Apogee Duet Input Calibration

Recording Techniques, People Skills, Gear, Recording Spaces, Computers, and DIY

Moderators: drumsound, tomb

Post Reply
User avatar
Jay Reynolds
carpal tunnel
Posts: 1604
Joined: Thu Feb 07, 2008 5:48 pm
Location: Austin, TX
Contact:

Apogee Duet Input Calibration

Post by Jay Reynolds » Thu Mar 28, 2013 3:31 pm

A friend just bought a Duet II and I thought I'd be helpful and look at the spec sheet for the input calibration.

http://www.apogeedigital.com/products/duet2.php#specs

No such. Max input = +20dbU. And an impedance measurement.

So my math was never that great to begin with, but is there way to distill how the inputs where calibrated from max dbU and impedance? Or does someone happen to know what levels the Duet IIs are calibrated to?

Thanks.
Prog out with your cog out.

matt_stevens
audio school
Posts: 6
Joined: Sat Mar 01, 2008 11:37 am
Location: Oxford
Contact:

Post by matt_stevens » Fri Mar 29, 2013 2:09 pm

Hello, you've already answered your question! The maximum input level specified is +20dBu, which will equate to full scale, or 0dBFS.

So in this case if you were to feed the Duet a signal that measured 0VU (+4dBu)on an analogue console, you would see -16dBFS in your DAW metering.

User avatar
Jay Reynolds
carpal tunnel
Posts: 1604
Joined: Thu Feb 07, 2008 5:48 pm
Location: Austin, TX
Contact:

Post by Jay Reynolds » Thu Apr 04, 2013 1:45 pm

So there is a correlation between voltage and binary? I thought that maybe I've had it wrong all these years, and maybe I do, but I don't think you can just take a number that applies to dbU and use it to manipulate the math as it applies to dbfs. Aren't the two exclusive without knowing how the manufacturer has calibrated their ADC inputs?
Prog out with your cog out.

matt_stevens
audio school
Posts: 6
Joined: Sat Mar 01, 2008 11:37 am
Location: Oxford
Contact:

Post by matt_stevens » Thu Apr 04, 2013 4:22 pm

Hi Jay,

Well, lets look at how that correlation between voltage and binary works...

An A to D convertor looks at an analogue signal level and outputs a binary value that corresponds to how high or low that voltage is. For a 24 bit convertor you have 24 bits... if there is no input signal at all these would all be "0", which isn't possible with analogue signals due to thermal noise and so on...

For higher level signals the converter will output a different binary numbers, and for a doubling of input signal level, you use another significant bit (eg 0100 is twice as much as 0010).

At some point though, the voltage that we send to our A-D convertor is going to be high enough that all our bits are set to "1"... This is our maximum input level as it is the highest number the convertor can represent, or 0dBFS. Even if we increase the analogue signal further, the output of the convertor is still going to be "all the 1s" as it has run out of bits to turn on... this is where clipping happens!

(DAWs can use floating point maths to shift the range the bits are working in, but at the convertor stage we're stuck in fixed point)


When we apply gain to a signal, be it analogue or digital, we're boosting or attenuating by the same amount... "6dB" is the same level of gain in either domain, but the signal we apply it to may be measured with respect to different reference points.

Going back to the Apogee, where you say "Max input = 20dBu", this means that the binary output from the convertor will be 0dBFS (all the bits will be "1") when 20dBu is sent to the input. -6dBFS will use one less significant bit (or 011111111111111111111111) when a voltage 6dB lower is applied (14dBu).

I guess the short answer to your question is that the manufacturer HAS told us how they have calibrated their inputs! "Max input = 20dBu" :D

Matt

User avatar
Jay Reynolds
carpal tunnel
Posts: 1604
Joined: Thu Feb 07, 2008 5:48 pm
Location: Austin, TX
Contact:

Post by Jay Reynolds » Fri Apr 05, 2013 8:56 am

Here's what's getting me, and it's probably really simple and I'm missing it, but I'm not sure how the math works outside of a system with a max input of 20dbu. I think I see how you got to 0dbu=-16dbfs (20-4 and then subtract from 0 to get -16). What if we're looking at the Symphony ADC, which can take a max input of 24dbu. When I plug that number in to the same equation, I come up with -20. Which can't be right. Obviously I could put on my best Nige Tufnel voice and say "this one's four higher", but then that means my first equation is only valid in that one case, meaning it's not really an equation.
Prog out with your cog out.

