Filtered impulse response - Page 3 - Home Theater Forum and Systems - HomeTheaterShack.com

Reply
 
Thread Tools
post #21 of 28 Old 03-16-15, 08:53 AM
Elite Shackster
 
Join Date: Mar 2009
Location: Arizona, USA
Posts: 1,562
My System
Re: Filtered impulse response

I can imagine a car environment is very different and challenging for measuring phase. I am a little concerned however regarding your comment on your ear indicating a phase problem. Most if not all reputable references to driver phase differences for a room setup indicates phase itself is a non issue. There apparently may be an extremely small detectable impact in a midrange XO with some signals. This presumes that there isn't a gross difference in time (several wavelengths) and the SPL has been controlled. I'm thinking that what you are hearing is not because of phase differences between drivers but something else.

We do know that we are more sensitive to SPL variations and those can be caused by modes, reflections, diffraction as well as phase differences between drivers.

We also know it is difficult to find a good EQ scheme that corrects issues that are correctable and avoids those that are not.

I'm just sharing my thought that you may be focusing on the wrong issue. It may be more related to car acoustics, speaker/LP placements, EQ related, or ??

To your question of which filter to choose: I would think either choice would work fine in room environment. If this method is not working in your case it should be good to just enter the delays from distance measurements and make any minor corrections based on achieving the smoothest SPL in the XO range.
jtalden is offline  
Sponsored Links
Advertisement
 
post #22 of 28 Old 03-16-15, 05:37 PM Thread Starter
Shackster
 
geopango's Avatar
 
Join Date: Jun 2012
Posts: 41
Re: Filtered impulse response

Good luck with you quest. Personally I wouldn't go anywhere near attempting this kind of measurement in such a chaotic environment where probably 60% of what you hear comes from reflections.
geopango is offline  
post #23 of 28 Old 03-18-15, 03:59 PM
Shackster
 
Join Date: May 2014
Location: Charleston, SC
Posts: 98
Re: Filtered impulse response

Here are some IR measurements I just took. I have not messed with the Filtered IR on these yet.

Here are the physical measurements to the LP micropone using a tape measure:


R HLCD = 73"
L HLCD = 62"
R Midbass = 61"
L Midbass = 46"
Subwoofers = 38"

My Right Horn-Loaded Compression Driver (HLCD) is the furthest speakers from the LP @ 73"


I've attached an MDAT file with the 0ms delay measurments for each driver, and then I used the unfiltered IR to delay everything so that the initial rise of the IR closely matches the Right HLCD.

Obviously the subwoofer is almost impossible to align this way due to the size of the IR vs. the other drivers.


Does this look like I am at least in the ballpark?


After this I can apply some crossovers and then try using the filtered IR technique again.


Thanks in advance.
subterFUSE is offline  
 
post #24 of 28 Old 03-18-15, 07:19 PM
Elite Shackster
 
Join Date: Mar 2009
Location: Arizona, USA
Posts: 1,562
My System
Re: Filtered impulse response

I wouldn't have confidence with any analysis based on this data. The problem is there is nothing to align until the XOs are chosen and active. The XO filters change the needed delays so any delay settings done without XOs in place doesn't help. If XOs had be applied your delays were correct for a reasonable RTA alignment starting point. You could just make that change and then set the delays in your preferred manner.

I would also recommend full range sweeps for all drivers. The XOs protect them and with full range sweeps the timing accuracy will be assured. [I am unsure if REW will always maintain relative timing accuracy when the sweep ranges are so different. I haven't really found a problem, but worry that SW sweeps may not always align at the same trigger point as a HF sweep. This concern may not be justified.]

Other changes that may help when setting the timing:
> The SPL response is extremely rough it would be better to apply some basic EQ to the drivers to remove the largest peaks to help smooth the response a little. The ragged response can make it more difficult to determine phase timing.

> I also would bring down the SW level to nearer the MW level. The large SPL discrepancy is a minor irritant when using trace arithmetic. If you don't use trace arithmetic method this probably makes no difference.

After the timing is set then the Levels and EQ can be reset as needed to fit your house curve. Those changes will not disrupt the XO phase timing.

