Graph psychoacustic

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Hi

Is the graph psychoacustic the frequency curve that one perceives with the ear? Should one use this smoothing to make corrections with the DSP?

Alpi
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
Psychoacoustic smoothing is an approximation to the perceived balance. Variable smoothing is a better choice for making corrections.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Thanks then I did it right.
What does the graph "Spectrogram" tell me in wavelet mode?
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
The wavelet spectrogram is useful for time alignment of drive units or speakers and subwoofers.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Hello John

Thank you for the explanation.
I have four measurements of my home theater here.
Measured is the left speaker + the double bass array (2x4) + my infrasub woofer 4x18 "Dayton.
The crossover from the infrasubwoofer to the DBA is 35HZ and the crossover from the DBA to the left main speaker is 100Hz.

what do the four measurements show you and why is the RT60 so bad even though the waterfall is good (<300ms) and RT60 (700ms)?

Thank you very much
 

Attachments

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
Please post mdat files rather than images. RT60 is meaningless below the rooms' transition band, which is around 200 Hz for domestically sized rooms.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Hi

Here are the measurements
If you want I can also measure the mains individually

thanks
 

Attachments

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
Thanks. The spectrogram shows the mids lagging the tweeter by 0.7 ms or thereabouts, then a much larger lag of about 9 ms to the bass region. The treble could be delayed to better align with the mid and both could be delayed to better align with the bass, though the audibility of such timing offsets is a topic of debate.

26300
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Thanks for the assessment
Tomorrow I try to improve the timing.
Can I also see that on other graphs?

thanks
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
Not so easily, the Group Delay plot when using a frequency-dependent window of 5 or 6 cycles width comes closest.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Hello John, where are you from? Texas?
I am from Tyrol :-)
With REW you've done something great, unfortunately I understand only a little. OK. I understand SPL, Waterfall, etc., but there are such great features in REW. is the waterfall under 100 Hz real? Or only calculated?

I'm writing something to my speakers, these are 2 way speakers with a 12 "TT and 1" HT horn.


Thanks and nice greetings from Tyrol!
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
I'm from the UK, Horst. All waterfall data is calculated from the impulse response, as is the case for most things including the SPL and phase plots. The impulse response reflects the behaviour of the system measured, so it is all as "real" as the measurement.
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
I don't understand your question, sorry, and there is nothing called DBA 1 or DBA 2 in that file.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Hi John,

Sorry for my incomplete question. The last two measurements.
Once just that inverted the back array without delay (DBA 2) and once the back array inverted and a delay of 15.5ms (~ space length) - DBA 1.
Crossover = 100Hz
Which is the better measurement or which setting do you find better?

Kind regards
 

Attachments

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
2,216
DBA 2 looks better to me at that measurement position, but to evaluate how successful a DBA is you should measure at all the positions of interest.
 

Horst

Member
Thread Starter
Joined
Feb 21, 2018
Messages
13
Thanks for the assessment and the tip.
I actually only have one reference place. :-). For the sake of interest, I will measure on several positions.

thanks
 
Top Bottom