A-B wave subtraction - Home Theater Forum and Systems - HomeTheaterShack.com

Old 01-07-13, 08:17 PM Thread Starter
Shackster

Join Date: Nov 2012
Posts: 9
A-B wave subtraction

Anyone...

I have a question relating to the subtraction of two measurements. I find confusing results when subtracting two waves. In the example shown in the attachment, the Green and Violet measurements are independent measures. The Yellow tracing is the result of subtracting the measurements. As one might expect, the A (Green) is almost overlapping the B (Violet) and where this occurs, I would expect a difference curve to show no difference or close to "Zero" on some scale. Also, in the high frequencies, there is a considerable difference between the response output of the higher A (Green) tracing and the lower B (Violet) tracing. I would expect an A-B curve to give a gain difference closer to 20 dB on the difference (Yellow) tracing.

Am I missing the boat here? How to I get a true difference response? How can I combine the data to reflect the difference in impulse responses? I think the IR has been corrected to "zero" but perhaps I need clarification on that as well.

Phil M cCA n.d

Last edited by pmccand; 01-07-13 at 10:39 PM. Reason: Grammatical corrections
pmccand is offline

Old 01-07-13, 10:58 PM Thread Starter
Shackster

Join Date: Nov 2012
Posts: 9
Re: A-B wave subtraction

I believe I found the problem with the difference equation. Seems as if there is a program bug. It looks at first glance that the A-B is actually A/B and the A/B tab performs a A-B. I will report the bug.

Phil M.cCan.d
pmccand is offline
Old 01-07-13, 11:49 PM Thread Starter
Shackster

Join Date: Nov 2012
Posts: 9
Re: A-B wave subtraction

Here is a photo of a PERFECT difference curve...Only thing is that you have to click on the DIVISION function (A/B) instead of the Difference function (A-B) as intended. Now everything works as it's supposed to!

Cheers,

Ph.il Mc.C.AnD

pmccand is offline
Old 01-08-13, 12:54 AM Thread Starter
Shackster

Join Date: Nov 2012
Posts: 9
Re: A-B wave subtraction

Aren't the functions swapped? A-B should be A/B, and A/B should be A-B?

Phil
pmccand is offline
Old 01-08-13, 07:57 AM
Senior Shackster

Join Date: Dec 2011
Posts: 239
Re: A-B wave subtraction

Quote:
pmccand wrote: View Post
I believe I found the problem with the difference equation. Seems as if there is a program bug. It looks at first glance that the A-B is actually A/B and the A/B tab performs a A-B. I will report the bug.

Phil M.cCan.d
This is not a bug. Differences in frequency domain are differences in correlation in time domain, which involves convolution.

Adding two copies of the same IR response just changes the amplitude of the response, not the frequency response. As phase and frequency response are shifted for one of these copies, the differences are a new phase and frequency response, not the difference in frequency response.

In REW multiplication symbol is used to represent convolution in the time domain. Division is effectively convolution of one IR with the inverse transfer IR of a second IR.

Frequency response is referenced to single point in time.

dspguide.com is great reference for studying fundamentals of signal theory and practice of DSP.

Last edited by Barleywater; 01-08-13 at 08:03 AM.
Barleywater is offline

 Bookmarks

 Tags ab , subtraction , wave

Message:
Options

## Register Now

Random Question
Random Question #2

User Name:
OR

## Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.