If you want me to provide recommended delays for your setup I will do that. To do that I would need new measurements.
> Apply the XOs
> Smooth the drivers SPL a little by removing the larger peaks.
> Use the REW loopback timing feature as before.
> Set all delays to zero
> It would be helpful to reduce the SW level to nearer the MW. [Scaling it in REW does not help.]
> Sweep each driver full range
jtalden is offline  
post #25 of 28 Old 03-18-15, 08:23 PM
Shackster
 
Join Date: May 2014
Location: Charleston, SC
Posts: 98
Re: Filtered impulse response

Quote:
jtalden wrote: View Post
I wouldn't have confidence with any analysis based on this data. The problem is there is nothing to align until the XOs are chosen and active. The XO filters change the needed delays so any delay settings done without XOs in place doesn't help. If XOs had be applied your delays were correct for a reasonable RTA alignment starting point. You could just make that change and then set the delays in your preferred manner.

I would also recommend full range sweeps for all drivers. The XOs protect them and with full range sweeps the timing accuracy will be assured. [I am unsure if REW will always maintain relative timing accuracy when the sweep ranges are so different. I haven't really found a problem, but worry that SW sweeps may not always align at the same trigger point as a HF sweep. This concern may not be justified.]

Other changes that may help when setting the timing:
> The SPL response is extremely rough it would be better to apply some basic EQ to the drivers to remove the largest peaks to help smooth the response a little. The ragged response can make it more difficult to determine phase timing.

> I also would bring down the SW level to nearer the MW level. The large SPL discrepancy is a minor irritant when using trace arithmetic. If you don't use trace arithmetic method this probably makes no difference.

After the timing is set then the Levels and EQ can be reset as needed to fit your house curve. Those changes will not disrupt the XO phase timing.

If you want me to provide recommended delays for your setup I will do that. To do that I would need new measurements.
> Apply the XOs
> Smooth the drivers SPL a little by removing the larger peaks.
> Use the REW loopback timing feature as before.
> Set all delays to zero
> It would be helpful to reduce the SW level to nearer the MW. [Scaling it in REW does not help.]
> Sweep each driver full range
Here are some new measurements with XOver and EQ applied.
Mic at LP, Driver seat.
All drivers @ 0ms delay.
No phase adjustments.
Everything confirmed wired in correct polarity.


Thanks for the help!
subterFUSE is offline  
post #26 of 28 Old 03-19-15, 11:10 AM
Elite Shackster
 
Join Date: Mar 2009
Location: Arizona, USA
Posts: 1,562
My System
Re: Filtered impulse response

Okay, here is the timing I suggest:

SW (Inverted) = 0ms
L MB = 6.643ms
R MB = 5.465ms
L HLCD = 5.715ms
R HLCD = 4.966ms

The SW polarity needs to be inverted and set at 0ms delay. The other drivers to be delayed as indicated.

> This will achieve close phase tracking through the 50HP/55LP lower XO and will result in good SPL support.

> I aligned the direct sound phase of the left channel 892LP/779HP XO and right channel 892LP/689HP XO without regard to the reflections. The right channel results in good SPL support through the XO. The left channel results in a deep null right at the XO due to a very strong reflection in the L HLCD. I do not know if the very unusual XO settings contributed to the issue.I would not be concerned that there is a null in that XO as there are several others scattered around the area in both channels anyway.

What was the reasoning for the overlapped and asymmetry of the XO settings? Is this something that is done in car setups?

Is locating the listening position at the driver also typical of car setups? I would have guessed better results might be achieved with a mic placement centered between the front seats and then EQ averaged in that area. I have no experience nor done any reading on the subject though.
jtalden is offline  
post #27 of 28 Old 03-19-15, 05:12 PM
Shackster
 
Join Date: May 2014
Location: Charleston, SC
Posts: 98
Re: Filtered impulse response

Quote:
jtalden wrote: View Post
Okay, here is the timing I suggest:

SW (Inverted) = 0ms
L MB = 6.643ms
R MB = 5.465ms
L HLCD = 5.715ms
R HLCD = 4.966ms

The SW polarity needs to be inverted and set at 0ms delay. The other drivers to be delayed as indicated.

> This will achieve close phase tracking through the 50HP/55LP lower XO and will result in good SPL support.

> I aligned the direct sound phase of the left channel 892LP/779HP XO and right channel 892LP/689HP XO without regard to the reflections. The right channel results in good SPL support through the XO. The left channel results in a deep null right at the XO due to a very strong reflection in the L HLCD. I do not know if the very unusual XO settings contributed to the issue.I would not be concerned that there is a null in that XO as there are several others scattered around the area in both channels anyway.
This is great. Thanks for the work you put into this. I can't wait to give these delays a try.

