Getting the most out of my audio system

[Adi777]

To know the RT60 target, I would need to know the dimensions of your room (L, W, H).

3,50 x 5,80 x 2,55m. Speakers are on the longer wall - 5,80m. I made measurements like you wrote - from listening position, on the middle - between speakers. Room is open for kitchen. Okay, first I will order carpet, a big one.

 

[Keith_W]

Not quite sure what that last measurement was supposed to show since there are no labels or description of any kind. But anyway:

The green line is your RT60, and the reddish-brown lines are the upper and lower targets calculated for your room size and a nominal RT60 of 500ms. You can see that your room is a bit too reverberant.

 

[Adi777]

Not quite sure what that last measurement was supposed to show since there are no labels or description of any kind.

From AJ, Soundfield Audio.

 

[Keith_W]

From AJ, Soundfield Audio.

That's reassuring. If that really is the response of the loudspeaker, then you should be able to diagnose and fix that dip. The book contains a procedure to diagnose the cause of dips. Dips that are wide and deep is usually the confluence of several dips occuring at the same frequency. The solution is to find a better position for the speaker. Push them towards the wall and re-measure, pull them away and re-measure, etc.

 

[Kaameelis]

Are you sure you had measured L and R signal? How you put mic so precisely on same distance from left and right speaker? If distance is different about 11 mm or more, then you must see fall in high frequencies and comb filter.

 

[Keith_W]

Are you sure you had measured L and R signal? How you put mic so precisely on same distance from left and right speaker? If distance is different about 11 mm or more, then you must see fall in high frequencies and comb filter.

You can answer this question by looking at the Energy-Time Curve.

See that double impulse at the start of the ETC? The second impulse is 0.492ms delayed compared to the first. This means that the microphone is (343/0.492) = 168mm closer to one speaker than another. This is from his first measurement, where he said he had his mic pointing at his left speaker and then he swept both speakers together.

That's his second measurement. It also shows a double impulse, but this time the peaks are much closer to each other. The time discrepancy is 0.092ms = 31mm closer to one speaker than the other.

 

[Mort]

@Keith_W can tell what ya had for lunch with an .mdat

 

[Kaameelis]

You can answer this question by looking at the Energy-Time Curve.

But why then comb filter is not visble in FR?

 

[Keith_W]

But why then comb filter is not visble in FR?

Probably because the comb filtering is occurring at pretty high frequencies and at those frequencies, there are a lot of reflections contaminating the measurement, thus obscuring the comb filtering.

A delay of 0.492ms should give you comb filtering with the first cancellation point at 1016Hz. The red measurement is the one with the 0.492ms L-R delay, and the green measurement has the 0.092ms delay. They look different from 1kHz up don't they?

 

[Cote Dazur]

How can I audibly improve the sound?

Speaker Placement | Cardas Audio

The Cardas Speaker Placement Guide. Setting up speakers in a golden ratio, golden cuboid, golden trapagon room, etc.. Speaker placement calculators.

www.cardas.com

 

[melomane13]

View attachment 500186
View attachment 500188

add a good subwoofer to fill the dip below 50Hz and improve the range, as Earl Geedes explained, then equalize the bass

 

[Adi777]

I hadn't thought about it before, but maybe I could move the speakers to the shorter wall, the one on the picture.
I can move fireplace.
I have a high WAF.

 

[Kaameelis]

Probably because the comb filtering is occurring at pretty high frequencies and at those frequencies, there are a lot of reflections contaminating the measurement, thus obscuring the comb filtering.
--
A delay of 0.492ms should give you comb filtering with the first cancellation point at 1016Hz. The red measurement is the one with the 0.492ms L-R delay, and the green measurement has the 0.092ms delay. They look different from 1kHz up don't they?

1016 Hz and 5435 Hz are not so high that comb filters can be invisible. This are more likely one channel measurement with big reflections. I have not real speakers comb filter measurements on hands as I avoid this kind of measurements, but there must be -10 dB and more holes on cancellation frequencies.

 

[Sokel]

1016 Hz and 5435 Hz are not so high that comb filters can be invisible. This are more likely one channel measurement with big reflections. I have not real speakers comb filter measurements on hands as I avoid this kind of measurements, but there must be -10 dB and more holes on cancellation frequencies.

Removing a bit of reflections (green) shows that what it needs is a better set of measurements, L and R separately but AFTER centering the mic properly and after finding the proper acoustic center for height.

This broad dip at 9kHz-10kHz is a common mistake about mic placement.
It's far easier to find the proper place by using MMM at first and play around with it.

 

[Keith_W]

1016 Hz and 5435 Hz are not so high that comb filters can be invisible. This are more likely one channel measurement with big reflections. I have not real speakers comb filter measurements on hands as I avoid this kind of measurements, but there must be -10 dB and more holes on cancellation frequencies.

I can assure you with 100% certainty that these are two speakers being measured.

Here is another view of the first measurement, showing the double impulse. Notice how the second impulse is an exact copy of the first, but with reduced gain? Now look at the photo of his room. If the second impulse was a reflection, you need a reflective surface very close to the speaker, e.g. a wall or maybe the speaker sat directly on a desk.

Re: the comb filtering, as @Sokel points out, all you need to do is a bit of FDW and it will be revealed. Thanks, I forgot about that one

 

[Kaameelis]

I made test on my system, put mic in way that time difference between channels is 0.1 ms and I get this. Comb filter is with more than -10 dB holes and first cancellation is at about 5kHz.