tutorial for good Time Alignement, Group Delay and Phase - Home Theater Forum and Systems - HomeTheaterShack.com

Reply
 
Thread Tools
post #1 of 29 Old 08-07-12, 05:27 AM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
tutorial for good Time Alignement, Group Delay and Phase

For the last few weeks I have learned a lot on this forum about 'subwoofer distance tweaking' and 'timing alignement'. Kudos goes to Jtalden (john) for this!! I have made a simple 'tutorial' for myself (and others) I would like to share. Of course I m open for improvement and maybe there are things that are not correct in the tutorial so please let me know. Together we know more!


1.
The first things that you need to do is set the perfect time alignement for all your main speakers (FR, FL, CC, SR and SL) . You can do this with a quite simple process. For this process you will need REW and a calibrated mic. (like the Cross Spectrum Calibrated ECM8000). Also, for the entire measuring process you need to set up REW for loopback timing reference (check REW help file how to do this).

a.
First thing that you need to do is set a reference distance for all main channels. You can let Audyssey do this or tape measure your speakers (remember to measure to the exact LP at ear height).
b.
After this you will have to check the IR peaks for all the main channels (not the subwoofer yet!) in REW by doing measurements with REW for all 5 main speakers separate. (If you measured correct the IR peaks will fall close to each other already). You can take you center channel as reference' and fine tune the distance settings of your FR, FL, SR and SL until the IR peaks fall the closest to each other (depeding on the limitations of the distance increments in your AVR it will likely not be possible to let them fall 100% (0ms) on top of each other but look for the best result.
c.
First make a measurement of the CC+SUB with the distance Auddysey (or tape measured) measured for your subwoofer. Look at the IR overlay in REW. The subwoofer and CC IR peaks will be pretty close to each other already. Change the subwoofer distance untill the IR peaks in the REW overlay panel fall as close on top of each other as possible.

Do this procedure for the rest of your main speakers also and write down all the 5 subwoofer distances where the IR peaks of the main channels+subwoofer fall on top off each other best. We will need them later.

It should look something like this if you did it right:



You now have all 5 main channels perfectly time aligned. Remeber that from this point on you DO NOT change the distance settings of your main speakers at all anymore!

The problem is that we still dont know how the phase acts at this moment. We do know that with this good time alignement we now have the Group Deplay will be at it s best.

3.
Next we are going to see how the phase acts per main channel+subwoofer. You can follow the instructions in the .doc in post 1 of this thread for this but in short:
a.
First make a measurement with the distances you found with the IR time alignement process you did in step 2. Look at the SPL overlay in REW. Change the subwoofer distance untill the SPL overlay graph in the Xover range looks best (highest SPL output with least dips).
b.
Compare this distance with the distance you measured before (with the perfect time alignement). If they are close to each other you are lucky and can take the average of both of them. It is also possible (like in my setup) that there a quite a big difference between them. You will then have to make a choice what you want: take the time aligned measurement length (so you will have perfect time alignment and Group Delay) or take the subwoofer distance which results in an good phase (you sacrifice time alignement and Group Delay).
c.
Do this for all main channels+subwoofer seprate too.


I think that if you want a compromise between time alignemet, group delay and phase for you total 5.1 speakers set you will have to average all subwoofer distances you found in step 2 and step 3 (this will be the average of 10 distances if you did everything correct).

Last edited by THX-UltraII; 08-08-12 at 04:50 AM.
THX-UltraII is offline  
Sponsored Links
Advertisement
 
post #2 of 29 Old 08-07-12, 08:54 AM
Senior Shackster
 
Join Date: Dec 2011
Posts: 239
Re: tutorial for good Time Alignement, Group Delay and Phase

Addendum: Preferences and soundcard need to be set up for loopback timing reference.
Barleywater is offline  
post #3 of 29 Old 08-07-12, 11:25 AM
Senior Shackster
 
Join Date: Jan 2012
Posts: 114
My System
Re: tutorial for good Time Alignement, Group Delay and Phase

Quote:
THX-UltraII wrote: View Post
For the last few weeks I have learned a lot on this forum about 'subwoofer distance tweaking' and 'timing alignement'. Kudos goes to Jtalden (john) for this!! I have made a simple 'tutorial' for myself (and others) I would like to share. Of course I m open for improvement and maybe there are things that are not correct in the tutorial so please let me know. Together we know more!

