Audio DACs, Instrumentation DACs.

[Brian Guralnick]

I'm having trouble figuring out what is the best sounding DAC out there.

Should I scrap the idea of looking for consumer grade Audio DACs & start looking at Instrumentation grade DACs?

Has anyone else out there looked into this?

[alvaius]

I think you will find that the relevant specification of audio DACS, for the purposes of audio are much better than instrumentation DACs, especially if you are talking sigma-delta.

[UrSv]

The question of which is the best DAC out there is probably as easy as to select the best wine or in my case the best single malt.

It is very much a question on personal taste, IMHO. There are many good ones and you can find information on quite a few here. Which one is best for you is a question only you can answer. I would however stay with the Audio type.

/UrSv

[peranders]

Originally posted by Brian Guralnick
I'm having trouble figuring out what is the best sounding DAC out there.

Should I scrap the idea of looking for consumer grade Audio DACs & start looking at Instrumentation grade DACs?

There is no instrumentation DAC which is better than any audio DAC for just audio. Am I wrong here?

[jackinnj]

There is (are) a treasure trove of information on the TI and Analog Devices websites.

In fact, if you are just getting into it, you can download an entire dSP primer from the analog devices website -- just like a university text, and well worth having in your library. (perhaps not a great idea, however, if you just have dial-up access to the web since it's a monstrous file.)

[tvi]

I thought the main problem with instrumentation DACs was they don't use an audio standard input scheme, like I²S, some use parallel data inputs?

My older Sony CD player uses PCM64 18bit parallel input DACs, implementing a custom Sony serial to parallel conversion IC.

A DAC I purchased about 2 years ago uses 2 PCM55s per channel, these are also parallel input devices, but the use of instrumentation DACs seems uncommon.

Regards
James

[Brian Guralnick]

The interfacing is not the issue here. I can easily convert I2S for anything I like.

The big question is, can I get a better sound?

[Jocko Homo]

Let's see.............PCM1704s cost over $33 for the "best" grade. I guess they are so expensive because they hope Krell will use them.

So why do you want to reinvent the wheel?

Jocko

[BrianL]

The whole world of converters has changed dramatically
since the early days of digital audio. The original "audio"
DACs, say from Burr-Brown, were slightly modified
"instrumentation" DACs, often with less rigorous testing.
The sort of customers that drove the "instrumentation"
DAC business in ways that were parallel to audio needs
seem from what I can tell to have been military-related.

These days it seems that audio DAC design is driven, spec-
wise, by the standard (Japanese-devised) audio specs and
other practical requirements like low glitch energy. They
seem to be evolving along a different path than "instrumentation"
DACs. Also, there seems to be a preference for delta-sigma
as it's lower cost to manufacture and test.

I've seen a lot of non-audio DACs (and ADCs) that look like
they'd might make interesting audio converters, but the
question is whether you have the time, money, ability,
inclination, etc., to build up all these things and try them out.
I know I don't. I'd be doing good to do a Jocko-recommended
replacement of the Philips TDA1541 DAC and associated
parts in my Magnavox CD player with the PCM1704.

Notice that even the high end audio brands are sticking
with "audio" DACs in their products. So either they don't
have the time or consider the effort not worth it.

bottom line: there are a number of interesting parts out there.
Do you want to take on this challenge? Is that the most
bang for the buck/effort in terms of things about your
audio system(s) that most need improvement?

[alvaius]

If you compare the bandwidth of audio DACs and the SNR, it makes for a combination that I have never seen in instrumentation DACS [and I have looked]. That is not to say they will sound better! I think the issue is that there are few "instrumentation" type processes looking for 100db SNR with 20-100K bandwidths. Most applications for 100dB SNR products are low frequency or specialized narrow band RF. To that end, I don't think there has been the marketing drive for parts that would appear be of interest both in the instrumentation world and Audio world, not that it will not happen.

However, even in the instrumentation world, sigma-delta ADCs are starting to take hold, though I have not seen much in the way of sigma-delta instrumentation DACS. While cost is definately an issue in doing a sigma-delta, achieving the linearity of a current high end 24 bit sigma-delta with a multi-bit DAC may be impossible. That does not mean the sigma-delta DAC will sound better, but it measures better with our limited measurement tools and hence looks good on a marketing sheet. [Check one of my threads for an oversampling DAC based on a DSP\PCM1704].

I have long considered doing my own sigma-delta DAC using a high speed video\RF oriented DAC.

Something to consider when I am driving to Montreal this week!