Post ringing

[PhilActiveAudio]

Hi Uli,

While I am testing the various functions of Acourate to improve my understanding, I am probably doing some mistakes when performing logsweep measurements, pulse processing or playing with room macros. So far, I am using several logsweeps performed at 1,2m that were intended to measure drivers delay. It is thus not a room measurement but it is enough to learn how to create a 3 voices crossover and then generate filters that are being used with Roon. The resulting filter is not that good but good enough to listen to music .

However, I found that if I was able to get rid off (or almost) pre-ringing, I am having post-ringing that I can not remove. Reducing or enlarging FDW in macro 1 and/or 4 does not remove post-ringing...

Post ringing.JPG (47.63 KiB) Viewed 2000 times

In general, what could be the cause of post-ringing in the step response?
What should I do to limit or remove post-ringing?

For your information, my room is large (5,5m x 8m) but not treated (it is my leaving room). And there are many sources of reflection (concrete walls, wooden floor, lot of windows)

[UliBru]

Do you mean the oscillations between time 0.062 and 0.064?
With high chance these oscillations are connected to to a rapidly falling frequency response at high frequencies.
Sometimes users do not run the logsweep to fs/2 = 24 kHz (q 48 kHz samplerate) but just to e.g. 20 kHz
This is equivalent to running a sweep to fs/2 but with a brickwall lowpass 20 kHz. The steep slope causes ringing. It cannot be removed by any filter setting.
So a picture of the corrected frequency response will tell more.

[PhilActiveAudio]

Yes, I mean these oscillations. I understand from your post that it is not post ringing.
I did the recording at 96 kHz but the sweep was limited at 30 kHz as the mic and the tweeter may not go beyond. Should I go at 48 kHz?

Here is the left pulse:

LPulse.JPG (251.06 KiB) Viewed 1972 times

One comment: the High Frequency Treatment in Room macro 1 was set to 25 000 Hz. May that be the root cause?

[UliBru]

The 30 kHz limitation causes the ringing.
But: if you measure up 48 kHz this does not necessarily help. Because the the tweeter and the mic also has to go up to 48 kHz.
Otherwise you see the ringing of the tweeter/mic lmitiation.

[PhilActiveAudio]

If I understand well, should I now perform logsweep measurement at 48 kHz with a measurement limited at 24 kHz ?
Can I convert 96 kHz measurement to 48 kHz with Acourate ?

[UliBru]

Hi Phil,

I would recommend to record again with 48 kHz samplerate and sweep end frequency 24 kHz.
This is what I always recommend.
A samplerate conversion is possible but it introduces an additional necessary brickwall filter. Which will also show up ringing.

Anyway it should be clear that you cannot correct what you cannot measure, e.g. high frequency limitation by tweeter, microphone or sweep end parameter.

[PhilActiveAudio]

Thanks Uli, I will do so.

[PhilActiveAudio]

I made some new measurements at 48 kHz / 24 kHz and then proceeded with Room macros. I still have the same issue with oscillations as you can see below:

Pulse 48.jpg (154.18 KiB) Viewed 1924 times

Any idea of what the problem is as the 48 kHz did not correct the issue?
I can't tell if I can hear it or not... but it is quite weird.

I look back to Mitch book or Archimago article. They sometimes measured at 96 kHz with a sweep end set at 30 kHz but did not seem to face any problem later on. I am a bit lost...

[UliBru]

Please also post the according frequency response of the test convolution

[PhilActiveAudio]

Recording was made at 48 kHz, 1,5m distance and 1m height (to prepare XO), from 10 Hz to 24 kHz, with standard values for fade in and fade out.

Pulse left:

Pulsu48L.JPG (278.88 KiB) Viewed 1920 times

Pulse right:

Pulsu48R.JPG (282.55 KiB) Viewed 1920 times