Now that we’ve figured out why we need an analog output stage on our DAC let’s discuss why it’s so important to the sound of the DAC.
We started this thread because one of my readers asked a really good question: why, if most of a DACS’ internals and technologies are devoted to decoding the digital audio data, does the final analog output stage make such a sonic difference?
I put forth the idea that IMHO, a mediocre DAC engine coupled with a great sounding output stage would always outperform a great sounding DAC engine with a mediocre output stage. One obvious answer to this question is the age old bottleneck problem. Because the analog stage is the very last stage of any DAC, if that isn’t up to par than whatever we have going on before it cannot come through in all its glory. That’s pretty obvious. But I believe there’s a lot more going on than just the classic bottleneck.
I’ll bet that the vast majority of analog output stages on DACS measure quite well and do not technically act as a bottleneck – at least not one that classic measurement techniques would identify. So if we’re not impeding the audio signal in a way that we can measure, how then do different analog stages sound so different? This question is, of course, a major can of worms – it involves many of the classis questions about measured results vs. what we hear. I am not going to fully open that can of worms in this series of posts, but how about if I let a few of them wiggle out?
I think the safest way to go about explaining this is to simply relate to you what I have learned about designing these stages and the audible results that come from making those decisions. That way you can get a reasonable idea of a design philosophy and perhaps we can debate the classic measurement vs. audible questions at a later time.
Tomorrow I’ll detail what the parts of the analog output stage are and what they do.