Home Theater Forum and Systems banner

1 - 8 of 8 Posts

·
Premium Member
Joined
·
21 Posts
Discussion Starter #1
I am new and I am interested in using REW. I have downloaded REW and am getting familiar by reading the help files and this forum. I have ordered the BFD 1124 and all of the cables I will need. Also, I ordered a SBLive external USB sound card, so I think all of my needs are ready. What a great forum for newbies?

I have question that I hope someone can answer. I have two M&K Subwoofers to integrate into my system. They a placed in my 12' W X 16' D, in the front left and right corners, about 5" from each wall. They are physically 11.5' from the PLP/SPS meter, and the mains/center are located 8.5' from PLP/SPL meter. I downloaded 1/3 octave test tones. I set subwoofer gain on each subwoofer individually, using a 50 hz test tone to 75db. I then connected both subwoofers and used a pink noise download to set my receiver volume to 75db at the PLP. I then ran test 1/3 octave test tones from 20 hz to 125 hz, and entered the result on a graph, I repeated the test tones and graph until I had an average of 75db on the 9 1/3 octave tone. I found +10 db peaks at 40 hz and 100 hz, crossover is set to 100 hz in receiver. I had a dip of -8 db and a dip of -15 db at 80 hz. I found information on the M&K speaker forum on how to blend the subwoofers and mains. The procedure was to run the test tones, previously listed in this post, and average the results to 75db, did that. Then he suggested adjusting my subwoofer distance in my AVR and rerun the test tones to get the best response. I ended up setting subwoofer distance from the 10.5' after I initially ran audessey, to 17.5'. After incrementing the subwoofer by 1', run test tones, graph, repeat, several times, and then rechecking the results on my graph, I arrived at the 17.5' subwoofer distance. The 50 hz and 80 hz dips were decreased to 50 hz -4db and -4db at 80 hz. After listening for the last two weeks, i was very happy with the resulting blend between the subwoofer and the mains. Even my wife said how everthing was blending together well.

Sorry to be so long winded, hope I wrote this so someone could understand.

My question is should I leave the SW distance at 17.5' when I start using REW? How does subwoofer distance set in Audessey affect bass response in a given room? Obviously something changed for the better using this poor mans method of setting subwoofer to mains. If this fixed my dips at 50 hz and 80 hz should I leave the SW distance alone. I still have +4db peak at 25 hz, +6db peak at 30 hz, +10db peak at 40 hz and +10db peak at 80 hz, that I hope REW and the BFD will help me fix.

I would really appreciate some help on this, I have spent a lot of time trying to learn how to best integrate my two subs and mains. Ready to move forward with guidance!
 

·
Super Moderator
Joined
·
9,264 Posts

Welcome to the Forum, dj!

Your subwoofer distance setting should be set before you start using REW. Sorry, can’t anwer your questions about Audssey, I don’t know anything about it. There’s a sticky thread at our Home Theater Receivers | Processors | Amps Forum where you can fine help with that.

When you get your BFD, the best way to equalize the subs is as a single entity, with the measurement taken at your listening position. Don’t bother to try to EQ them separately, it’ll drive you nuts.

Regards,
Wayne
 

·
Registered
Joined
·
678 Posts
Welcome to the forum.

I infer that you opted for 1/3 octave test tones because you are still waiting to use REW. With REW, you can use the sweeps to measure the curve in detail, not just looking at individual frequencies separated by 1/3 octave.

As I understand the logic of adjusting the subwoofer distance by looking at the dips in the response curve, the idea is that, when the distance setting is correct, the sub and the mains should be in phase at the crossover. Of course, multiple distance settings can give the correct phase relationship at the crossover; only the correct distance setting will give the correct phase relationship across a wide range, say, ±1 octave around the crossover. Things get complicated when you take into account the phase changes in the signals themselves introduced by the crossover, and the phase changes from reflections in the room.

If you want to look at another technique, I wrote an article discussing techniques now available with REW v5, looking at the group delay and impulse curves. It is a little complicated, though, and you may want to experiment a while just taking frequency response measurements with REW before you try timing measurements. Audyssey uses an automated technique similar to this to calculate the distance based on the time to the impulse; its procedure is generally accurate but occasionally gives distances that are too low.

Bill
 

·
Premium Member
Joined
·
21 Posts
Discussion Starter #4
Welcome to the forum.

I infer that you opted for 1/3 octave test tones because you are still waiting to use REW. With REW, you can use the sweeps to measure the curve in detail, not just looking at individual frequencies separated by 1/3 octave.

As I understand the logic of adjusting the subwoofer distance by looking at the dips in the response curve, the idea is that, when the distance setting is correct, the sub and the mains should be in phase at the crossover. Of course, multiple distance settings can give the correct phase relationship at the crossover; only the correct distance setting will give the correct phase relationship across a wide range, say, ±1 octave around the crossover. Things get complicated when you take into account the phase changes in the signals themselves introduced by the crossover, and the phase changes from reflections in the room.

