[kohno]

kohno

2 December 2020 at 8:14 pm #299

Hello again,
This is just a report that the problem was resolved.
Simply it was my mistake.
I normally use 1/6 Oct smoothing on REW, but 1/48 Oct plot showed fine structure of measured phase that is different from 1/6 plot.
I imported it into FIR Creator and generate a filter that makes it linear, and then it works well with OpenDRC.
I am satisfied with the outcome.
Regards,
Kohno

[kohno]

kohno

7 August 2020 at 3:14 pm #259

Thank you for your advice. Based on your input I took several measurements and it was appeared that the rotation was happened sometime and not happened sometimes. This loudspeaker has 300 cm long horn so it is not usual as you pointed. Anyway thanks again!

Regards,
Kohno

[kohno]

kohno

6 August 2020 at 11:29 pm #257

Hello EA support,

Thank you for your advice. This loudspeaker is back loaded horn type. I use up all 8 x 2nd order all-pass filters (maximum phase) for phase rotations in lower frequencies, so I cannot do it unless I upgrade to FIR Designer. FIR has 6100 taps.
I have 3 questions.
1. That REW shows 360 degrees phase jump means Auto Phase tab corrects 2 x 360 but cannot remaining 360, therefore it is appeared in REW measurement, correct?
2. You mentioned “depending on the amount of smoothing in use”. I use 1/6 Oct smooth. Which one (1/3 or 1/12) will be better for the case?
3. Auto Phase tab and Export tabs show that phase is corrected to linear. Why it does not show 360 degree rotation in these tabs as REW says if it cannot correct?

Regards,
Kohno