February 10, 2011

DAC Marketing Malice?

For the CAS DIY crowd, the above pictured ESS Sabre DAC kit is one of the most popular today!

As we rush head long in to the CAS (Computer Audio System) scene, we need to look very carefully in to many details for consideration. After all CAS is a new musical format that emerged out of the MP3 sector only as recent as three years ago or so. There is still many areas of uncertainty, mostly concerning the lack of standardization or common compatibility and interface. One area that I've found to have more grey description than others is in the DAC product specs. I'd be very wary about manufacturers claiming it's DAC capable of "Up To 24/192" or any other higher sampling rates.

This is due to the fact that after my extensive research(which is not that advanced by the way, and many are way ahead of me in CAS), that many DAC manufactures claims, on spec sheet are merely frivolous and are mostly marketing, and numbers game driven to trick gullible spec sheet brigade among the consumers.

After all, why say "Up to 24/192"? Which I find the statement suspicious to say the least. Other manufacturers even claim to offer "Up to 32/384 or 32/798" too! When I see spectacular looking numbers like that, my mind goes on RED ALERT!!! I hope yours does too, especially when the DAC manufacturer claims to be able to do so via the USB input, asynchronous or otherwise!

There are two questions that immediately spring to my mind on such claims.

1) If I find any Burr-Brown or other audiophile approved DAC chips from the older days, inside such a DAC, I'd immediately know there's a problem. The problem is that most established audiophile approved DAC chips are at best limited to 24/96 sampling rates, for use with the then hi-rez format of DVD-A and SACD. If any DAC claiming to have BB DAC chips inside, it is highly likely, the signal goes thru a post conversion internal up sampling to 24/192, 32/384 or what ever impressive figures published! This means that the signal is no longer native, but internally up sampled. It just means the up sampling chips just randomly fills up 0s in to the signal path, to make it whatever sampling frequency you want(it's actually a bit like the pre 2008 pre crisis US economy, where the stock brokers just add zeroes behind any number, which in return is translated as gain or profit. However, it's all just bits from the computer system in Wall Street, because the gains are not supported by any substantial form of solid asset, not even paper, so you can't even call it paper gain! Now we all know what happen after that?). It's the same with digital audio, which I on the other hand, am a firm believer of doing things native. There are only 2 or 3 DAC chips known at the moment that'll genuinely do 24/192 or higher sampling rates at the moment, but I believe technology is not stagnant, and more capable chips will be on the way soon.

2) Many form of digital audio connection only supports sampling rates of up to 24/96, such as the co-ax, USB version 2.0 and perhaps even toslink. My list is not particularly exhaustive, as I've not tested all the means of connection my self just yet. As it stands at the moment, my favoured mode of digital connection for hi-rez (meaning for sampling rates above 24/96) remains with either firewire, I2S and AES/ABU only.

While the PC or the computer set up details is just as important, which I've overlooked those, only because of my chosen safe route via the Bryston BDP-1 digital media player. However, if it helps, it always pays to beware when looking at such things.

No comments: