Optimizing Digital Volume Control


Hi,

First I would like to state that I am relatively new to audiophilia (sounds like a disease), however I have a relatively good engineering background.

Presently I am using a traditional CD->DAC->pre->power->speakers setup. I am using a Hegel HD20 DAC that has a digital volume control feature. I recently noticed that without the pre (ARC LS2) in the chain, using the DAC for volume control, I get better results.

Since the DAC is 24b, having a -140db noise floor (close to 144db theoretical limit) this makes sense to me. Since a CD, having 16bit resolution, supports a theoretical maximum of 96db (practical implementation are below 90db). So s properly designed digital attenuation of over 50db (probably in this case over 60db) should not degrade the sound.

Now to my question, assuming a computer as the source, followed by a 24b DAC that doesn't support volume control, one can in theory achieve the same results if the computer converts the 16b original data to 24b and then apply digital volume control. In this case the computer should output a 24b signal to the DAC.

Does anyone know if this is something that a JRiver or Foobar solution is capable of doing? or in general, does anyone know how volume control would work using JRiver or Foobar?

Thank you in advance for your attention.
oferi

Showing 4 responses by oferi

Thank you. Just found the write up on this topic. Seems like they know what they are doing :-)

BTW, this leads me to believe that a preamp is practically no longer required. The best reference preamps at best can match a good 2K DAC. So most likely having a preamp in the chain simply degrades the result (unless of course you view the preamp as a filter that colors the sound to ones liking)
Let me qualify my comments by stating that my conclusions are a combination of about 6 months of listening and many years of engineering studies. I listen mostly to classical music, so this may be coloring my assessment. My speakers are Magnepan 3.6 and Willson CUB2s (I alternate between them).

During my brief journey so far I had the opportunity to work directly from the CD through an ARC LS2, Audible Illusions M3A, McCormak LD-2, and through the Hegel HD20 DAC (for the past 3 months). Today I can say with confidence that with the equipment I use, without the preamp, using digital volume control provides better result to my ears. I started with the expectation that a preamp is a good idea (based on the majority opinion), so my prejudice was to confirm that belief. But as I went through my music library I couldn’t ignore the fact that the sound without a preamp is better.
It so happened that this conclusion is aligned with sound engineering principles. The modern DACs are incredibly good. Their dynamic range and noise figures are amazing (to an engineer). This is a result of advances in semiconductor technology, which makes it practical to build such works of art at a low price. For example, Hegel just came out with a newer version of the HD20 that utilize a 32b DAC with a noise figure of -145db. With such a device I doubt an analog preamp (at least at the same price range) can compete, even up to a 50db volume range.

BTW, I contacted Hegel regarding this observation and Anders wrote me back the following: “At the top 50% it (the HD20) still outperforms our P4Amk2, that used to be our 4600 USD pre amplifer”.
Regarding preamp drive or DAC drive, the argument that the upstream component driving the amp should have a "low impedance with a lot of drive capability" while technically correct, in practice should be a non issue for any reasonable designed component. The input impedance of most power amps is 50K or higher (I know some are down to 10K), and with current state of electronics it is trivial to design an output stage with 200 ohm or less impedance. In fact if you look at reference preamps schematics they typically have a resistor in series at the output of roughly 200 ohm.I am guessing this is for protection against shorts. Regardless this tells me that reference level designers are not too concerned with having a very low output impedance. Otherwise removing that resistor would be the first thing an audiophile should do to get the best sound. I know that some of you may say that this has to do with impedance matching. Unfortunately physics tells us that for up to 20KHz bandwidth with short distances there is practically no impedance matching involved (unlike in the case of microwave or RF where the wavelength are on the order of the interconnect length or smaller). I believe the critical functionality for a preamp is minimizing noise and distortion that may be introduced due to attenuation or amplification, while presenting a high impedance to the source and reasonably low impedance to the load. I suspect that this is what differentiate a good (line stage) preamp from an average one.
Audioengr, I agree that the input impedance of a power amp is probably slightly frequency dependent, and hopefully not amplitude dependent. So I agree that a low impedance source is a good idea. My point is only in regards to how low is good enough. A 200 ohm source into a typical amp would be plenty good (as evident by the design of reference preamps having a 200 ohm or bigger series resistor at the output). On the other hand a passive (pot) would probably degrade the quality if its impedance is on the order of the amp input impedance.

BTW, a passive volume control with a 20K ohm into a 100K good amp seems to do an excellent job (but not as good as the best preamps).