A/D/A Converter...

Posted: Sun Mar 23, 2008 9:44 pm

I'm looking for a somewhat specific A/D/A converter of comparable quality to the fireface. Are there any 192khz sample rate converters out there that have between 10-20 line inputs, and no mic pres? I can't seem to find any that have exactly what I'm looking for. I might just opt for a fireface 400 and a multiface...

Thanks,

Conor

Posted: Sun Mar 23, 2008 9:58 pm

ehh i forgot to mention that I don't want to mess with any pci cards or anything... strictly firewire

Posted: Mon Mar 24, 2008 12:54 am

First thing is that Im not sure what if anything is available for 192kbs convertors multi channel into firewire .
As most computer manufactures have yet to adapt all of their firewire designs to the faster firewire800 port chips .
Next how many channels.
If you really want proper AD's and DA's you might want to look into
the following at EMM labs.
see link;http://www.emmlabs.com/html/about/ed.html
Ed Meitner designed SONY's latest AD's and DA's for thier professional console systems , he also re-wrote the book on why most digital audio sounds as bad as it does.
The biggest reason for bad digital is a few things combined;
A/ Too low of a sampling rate . 44.1 doesn't cut it unless your only recording pure sinusoids. aka Nyquist only applies to standard sinusoidal waveforms . Unfortunetly a few good engineers got fooled by that one.
And for reall sonic quality you want to go as high as possible (read at least 10 times the highest possible frequency) well thats going to be really interesting if you consider that a cymbal has harmonics that go ultrasonic right up to 100Khz. Then your sampling frequency would have to be at least 2Mhz for a 1 Mbs sampling rate. COnfused yet. Laughing

Just see what James Boyk (Profeesor at Caltech) has done in research in this area. http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
Anyway , if you really want to teach audio , please do yourself and your students a favour and know the subject.
anyway enough of that 192kbs is usually the minimum that any "real" studio will deal with , its does provide substantially better audio that lower , but pay attention to phase when mikeing (ie make sure mike distances are exactly the same for left and right.
I'll s3ee whats out there for firewire.
Cheers...ampmaster.
Cool

Posted: Mon Mar 24, 2008 1:15 am

Oh yeah ,, I didn't even get to B & C on my little list of why digital sounds bad.
B/ Biggest problem with most convertors is that they use cheap clock sources (IE cystal chips with marginal stability ) , even with the best methods of temperature controlled X-Tals , thier is always a Jitter issue within the converter circuit as Ed Mietner found out.
C/ Next is the actual conversion method (IE method of converting analog to digital and vice versa) , many of the topologies have either timing issues associated with them or linear translation problems ( ie The bit doesn't go high when its supposed to) Ever a lot of the best sigma delta convertors suffer some kind of stability or linearity issues.
Ya I know too much tech talk , OK first things first.
A lot of issues can be cleaned up with just plain clean power . In other make sure your A to D / D to A box or boxes are on a high reliability ultra clean power source , the more stable your get your AC , the more accurately the converters will track .
Next is make sure there are no computers near your AD's and DA's , despite all the so-called shielding that the comp mgft's say they have , most computers radio out massive amounts of radio frequency digital energy which always gets injected into the convertor one way or another (IE cables , leaks in chassis against RF shielding, etc,etc,etc)
If you could put them into thier own little Faraday cages it would help.
Next part is grounding , all electronics derive thier stability from some type of ground , (ie the reason why all mike inputs use a drain or grounding pin) the better your ground the better your audio is going to sound. (IE less resistance to ground is best)
From there its a matter of trying something and listening to see what works and what doesn't.
Cheers..ampmaster.

Posted: Mon Mar 24, 2008 8:59 am

ampmaster wrote:

Anyway , if you really want to teach audio , please do yourself and your students a favour and know the subject.
anyway enough of that 192kbs is usually the minimum that any "real" studio will deal with , its does provide substantially better audio that lower , but pay attention to phase when mikeing (ie make sure mike distances are exactly the same for left and right.
I'll s3ee whats out there for firewire.
Cheers...ampmaster.
Cool

Irony....

Uhhh, ampmaster, it's 192kHz, not 192kbs (sic).

Second,
"192kbs is usually the minimum that any "real" studio will deal with..."

Is a completely bogus and unrealistic statement. Most any studio I've ever seen and/or worked with tries to stay at 44.1kHz except on special projects or projects in which they get paid extra for the higher sample rates.

Also, suggesting Meitners and a Faraday cage for a young person's studio (or even an advanced, high-end studio) is a bit excessive.

Conor -

Consider the TC DigitalKonnekt32 plus an external converter.

Also, the Lynx Aurora 16 with the firewire cable would be an option as well.

