Impulse form change on estimate time delay.Why?

asilas

Registered
Thread Starter
Joined
Aug 13, 2018
Messages
5
Hello,
If i do impulse response offset it just shifts impulse without changing it. Phases changes as should be, but what causes impulse response waveform change if i do estimate time delay or set t=0 at cursor. What is the math behind this and why is that? What is happening here? REW version 5.20 beta 6

Thanks in advance,
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
Please attach an example mdat file which shows the behaviour you described and say what differences you are seeing.
 

asilas

Registered
Thread Starter
Joined
Aug 13, 2018
Messages
5
attached the file and two screenshots before and after clicking "set t=0 at cursor
Screenshot 2019-03-14 at 23.00.50.png Screenshot 2019-03-14 at 22.52.04.png "
 

Attachments

  • fullrangeimpuls.mdat
    814.5 KB · Views: 2

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
The problem is that most of your impulse response is missing, there is just a small portion from about -6 ms to 16 ms, then a discontinuity at each side. Fractional sample shifts are achieved using the timeshift property of the FFT, but the discontinuities in the response generate large oscillations at frequencies at and near Nyquist. It would be better to import a complete response and apply any windowing in REW.
 

asilas

Registered
Thread Starter
Joined
Aug 13, 2018
Messages
5
Thanks,
Understood. I measure impulse response with other programs, because on my linux pc i must use jackaudio server, but REW is not jack aware program unfortunately. But yes REW is the best on user interface compared to others!
 
Top Bottom