[scubasteve2365]

First, I want to apologize if a thread like this exists. I'm not aware of one.


I'm looking for a sort of high level or a 5000 foot view (a roadmap) of where to begin in tuning my audio system in my theater. I'm not looking for information on treatments, information on measurements or information on EQ. That information is out there. I have a nice speaker system, a nice subwoofer system, a measurement microphone, an EQ ability in the AVR and a separate EQ for the sub LFE. What I'd like to know is really where to begin. I know measurements are sort of an ongoing process used to constantly identify weak points and items that need addressing. Up until now I've always just used my AVRs auto calibrate and then adjust for my own dynamic taste (a little hotter on the center, rear and LFE)


Now, I think I'm ready for a more advanced approach to the audio system. I have a Dayton UMM6 measurement microphone and a PC with REW/HDMI


Where does one want to begin? One area that I feel that I'm not very informed on is the upper LFE region and blending the subs with the mains.


Do most people still set their levels by using the RatShack meter at 70dbs? Since I will now be looking a frequency response as well as SPL, how does that impact the general level setting of each speaker? My yamaha AVR has a built in PEQ that is adjustable for each individual speaker.

 

[Nyal Mellor]

I might suggest first doing a basic calibration on your system using REW and your SPL meter so you understand how to use everything.


- speaker levels set to 75dB with C weighting

- speaker distances set using tape measure (NOT sub distance)

- sub / mains integration (set sub distance / phase / polarity for smoothest integration at XO point)


Next try moving speakers, sub and listening position around to see if the bass response can be improved. You might also experiment with slightly lower or higher crossover points for your bass management.


Finally I'd try some EQ on some of the room mode related peaks that stick around on the frequency response measurement.


Here's the process we use for system setup (rotating graphic bottom right): home theater calibration process and checklist .

 

[scubasteve2365]

Thanks for the feedback. I fired REW up for the first time and took some first measurements of my front speakers. I think I may have forgotten to turn off the Yamaha YPAO Auto-EQ as there is a weird drop around 175Hz. Regardless, it was more of an exercise to just get familiar with REW than anything.


Here's my roadmap, I think


1) Set Distance of each speaker according to tape measure

2) Measure FR of each speaker and subwoofer

3) EQ each speaker and subwoofer independently

4) SPL Level match each speaker. (I'm not sure how to accomplish this with the REW tools, in the past I just used the AVRs tone generator and my RatShack SPL meter, is that still the best method)

5) Check Subwoofer blend to Mains with regards to FR and Phase (what do I look for in regards to phase?)

6) House curve and levels to taste (In the past I've always found that I've preferred a center channel bump for dialog clarity/presence)


#5 is the biggest unknown to me on how to accomplish.


This is more of an immediate near term tuning/setup. I will also be adding treatments and then repeating this process I assume. I currently have some treatments on the first reflection points.

 

[Nyal Mellor]

I would not EQ until you have done 4) and 5) at least once. Calibration is iterative, and the first step is to do a basic calibration - levels, distances, sub/mains integration before doing anything more advanced.


With respect to levels REW has a SPL meter in the software though you will need to calibrate the SPL levels in REW with your meter. All this is explained in the instructions. A more advanced way to set levels is to SPL match each channel over a frequency range. I use 100-1000Hz typically using 1/3rd or even 1 octave smoothing.


For sub / mains integration what you want to do is adjust the simplest way to do it is to input a signal to the sub and center channel, set the sub polarity and phase to 0 and adjust the distance 1ft at a time in the AVR, starting from 0ft, until you have the highest SPL at the crossover point. See Home Theater Calibration 101 and Subwoofer Setup 101 .

 

[scubasteve2365]



Quote:
Originally Posted by Nyal Mellor  /t/1470721/roadmap-for-tuning-your-theater#post_23270810


I would not EQ until you have done 4) and 5) at least once. Calibration is iterative, and the first step is to do a basic calibration - levels, distances, sub/mains integration before doing anything more advanced.


With respect to levels REW has a SPL meter in the software though you will need to calibrate the SPL levels in REW with your meter. All this is explained in the instructions. A more advanced way to set levels is to SPL match each channel over a frequency range. I use 100-1000Hz typically using 1/3rd or even 1 octave smoothing.


