Home Theater Forum and Systems banner

1 - 20 of 33 Posts

·
Registered
Joined
·
452 Posts
Discussion Starter #1
I have used REW a bit and can fumble my way through most of what I need to do. But I am certainly am not an expert.

I would like to find the distance/time delay from various speakers to my listening position. Is this something REW can do? If so how?

My reason is, I am running my subs wireless. The wireless system inserts a delay of n ms. I would like to solve for n.
 

·
Premium Member
Joined
·
1,562 Posts
If using a USB mic the REW feature to use is; 'acoustic timing'.
If using an XLR mic the REW feature to use is; 'loopback timing'.

Using either of these features will result in all measurements being set to a consistent reference time. The impulse positions in time then can be used as good indicators of the relative timing between main channels.

Relative impulse locations can also be used a rough indicator of the timing between the SW and the mains, but it is best to then fine tune that timing to provide the best SPL support in the XO range.

Each setup is a little different Here is an example of someone working through this process using the phase tracking approach.

I can help with the process, if this is the method you decide to follow.
or:
Given the needed measurements, I can also just advise the recommended timing, or just confirm the timing you have arrive at using other methods.

A web search will provide other simpler methods for subwoofer timing that may provide good results also.
 

·
Registered
Joined
·
363 Posts
Hi Jtalden,
not aceinc but could I have your advise and correct me if I am wrong. I use the acoustic timing reference since a while with success for the 5 satellites (mdat file joigned here). But I finf that it is more difficult to understand the response with the sub.
You thaught me how to fine tune the timing between the sub and satellitees (LFandC) using measurements to get the best SPL in the xo range. With this method, there is no significant differences between delays from 315cm to 390 into Rotel receiver. But the curves are different.
Which one should be the best?
I have chosen a delay of 360 cm Am I right?
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
Yes, the 360 setting is the starting timing that I normally recommend when using a trial and error method to set timing.

I can't tell much though with these particular measurements. The measurements show strong room modes in the XO range. This can make choosing the best timing very difficult. If there is no significant SPL change with the other settings then fine tuning using this method is not possible. When evaluating the SPL in the XO range I suggest using 1/3 octave smoothing on the overlay chart. Choose the setting with the best SPL fill in the XO range.

If you want to investigate this further I suggest:
Using the current 360 setting, invert the polarity of the SW. If the SPL looks better in the XO range that way then use that as the starting place and repeat the same series of distance setting options.

If you want me to try to see if I can find a better setting then post an mdat with 4 20-20k Hz measurements:
  • G
  • D
  • C
  • SW
The acoustic timing needs to be active and set to the same main speaker for all measurements, e.g., the G main channel. The SW needs to be muted/off for the 3 main channel measurements. The SW measurement needs to use one of the main channel outputs, e.g., the G channel and with that G main speaker muted/disconnected. That way the XO and proper SPL levels are active and correct. The modes in the room will still make the analysis difficult so there is no guarantee that I will find a better setting.
 

·
Registered
Joined
·
363 Posts
Yes, the 360 setting is the starting timing that I normally recommend when using a trial and error method to set timing.

I can't tell much though with these particular measurements. The measurements show strong room modes in the XO range. This can make choosing the best timing very difficult. If there is no significant SPL change with the other settings then fine tuning using this method is not possible. When evaluating the SPL in the XO range I suggest using 1/3 octave smoothing on the overlay chart. Choose the setting with the best SPL fill in the XO range.

If you want to investigate this further I suggest:
Using the current 360 setting, invert the polarity of the SW. If the SPL looks better in the XO range that way then use that as the starting place and repeat the same series of distance setting options.

If you want me to try to see if I can find a better setting then post an mdat with 4 20-20k Hz measurements:
  • G
  • D
  • C
  • SW
The acoustic timing needs to be active and set to the same main speaker for all measurements, e.g., the G main channel. The SW needs to be muted/off for the 3 main channel measurements. The SW measurement needs to use one of the main channel outputs, e.g., the G channel and with that G main speaker muted/disconnected. That way the XO and proper SPL levels are active and correct. The modes in the room will still make the analysis difficult so there is no guarantee that I will find a better setting.
There is 3 files (because of the site limit for downloads. Measurements following your instructions. No EQ, timing ref is right front channel, satellites set to small. For the "fun" I add xo 50hz measures. I can not switch polarity on the sub. Should I switch polarity of the center? Should it be the same? I can switch phase on the sub 180 degrees but if I am right it is not the same as polarity, isn't it?
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
Thanks. With the 70Hz XO you provided the 4 measurements that I needed for analysis. The SW was not measured with the 50 Hz setting so I was not able to evaluate that case.