If you want to look at another technique, I wrote an article discussing techniques now available with REW v5, looking at the group delay and impulse curves. It is a little complicated, though, and you may want to experiment a while just taking frequency response measurements with REW before you try timing measurements. Audyssey uses an automated technique similar to this to calculate the distance based on the time to the impulse; its procedure is generally accurate but occasionally gives distances that are too low.

Bill
So, does that mean that even though my 17.5' distance setting, eliminates the dips by +10db, that my speakers are not in phase?

Thank you for your reply, I have a lot to learn going forward!
 

·
Registered
Joined
·
678 Posts
... does that mean that even though my 17.5' distance setting, eliminates the dips by +10db, that my speakers are not in phase? ...
The only way you will know that is by taking measurements yourself. REW v5 does let you take absolute timing measurements to compare speakers in time (via impulse or group delay graphs) or in phase.

If the distance calculation is off, you can be in phase at one frequency (at or near the crossover, say) and out of phase at other frequencies. Imagine what happens if you start with everything right, then you increase the distance by 14'. You will still be in phase at 80Hz, but you will now be out of phase at other frequencies, more out of phase the farther you go above or below 80Hz.

Something just occurred to me: you changed your distance setting by ~7' from the Audyssey setting. That happens to be 1/2 wavelength at 80Hz. It could be that Audyssey calculated the distance correctly, but did not tell you that the phase/polarity of the speaker needed to be inverted. Most subs have a +/- polarity switch, frequently labeled 0°/180° phase. Or it could be that you indeed have it set correctly now. You'll just have to make some measurements to find out.

Have fun,
Bill
 

·
Registered
Joined
·
74 Posts
I agree with Bill: you might want to play with the phase settings to see if that helps your distance conundrum. Normally Audyssey is pretty good at getting the relative distances right, so it might be it just got confused by the phase being reversed.

In the end, if adjusting phase doesn't help, as long as the FR looks good and you are happy, I wouldn't let the odd value throw you a curve. Sometimes things are weird but OK anyway...

shinksma
 

·
Premium Member
Joined
·
21 Posts
Discussion Starter #7
The only way you will know that is by taking measurements yourself. REW v5 does let you take absolute timing measurements to compare speakers in time (via impulse or group delay graphs) or in phase.

If the distance calculation is off, you can be in phase at one frequency (at or near the crossover, say) and out of phase at other frequencies. Imagine what happens if you start with everything right, then you increase the distance by 14'. You will still be in phase at 80Hz, but you will now be out of phase at other frequencies, more out of phase the farther you go above or below 80Hz.

Something just occurred to me: you changed your distance setting by ~7' from the Audyssey setting. That happens to be 1/2 wavelength at 80Hz. It could be that Audyssey calculated the distance correctly, but did not tell you that the phase/polarity of the speaker needed to be inverted. Most subs have a +/- polarity switch, frequently labeled 0°/180° phase. Or it could be that you indeed have it set correctly now. You'll just have to make some measurements to find out.

Have fun,
Bill
Bill,

I got the following quote from Audioholics Online A/V Magazine, post titled "Subwoofer Calibration Using Rives Audio Test CD 2"
written by Clinton DeBoer. This is where I got the idea of playing with subwoofer distance to blend subwoofer to mains. I found others that used this same method with success on the M&K speaker forum. Whether right or wrong here is the post. Before I continue with my EQ with REW and the 1124, I would be interested in discussion from this forum! Subwoofer distance has got me bugged, because the additonal distance of 7', sounds better to me, since the only part of my manual test tones that changed were the dips at 50 hz and 80 hz, where the distance brought them 5 and 10db closer to the 75db level. Nothing sounds out of phase, with the mains, phase switch was tested in both positions.

Here is the quote from the post that I used.

"Subwoofer Calibration Procedure
Step 1 : Download and print the chart for plotting room response. It is available for free on their website. ( Chart )

Step 2: Set the SPL meter on a tripod in your primary listening position. Set the meter response for "Slow C-Weighted" and rotate the dial to the 80db setting. 80 db is a good level to use when evaluating bass response. Be careful when listening to the higher tones at levels above 80db. If your speakers are not up to the task, you could be looking at blown drivers. I have seen it happen. Not with this disk and these tones but be careful nonetheless. Rives mentions this as well.


Step3: Starting with your volume down low, play the 1000Hz tone from the compensated set of tones and raise it until you reach the 0 point on the meter. This will be 80db. It may help to loop the track until you have it correct.


Step 4: Starting with the first compensated tone run through all of them and mark them on your chart. You will now have a graph or your room response with 1/3 octave resolution.


Step 5: Now you will need to determine if your sub is in an adequate location to begin the process of lowering the peaks with the EQ. If your peaks are not too severe (less than +10db compared to the rest of the bass spectrum), run through the tones again and begin adjusting with the EQ to minimize peaks. When you are satisfied repeat step 4 by charting a new graph. If you have some problems with the first pass, then try relocating your subwoofer and start again. I said this was affordable, I never said it would be easy.