For sub / mains integration what you want to do is adjust the simplest way to do it is to input a signal to the sub and center channel, set the sub polarity and phase to 0 and adjust the distance 1ft at a time in the AVR, starting from 0ft, until you have the highest SPL at the crossover point. See Home Theater Calibration 101 and Subwoofer Setup 101 .

Thanks a ton, those blog links were well written (wish it was with REW, but no biggie) and is the type of information I was looking for.

 

[commsysman]

JOB 1 IMO is to get the front main speakers to sound right.


Disconnect everything but the front speakers and experiment until you get the best placement in the room for them.


You cannot do this with other speakers hooked up, and your receiver's software program cannot make up for poor placement.


One JOB 1 is completed, THEN you can proceed to fine-tune the rest of the system.


The front L/R speakers are the engine of the system; don't neglect JOB 1!!!

 

[scubasteve2365]

So I did a basic calibration from scratch last night, has been awhile since doing that.


1) I reset everything in the AVR (levels to 0, distances to 0, ect)

2) Set the distances for for all 7 speakers according to tape measure from speaker to listening position

3) Set levels using the AVRs test tone and my Radio Shack SPL meter.


I'm still no feeling good with regards to sub integration (phase, distance timing). I used some of the steps in Nyal's blog/guides but using the RTA in REW I wasn't able to notice any level change at the cross over frequency when I flipped the phase from 0 to 180 (normal or reverse). I have a dedicated sub EQ (a mini-dsp) that is not in the signal chain at the moment. It may have the ability to do some phase adjustments other than just 0/180. Not sure.


My measurement microphone is a USB mic (Dayton UMM6) so I don't think I can do any timing measurements with it. At least so I've read that's a limitation read with USB microphones. What would be another suggested alternative for setting the sub timing/delay (distance)?


I would like to mention that my sub is an infinite baffle system in case that ultimately matters.


I have attached some measurements that I took last night.


Sub (No Smoothing) (No EQ)



Sub Waterfall (No EQ)



Front Left and Right (No EQ) 1/6 Smoothing



Center (1/6 Smoothing) (No EQ)

 

[Nyal Mellor]

Good work!


But don't set sub distance using a tape measure!


You won't notice any change from flipping the polarity switch between 0 and 180 on the sub UNLESS you are playing a signal through BOTH the sub and mains at the same time.


To calibrate the sub you need to play a test signal through the sub and all three mains or a reference speaker (e.g. the center) and then use the time delay / polarity / phase controls to get maximum SPL at the crossover frequency.

 

[Roger Dressler]



Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23263250


Do most people still set their levels by using the RatShack meter at 70dbs?

The SPL depends on the test signal or test disc. If using the Yamaha's internal test, then 75 dB most likely, as that is the standard. Also true for many test discs, especially THX Optimizer signals. But it does not matter too much unless you like to compare notes about how loud you play movies.

Quote:
Since I will now be looking a frequency response as well as SPL, how does that impact the general level setting of each speaker? My yamaha AVR has a built in PEQ that is adjustable for each individual speaker.

After you have achieved the desired EQ settings, you need to recheck the SPL and readjust if necessary.


Be careful. Some processors fail to pass their cal noise through the internal EQ, so then you'd want to check with a test disc.


I looked at your speaker plots and the center isn't happy. That notch at 8 kHz is pretty severe. I wonder if it is because the speaker is both low and aiming parallel with the floor. I'd suggest raising it up so the top is flush with the screen, and tilt the speaker up so it points to listener heads. Then run the measurement again. It may also reduce some of its bass rise.


Speaking of which, the L/R have quite a bit, and maybe the EQ can help. Have you tried YPAO, assuming your model has it? Just for fun/learning, you might try pulling one of the L/R speakers down off the riser, and position it a little closer to the screen edge (still out of sight for viewers in the end or row seats). It looks like it is really close to the front wall. Then see if that helps the hot bass response at all.


Lastly, try a little toe-in on the mains and see if that affects the HF response.

 

[scubasteve2365]



Quote:
Originally Posted by Nyal Mellor  /t/1470721/roadmap-for-tuning-your-theater#post_23275993


Good work!


But don't set sub distance using a tape measure!


You won't notice any change from flipping the polarity switch between 0 and 180 on the sub UNLESS you are playing a signal through BOTH the sub and mains at the same time.


To calibrate the sub you need to play a test signal through the sub and all three mains or a reference speaker (e.g. the center) and then use the time delay / polarity / phase controls to get maximum SPL at the crossover frequency.