I actually have the ability in my DSP to adjust the phase angle in 11.5 steps. So I can invert the sub 180 degrees, or I could rotate the phase to almost any angle desired.

Quote:
What was the reasoning for the overlapped and asymmetry of the XO settings? Is this something that is done in car setups?
I was trying for an acoustic crossover of 900 Hz for the horns. Using REW's EQ section and a house curve, I set the crossover for 900 Hz @ 24 dB slope and then measured sweeps at different crossover settings in the DSP until I found the setting that made the response follow most closely to the target line. In this case, that meant a different crossover frequency point for the right horn vs. the left.

The horns can play lower than 900Hz, but when I was tuning by ear I noticed that below 900Hz the horns began to pull the stereo image lower. With an 800Hz crossover, pink noise samples were coming from the dash below the air vents. When I moved the crossover to 900Hz they moved upward to the center of the windshield where I want them.

I might need to play with the crossover points some more to get them right. Do you think I have too much overlap with the midbass?


Quote:
Is locating the listening position at the driver also typical of car setups? I would have guessed better results might be achieved with a mic placement centered between the front seats and then EQ averaged in that area. I have no experience nor done any reading on the subject though.
Yes, driver seat as the LP is common for car audio. The asymmetrical distances to the speakers make it difficult to achieve good stereo imaging for both passengers simultaneously. It can be done, but installation locations of speakers becomes a huge deal for 2 seat listening.

With 1 seat listening, DSP with time alignment can be used to overcome the less-than-ideal speaker placements in a car. But when you use time alignment to fix the stereo imaging for 1 person, it usually has a detrimental effect for the other seat.


When I am measuring for frequency response, I usually take 6-8 measurements around the head area and then average them together. But for time alignment, I just used a stationary microphone at the center of when my head would be. I figured taking an average for impulse response TA would probably make things more difficult.
subterFUSE is offline  
post #28 of 28 Old 03-19-15, 09:54 PM
Elite Shackster
 
Join Date: Mar 2009
Location: Arizona, USA
Posts: 1,562
My System
Re: Filtered impulse response

Setting a variable phase control to 180 is not the same effect of reversing the polarity. If the DSP does not provide an option to invert the polarity then just reverse the wire connection at the SW driver or at the SW P-Amp. The variable phase control should be left set to 0.

There is nothing wrong with setting XO filters to under-lap or even overlap if careful experimentation shows that is advantageous. I was just wondering how you tested it. Unless phase is accounted for in the experiment it may be misleading. I cannot really help with that question as there is significant work to setup each timing correctly to properly evaluate different XO settings.

Setting the XO filters differently on the 2 channels is something I have never done. It is considered good practice in having a symmetrical room, identical XOs, and even similar EQ for both channels in the MF and HF. With your car situation it's possible there is something to gain in having XO differences. I wouldn't know.

I agree that an average near the LP for EQ purposes is a good idea, particularly so when there are so many strong reflections.
jtalden is offline  
Reply

Bookmarks

Tags
filtered , impulse , response

Quick Reply
Message:
Options

Register Now




PLEASE COMPLETE ALL REQUIRED FIELDS BELOW... THANKS!

REQUIRED FIELDS ON THIS PAGE
YOU MUST COMPLETE ALL OF THESE

Username
Password
Confirm Password
Email Address
Confirm Email Address
Random Question
Random Question #2




User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.

Password:


Confirm Password:
Email Address
PLEASE READ BELOW PRIOR TO ENTERING AN EMAIL ADDRESS!

ATTENTION!

YOU MUST ACTIVATE YOUR ACCOUNT!

Activation requires you reply to an email we will send you after you register... if you do not reply to this email, you will not be able to view certain areas of the forum or certain images... nor will you be able download software.

AN INVALID EMAIL ADDRESS WILL CAUSE YOUR ACCOUNT TO BE DELETED!

See our banned email list here: Banned Email List

We DO NOT respond to spamcop, boxtrapper and spamblocker emails... please add @hometheatershack DOT com to your whitelist prior to registering or you will get nowhere on your registration.


Email Address:
OR

Log-in










Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML is not allowed!
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

 


For the best viewing experience please update your browser to Google Chrome