Can I save a predicted EQ result as a measurement?

[John Warden]

Hi

I've been using REW for some time in various ways, and I've found it really useful. I'm looking at a project to model how best to combine four subs. Being able to model is much faster than trying something and remeasuring. I'm trying to do something that I'm not sure if REW handles.

If I want to examine the interaction of more than one measurement, the Alignment tool and Vector average function are really useful. They also preserve the decay information, which is also of interest. However, in using these two functions there are limits to the changes you can make to the original measurements - amplitude and timing are the only two that I've found.

What I'd like to be able to do is apply EQ to a measurement, and then save the predicted result as a measurement for further analysis. Is that possible?

Many thanks.

 

[John Mulcahy]

Yes, though it is a bit roundabout. You can export the EQ filtered impulse response as a WAV file (File -> Export impulse response as WAV) then import that WAV.

 

[John Warden]

Yes, though it is a bit roundabout. You can export the EQ filtered impulse response as a WAV file (File -> Export impulse response as WAV) then import that WAV.

Hi John

Thanks for the reply. I've found what you pointed me to, and I've just done a quick test to export the impulse response of an unmodified measurement, import the impulse file and then compare it to the original. I noticed two things:

• The frequency response is slightly altered (see below)
• The timing moves by over 50 ms according to the alignment tool, which is more than can be adjusted

Am I missing something here? For my purposes the timing alignment is important, so I need to understand how to handle this correctly.

Many thanks.

 

[John Mulcahy]

Window widths and reference time will affect the responses, and you will need to select any cal files that should be applied using the Change Cal button. A WAV file has no concept of absolute time, but some fixed alignments can be applied by exporting a windowed response (see the help) and changing the analysis preference for where t=0 is set for imported impulse responses. You can also shift the t=0 point for an imported response using the various controls on the Impulse graph.

 

[Clayton Williamson]

The new "generate measurement from predicted" is an excellent addition to make this a simple and direct procedure. I find it very handy to compare before / after / various EQ iterations. Thanks for continuing to make improvements!