Thanks for taking time to provide all the steps. I am still digesting. If you don't mind can you post with .mdat of final resulat as attachment. It will help to look at the graphs as base line.
chinni123 is offline  
 
post #4 of 29 Old 08-08-12, 01:31 AM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
Re: tutorial for good Time Alignement, Group Delay and Phase

Quote:
Barleywater wrote: View Post
Addendum: Preferences and soundcard need to be set up for loopback timing reference.
thxz. Edited first post.
THX-UltraII is offline  
post #5 of 29 Old 08-08-12, 01:33 AM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
Re: tutorial for good Time Alignement, Group Delay and Phase

Quote:
chinni123 wrote: View Post
Thanks for taking time to provide all the steps. I am still digesting. If you don't mind can you post with .mdat of final resulat as attachment. It will help to look at the graphs as base line.
I bought a new processor and need to re-measure everything. I ll post graphs and .mdats as soon as possible
THX-UltraII is offline  
post #6 of 29 Old 08-08-12, 02:21 AM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
Re: tutorial for good Time Alignement, Group Delay and Phase

Despite that I made this tutorial I still have some some questions myself too about this.

I did some new measurements last night with my new processor (Onkyo TX-NR876 with Audyssey MultEQ XT). I measured all my speakers with Audyssey and after this checked the IR peaks with REW for the 5 main speakers (FR, FL, CC, SR and SL). The IR peaks fall pretty close to each other so there s no need to adjust this any further (this assumes that Audyssey does a good job for setting the delays). After this I started measuring all the separate main speakers+subwoofer. I m already 'stuck' with my second measurement. To be more specific:

I first measured the FR+SUB with REW with the distances that Audyssey measured. There where some big dips at the XO frequency range. Like I said in my 'tutorial' a few posts back you only want to adjust the subwoofer distance to maintain proper time alignement for all 5 main speakers (FR, FL, CC, SR and SL). So I adjusted the subwoofer distance to a point where the SPL output in the OX range was best (highest SPL and least dips). I had to change the subwoofer distance from 3.60m (what Audyssey measured) to 6.15m to get the best SPL in the XO range (this proves that the Audyssey measurement is not utilized to determine or check the integration of the speakers at the critical crossover frequency. Audyssey focuses on response and time domain correction for each speaker in isolation; it does not measure the combined response of the subwoofer and main speakers together). With the subwoofer @6.15m my FR+SUB measurement gives me a pretty nice/flat SPL response.

Next thing I did was measuring the next main channel. The order of doing all the main speakers does not matter; I chose the CC. As expected the CC showed a pretty bad SPL response at the XO range with the 3.60m subwoofer distance that Audyssey measured (I had my subwoofer distance set back on it s default 3.60m that Audyssey measured). I was hopinh that a [email protected] measurement would give me the best SPL reponse. Of course this 6.15m was already much better than the [email protected] but I noticed that there was room for further improvement in the CC+SUB SPL response. With some trail and error measurements I finally found the best subwoofer distance setting for the CC+SUB which is 7.05m. This is a pretty big difference with the best subwoofer distance for the FR+SUB which is 6.15m. Before going further with measuring the FL+SUB, SR+SUB and SL+SUB I want to know what will I need to do with this information. I checked the [email protected] (best CC+SUB distance) but the SPL response is much worse then [email protected]
With this information I can draw a few different conclusions/possibilities which I like to discuss with the REW guru's here (Barleywater, JohnM, Jtalden etc ):