I ended up using the center since it was the last speaker I EQ'd. See the REW plot below. The graph limits are focused in on the 80Hz X-over point. I took many sweeps with just the sub and center playing with varying degrees of delay. I can adjust with the AVR distance setting or by adding a delay in my MiniDSP EQ that is handling the subs.




I ended up settling on the dark blue line - which was oddly set at the AVR default (minimum) of 1ft and 0mS of delay in the MiniDSP. That blue line technically has the slightly higher SPL at the exact X-over point. Since taking the measurements I've noticed that the green line is nearly as high SPL at the X-over point but has a notably higher SPL from 10Hz below X-over all the way down. The values for that sweep is 3ft distance and 0mS. Do you think I should use that value or am I over-analyzing?

 

[scubasteve2365]



Quote:
Originally Posted by Roger Dressler  /t/1470721/roadmap-for-tuning-your-theater#post_23277336


The SPL depends on the test signal or test disc. If using the Yamaha's internal test, then 75 dB most likely, as that is the standard. Also true for many test discs, especially THX Optimizer signals. But it does not matter too much unless you like to compare notes about how loud you play movies.

After you have achieved the desired EQ settings, you need to recheck the SPL and readjust if necessary.


Be careful. Some processors fail to pass their cal noise through the internal EQ, so then you'd want to check with a test disc.


I looked at your speaker plots and the center isn't happy. That notch at 8 kHz is pretty severe. I wonder if it is because the speaker is both low and aiming parallel with the floor. I'd suggest raising it up so the top is flush with the screen, and tilt the speaker up so it points to listener heads. Then run the measurement again. It may also reduce some of its bass rise.


Speaking of which, the L/R have quite a bit, and maybe the EQ can help. Have you tried YPAO, assuming your model has it? Just for fun/learning, you might try pulling one of the L/R speakers down off the riser, and position it a little closer to the screen edge (still out of sight for viewers in the end or row seats). It looks like it is really close to the front wall. Then see if that helps the hot bass response at all.


Lastly, try a little toe-in on the mains and see if that affects the HF response.

I'm not sure why the center had that response before. I redid the non-EQ measurement a couple of days later just prior to delving into EQ for the first time on that speaker. Here are the results. The center already angles up, might just be hard to see from the images. When sitting in my listening position it's firing directly at my ear level. I would like to refrain from raising it because I notice the slight reflection off of the cabinet when viewing brighter material.


Center with and without EQ. Black = No EQ, Red = With EQ (Smoothing at 1/6)

 

[LastButNotLeast]



Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23291470



I'm not sure why the center had that response before. I redid the non-EQ measurement a couple of days later just prior to delving into EQ for the first time on that speaker. 

 

Unless EVERYTHING is in EXACTLY the same position, your results will be different. Try a couple of runs just moving the mic a few inches and you'll see what I mean. Part of the frustration. I mean fun! 

 

Quote:
Do you think I should use that value or am I over-analyzing?

 

Graphitis Nervosa (thanks, Bill Fitzmaurice).

 

Michael

 

 

[scubasteve2365]



Quote:
Originally Posted by LastButNotLeast  /t/1470721/roadmap-for-tuning-your-theater#post_23291538


Unless EVERYTHING is in EXACTLY the same position, your results will be different. Try a couple of runs just moving the mic a few inches and you'll see what I mean. Part of the frustration. I mean fun! 

Graphitis Nervosa (thanks, Bill Fitzmaurice).

Michael

Yeah, well, even when the mic was in the exact same position I'd sometimes (maybe 1 out of every 6 or so) sweeps would have this dramatic high frequency roll-off (12kHz and up).


I attributed that to a grasshopper farting in my lawn.

 

[Nyal Mellor]



Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23291464


I ended up using the center since it was the last speaker I EQ'd. See the REW plot below. The graph limits are focused in on the 80Hz X-over point. I took many sweeps with just the sub and center playing with varying degrees of delay. I can adjust with the AVR distance setting or by adding a delay in my MiniDSP EQ that is handling the subs.




I ended up settling on the dark blue line - which was oddly set at the AVR default (minimum) of 1ft and 0mS of delay in the MiniDSP. That blue line technically has the slightly higher SPL at the exact X-over point. Since taking the measurements I've noticed that the green line is nearly as high SPL at the X-over point but has a notably higher SPL from 10Hz below X-over all the way down. The values for that sweep is 3ft distance and 0mS. Do you think I should use that value or am I over-analyzing?

