Suggestion for improving harmonic distortion analysis in REW - Home Theater Forum and Systems - HomeTheaterShack.com

Old 11-08-15, 09:18 AM Thread Starter
Shackster
Etienne Dechamps

Join Date: Mar 2010
Location: London, United Kingdom
Posts: 22
Suggestion for improving harmonic distortion analysis in REW

Hi,

I would like to point the attention of REW developers to this fresh AES paper published a few days ago: Synchronized Swept-Sine: Theory, Application, and Implementation

The paper claims that using their own swept sine function and analysis method leads to better HHFR (Higher Harmonic Frequency Response) measurements. Note that most of the paper deals with correctly measuring the phase of harmonics, which is probably not something REW users care about, but the paper also mentions a hidden gem in section 4, illustrated by figure 6:

Quote:
The deconvolution using [the traditional method] set the upper bandwidth limit to f2. (b) The deconvolution using [our method] does not limit the frequency bandwidth of HHFRs.
What I believe this means in practice is: using this method, it is possible to, say, measure using a 20-200Hz signal, and still be able to measure harmonic distortion for harmonics higher than 200Hz (f2). Currently, when using REW to measure distortion, harmonics higher than the measurement limits are not shown. Using that new method should hopefully remove that limitation.

The authors kindly provide a MATLAB implementation of their method, so it should be relatively easy to translate it into usable code.
e-t172 is offline

Old 11-10-15, 04:38 PM
REW Author

John

Join Date: Apr 2006
Location: UK
Posts: 6,296
Re: Suggestion for improving harmonic distortion analysis in REW

The AES email reached my inbox yesterday and I read that paper last night. Interesting. Use of an analytic expression for the FT and allowing that to extend beyond the frequency range of the actual signal used seems equivalent to synthesising a sweep that extends beyond f2 and using the FT of the extended sweep in the conventional calculation. The way they suggest applying the fade in and fade out windows seems wrong to me as it would alter the derived response, perhaps I am misunderstanding their wording. I'll try an implementation at some point and see how it performs.
JohnM is offline

 Bookmarks

 Tags analysis , distortion , harmonic , improving , rew , suggestion

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.