAC for each listening session is doable. I will wait for the offramp 5 I ordered to arrive first, but my next project may well be a shootout of your DAC against the Perfectwave MKII.
Mr tennis - The "analog" volume control on the DAC that I am referring to is not analog at all. It does not suffer from the gain issues. This allows the digital volume to be at or close to max, which is optimum.

Volume control using analog "gain" adjustment is one of my least favorite methods. Best to avoid. My own modded ML preamp is a dust-gathering boat anchor now.

Steve N.
Empirical Audio
"my next project may well be a shootout of your DAC against the Perfectwave MKII."

No problem, we can do that. Looking forward to it. It will probably be after Newport Beach show June 1 however.

Steve N.
Empirical Audio
a better test would be a comparison between the original zanden dac steve's latest dac.

i consider the first zanden dac as my favorite dac.
MRtennis - No need to do this. My older discontinued DAC the Spoiler, which is not as good and used Adaptive USB reclocked, was compared to the Zanden5000 here:



Steve N.
Empirical Audio
Audiofreak32 - In stating several times that you recommend the W4SDAC2 - you finally admit that YOU HAVE NEVER EVEN HEARD THE BENCHMARK DAC1!

And, further, you contradict your own statements in which you strongly state that "I am sick and tired of all this talk about what sounds better...I prefer to try it and determine for myself."

Really? I therefore recommend you follow your own advice and listen to the DAC1 in a good system and then decide.

I have heard BOTH and prefer the Benchmark DAC1, as have others.

I'm tired of vocal naysayers who diss something without even hearing it, esp. re the DAC1.

I find it interesting you prefer digital attenuation with its lack of detail over analog attenuation with its high compression. I would guessed the analog attenuation would be preferable. I assume you mean down to just -9db or even lower?

From my own experimenting with my V-DAC II *I think* I prefer cutting back my amp's gain vs. foobar's. It's something I will have to experiment with, but I plan on getting all new equipment in the next few months anyways.

Maybe your Off-Ramp 5 with a benchmark will work for me?

I said that I have been told that the W4S DAC-2 sounds better than the Benchmark units. Correct, I have not owned the Benchmark unit. I am sorry if you own the Benchmark unit... you are missing out.
"I find it interesting you prefer digital attenuation with its lack of detail over analog attenuation with its high compression. I would guessed the analog attenuation would be preferable. I assume you mean down to just -9db or even lower?"

Correct, certainly no more than -12dB.

"Maybe your Off-Ramp 5 with a benchmark will work for me?"

It will certainly improve on the Benchmark. I modded them for almost 10 years, but stopped in 2009. For the same money you could have Metrum Octave. I have a few customers that use this combo with great results.

Steve N.
Empirical Audio
hi steve:

the zanden which was used in the comparison with the spoiler, was the latest zanden dac, not the original. having heard both, i can say that i don't like the latest zanden dac. it doesn't sound like it has a tube in it when mated with the zanden transport.

so, there are two variables here, the version of the sanden dac and the transport. perhaps there is a better transport for the zanden dac.

it is virtually impossible to compare dacs, because of the interaction with the transport. i wasn't sure if the spoiler was mated with the zanden transport.

one is really comparing a dac/transport combination with another. there are 4 variables.

one cannot conclude the superiority of one dac or another, definitively, unless one can mate it with every transport--impossible.

one can, inductively have confidence if one samples many transports, using two dacs. also, is there not the issue of the digital interface , another variable ??

i think it is pointless to compare dacs, unless one can construct a valid experimental design.

by the way, there are problems comparing components, in general, because of interactions with other components. there are many variables and it is almost impossible to isolate the contribution of each one.

all you can do is specify the variables and let your ears decide which you prefer.

certainly there are mismatches between amps and preamps, and i would conjecture that there are mismatches between dacs and transports.

the fact that a company fabricates a dac and transport does not ensure the combination is optimal
Mrtennis - If you read the last 3 or 4 DAC reviews in TAS, you will see that once you have a really low jitter digital source (not a Transport), the differences in DACs are much smaller. This is the only way to compare DACs. Otherwise, you are mostly hearing the DACs ability to reduce jitter or replace it with other jitter as in ASRC. Without this low-jitter source, many older ladder and NOS DACs will sound really bad.

Its a lot like comparing decent turntables using a really cheap cartridge. Yes, the really expensive turntables/tonearms will sound a bit better, but they a will all sound relatively poor due to the cartridge. With a really good cartridge, the differences are much smaller I think. The cartridge is the most important part, as the Master Clock is in digital.

Steve N.
Empirical Audio