After the first pass, it was apparent that I had some serious issues. I had a huge peak (+20db) at the 63Hz point, one at 40Hz and some peaks in the 100-160Hz range. This is something I would have never known without running these test tones.


This meant when I set my subs level with the SPL meter and the pink noise from my receiver, what I was really in effect doing was setting the bass level to the peak at 63Hz and all the other frequencies were being obscured.


In reality I was down 21db at 30Hz. That also explains why the gain on my subwoofer was always so low.




Locating the SubwooferNow I could begin the process of finding a better location for the sub. This is what I did to get a flatter response in my room. There are several schools of thought on this and since every room is different your steps and conclusions will no doubt be different too.

I initially used the crawling for bass technique when I placed my sub, but because of my room I found this was not ideal. I attached long cords on my sub, placed the offending 63Hz tone on repeat and started moving my sub around the room until I got the tone into a more manageable range.

After trying no less than eight places, I settled on the front right corner. Incidentally, I had to turn my front firing sub into a down firing one. Hockey pucks worked great for this and they are only .99 cents each. Also, I had to increase the amount of delay to 20 feet even though the subwoofer is only 13 feet away. I could watch the peak at 63Hz rise and fall as I played with the delay settings. Without using a SPL meter and this CD, I would have never been able to see this either.


Remeasuring the Sub
Now according to the Rives CD and the tones, I was only up 8db at 63 and dead on at 31.5Hz. Now I had good subwoofer positioning to begin tweaking the response with the EQ. Since my sub is only rated flat down to 27Hz, I did not touch any of the EQ settings from 25hz and below and I lowered the bands from 80Hz to 160hz to account for any overlap in the receivers cross over. My final results were pretty respectable from 31 to 160.

My bass was now almost ruler flat according to the meter and the Rives tones and I only had to boost ever so slightly at a few points. However, I am only using the EQ below 80Hz. To fix the upper areas I had lower my woofer sections overall level. I have my receiver bi-amped so I just lowered them a few db's this also helped out from 600 to 2500 as well because it was a little high there too.


If you are not bi-amping your receiver, this is a good reason to do so, providing both your speakers and receiver are capable. If not then you will have to find the best compromise.


The Rives CD does not have tones in between those on the graph below so I am not sure what the response is in between those frequencies. I would like to see just how flat is in smaller steps 30,35,40,45 Hz etc. That is my second complaint. Rives doesn't mention it so maybe I am worrying for nothing but it would be a comfort for people like me to know. I also found out that my room has some very strange things going on when have the sub facing forward. If I changed the position of the meter slightly I would get totally different readings and it was hard to tell just where it should be facing. I was able to correct this by down firing the sub. Once it was facing down this all but disappeared on the low frequencies. Again, without these tones I would have never of known this. So thank you Rives!


After going through the rest of the tones out to 20Khz I found out my room wasn't to bad, aside from the expected few nulls and a peaks. Within the limits of this test, I found no peaks over 6db and only one null of -10db at 250Hz. All this is according to my SPL meter and the Rives tones. Since I have no other equipment to verify my results with, I can only go with what I have. I am sure it is not perfect but I am also sure it is closer than it was before. If the end result is better sounding bass for less money, than I have achieved the goal this CD was intended for. Rives mentions not to worry too much about the nulls and I have read several other sources that say the same thing. Not much you can do about them anyway until you get serious about treating the room, and that is something that I will do when my wife and I buy a new home."
 

·
Registered
Joined
·
678 Posts
I'm not sure there is much I can add to what I already wrote.

... Nothing sounds out of phase, with the mains, phase switch was tested in both positions. ...
You did test the polarity at each distance? That's good.

I've seen reports of people using REW to set their sub distance, by varying the distance and looking for the best response curve of the sub + main. In most threads, only the final curve is shown, if any. Here is one where cyberbri compared curves at two different distances (and polarities) trying to determine which is best. I did find another thread, where someone was comparing just the polarity, without varying the distance. You can see that REW is a finer tool than measuring discrete frequencies, the continuous sweep gives much more information.

... Whether right or wrong here is the post. ...
I'm not say one technique is right and another is wrong. I merely offered you another approach, if you are interested, and to answer your question as to why Audyssey might have calculated one distance where you obtained another. To my mind, if two valid techniques give me the same solution, it makes it more likely that the found solution is correct. Also, I was always queasy about varying the distance and looking at the curve, as I was concerned I might converge on a local optimum that was not the best solution. So I was happier looking at the impulse curve, then verifying that the end result gave me a better response curve than the starting point.

The challenge, once you start using REW, is that it can give you so much data, you can spend lots of time looking at things different ways to see what you can learn. It becomes addicting.

Have fun,
Bill
 
1 - 8 of 8 Posts
Top