Ran REW, now what?
- Thread starter sud98
- Start date
sud98
Well-Known Member
Made the measurements post moving forward by 3 feet and backwards by 1 feet.
I dont see any significant differences.
sud98
Well-Known Member
Both speakers plus sub
The peak at 37 and 74 Hz is more of a measurement artifact due to comb filteringhappening from floor reflections. If you can also post the step response, we can predict the first reflection point causing the combing. Gating the response for this measurement can smoothen the curves, but you will not get much information below 200 Hz.
sud98
Well-Known Member
Thanks to elango understood some other issues in the measurement. I redid all the measurements with individual speakers. The asio drivers were not configured properly and I was using java drivers which have its own limitation.
Let me know if you need more inputs.
Hari I can post the mdat file and if you have REW, you can check for the step responses, as I have not idea of the same, or else you will need to tell me how you need to get it.
Let me know if you need more inputs.
Hari I can post the mdat file and if you have REW, you can check for the step responses, as I have not idea of the same, or else you will need to tell me how you need to get it.
I downloaded your mdat file from the first post and this is the step response that i get,
https://ibb.co/hp26Xb
This looks a bit wierd to me. Can you confirm the following,
- have you measured the speaker with the mic at the tweeter axis?
- is the mic atleast 1 meter away from the tweeter axis?
The ideal step response looks like this https://ibb.co/nbrN5w
Your first reflection point looking at your step response is around 11.5ms. Converting this to distance this is around 13 feet and you are quite good with this value. But i still feel there is some measurement error hear as the floor reflection would be around 3 feet which is missing in the step response.
Can you do the following,
Measure only one speaker keep it in the center of the room, mid at approximately 1 meter distance from the tweeter axis and post its step response for analysis.
Cheers,
https://ibb.co/hp26Xb
This looks a bit wierd to me. Can you confirm the following,
- have you measured the speaker with the mic at the tweeter axis?
- is the mic atleast 1 meter away from the tweeter axis?
The ideal step response looks like this https://ibb.co/nbrN5w
Your first reflection point looking at your step response is around 11.5ms. Converting this to distance this is around 13 feet and you are quite good with this value. But i still feel there is some measurement error hear as the floor reflection would be around 3 feet which is missing in the step response.
Can you do the following,
Measure only one speaker keep it in the center of the room, mid at approximately 1 meter distance from the tweeter axis and post its step response for analysis.
Cheers,
Last edited:
That measurement is the subwoofer alone, like I asked? Also, why just 1 foot back? Is your seating close to the back wall?
No, they're not. Boundary interference (including floor bounce) shows up as a cancellation notch in the frequency response, not a peak.
Those peaks are from room modes. Rather than take my word for it, you can plug Sud98's room length into any modal calculator and see that those frequencies are resonances for a 15' dimension.
There are two rows of numbers in the graph above. The bottom row shows the first 4 length modes of his room.
sud98
Well-Known Member
Yes. Seating is next to the back wall.
https://drive.google.com/folderview?id=0B8SRo782zUzxV1ZWNDhJTi1MMDQ
Above link has all the pictures of the room for reference.
Sent from my ONE A2003 using Tapatalk
sud98
Well-Known Member
Here is the sub in different positions. let me know your thoughts.
That explains the peaks at 37Hz & 75Hz. All modes peak at room boundaries, which is where your seating is located (circled in green in the graph below):
The good news is that it is a length mode (peaks & dips change along the length of your room). So any seat that is the same distance from the back wall should have those same two peaks. Each of those peaks can be pulled down with a single band of PEQ. If you fix the problem in one seat, it will be fixed in all seats.
elangoas
Well-Known Member
That explains the peaks at 37Hz & 75Hz. All modes peak at room boundaries, which is where your seating is located (circled in green in the graph below):
The good news is that it is a length mode (peaks & dips change along the length of your room). So any seat that is the same distance from the back wall should have those same two peaks. Each of those peaks can be pulled down with a single band of PEQ. If you fix the problem in one seat, it will be fixed in all seats.
If the user shifts his listening position a feet forward from his present position, would it be of any benefit?..
The minimum and excess phase group delay can shed information about if these boom can be EQed. Any peak frequency in the excess phase group delayed can be easily EQed. But room treatments take precedence to EQing imo.
sound_cycle
Well-Known Member
Hi,
Yes of course, two plus ones. The trick would be to preserve phase and tonal balance in eq. I recently read about Equalizer pro which is free but windows only. I have not tried it out though.
Ciao
GR
Yes of course, two plus ones. The trick would be to preserve phase and tonal balance in eq. I recently read about Equalizer pro which is free but windows only. I have not tried it out though.
Ciao
GR
sud98
Well-Known Member
The big issue is the solid concrete walls which will be a cause of serious reflections. Not sure if I have to convert the entire section with panels. I would assume using wall carpet wont work as it would be too thin to be of some effect.
Once I get the other problems solved, I will look at the PEQ.
I have a third more expensive option of breaking my front wall and moving my wall about 3 feet. I planned to do that a few years down the line but that will change my room mode to slightly more manageable frequencies as also I can build a brick rather than concrete wall for atleast one surface.
I am assuming a device like this would be needed for PEQ, right?
https://www.minidsp.com/products/ht-series/nanoavr-hd
Once I get the other problems solved, I will look at the PEQ.
I have a third more expensive option of breaking my front wall and moving my wall about 3 feet. I planned to do that a few years down the line but that will change my room mode to slightly more manageable frequencies as also I can build a brick rather than concrete wall for atleast one surface.
I am assuming a device like this would be needed for PEQ, right?
https://www.minidsp.com/products/ht-series/nanoavr-hd
sound_cycle
Well-Known Member
Hi,
I am sure such drastic steps are not going to be necessary. IAC you will still have room modes and your walls will still be parallel after the expense.
For that money you could get WAF yukt room treatment.
Before you invest in any PEQ/ DRC you can audition them with a PC and the trial versions.
Have you seen this Perfect bass with Equalizer APO and REW Tutorial
ciao
gr
That won't help. You'll still have large peaks & dips, just at different frequencies.
Right.
Meanwhile, if you move your seating forward so that the listeners' ears (NOT the back of the seating) are 3 feet away from the back wall, it should improve frequency response. Best part: it's free (doesn't cost anything).
Other way 'round: flat regions (NOT peaks) in the excess group delay are minimum phase regions, making them good candidates for EQing.
https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html
Impractical at very low frequencies. Treating 37Hz with passive absorption would require several feet of thickness to absorb a 31.5-foot wavelength. By comparison, a single bad of EQ could flatten out that peak.
Follow HiFiMART on Instagram for offers, deals and FREE giveaways!
Similar threads
