Software audio analyzer to mic pre... impedance issues?

[hobiesound]

and another bump coz i'd like to see that picture too!

Pretty please with a cherry on top!!!!

 

[skipwave]

I've been looking round for it since Ilya PM'd me about it. I know I have it on a backup disc somewhere. :? I'll find it soon enough.

 

[SonsOfThunder]

I just noticed this thread. Now I feel guilty about taking my Symetrix comp that I stuffed with THAT2180 to work on a Sunday in order to trim the THD on an AP2712 at work...and I had to choose which one to use...among several...

The sad thing is when I work with one all day, sometimes I don't want to look at one on my off time to do my DIY.

Peace!

 

[skipwave]

Well, I still can't find Dave's drawing. IIRC, it was basically a -20db U-pad, 1k resistors in series and 220 ohm shunt resistor. There may have been more to it, but that would get you a functioning test jig, with proper impedance and level.

 

[clintrubber]

[quote author="skipwave"]Well, I still can't find Dave's drawing. IIRC, it was basically a -20db U-pad, 1k resistors in series and 220 ohm shunt resistor. There may have been more to it, but that would get you a functioning test jig, with proper impedance and level.[/quote]
I must have it somewhere, give me some days - but why is it down ? If I put it up again that musn't be because of NYD taking it down on purpose.

Bye,

Peter

 

[skipwave]

Oh, I don't think Dave purposefully removed it. I think it was lost in the Twin-X crash awhile back.

Anyway here it is.

 

[clintrubber]

Cool, wasn't aware (anymore) of a crash there.

OK, done :thumb:

 

[Ilya]

Thanks, skipwave!

clintrubber, I don't know exactly what has happened, but it seems that they changed the way you can link to files you have uploaded. This lost picture is not the only one, I've noticed that some other pictures from Twin-X don't display because of the broken link.

 

[letterbeacon]

Oh, I don't think Dave purposefully removed it. I think it was lost in the Twin-X crash awhile back.

Anyway here it is.

If I was coming out of balanced output on my Fireface 400, would I need to place the 1K on the tip and the ring connections, with the shunt across the tip and the ring?

 

[alexc]

Something to keep in mind when viewing the reported stellar noise floors of -120db and better
is that audio analysis software is generally 'calibrated' to digital full scale for the sound card,
not an absolute voltage scale.  ie. 120db below FS, where FS may be 12Vpp or whatever

I find it useful to compare the results reported by audio analysis software to a cro in order
to 'calibrate' my interpretation of what the audio software reports.

My setup loopback test, in one audio program's calibrated cale, gives an overall SPL for terminated inputs and outputs
of -83dBu with worst individual freq bands around -90dBu. I can calibrate my level of confidence with the software
by input/output of known, CRO-measurable signals and comparison with the results of audio software.

The same system reports wildly different results with other audio software  ie. rightmark

So my advice is to be skeptical and try to compare results with known standards

 

[ruffrecords]

Something to keep in mind when viewing the reported stellar noise floors of -120db and better
is that audio analysis software is generally 'calibrated' to digital full scale for the sound card,
not an absolute voltage scale.  ie. 120db below FS, where FS may be 12Vpp or whatever

It is even worse than that. The spectrum these programs give are per root Hertz. What this means is that if it is flat along the -120dB level you have to integrate this over the 20KHz bandwidth of interest. This basically means adding the square root of 20,000 expressed in dB to the measurement. The root of 20000 is 141and 20log 141 is 43dB so over a  20K bandwidth a -120dB reading really means  S/N of 120 - 43 dB = 77dB.

Cheers

Ian

 

[alexc]

This basically means adding the square root of 20,000 expressed in dB to the measurement

Yikes! Didn't know that. Need to go over my cal procedure and try to 'integrate' that into my head!
And review some of my dusty theory notes.

 

[ruffrecords]

This basically means adding the square root of 20,000 expressed in dB to the measurement

Yikes! Didn't know that. Need to go over my cal procedure and try to 'integrate' that into my head!
And review some of my dusty theory notes.

Some of the better programs even include a little bar graph at the side that shows the equivalent rms noise over a specified bandwidth.

Cheers

Ian