As suspected, the room modes make any analysis difficult. I found 2 different SW settings that were marginally better for phase tracking of the direct sound. The room mode differences between the SW and the mains however result in only very marginally improved SPL response from your current good setting. The improvement is limited to the 50 Hz range and that applies to only the FL, FR channels. The CC is actually worse through much of the XO range as it is apparently a different speaker design resulting in a different phase rotation (2-way vs 3-way?). This is common for CC speakers and I normally would tend to set the SW delay ignoring the impact on the CC. In this case the negative impact to the CC is significant however. If you do decide to try the optional settings below it is very likely to be advantageous to invert the polarity of the CC.

In summary there are 3 options and it is not possible to say one of them will work better from the analysis I did. There easily may be a notable difference in the perception of the sound quality, so it may be worth experimenting with the other options.

The 3 cases:
  1. Your current settings.
  2. SW delay increased by about 5ms (-170cm) [likely best to invert the CC also.]
  3. SW inverted and delay decreased by about 2ms (+70cm) [likely best to invert the CC also.]
Other comments:
You indicated there is no EQ applied here. Be advised that some of the automated EQ systems may change the SW timing. That would make this effort moot. If the EQ is done manually the timing here is likely to hold up very well.

I don't know if a SW phase setting of 180° has the same impact as reversing the polarity. If you measure both ways with the conditions above (70Hz XO and FR reference channel) we can find out what it does. A phase filter will change the phase rotation and/or possibly the timing and a polarity change will just invert the phase.
 

·
Registered
Joined
·
363 Posts
Jtalden,
thank for your help. I hope that our discussion can help aceinc too.
You are wright, the center is a 2 ways but the left and right are 3 ways. All three speakers are front ported.
I checked the third file joined (fronts and xo 50 ) and the forth measurement in it is the sub at xo 50. I did not put the mains only (sub disconnected) in this file. If you need it , I will provide.
I will try the suggested settings and report but to be sure, to increase sub delay I add (+) (not -170) 170cm in Rotel to 360, no?


If I may, I have a choice to place the sub I can put it at the left side of the TV . Here are the measurements for this place, same xo to 70 but a little difference in the delay (315cm instead of 360cm). I thought it could help to show the difference espacially for the room mode at 60hz. I can't decide which place should be best.
If you were me which one would you prefer?
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
I will try the suggested settings and report but to be sure, to increase sub delay I add (+) (not -170) 170cm in Rotel to 360, no?
No, we need to decrease the Rotel distance setting. That will increase the delay. Delay and distance work counter to each other.

If I may, I have a choice to place the sub I can put it at the left side of the TV . Here are the measurements for this place, same xo to 70 but a little difference in the delay (315cm instead of 360cm). I thought it could help to show the difference espacially for the room mode at 60hz. I can't decide which place should be best.
If you were me which one would you prefer?
There are tradeoffs in moving the SW to the left. It changes the frequency of the major null of the SW from 106 to 60 Hz. The overall issues of the room modes in the XO range is still there however. The mains nulls are located at different frequencies from each other and the SW. We could probably find a setting that approaches the SPL results with the SW on the right, but I only tried the one alignment. The one that provided the best phase tracking for the direct sound. That SPL looked significantly worse than any of the 3 settings recommended initially however. Timing that helps one channel makes the others worse. Many of us have room and speaker location litations that limit our ability to approach an ideal setup.

SW Right vs Left.jpg
 

·
Registered
Joined
·
363 Posts
Ok report. Suggestion 1- (add 170cm and reverse polarity of the center sounds a little better that current setting. More listening to do before to decide. Also, I will try left side of the tv for the sub just to ear it and come back here. Thanks again.
 

·
Registered
Joined
·
278 Posts
I was wondering because when I ran it or read it somewhere it states the Reference speaker MUST have a Tweeter in it..

What happens if the reference speaker I used did not have a tweeter in it are the measurements void??
 

