DAC specs versus sound quality potential

You have that so right cedargrover

That’s why I always state "well implemented" when ever I talk about differences.

Most mainstream big or small dac makers copy the implementation straight of the DAC chip manufactures data sheet, usually in the last 3 or 4 pages, it’s sometimes only diyers with knowledge (not the voodoo'ists) a couple the boutique manufactures, that go to the trouble of eg: discrete I/V, discrete buffer outputs, and better than the norm power supplies ect ect.. Which would add huge extra cost to those production units, that maybe big dollars and look the goods, but are just data sheet copies on the inside.

Cheers George

When it comes to DACs, it is all about implementation of a DAC chip in the overall design.  Some designers use only portions of the functionality available in a DAC chip, and design their own filters, oversamplers, etc. to achieve their design goals.  And then you have the whole output stage to worry about too, and that makes a huge difference in sound quality....  

The differences between dac chips are hard to tell apart given the same topology and specs.

But the differences in topology is another story when both are converting the same media.
The different topologies are Multibit and Single Bit (Delta Sigma) These two when compared can be quite different if both are implemented well.

To me Multibit does PCM (dxd, redbook 16/44 24/96) the best, there are some very expensive ones that can also do DSD as well I have not heard these, doing dsd.

Where Delta Sigma does DSD well, it falls short on doing pcm right when compared to Multibit, it’s very smooth and non intrusive and to me a bit boring, it lacks that live dynamic jump factor that well implemented Multibit is capable of and can put out in spades.

Cheers George

Having gone through the 42 page datasheet for the AK chip,thanks to almarg's post above,I have a better understanding of the myriad functions dac chips perform as well as their specs and tolerances. The spec sheet is obviously not comprehensive as it lists functional parameters/characteristics but not manufacturing details and material specifications. I'm not the least qualified to evaluate the chips even provided all relevant data. I remain interested in a comparison of the chips in a way I can understand. For example; which would be like the best Porsche motor and which like a Mercedes Formula 1 engine.

I have two non-oversampling DACs, mhdt Paradisea and Constantine. I have no idea how they measure, but the sound quality speaks for itself and both compete well with anything at any price that I have heard to date.

The Paradisea has a tube output stage and the resulting sound will vary greatly depending on specific tube used, I have found.

No doubt there are many implementation details with DACs that matter, but frankly, most newer DACs I hear these days tend to sound quite good overall. That was not so much the case I think just 5-10 years ago. SO maybe DAC designs have matured quite well of late and excellent performance not hard to find? Kind of like how most modern computers are well suited to leverage the internet? OR most digital cameras can take very high quality photos that would have cost a lot more to produce in teh past with older technologies?

DACs are pervasive and are used everywhere similarly these days, so that would not surprise me at all if good sounding DACs are commodity items now these days for the most part.

Not to say that any two designs will sound exactly the same, so there still is some decision making involved there for discriminating ears I would say.

I'll second the responses by the others. I'll add that in situations where some of the specs on a given part may be "superior" to those of a competitive part, it will often be the case that both specs are sufficiently good that the difference is sonically irrelevant.

As to why Esoteric chose the Asahi Kasei part instead of an ESS Sabre part, I wouldn't begin to want to speculate. A quick scroll through this 42 page datasheet for the AK part will give you some idea of the complexities that are involved. (ESS doesn't appear to provide detailed datasheets at its website). Also, I would expect that experimental evaluations were likely to have been performed, perhaps using circuit boards provided by AKM, ESS, etc. for evaluation purposes. And non-technical factors may also have been involved, perhaps including things like working relationships among the particular companies, and how much familiarity and experience Esoteric's circuit designers may have with the products of the respective chip manufacturers. Perhaps it is even relevant that AKM is a Japanese company (as is Esoteric), while ESS is headquartered in California and does a lot of their work in Canada.

In general a great many things, both technical and non-technical, can factor into the selection of a key part in a design, and I would not want to speculate in any given case.

Regards,
-- Al

+ 1 Arh.
If given the opportunity, you have to judge components by listening. Regarding DACs, their analog stage implementation and power supply quality are likely more important determinates of sound than what specific chip is used.

What specs are you referring to?

A lot of specs taken in the measurement of dac chips do not translate directly into sound quality. It is the overall design of the player that will determine the sound. Audio Note dacs are highly praised for there sound quality yet they measure poorly compared to other dacs because they use non oversampling chips that measure not as good as the Sabre or other Delta Sigma chips but still sound very good. It is like SET power amps that measure terrible but many people love there sound.. If you buy stuff based on measurements you will wind up with a system that measures great but may or may not sound good. Use your ears, not measurements to evalulate components.
Alan