Thread: ADCs, what is omportant?

I've heard the same thing, K. Don't know the answer, but yeah, I've talked to cats who said the same thing to me.

The answer is to be found one step back from that. What you have to compare is the overall design of the equipment. It's no good saying that a well-regarded piece of gear uses the same chip as a cheap far-eastern import and therefore must sound the same. Every bit of an equipment's design has an effect on its performance, however small, and top designers go to great lengths to get as many as they can of these small effects working in the same direction within the product budget. In the right hands, standard chips can sound stellar; in the wrong hands the same chips can sound like mud.

I have done contract designs for a number of medium to high-end companies, and it's very surprising how different the attitudes are to the concept of overall design. I have been asked why I bothered about a particular detail when it would be cheaper to leave it out, and sometimes I have had to say "I can't demonstrate it to you, but I know it makes a difference." The companies that then remove that detail in manufacture to save the pennies I do not accept work from again.

So then, what you're saying is in support of the old saying that a chain is only as strong as its weakest link.

I don't think that applies directly in this case. A good design can often tolerate some weaker links that may be in there for cost purposes ("value engineering") without greatly compromising their performance or specifications. However, the very top designs have no discernible weak links, and, often as a consequence, have no price limit.

I'm gonna chime in here for just a moment because I experienced this recently.

Of course the build quality throughout the device will be the key to it sounding better than a bag of the very same components assembled in a cave by blind monkeys...

To the 'clocking' thing.

I just finished a record that was recorded and mixed here at The DroolinDogg Ranch. Usually i go outside and mix at another facility with 'more stuff'. Not this time. Understand that my conversion isnt up to par with most of you. Really, it has been the last point on my list to upgrade because things simply sounded 'good enough' to my ear for release and playback on most earbud and iPod systems as well as computer stations and car stereos......isnt that where we deliver most of the goods too??
Anyway, I had a younger fellow come in and edit some tracks for me. (This IS a skill I ABSOLUTELY MUST HAVE!!!!) He was fast and accurate and actually graduated from a well known recording school....His take was all about the clocking at mix. I have a Digi 003 factory that I use for my fingers-on automation primarily and thus am stuck with the Digi converters and clock........but not anymore. I also have an Alesis HD24 that I use as my ADAT conversion when I'm running a lot of tracks at tracking. In doing this it makes the HD24 the master clock and I used it almost exclusively throughout the tracking stages so the only analog through the Digi converters happened at basic tracks BUT ALWAYS with the HD24 as the clock.

So he tells me that simply having the ADAT pipe hooked up at mixdown and selecting the ADAt as my external clock makes this the master clock no matter what I'm doing in Pro Tools. The difference is audible immediately. It happens that the clock in the HD24 is 'better' in some way that my aged mind can't conceive but is willing to acknowledge.


So next is a clock. Before conversion. Black Lion maybe? Others I know will swear by the really really cheap ADA8000 conversions this company does to that coverter and I'm starting to believe. Its true that only a few companies make chips used by the companies building conversion these days.And of these only a small number take this past a certain level of development.

I dont know why......but I do know that the clocking makes a big difference....especially when you are using cheap crap like I have here.

well.....its not ALL cheap.+

hmmm intersting stuff as always fellas. even before i read dave's post, which re-inforced my next thought. lol, i think every few days.

very very hypothetically this would be a question of what makes more a drastic effect, given all other parameters the same. conversion or clocking.?

like if say there was a 1-5 scale for both, say a new design employed a level 4 all-round, but the suits said hey we need to save some money, lets go w/ a level 3 converter, or clocking (thing) (i'm very rudimentary in my understanding of circuits pretty much noob level so thanx for bearing w/ me). which level drop would preserve the overall quality level if it was between only those two components.?

interested in more depth of the topic. more than just yes or no, some basic explanation of why would really help me start wrapping my head around. also as sample rates and bit rates are becoming higher, how does that relate to conversion/clocking?

cheap, lol, what about audio is? lol i wish i was genius enough to be able to sell chunks of mahogany for 2500. ya know ber-copy-inger really does get it rite once in a while, def a try before ya buy company, but i don't don't their ability to get by patent issues. cheers all!

They are both important. The most expensive ADC in the world will sound bad if you force it to run on a rubbish clock. Looked at from the other side, the highest quality, most stable clock you can generate will allow a poor ADC to run to the best of its ability, but it won't turn it into a good ADC.

I don't think you can judge them by levels 1 - 5, as you have to look at the wider picture, including the purpose of the recording. Deficiencies introduced by lower quality clocking may be deemed acceptable for recording MP3s for YouTube, but might sound horrible in the studio.

There is a qualification to all this. Some of the higher-quality interfaces have clock flywheel circuits that can straighten out shortcomings in external clocks. Even boxes like the RME FireFace800 have sophisticated clock-regeneration circuits that result in there being little difference between moderate and high-quality external clocks. So I suppose one answer to your question is that if you go for the right type of ADC and DAC unit, you do not have to spend huge amounts on high-quality external clocks, and a simple box that keeps everything in sync could be all that is needed until the next spending round.

I covered a lot of stuff regarding clocks in the "What is clock jitter?" thread in the DIY pro audio forum which is relevant here. Particularly where external clocks are used or soundcards are slaved to others. In that thread I mentioned that for a given A-D chip, we know, from the datasheet, what the best possible performance is going to be. Once we start putting this chip on a soundcard PCB all we can do is detract from this optimum. We certainly can't improve on it as the chip makers spend a small fortune on getting the last drop of performance for their datasheets!

I am firmly with Boswell here, yes the clock is important but so too are the input conditioning amplifier/buffer (we cannot just connect a line input straight to the chip), internally generated power supplies etc. Poor design of any one item can throw away performance hand over fist and no circuit element should be the weakest link. This is all down to the design, both electrical and mechanical (good PCB layout is crucial).

In the past, some clock circuits have not received the attention they deserved in the design process (my gear included) and we end up with cautionary tales about clocks such as that given by Davedog. Certainly it is important to understand how clocks can affect our converters but, even if we may witness some stellar improvement on a particular set up by switching clocks around, this wisdom may well not hold true with other equipment.

This formed a large part of what I tried to suggest in the other thread as I believe a good understanding is more important than anecdotal evidence which may not hold up to closer scrutiny. Anecdotal evidence also has a nasty habit of "going viral" and being taken as gospel.

So yes, we can all probably come up with some surprises (as I did with my own set up in the other thread) and this should not cloud the facts but be used wisely within the set up that created the surprise.