FSAF (Fast subband adaptive filtering) measurement

[user44455555]

Changing the tweeter delay does not change non-linear distortion, it changes the linear response at the measurement point from the sum of the tweeter and woofer signals. That sum is usually not flat, but you can see how it looks by taking any measurement, sweep or FSAF.

maybe when use music which contain transients and timing it is diffrent. https://www.avnirvana.com/threads/f...ltering-measurement.13810/page-15#post-125924
here i have the measures of speaker with diffrent tweeter delay. groupdelay show over 3 khz 0 ms for tweeter but it is 1 ms on the blue line. only in crossover range(when both speakers send enough output to microphone), the 1 ms delay tweeter measure show more jitter in crossover range.
the measure was on hearing distance. in treble the group delay is 0 and on woofer is worse jitter see but not a constant offset of 1 ms. normaly this test need repeat, more near, but it cost time. So when REW is able to load original and recordet that is digital processed, nobody can say it is the room influence or speaker influence that let it fail.

at 4 khz or later when only tweeter play it show not 1 ms, time is near 0 . the jitter in crossover range happen, because the waveform in this range is very worse when both speaker output. coee have problems to detect the cycles

but when have music and shift tweeter 1 ms, then on a simple diff compare huge diffrences happen. and FSF is able to test timing

 

[user44455555]

I get now with newest beta good results in mid and high freq. All i test currently of my speakers can measure now in nearfield few cm to woofer ..only i see when level is 76 db at 1 meter (measures with check level and a hand db measure instrument) at 50 hz distortion is 2%. when go to 80 db at 1 m distortion is 20%. there can not hear increase of distortion in bass, but the distortion view and play residial have a huge increase. on 83 db is even more. the microphone is not clipping. speaker position and microphone position was always same with stative. i also attach the wav files and files with save FSAF residual so can hear. same happen also with the large speakers, you can see in rew measure mdat i attach too from all . the test is drumtest 2 in the file burst test.zip here https://www.avnirvana.com/threads/f...ltering-measurement.13810/page-14#post-125864

 

[dcibel]

I think I wouldn't be surprised that a little micro-monitor with 3.5" driver struggles will bass, even at a relatively low volume. Potentially the little passive radiators are reaching their limits as well. Most/all drivers produce distortion (especially intermodulation) from bass, even high end ones and at low excursion levels. The modulation travels well into the midrange.

Give your residual audio a listen, I think there is nothing out of the ordinary in your measurement.

 

[user44455555]

I noticed this worse results on the larger speakers too. but i only meaasure at more levels with the adam d3v . with the nearfields microphone position i use need from the SPL value subtract 15 db to reach 1 meter value. I add a calcued level to the other speakers too.That the kali have much fewer SNR also in high freq is because the Kali do much hiss.

there can also look at distortion at 80 hz which should be no problem also for small speakers. At 80 db 1 m it is 2.86%. at 83 db at 1m it is 7.91%