Dirver linearisation with measurement using 96K Sampling rate

Post Reply
User avatar
Ackcheng
Posts: 62
Joined: 19 Oct 2019 18:31

Dirver linearisation with measurement using 96K Sampling rate

Post by Ackcheng » 15 Aug 2020 14:11

Dear Uli and friends,

I am now trying to linearize my drivers again using a sampling frequency of 96kHz.

The XO is generated in 96K and the measurement are all done at 96K.
I went through FDW, phase extraction, inversion, convolution
Now comes the step i need to cut and window. I am not sure if I should choose the length as 65536 or 131072



The screen shots showed the current length as 196608 (a bit strange) and peak at 65536?
User avatar
Ackcheng
Posts: 62
Joined: 19 Oct 2019 18:31

Re: Dirver linearisation with measurement using 96K Sampling rate

Post by Ackcheng » 15 Aug 2020 14:36

I think I work out the reason. I tried to combined different crossover types and convolved them together . After the convolution, I need to Cut and window to the 65536 peak symmetric filter tabs for each XO first
User avatar
UliBru
Posts: 434
Joined: 19 Oct 2019 13:58

Re: Dirver linearisation with measurement using 96K Sampling rate

Post by UliBru » 15 Aug 2020 14:55

Arthur,

by default Acourate uses 64k = 65536 samples/taps.
You can also select a length of 128k = 131072 which gives a two times better frequeny resolution (but also a higher CPU load). Anyway in this case select the result length of 131072 in the logsweep recorder.
And you may also select this length in the XO generator.
Stay well tuned
Uli

Moderator
Acourate system: JRiver/Roon -> AcourateConvolver -> miniDSP U-DIO8 -> TacT M/S2150 amps -> DIY horn speakers
Post Reply