1.
IF my speakers where only the FR, CC and SUB in my system (this is only theoretically to make it easier to explain) the best thing I could do is choose 6.60m (average of 7.05m and 6.15m) for the subwoofer distance to achieve the best compromise.
2.
IF my speakers where only the FR, CC and SUB in my system (this is only theoretically to make it easier to explain) the best thing I could do is choose 7.05m for the subwoofer distance because a correct phase between the CC and SUB is most important for movies to get the best dialog sound representation.
3.
IF my speakers where only the FR, CC and SUB in my system (this is only theoretically to make it easier to explain) the best thing I could do is choose 6.15m for the subwoofer distance because a correct phase between the SR and SUB is most important for movies to get the best overall movie sound representation.
4.
In orde to achieve the best SPL response for both SR+SUB and CC+SUB I could stick with a fixed subwoofer distance of 6.15m, choose the default FR distance that Audyssey set (which results in best SPL for FR+SUB) and CHANGE the distance setting of the CC (so leave the subwoofer distance @6.15m). This would mean that I m changing the distances of the main speakers after the Audyssey calibration. This would mean the best SPL response for both FR+SUB and CC+SUB but losing time alignment between FR and CC.
EDIT: The big question with this method is how much time difference in ms is reasonable. Someone said that max. 5ms is reasonable. 5ms is approx. 171.5cm (assuming the speed of sound is 343 m/s) so this would mean there is quite some room to adjust all the main speaker distances. But others say that you don t want to change the distances for the main speakers at all......

What are the best things to do......

Last edited by THX-UltraII; 08-08-12 at 04:45 AM.
THX-UltraII is offline  
post #7 of 29 Old 08-08-12, 08:18 AM
Senior Shackster
 
Join Date: Dec 2011
Posts: 239
Re: tutorial for good Time Alignement, Group Delay and Phase

For all the science, and all the engineering, in the end much of the sound stage remains an illusion within the mind. Sound mixing and engineering is human driven, and occurs in diversity of studio setups, all involving human interpretations. Compromise must happen at every step of the way. In the end, if most music and soundtracks you listen to bring enjoyment, that's all that really matters. It is really still about the message, not the messenger.

In perfect surround sound set up all speakers are exact same distance from single sweet spot listening position. If sub is only element off of circle, and is crossed at frequency below that for room's single mode behavior (pressurization only mode), then a single delay setting works. As more speakers diverge from circle, and sound at frequencies with directional characteristics are involved, then one perfect solution does not exist, but many possible partial (compromise) solutions do.

Don't let knowledge of real physics destroy enjoyment in listening.

Since primary focus in movies is concentrating on screen, and for most music focus is concentrated on grouped performers being watched, setup should be focused on front speakers. As long as front speaker direct sound arrives several milliseconds before first reflections, and arrives before correlated sounds from other speakers (rears, sides), the mind will assign direction of sound based on cues from front speakers and sound appears in front. Once this is achieved, gross defects in low frequency behavior are addressed with phasing and delay tweaks (with lots of compromise), room treatment, (typically compromise in added expense, and potential aesthetic impact), and finally equalization, for which various limits (compromise) also exist, but has greatest flexibility in quickly controlling perceptions about listening space and controlling perceived defects of particularly disturbing source material.

Sometimes reading is more enjoyable than watching and listening.

Regards,

Andrew
Barleywater is offline  
post #8 of 29 Old 08-08-12, 08:43 AM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
Re: tutorial for good Time Alignement, Group Delay and Phase

thxz for your reply Andrew. I know I should enjoy it but would you please give your (technical) opinion and reaction on my 4 conclusions? That s the last thing I m asking from you
THX-UltraII is offline  
post #9 of 29 Old 08-08-12, 01:51 PM
Elite Shackster
 
Join Date: Mar 2009
Location: Arizona, USA
Posts: 1,562
My System
Re: tutorial for good Time Alignement, Group Delay and Phase

Quote:
THX-UltraII wrote: View Post
What are the best things to do......
Whatever you want!
You are laboring under the idea that there is a “best” alignment.

As I have stated there are several good solutions and those are not the same for all situations or people. You now have a reasonably good understanding and are starting to see this.