·
Registered
Joined
·
363 Posts
No, we need to decrease the Rotel distance setting. That will increase the delay. Delay and distance work counter to each other.



There are tradeoffs in moving the SW to the left. It changes the frequency of the major null of the SW from 106 to 60 Hz. The overall issues of the room modes in the XO range is still there however. The mains nulls are located at different frequencies from each other and the SW. We could probably find a setting that approaches the SPL results with the SW on the right, but I only tried the one alignment. The one that provided the best phase tracking for the direct sound. That SPL looked significantly worse than any of the 3 settings recommended initially however. Timing that helps one channel makes the others worse. Many of us have room and speaker location litations that limit our ability to approach an ideal setup.

View attachment 146169
Jtalden,
would you mind to explain how did you find the delays to add or to substract with REW (-170cm or +70? Is it too long to explain or could you suggest a reading for me ?
 

·
Premium Member
Joined
·
1,562 Posts
It is too complicated for me to detail at this time, but at the basically:
  1. Measure the SW and measure the L+R main speakers with acoustic/loopback timing applied.
  2. If necessary, offset the impulse of both measurements by the same amount such that the L+R speakers impulse (higher freq channel of the 2 measurements) is located near 0 ms.
  3. Use REW to duplicate the SW measurement. This SW-Copy is the measurement impulse that is adjusted for timing. The original SW and the L+R impulses will not be moved.
  4. Note the XO range for 25 dB rolloff (or more as desired), for example; 50-200 for a 100Hz LR-24 XO. That is the range we focus on for phase tracking.
  5. Set the smoothing to 1/48 octave for all 3 measurements.
  6. Open the IR Window panel and reset the ref time to 0.0ms (even if it already reads that). Also set FDW to 6 cycles/octave and then apply to all measurements.
  7. Open the phase measurement overlay window and uncheck the SW measurement and thus view the phase tracking of the L+R and the SW-Copy in the 40-300 Hz frequency range of interest.
  8. Now using trial and error, move the SW-Copy impulse location manually until the closest phase tracking is found. After each move of the SW-Copy impulse it is necessary to reset the window ref time to 0.0 ms. It may happen that the best tracking is found with the SW-Copy impulse inverted, so it is best to try both ways.
  9. When the closest phase tracking is found, the change in timing offset that is needed can be measured on the impulse overlay chart. Use 'Ctrl-right button drag' to measure the distance from the original SW measurement impulse to the new location of the SW-Copy impulse.
-----

  • With these low frequency LP measurements, the room modes can make this process very difficult. There may be no significant improvement using this method verses any of the other common timing methods.
  • I do this process for each main speaker individually rather than L+R together. That way it is easy to see the tracking issues in each main speaker. There is very often a big difference in the phase tracking and thus the SPL support for the 2 channels, so a compromise timing is needed. I am hoping that measuring the L+R together as suggested here results in that good compromise. I have never investigated this directly however.
 

·
Registered
Joined
·
363 Posts
It is too complicated for me to detail at this time, but at the basically:
  1. Measure the SW and measure the L+R main speakers with acoustic/loopback timing applied.
  2. If necessary, offset the impulse of both measurements by the same amount such that the L+R speakers impulse (higher freq channel of the 2 measurements) is located near 0 ms.
  3. Use REW to duplicate the SW measurement. This SW-Copy is the measurement impulse that is adjusted for timing. The original SW and the L+R impulses will not be moved.
  4. Note the XO range for 25 dB rolloff (or more as desired), for example; 50-200 for a 100Hz LR-24 XO. That is the range we focus on for phase tracking.
  5. Set the smoothing to 1/48 octave for all 3 measurements.
  6. Open the IR Window panel and reset the ref time to 0.0ms (even if it already reads that). Also set FDW to 6 cycles/octave and then apply to all measurements.
  7. Open the phase measurement overlay window and uncheck the SW measurement and thus view the phase tracking of the L+R and the SW-Copy in the 40-300 Hz frequency range of interest.
  8. Now using trial and error, move the SW-Copy impulse location manually until the closest phase tracking is found. After each move of the SW-Copy impulse it is necessary to reset the window ref time to 0.0 ms. It may happen that the best tracking is found with the SW-Copy impulse inverted, so it is best to try both ways.
  9. When the closest phase tracking is found, the change in timing offset that is needed can be measured on the impulse overlay chart. Use 'Ctrl-right button drag' to measure the distance from the original SW measurement impulse to the new location of the SW-Copy impulse.
-----