I would use the green. What we're looking for is maximum SPL at the crossover but also the smoothest response either side.


Also I would try adjusting your subwoofer distance the other way i.e. increase the distance relative to the speakers. This will delay the speakers relative to the sub. Normally the sub is acoustically behind the speakers i.e. the low frequency sound waves take longer to 'get through' the sub than they do the mains due to the group delay introduce by the EQ used in the sub to prop up the low frequencies.

 

[scubasteve2365]



Quote:
Originally Posted by Nyal Mellor  /t/1470721/roadmap-for-tuning-your-theater#post_23293798


I would use the green. What we're looking for is maximum SPL at the crossover but also the smoothest response either side.


Also I would try adjusting your subwoofer distance the other way i.e. increase the distance relative to the speakers. This will delay the speakers relative to the sub. Normally the sub is acoustically behind the speakers i.e. the low frequency sound waves take longer to 'get through' the sub than they do the mains due to the group delay introduce by the EQ used in the sub to prop up the low frequencies.

Thanks for the feedback, I admit was very skeptical of the default setting being the best.


Here's a better look at the same sweep. I turned off all of the irrelevant plots. The three are1ft 0mS (Blue), 3ft 0mS (Green), 25ft 0mS (Red). My AVR will go up to over 100feet, I believe.


As I increase in distance I notice that the SPL at the X-over point is cyclic, that is to say the SPL drops and drops with more distance until it turns back and is peak again. You can see in this image that at 25 feet (versus 1 foot) the SPL at X-over are essentially the same but the curves leading into the X-over point are not. Maybe this 25ft delay is closer to 180degree out of phase? Maybe adjusting to 50feet would put it back in time but as you said the mains will have an additional delay.


I'm assuming with the sub properly time aligned that the bass response will appear a little tighter?

 

[Nyal Mellor]

Good work! I think you are ready for this: http://www.soundoctor.com/whitepapers/subs.htm

 

[scubasteve2365]

Getting there: From the test CD link in the link:

Quote:
All modern, powered, sealed subs have a phenomena called group delay so to best integrate sub(s) you must fix that timing issue so the sub lines up in time with the mains at the crossover point. Since you cannot remove this inherent delay in the sub you must add this delay to all the top channels. You do this by manually setting the speaker distance settings in the setup menu. I suggest setting all the distance settings THE SAME and to a LOW number (for example 7 feet; then add 12 feet to the SUB distance only (so the sub distance now = 19 feet). Now you have added the correct amount of delay to the REST of the system (the L C R Ls Rs) so you can then properly FINE TUNE the sub's phase with the mains by adjusting the phase knob and polarity switch on the sub. Therefore you have added just a bit more delay to the tops to give you a space for a fine-tuning adjustment on the sub to sharpen the focus. This will give you the best possible impulse response through the entire system; the imaging and focus should then should be uncanny, and the bass focused and as tight as possible. If this is done correctly, even if the subs are behind you, you will NOT localize them; it will seem as if the bass is playing from the front of the room, where it belongs, and this is true even if the crossover frequency is as high as 120 Hz.

I think I get it, what I don't get is that author went through a very detailed, yet simple, scientific based manner to explain the problem but then offered what I think might be a sort of layman or feedback-less manner to solving the problem. Why would 12 feet be the magic number? Wouldn't different subwoofer driver mechanics or enclosure types (sealed, ported or something even more rare like I have an infinite baffle) introduce a variance in electro-mechanical delay in the output chain?


If I remember correctly (I'm not a home), my FL distance was 10 feet, my FR distance was 10 feet and my center was 9 feet. If I add 12 feet to these distances I'm at 21 to 22 feet. As I noted in a previous post it seems that as I drifted away from 1ft the SPL @ X-over was reduced but started coming back as I got into the upper low to mid twenties. Unfortunately I did not take any measurements at 20, 21 or 22 feet. I believe the closest measurement other than 25 feet was 18 feet. and 18 feet did not look good. 25 feet looked better.


Would it be your recomendation at this point to start with a sub distance of 12 feet greater than my mains and then tweak a few feet from there according to my REW measurements?


Also, why would I set all of my "other" speakers to the same distance setting? Luckily my front soundstage is very close in distance (within 1 foot tolerance) but when it comes to the surrounds, particularly my back surrounds a larger disparity occurs.


