How does firmware affect the sound quality of a DAC?

[esimms86]

The debate going on on the PS Audio forum regarding the sound quality of one firmware version versus another raises athe question in my mind as to how much the DAC firmware affects the sound quality. For instance, when auditioning a given DAC how much of what we're hearing is a matter not purely of the DAC itself, but rather what firmware the manufacturer of the DAC has settled as on the best for his/her hardware product. Would, for example, a Berkeley Audio or Weiss DAC sound subjectively better or worse with tweaks of the software? To give an example, there are posts on the Phasure NOS1 forum about the sound quality of HDCD on the Phasure not equaling the level of the same HDCD files played through the Berkeley or Pacific Microsonics DACs. This is not a knock on the Phasure(which I have to admit I have never heard) as much as it has to do with the lineage of HDCD as Pacific Microsonics creation.

 

Esau

 

[The Computer Audiophile]

Hi Esau - Hopefully some of the component manufacturers who frequent CA can jump in here. I believe a firmware change can do an unlimited number of things to the sound of a DAC. Some firmware updates simple add or subtract features while others may introduce new filtering options.

 

 

 

[PeterSt]

Let alone (indeed) the features, quite some things can be done, also dependend on the hardware setup. For example, the Phasure NOS1 does *not* have any digital filtering section, and thus nothing can be done to that (this is meant to happen outside of the NOS1, likt btw HDCD is). But, if it would have been inside, it would have depend on the chip used whether it was programmable or not. If it is programmable, it is still not "hot" to allow upgrades to it, although for obvious reasons I would have allowed it.

 

What does happen in the NOS1 is the setup of the D/A chip configuration. So, since 8 of those chips are in there, some variations are possible. This is about parellelling them, set them up differentially, or apply a mix of that.

This is not common of course, but it *is* an application of the firmware which not only influences the sound (quality) but in this case also allows for balanced operation vs. more channel output (which latter is again another possibility).

 

Quite some more SQ influencing things are arranged for in the firmware, like the output timing may be an example.

 

I can say quite some more, but I better leave it to this.

Peter (merely a component selector than a manufacturer of that)

 

 

[Sasaki]

I'd say the software importance inside of the DAC has been increased recently. For example, XMOS chip is actually a small computer with dedicated software.

 

I recently had a chance to audition the SQ difference of XMOS software(ie. firmware of class2 USB DAC).

A Japanese manufacture Phasemation developed class2 USB DAC called HD-7A 192. Here is the product page. HD-7A employs XMOS as class2 USB controller.

http://www.phase-tech.com/digital/productspage_HD-7A192.html

 

 

They told me the default control software is included in XMOS however they developed their custom codes for the XMOS to improve sound quality.

They added a A/B switch to test the difference to this developer version of HD-7A so we can hear the difference between "XMOS default" version and their custom version easily.

We used Audirvana on Macbook pro (integer mode engaged) with Helge Lien trio's new  natsukashii album at 192k as a source.

The result was quite obvious. Their custom code is more transparent and more detailed than XMOS default one. Most notable difference is improved frequency response I think. Lower region is more deeper and highs are more sharper, crisper. Especially I felt wood bass sounded more life-like.

They told me they studied the behavior of XMOS and optimized the software code. They also told the other point is to reduce computation task on PC side by the optimization. This lead to more smooth audio stream flow hence total sound quality improvement.