Is there any such thing as a bad sounding DAC these days?


I think the problem of DAC for quality audio has been pretty much universally solved.  Not to say all DACs are equal, they aren’t, but do any that really matter these days not sound “good”?

128x128mapman

My primary current setup is Roon->Cambridge Audio mxn10 streamer/DAC analog out-> Schitt Freya balanced XLR out -> Class D Audio Premium GaN 6.5 Balanced 2-Channel Amplifier 200W RMS into 8 OHM -> KEF ls 50 meta + powered sub or Ohm Walsh F5 series 3.

I listen to the KEFs nearfield in a small 12X12 room and the Ohms non-nearfield in a much larger L shaped room.

I couldn’t ask for anything more regarding sound quality and performance in either listening scenario. The nearfield setup with the KEFs in particular allows one to really listen in detail into the recording. The detail, soundstage and imaging, and dynamics at all levels is just incredibly good on the grand scale of things. Tone overall is very neutral, which is what I like. Every nuance seems to be delivered...things I never noticed before even with old familiar recordings. Different remastered releases of the same content sound clearly different as well. From listening experience, I can say with confidence the new GanFET amp in particular is a big contributor, but the DAC is certainly holding up its part very well indeed.

I may play with the DAC part a bit just for fun to see what differences I might hear.

Open for suggestions!

 

 

 

 

 

The digital part is thoroughly transparent, always. The only thing that matters is the analog output, and that's where we measure the transparency (or not) of a DAC. Of course there could be some sort of contamination (as with everything analog), but it would be captured in measurements. DACs imo have been a thoroughly solved problem for a long time. It's not where I'd want to tune my system to have a signature... amp and speakers... sure. I am not saying I believe utter transparency should be everybody's ideal. :-)

You might be surprised by hearing some very nice sounding DACs / players from long ago. They will absolutely measure like dog crap compared to modern. But our ear-brains do not process information via Fast Fourier Transform!

Perhaps surprisingly, our ears and brain DO process sound almost exactly like a Fast Fourier Transform.

Our cochlea and our auditory cortex are tonotopic.  When a complex wave gets to our inner ear, different frequencies within the complex wave have peak resonant points at different physical locations along the basilar membranes of our cochlea which have different stereocilia bundles connected to them. Our inner ear actually breaks down the complex wave into component frequencies based on where each component frequency maximally excites the basilar membrane, and we have separate nerve firings for each of those component frequencies based on the hair cell bundles connected to the basilar membrane at those locations. The tonotopic geography continues into the auditory cortex in the brain, which Dr. Nina Kraus of Northwestern likens to a piano, where you see different physical regions in the auditory cortex responding to different frequency components of the complex waveform.

So, yeah, actually, our ears and our brains are breaking down incoming complex waveforms into their component frequencies very much like an FFT, and, further, actually converting them into binary-like neural spikes -- when a stereocilia bundle is deflected it creates a nerve spike or no spike, functionally like a 1 or a 0 -- and those go up to higher centers of our brains where the physical separation continues until other processes take place to create a perception of an integrated sound (or multiple separate sounds).

It's a sidebar to matter of what people prefer in terms of the particular sound of a particular piece of equipment. But our ears and are brains, when we hear, are very much doing something very like an FFT.

@chervokas 

Thanks, very interesting, I did not know that.

our ears and our brains are breaking down incoming complex waveforms into their component frequencies very much like an FFT, and, further, actually converting them into binary-like neural spikes -- when a stereocilia bundle is deflected it creates a nerve spike or no spike, functionally like a 1 or a 0

So, if I'm reading you correctly you are saying that our brains re-convert (analog) sound waves into digital?