  • With these low frequency LP measurements, the room modes can make this process very difficult. There may be no significant improvement using this method verses any of the other common timing methods.
  • I do this process for each main speaker individually rather than L+R together. That way it is easy to see the tracking issues in each main speaker. There is very often a big difference in the phase tracking and thus the SPL support for the 2 channels, so a compromise timing is needed. I am hoping that measuring the L+R together as suggested here results in that good compromise. I have never investigated this directly however.
Hi,
I am doing my homework (thank you) and stuck at step 8. How or where do I manually move the impulse location of the sub?
 

·
Premium Member
Joined
·
1,562 Posts
Step 8 - Manually move the SW Impulse position:

We can use the controls shown below. I use the ms shift option as this is the way my delays are entered, but the other options 2 can be used also.
For trial moves to a SW impulse I suggest initially using ~1 ms move increments. We could optionally adjust the distance instead using 0.3 m or, 1 ft. increments for a similar effect. After we get close we can fine tuning more closely if we like. With experience it's possible to find tune in fewer steps, but, to start, it's best to proceed more systematically.

To add delay we need to enter negative time or distance. This will move the impulse to the right on the chart. The effect on the phase overlay will be to shift the right tail of the SW phase trace downward. [Don't forget to enter 0.0 ms into the ref time before looking at the resulting phase overlay chart. Sometimes REW resets ref time to 0.0 ms automatically and sometimes it doesn't.]
 

Attachments

·
Registered
Joined
·
363 Posts
Step 8 - Manually move the SW Impulse position:

We can use the controls shown below. I use the ms shift option as this is the way my delays are entered, but the other options 2 can be used also.
For trial moves to a SW impulse I suggest initially using ~1 ms move increments. We could optionally adjust the distance instead using 0.3 m or, 1 ft. increments for a similar effect. After we get close we can fine tuning more closely if we like. With experience it's possible to find tune in fewer steps, but, to start, it's best to proceed more systematically.

To add delay we need to enter negative time or distance. This will move the impulse to the right on the chart. The effect on the phase overlay will be to shift the right tail of the SW phase trace downward. [Don't forget to enter 0.0 ms into the ref time before looking at the resulting phase overlay chart. Sometimes REW resets ref time to 0.0 ms automatically and sometimes it doesn't.]
HI,
I made a lot of trial since the last post. I want to thank you very much for your advise.
I think that the + 70cm is the best option, finally, with polarity not inversed. For this, I put Rotel sub xo setting to 100hz (lpf 24 db) and a HPF of 40hz for the fronts and the center.
I tried also to play with the phse button of the sub without convincing (for me) success.

P"s": I still working how to find delays for the sub folllowing your procedure!)
Thanks again.
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
I suspect that the delay setting you found will work as well as any of the alternatives. Only significant changes to the room and setup can help avoid a null in the important 80-120 range. This is a situation that many of us live with.
 

·
Premium Member
Joined
·
6,711 Posts
Now using trial and error, move the SW-Copy impulse location manually until the closest phase tracking is found. After each move of the SW-Copy impulse it is necessary to reset the window ref time to 0.0 ms.
There is a bug fix in V5.19 beta 8 which affects that:

Bug fix: When a Frequency Dependent Window was being applied the phase response was incorrect if the win ref time was not 0
 

·
Registered
Joined
·
8 Posts
Hi jtalden,

I am trying to figure out how to measure time delay with my UMIK-1 mic. I knew about the acoustic timing reference thing, but since I don't have an external soundcard, I connect my laptop, which is with REW installed, directly to my receiver, and REW recognized my system as 2.0. I can't measure the rest of my speakers. Any chances to solve this questions without purchasing an external soundcard?

And you mentioned "XO range" a few times but I don't know what is means, it would be very nice that you explain it to me.
 

·
Premium Member
Joined
·
6,711 Posts
Have you looked at the guide in this thread about using an HDMI connection to a system with the ASIO4All driver to allow measurement of different channels?
 
1 - 20 of 33 Posts
Top