etc smoothing plot normalized have in overlays diffrent heights

[user44455555]

Hello

can this with smoothing also normalized so peak is same height or maybe less smoothing possible as 0,1 ms help ?

 

[John Mulcahy]

Comparisons would be very difficult if every trace was normalised to a different level depending on how smoothing affected it. 0.1 ms is already a very small amount of smoothing, if that is too much use unsmoothed.

 

[user44455555]

when i use unsmoothed the best sounding speaker have the lowest height of the peak. when smoothed. then can say that the fewest height of the impulse is the best it have the fewest resonances ?. i have treated room and measure on same position er and microphone at 80 cm away. result stay same when i measure left and right speaker. this is unsmoothed. the worser speakers have another peak with 70% level and 55% level of peak

 

[John Mulcahy]

How much a peak is affected by smoothing depends primarily on how narrow it is. For the initial peak (corresponding to the direct sound) higher bandwidth (more HF extension) will produce a narrower peak, but very early reflections will affect that.

 

[user44455555]

there is huge diffrence between speakers. maybe you like do a new designed measure test that show automatic the distortion in compare to a burst wave. i measure 1 cm away in nearfield and 10 cm away.
so no room problems. I allign left to beginning of impulse because the middle of wave can not easy see. because of this the best speaker is this with the most left peaks in compare to originalt. On 160 hz or higher the focal alpha 65 is much better as the kali lp 6 but on 90 hz the focal is worser. on 160 hz there is not so much diffrence in focal. This i hear to. mid is better on focal but bass is more muddy as in the kali. there are speakers that cost thousands of $. nobody know whats better with them because current available tests let all speakers look nice .

maybe you have a idea for a better test that show the truth of a speakers in more realistic signals to music ?