User avatar
Nick Sevilla
speech impediment
Posts: 4988
Joined: Mon Mar 03, 2008 1:34 pm
Location: Lake Arrowhead California USA
Contact:

Post by Nick Sevilla » Fri Apr 05, 2013 12:17 pm

Jay Reynolds wrote:Here's what's getting me, and it's probably really simple and I'm missing it, but I'm not sure how the math works outside of a system with a max input of 20dbu. I think I see how you got to 0dbu=-16dbfs (20-4 and then subtract from 0 to get -16). What if we're looking at the Symphony ADC, which can take a max input of 24dbu. When I plug that number in to the same equation, I come up with -20. Which can't be right. Obviously I could put on my best Nige Tufnel voice and say "this one's four higher", but then that means my first equation is only valid in that one case, meaning it's not really an equation.
Jay, have you ever used a "slide rule" for engineering? You know, the one with different SCALES in each side, which you slide back and forth to get a RELATIONSHIP?

This is very simple once you realize that dBFS (Full Scale) which is used for digital converters is an ABSOLUTE scale, and everyone else has to relate to it like that, since there is no way to change 0dBFS, as that is the absolute maximum in this scale.

The other scales can and do have voltages ABOVE their zero point.

Do the following on a piece of squared paper :

Make one line going vertically, and at the top cross line, write down 0dBFS. This is the top of the digital converters scale.

Then to the right of this line, make another line, and make the top cross line be +18dBu, which is a very common console maximum output.

Then, to the right of this line, make another line, and make the top cross line be +20dBu, which is you Apogee converters maximum input, as referenced to the dBu scale method.

Now, on the second line, count crosslines down until you find 0dBu, and make a line to the left to the dBFS scale. How many dBFS does this equal? -18dBFS.

On the third line, count down the crosslines until you find your zero point. Then make a line to the right to the dBFS scale. Where does this lay? At -20dBFS.

Does this help you understand the relationship between dBFS and dBu?

dBu is a scale that is dependent on the actual analog equipment maximum operating levels, which vary from equipment to equipment, related to the industry standard dBu of 0.774 volts RMS = +0dBu.

In the digital world this dBFS scale is NOT dependent on that standard, instead it only is related to the clipping point of a given digital device, which is the same for any ad converter. This point is where all the bits are filled to maximum. And this does not vary from converter to converter, they all clip at some voltage.

The Apogee one clips when you feed it more than 20dBu above that 0dBu point, 16dBu above the standard +4dBu operating level you see on manuals.

You can read more on loudness scales if you search for Bob Katz.

Cheers
Howling at the neighbors. Hoping they have more mic cables.

User avatar
Jay Reynolds
carpal tunnel
Posts: 1604
Joined: Thu Feb 07, 2008 5:48 pm
Location: Austin, TX
Contact:

Post by Jay Reynolds » Fri Apr 05, 2013 12:49 pm

Riiiiiight. Because the higher max input represents headroom between 0dbu and the point at which the input clips. Makes sense now.

Thanks guys.
Prog out with your cog out.

matt_stevens
audio school
Posts: 6
Joined: Sat Mar 01, 2008 11:37 am
Location: Oxford
Contact:

Post by matt_stevens » Fri Apr 05, 2013 1:11 pm

Jay Reynolds wrote: I think I see how you got to 0dbu=-16dbfs (20-4 and then subtract from 0 to get -16).


I'm not sure if that's a typo or if you misread my post, I said that with the duet 0VU (which equals 4dBu) would equal -16dBFS.

4dBu Analogue = (20dBu - 16dB) = (0dBFS - 16dB) = -16dBFS Digital

The underlined bits just show the offset that is present... When you know the reference (20dBu = 0dBFS) you can work out the levels as the offset will always be the same.

Jay Reynolds wrote:What if we're looking at the Symphony ADC, which can take a max input of 24dbu. When I plug that number in to the same equation, I come up with -20. Which can't be right.
In the case of the Symphony, the maximum input is 24dBu, which will equal 0dBFS.

So in this case a 4dBu signal will be converted to a -20dBFS signal.


Here's another example, I have a MOTU 828 convertor, which is calibrated so that 18dBu = 0dBFS. My mixer has a tone generator built in, which outputs 0dBu. When I send my mixer's tone generator out to the MOTU, it will show me a signal level of -18dBFS in the DAW. (The Mixer will also show me PPM 4 on my BBC Peak meter, but that's a whole other story! :D )

Matt

Post Reply

Who is online

Users browsing this forum: No registered users and 26 guests