Regarding timing alignment choices:
Do the choices affect the sound; are they important?
I find more difference than I would expect, but 90% of my experience is with significantly higher order XO filters than the AVR provides. Possibly I would not find the difference using only the standard AVR SW XO filters. This is why I suggested you check it out for yourself and pick the one you prefer. Based only on my limited experience I suspect that most would to prefer (in order):
1. Reduce the SW delay by 1/2 WL. [by moving the SW distance to a larger number and inverting the SW polarity.]
2. Initial rise of SW and main IRs aligned. [SW and MR drivers must be the same polarity. This will provide a smoothly changing GD and phase through the XO handoff.]
3. Increase the SW delay by 1/2 WL. [by moving the SW distance to a smaller number number and inverting the SW polarity.]

Regarding “1.”, the reduced SW delay will flatten the GD in the lower freqs and bring the GD more in line with the high frequencies. The tradeoff is that GD and phase in the XO range is no longer smoothly changing. The phase at the XO point will still be the same, but the SW phase will be crossing the main phase at an angle making the phase at surrounding frequencies diverging. I am not sure that others will hear the difference or have the same impression of which is better. Most of my setups use significantly higher slope XO filters than the conventional AVR handoff. That may make a significant difference in the audibility of the differences. My SW delay is currently reduced a full WL because with the steep XO filters, that is what is needed to bring the GD in line through the entire frequency range except for the narrow XO region.

Regarding SPL EQ concerns:
[Assuming acoustic treatments speaker and LP locations have been fixed.]
> The most important aspect of the 20-200 range is a smooth SPL response and that should be the overwhelming criteria.
> If the phase is unfavorable at some point in the XO range it is less effective to EQ a dip at that point.
> It’s fine to EQ as needed within reason. Very larger EQ magnitudes are best avoided for various reasons.

Now let’s take your situation as an example:
> Audyssey aligned the main speaker IR distances effectively as expected.
> It appears Audyssey aligned the SW IR to the mains to approach alignment “2.” but did not provide you a warning to reverse the polarity.
[Your “2.” continuous handoff would be achieved by inverting the polarity of the SW (or the 5 mains).
Note that, your MR drivers are inverted relative to the TW and SW and therefore in an anechoic environment the polarity is 180 deg out of phase when the IRs are aligned as Audyssey provided. In a room, the room modes play havoc anyway, so you never know the situation until you measure.]
> There was a significant dip in the SPL response from 50-90 Hz for the SW+FR main speaker. This dip did not center itself at the XO freq of 80 Hz. Again this is the effect of room modes. The dip was large enough that EQ boost and cuts would likely be larger than desired. [Maybe not, but we didn’t try.]
> With some adjustment to the timing/phase we aligned the phase near 65 Hz instead of the 80 Hz XO setting and were able to reduce the dip prior to EQ. Since now it is in phase there, it will also respond better to EQ.
> Looking at your other main speakers you found that there are different timing decisions that would be made for each speaker.

That brings us back to your initial question about what timing to select. The short answer is:
First choose which overall alignment range you want to use; either 1, 2 or 3. Then adjust as little as possible from that alignment to compromise for the SPL differences between the other main channels. Keep in mind that if the EQ is effective in smoothing the response in through the XO with reasonable boosts and cuts then that timing is good.

Again:
Any timing where EQ can be effectively implemented to is okay. I see lots of SPL curves on this forum that suggest that the timing adjustment is unfortunate. Yours would have been one of those if you had accepted the AVR’s automated setting.
jtalden is offline  
post #10 of 29 Old 08-08-12, 02:53 PM Thread Starter
Senior Shackster
 
THX-UltraII's Avatar
 
Join Date: Nov 2008
Posts: 263
Re: tutorial for good Time Alignement, Group Delay and Phase

I ll come back on your extensive reply tomorrow John. Here s a lot of interesting new measurements I did tonight:
Attached Thumbnails
tutorial for good Time Alignement, Group Delay and Phase-fl-only-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-fl-only-no-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-fl3.9-sub3.6-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-fl3.9-sub3.6-no-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-fl3.9-sub6.15-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-fl3.9-sub6.15-no-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-sub-only-no-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-sub-only-aud.jpg  

tutorial for good Time Alignement, Group Delay and Phase-3-5.jpg  

tutorial for good Time Alignement, Group Delay and Phase-4-6.jpg  

THX-UltraII is offline  
Reply

Bookmarks

Tags
alignement , delay , good , group , phase , time , tutorial

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