Is it just me or do most threads here focus on frequency response only and not these timing issues?

 

[LastButNotLeast]


Steve: Yes, most of the discussion is about FR since it's easy to measure and easy to "fix."

Nyal: So I should plug up the ports, even if the speaker/sub is designed as a "ported" speaker/sub?

Michael

 

[Roger Dressler]



Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23296193


I think I get it, what I don't get is that author went through a very detailed, yet simple, scientific based manner to explain the problem but then offered what I think might be a sort of layman or feedback-less manner to solving the problem. Why would 12 feet be the magic number? Wouldn't different subwoofer driver mechanics or enclosure types (sealed, ported or something even more rare like I have an infinite baffle) introduce a variance in electro-mechanical delay in the output chain?

Yes. And some subs also have electronic EQ, perhaps digital, and that adds to the uncertainty. In my system the mains are 8' and the sub is 27' even though they are all at the same physical distance.

Quote:
Would it be your recommendation at this point to start with a sub distance of 12 feet greater than my mains and then tweak a few feet from there according to my REW measurements?

Seems reasonable.

Quote:
Also, why would I set all of my "other" speakers to the same distance setting?

You wouldn't necessarily. Set them based on the actual distances.

Quote:
Is it just me or do most threads here focus on frequency response only and not these timing issues?

The timing issues are "assumed" to be covered by the auto EQ or even the tape measure. But as you can see, some finesse may still bring benefits.

Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23295390


As I increase in distance I notice that the SPL at the X-over point is cyclic, that is to say the SPL drops and drops with more distance until it turns back and is peak again.

Adding more delay will just move the sine wave to the next cycle, so the pattern repeats. The problem is that this offsets the sound arrivals whenever you decide to listen to something other than sine waves. It also causes combing interference at other frequencies. So you want the shortest delay offset that gives the desired relationship at the crossover point.


Be advised that mains at 10' and sub at 22' is exactly the same delay as mains at 0' and sub at 12'. The AVR looks at the difference in the distances to calculate the applied delays, not the absolute distances.

 

[Nyal Mellor]



Quote:
Originally Posted by scubasteve2365  /t/1470721/roadmap-for-tuning-your-theater#post_23296193


Getting there: From the test CD link in the link:

I think I get it, what I don't get is that author went through a very detailed, yet simple, scientific based manner to explain the problem but then offered what I think might be a sort of layman or feedback-less manner to solving the problem. Why would 12 feet be the magic number? Wouldn't different subwoofer driver mechanics or enclosure types (sealed, ported or something even more rare like I have an infinite baffle) introduce a variance in electro-mechanical delay in the output chain?


If I remember correctly (I'm not a home), my FL distance was 10 feet, my FR distance was 10 feet and my center was 9 feet. If I add 12 feet to these distances I'm at 21 to 22 feet. As I noted in a previous post it seems that as I drifted away from 1ft the SPL @ X-over was reduced but started coming back as I got into the upper low to mid twenties. Unfortunately I did not take any measurements at 20, 21 or 22 feet. I believe the closest measurement other than 25 feet was 18 feet. and 18 feet did not look good. 25 feet looked better.


Would it be your recomendation at this point to start with a sub distance of 12 feet greater than my mains and then tweak a few feet from there according to my REW measurements?


Also, why would I set all of my "other" speakers to the same distance setting? Luckily my front soundstage is very close in distance (within 1 foot tolerance) but when it comes to the surrounds, particularly my back surrounds a larger disparity occurs.


Is it just me or do most threads here focus on frequency response only and not these timing issues?

Quite right, and that's because most people don't use an acoustic measurement system. For those that can't measure Barry (the Soundoctor) recommends 12ft.


For time alignment your sub generally wants to be further away in the AVR settings than the mains. Without getting into other ways of looking at sub / mains interaction such as examination of the group delay plot, impulse response etc (which are difficult to do), I suggest taking sub measurements at 1ft increments past your 9ft for the center channel (i.e. 10, 11, 12, etc) and see what looks best. If you wanted to speed things up you could measure in 3ft increments and then bracket down to 1ft once the response starts getting worse and not better.


With JL F112s I have needed to add everywhere from 4 to 16ft of additional distance to get them properly integrated depending on the main speakers low frequency alignment (ported, sealed, etc) and sub placement.