Posted: Mon Mar 24, 2008 1:00 pm

ampmaster wrote:

Anyway , if you really want to teach audio , please do yourself and your students a favour and know the subject.

I think this is very good advice. It's also good to read contributions from people who have long, direct experience with studio and/or live recording, who know the technical background to their art, demonstrate attention to detail in their replies and who don't just reproduce the opinons of others, mis-quoted and out of context.

Posted: Mon Mar 24, 2008 2:47 pm

ampmaster wrote:

First thing is that Im not sure what if anything is available for 192kbs convertors multi channel into firewire .
As most computer manufactures have yet to adapt all of their firewire designs to the faster firewire800 port chips .
Next how many channels.
If you really want proper AD's and DA's you might want to look into
the following at EMM labs.
see link;http://www.emmlabs.com/html/about/ed.html
Ed Meitner designed SONY's latest AD's and DA's for thier professional console systems , he also re-wrote the book on why most digital audio sounds as bad as it does.
The biggest reason for bad digital is a few things combined;
A/ Too low of a sampling rate . 44.1 doesn't cut it unless your only recording pure sinusoids. aka Nyquist only applies to standard sinusoidal waveforms . Unfortunetly a few good engineers got fooled by that one.
And for reall sonic quality you want to go as high as possible (read at least 10 times the highest possible frequency) well thats going to be really interesting if you consider that a cymbal has harmonics that go ultrasonic right up to 100Khz. Then your sampling frequency would have to be at least 2Mhz for a 1 Mbs sampling rate. COnfused yet. Laughing

Just see what James Boyk (Profeesor at Caltech) has done in research in this area. http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
Anyway , if you really want to teach audio , please do yourself and your students a favour and know the subject.
anyway enough of that 192kbs is usually the minimum that any "real" studio will deal with , its does provide substantially better audio that lower , but pay attention to phase when mikeing (ie make sure mike distances are exactly the same for left and right.
I'll s3ee whats out there for firewire.
Cheers...ampmaster.
Cool

Uh.......nope.

The human ear can't typically distinguish a tone above ~20kHz. Above that, there is some evidence that frequencies (maybe even as high as 30kHz, but that's a stretch) can influence perception of tone, but that's on an incredibly small level. On top of that, most mics cannot pick up anything past 20-22kHz.

192k is simply overkill. I have never heard of anyone using a sample rate that high. It's simply not necessary.

44.1 is fine for most applications. I'm not sure why but you have to divide that by 2 and that is the highest frequency it can represent (someone correct me if I'm wrong about that). So 22.05kHz at 44.1, basically higher than most people can hear on a good day. I use 48kHz because...why not? For recording where sound quality is of the utmost importance, 88.2 is the highest people tend to go. But even that is like swatting a fly with a bazooka. And you'd better be using something like those Earthworks mics that go up to like 30-50kHz to justify it.

Posted: Tue Mar 25, 2008 5:54 am

I use sample rates that high DanBronson. I do it all the time. Especially for pipe organ and major symphony works.

Many mics pick up well beyond 20 kHz. They're response charts don't depict much past there in most cases and true, in many cases, there is a rolloff which begins well before that, but it's not all about just what can be heard and picked up by the mics. There are anti-aliasing and brick wall filters that come into play (which have admitedly gotten much better since the advent of the CD in the early 80s) and latency numbers as well.

It's not all as simple as "we can't hear that high, so why go there..."

Posted: Tue Mar 25, 2008 9:24 am

Yeah most big professional studios nowadays are using 192 khz sample rates and it does sound better

Posted: Tue Mar 25, 2008 9:27 am

I wouldn't go in that direction yet Conor -

Some studios use 192. I wouldn't say most big pro studios do. I would say many have the capability and many occassionally do projects in 192, but certainly most of the projects being put out are still done at 44.1 or 2".

Posted: Tue Mar 25, 2008 9:42 am

From what I understand, one of the big reasons for using much higher sampler rates than 44.1khz is that in order to avoid aliasing you have to have a high quality low-pass analog filter that activates at whatever sampling rate you're using. The roll-off of most filters is pretty um, long? slow? Even for high quality filters. So, if you're sampling, say, at 96khz and your low-pass filter isn't that great of quality the aliased signal will still only fold back to around 44khz, which is definitely inaudible.

And so, you take the expense of a very high quality low-pass analog filter out of the equation with sampling rate. Seems like there are a lot of conflicting opinions on this, so please don't take my word as an "authority."

Posted: Tue Mar 25, 2008 5:31 pm

anyway... back to the main topic.... im leaning towards a fireface 800 so i'll have the option of adding some more channels through adat at a later time

This topic sponsored by: