Home Theater Forum and Systems banner

1 - 18 of 18 Posts

·
Registered
Joined
·
363 Posts
Discussion Starter #1
Hello,
see my system and my room dimensions in my profile.
I use Antimode to eq the sub. All measurements were taken with Antimode on. YPAO EQ the 5 channels only. Ypao filters was to flat.
Distance from mains to MLCP= 2,70m; same from the center channel and the SW. Although YPAO set the same distance for the mains and the surround, it sets probably correctly the sub at 5,40m (antimode added a delay to the signal).
Crossover is set to 80hz in the receiver.
I am beginning to use REW and UMIK-1 with hdmi connections. So, maybe I have made mistakes. Like someone says elsewhere in this forum: I know REW just enough to be dangerous !!

What I am trying is to see if the time/delay SW-Fronts with the choosen distance by YPAO for sw is optimum. Since I have a usb mic, I understood that I can not use the group delay option in REW.

I followed Jtalden advise (post#5 in this thread:http://www.hometheatershack.com/forums/rew-forum/115009-seeking-some-input-help-interpreting-my-rew-results.html#ixzz3jHQ4CgeE) for this.
Followings are for the center channel only, for the moment.

Centre xover 80hz avec sans sub2.jpg

Any comments will help. Specially is the result already good or is there place to change YPAO settings distance for the SW?
BTW, every places possibles for the mains and the sub has been tested and retested with REW. For the moment, there is no room for changes in distances or placements without risking wife's clipping (wife's headroom now set to 0 !!)
This forum is the best.
 

·
Premium Member
Joined
·
1,562 Posts
The SW distance setting is questionable. We expect to see the distance setting in the AVR to be very near the actual distances for all main speakers. I understand that was achieved. We also expect the SW distance setting to be very near the actual distance or possibly a little larger because of the Antimode and SW delays. This add-on is expected be small however; less than 1m.

Was the 80Hz XO active for the CC measurement? I ask because the there is more CC SPL output below 80Hz than I would expect with the XO active. Assuming it was, then the your chart accurately shows the current condition. If it wasn’t active then this data shouldn't be used.

We want the CC+SW SPL trace to be higher/greater than the CC and the SW SPL alone throughout the XO range. Here it is lower through that range with a null at 80Hz XO point. This suggest that possibly timing would be correct if the SW polarity is reversed from the current setting.

1. You may want to give that a try. Just reverse the SW polarity and remeasure CC+SW.

2. With the current polarity of the SW you could also reduce the SW distance setting about 2.2m and take another CC+SW measurement. This would put the distance setting back where we expect it to be and should have a similar SPL impact as '1'. It's a distance reduction of about 1/2 wavelength at 80Hz.

It is very risky however to suggest anything looking at only one chart and having limited understanding of the entire measurement setup and test conditions.

If you want to confirm the settings there are several ways to do that. One is to start with the SW distance setting at the actual distance + 2m. Measure CC+SW at that distance and also at decreasing distance settings using 0.4m increments until you have 5 measurements. The setting with the highest SPL is the best setting.

If you want me to offer a setting just post an mdat here containing 3 measurements.
> CC
> SW
> CC+SW

Your current file may be fine so long as it adheres to the following requirements.

> The XO was active for all 3 measurements.
> The CC is identical speaker design to the FL and FR. If not, then measure either the FR or the FL in place of the CC.

Given this mdat, I will be able to use REW analysis to provide a SW distance setting change needed in the AVR to provide the best SPL support.

 

·
Registered
Joined
·
363 Posts
Discussion Starter #3
The SW distance setting is questionable. We expect to see the distance setting in the AVR to be very near the actual distances for all main speakers. I understand that was achieved. We also expect the SW distance setting to be very near the actual distance or possibly a little larger because of the Antimode and SW delays. This add-on is expected be small however; less than 1m.

Was the 80Hz XO active for the CC measurement? (...)

(...)

If you want me to offer a setting just post an mdat here containing 3 measurements.
> CC
> SW
> CC+SW

Your current file may be fine so long as it adheres to the following requirements.

> The XO was active for all 3 measurements.
> The CC is identical speaker design to the FL and FR. If not, then measure either the FR or the FL in place of the CC.

Given this mdat, I will be able to use REW analysis to provide a SW distance setting change needed in the AVR to provide the best SPL support.

Dear jtalden,
of course I would like you to offer settings. Your skills make your post easy to understand (I think, hopefully!).

The crossover was active at 80 for each measurements. The SW distance measured with a tape is 2,75m (center of the cylinder of the PC-2000).The fronts tape measured are at 2,70m as YPAO found.

YPAO set the sub at 5,40m as already mentionned. (Also I need to say that it is impossible to add an other sub becaue I have neither the budget nor the place) I want to get the best I can in those circumstances.

About the center speaker design it is a PSB matched with the PSB Alpha B. Alpha B fronts and surround have one 5 inches woofer, one tweeter and back ported. The center speaker has two 5 inches woofers one tweeter, back ported. It is place above the TV not under. Also, you might want to know that the front left speakers is in the corridor leading to the other rooms of the basement (no door on the left side wall). There is a full wall at the right side. So there is no symetry between left and right (compromises must to be done )

Knowing all that, do you prefer FR and FL instead of CC measuremnts?
Thank you so much.
 

·
Premium Member
Joined
·
1,562 Posts
You have the data then for the CC. So post it and I will evaluate it. It is very likely the design comparable to FL and FR so the results should be okay for all. I don't mind looking at the FL and FR also if you like. Two similar files for them would be needed.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #6
You have the data then for the CC. So post it and I will evaluate it. It is very likely the design comparable to FL and FR so the results should be okay for all. I don't mind looking at the FL and FR also if you like. Two similar files for them would be needed.
Hello,
hopefully I have what you need.

View attachment Center_sub_xover80.mdat

View attachment LEFT_xover_LEFT and sub.mdat

View attachment RIGHT_xover80_Right and sub.mdat

Again, many thanks for your help.
 

·
Premium Member
Joined
·
1,562 Posts
Sorry, this data can't be used for the analysis.

The technique requires all 3 measurements to be with mic at the LP and the sweeps range to be identical; preferably wide range. This data has a 15-300Hz sweep range for the individual driver measurements and a 15-20k Hz sweep for the combined measurement in each channel.

Please sweep 15-20k Hz for all measurements. Place mic at the LP and do not move the mic position when taking the measurements.

Had the combined measurements also been 15-300, I may have been able to use the data. I'm not sure though, having never tried. Since it needs to be repeated, please use the wide range sweep. That has worked well in the past and the accuracy will probably be greater.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #8
Sorry, this data can't be used for the analysis.

The technique requires all 3 measurements to be with mic at the LP and the sweeps range to be identical; preferably wide range. This data has a 15-300Hz sweep range for the individual driver measurements and a 15-20k Hz sweep for the combined measurement in each channel.

Please sweep 15-20k Hz for all measurements. Place mic at the LP and do not move the mic position when taking the measurements.

(...)
Sorry:sad:. For a beginner it is not so easy to find on the web which procedure fits for which problem to analyse.

Just to be sure ( I do not want to waste your time) please check my steps (stages?):

1- Center speaker and sub, play pink noise, adjust AV volume with sound meter to get SPL to 75db;

2- Center channel 1.3 via hdmi, Antimode dual core activated (please note that no house curve is added for the low fr, I use the typical room calibration by Antimode as it is),YPAO to off (set to direct in the receiver), xover in AVR set to 80hz, center speaker set to SMALL in the receiver, SUB swich to OFF, sweep 15-20k = C only;

2- disconnect center speaker, SUB swich to ON and repeat sweap for measurement = sub only;

3- reconnect speaker, repeat sweep = C+sub

And repeat for front left and right. Since I will send the files, smoothing or not is irrelevant.

Isn't it?

Hopefully, I will do my homework tonight, if not, tomorrow night.
Thank for your patience (keep in mind that I am a computer dinosaure trying to survive in the 21st century!!:sweat:)
Have a nice day.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #10
That seems fine.
Hello,
here are my meaurements done as described in post#8.
In the file of the center speaker (integration centre et sub xover80) I added one more measure with sub settings to "inversed" in Yamaha RX-V773 receiver; I don't know if this help.

Center speaker
View attachment integration centre et sub xover80.mdat
Front Left
View attachment integration gauche et sub xover80.mdat
Front right
View attachment integration droit et sub xover80.mdat
Looking forward hearing from you.:whistling:
 

·
Premium Member
Joined
·
1,562 Posts
I have analyzed the data.

Regarding the 3 data sets your provided; the SW Polarity for the FR channel measurements were reversed from the SW Polarity in the FL and CC data sets.

I am assuming the FL and CC contain the correct 'original' SW polarity and the FR was inadvertently reversed when measuring that channel. Possibly this was just the result of the forgetting to set it back after reversing it for that extra reversed SW measurement contained in the CC data set?

If this assumption is correct the results are:
> The distance settings are good as they are now set.
> The SW polarity is needs to be 'reversed' from the 'original' setting.
> With the 'reversed' SW polarity the phase tracking and SPL support through the XO range is optimized.
> You can see the impact of the SW reversal in your original vs reversed CC+SW measurements. Just overlay the SPL traces to see the improved SPL through the XO range. A similar impact is present in the other 2 channels as well.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #12
I have analyzed the data.

Regarding the 3 data sets your provided; the SW Polarity for the FR channel measurements were reversed from the SW Polarity in the FL and CC data sets.

I am assuming the FL and CC contain the correct 'original' SW polarity and the FR was inadvertently reversed when measuring that channel. Possibly this was just the result of the forgetting to set it back after reversing it for that extra reversed SW measurement contained in the CC data set?

If this assumption is correct the results are:
> The distance settings are good as they are now set.
> The SW polarity is needs to be 'reversed' from the 'original' setting.
> With the 'reversed' SW polarity the phase tracking and SPL support through the XO range is optimized.
> You can see the impact of the SW reversal in your original vs reversed CC+SW measurements. Just overlay the SPL traces to see the improved SPL through the XO range. A similar impact is present in the other 2 channels as well.
Seeing all that in my graphs ! You must be a pro or a magician :clap:!
So many things to think about at the same time! So many variables.
I will check this week-end the "forgetting to set it back after reversing it for that extra reversed SW measurement contained in the CC data set" by doing a before and after reversing polarity session for the 3 channels and will give you feedbacks.
So, I let the distance settings by YPAO for the sub at 5,40m.

I f I need to run an EQ with YPAO in the future, the way I understand it, I must first set the polarity of the sub to normal, run YPAO, and after the new filters beeing set by YPAO, reverse again the polarity of the sub isn't it ?

Thank you so much.
 

·
Premium Member
Joined
·
1,562 Posts
...

So, I let the distance settings by YPAO for the sub at 5,40m.
Yes, the phase tracking and SPL support through the XO range is optimized with that setting.

I f I need to run an EQ with YPAO in the future, the way I understand it, I must first set the polarity of the sub to normal, run YPAO, and after the new filters beeing set by YPAO, reverse again the polarity of the sub isn't it ?
I don't know, but would guess that is not the better option. It would have worked for this last setup, but may not work for the next. I would be tempted to reverse the polarity of the SW next time or possibly reverse the polarity of all the main speakers and leave the current polarity of the SW. YPAO may be more likely to find the correct settings that way?

The FL, FR, CC tweeters all appear to be connected with negative polarity. I also think the midwoofers are connected negative polarity, but that is difficult to tell for sure without measuring the midwoofer driver directly, i.e., without the tweeter. If both drivers are negative polarity in the mains then possibly those 3 main speakers were all connected/wired incorrectly to start with. The SW appears to be connected with positive polarity so the mains and SW were opposite of each other originally.

This may be why YPAO had a problem? I have no experience with it so I don't have any idea how well it identifies and deals with such issues.

Below are 6 SPL traces. The 3 with the lower SPL level in the XO range are the results from the original SW setting. Look at range 70-100Hz. This is left-hand column in the legend. The 3 higher traces (right-hand column in the legend) are the response with the SW reversed. This should be very similar to what you find when you do your experiment. I smoothed the responses 1/3 octave here just to more easily see the SPL trend.

SPL with vs without SW inverted.png
 

·
Registered
Joined
·
363 Posts
Discussion Starter #14
Thanks for all the info.

(...)

The FL, FR, CC tweeters all appear to be connected with negative polarity. I also think the midwoofers are connected negative polarity, but that is difficult to tell for sure without measuring the midwoofer driver directly, i.e., without the tweeter. If both drivers are negative polarity in the mains then possibly those 3 main speakers were all connected/wired incorrectly to start with. The SW appears to be connected with positive polarity so the mains and SW were opposite of each other originally.
(...)
I suppose that you refer to internal wiring of the speakers. I can't go there of course. I checked all cables and connections and everthing seems OK. YPAO did not reported bad connections too.

Just to let you know. I tried to lower the distance settings for the SW choosen by YPAO as suggested in your post #2. After some tested distance, I found 4,40m and SW polarity to normal was giving very closed FR to 5,40m+SW inversed polarity for the mains and the center.
I checked the surrounds. For them, the FR of 5,40m+SW inversed setting was less good than the 4,40m+SW normal polarity setting.

Is this could be an indirect confirmation of what you say about the polarity of my mains ? I do not know.

My instinct (surely not my knowledge) tell me that it could better if all speakers have the best integration with the sub i.e. to choose 4,40,+SW normal instead of 5,40m+SW inversed?

Or should I choose the 5,40,+SW inversed AND change the connection of the mains and the center (negative on red plug and positive on the black?) and leave the surrounds as is.

I can send images or files if you want but I do not want to use too much of your time.
 

·
Premium Member
Joined
·
1,562 Posts
To confirm:
If we deem the data the Post 10 was measured with SW and Mains polarities as 'Positive' and that the distance setting of the SW was 5.4m then the best settings are:
1. SW negative polarity, Mains positive polarity, 5.4m SW distance setting
-or-
2. SW positive polarity, Mains negative polarity, 5.4m SW distance setting

There are no other settings that will provide better phase tracking throughout the XO range.

Below is the calculation for the SW distance setting at 4.4m with a positive polarity setting for both SW and FR. That setting is reasonably similar to the original settings and shows the poor SPL support that would be expected. I can only presume that there was some confusion as to what the settings were if good SPL result was achieved.

Temp1.png

Possibly you are troubled with the thought of the mains and the SW being set to 'opposite' polarities per the wires/switches/color-coding. This should not be a concern. There is only correct or incorrect relative polarity. Whether it is the same or opposite depends on the choice of speakers, XO filters, speaker box designs, wiring, electronics box settings. REW allows us to find the correct acoustic setting regardless of all the stuff in between.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #16
Thank you for your much appreciated guidance.

The best is that I can hear a difference, some details are easier to hear and the bass are more immersive (may I said that ?). I understand better what is a good tranistion and SPL support in the XO range.

One last question, if I may.

Just for my own curiosity, I also had made the same measurements with YPAO filters set to FLAT. The curves in the XO range were always looking worst than with YPAO sets to off (at least to my untrained eyes).

Should I be concerned by that? Although, to tell the truth, I do not hear any difference.
 

·
Premium Member
Joined
·
1,562 Posts
As stated, I have no experience with YPAO.

I know it provides distances and some EQ, but don't know the necessary detail of how it is implemented.

If you run it after manually aligning the timing and it changes the distances or it provides significant EQ to the mains and SW separately within the XO range then that can change the timing.

Small EQ adjustments there to achieve a target acoustic XO roll-off should not be a significant impact. If it applies heavy EQ to try to make the mains and SW SPL both flat for an octave or more above and below the XO frequency then that may significantly impact the timing.

If that is the problem we could check and adjust the timing as needed after YPAO does its thing.
 

·
Registered
Joined
·
363 Posts
Discussion Starter #18
As stated, I have no experience with YPAO.

I know it provides distances and some EQ, but don't know the necessary detail of how it is implemented.
(...)
Small EQ adjustments there to achieve a target acoustic XO roll-off should not be a significant impact. If it applies heavy EQ to try to make the mains and SW SPL both flat for an octave or more above and below the XO frequency then that may significantly impact the timing.

If that is the problem we could check and adjust the timing as needed after YPAO does its thing.
Just in case it could help somebody with YPAO like myself, I ran YPAO with sub inversed. YPAO sets the sub distance at 5,35m instead of 5,40m. All others distances remain the same.
To EQ, YPAO choose 7 filters among a choice of 26 predefined frequencies (49,6hz, 62,5hz, 78,7hz, 99,2hz etc.) with Q and dB for each 5 channels.

With the sub inversed, many filters are different, the sound stage and the imaging changed also for better I think. (not enough time to listen).

I provide a file with 3 curves. The black is sub inversed only. The blue is the result of YPAO filters to FLAT with sub polarity to normal (measured the day before yesterday) and the red is YPAO new set of filters to FLAT with sub inversed.
I am not sure if it is helpfull.

View attachment sub XO 80 comparaison YPAO FLAT sub normal et FLAT sub inversé.mdat
 
1 - 18 of 18 Posts
Top