AVS Forum banner

Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs

2M views 38K replies 1K participants last post by  LastButNotLeast 
#1 · (Edited)
The purpose of this thread is to explain how to both physically hook up the connections between your computer and AVR/Pre-Pro to get started with REW (Room EQ Wizard) and to share information on the proper use of REW including proper techniques for both measuring and interpreting graphs, what you should be looking for in each graph, how to fix problems using room treatment, speaker placement, listener placement, and other tweaks, both before and after EQ, and delve more into topics not often talked about which when dealt with properly can drastically change (for the better) your listening experience to become true audio nirvana.

I highly recommended reviewing the very thorough guide created by AustinJerry (My heartfelt thanks to you, Jerry, for taking the time and effort to put this together and keep it updated) which contains information contributed by myself and many other AVS'ers on both how to connect your equipment and how to use REW in (post 275).
 
#7,061 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...ow-to-interpret-graphs/7000_100#post_23958590


It's the center that matches the left and right, first generation Pioneers, aimed right at the MLP. I can certainly move it around some.

So I should use PLII movie/game and get FR to be close to stereo/flat?

Thanks.

Michael

Is the center sitting on a shelf or ledge of any kind? If so, do you have it sitting over the edge?
 
#7,062 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/7050#post_23958590


It's the center that matches the left and right, first generation Pioneers, aimed right at the MLP. I can certainly move it around some.

So I should use PLII movie/game and get FR to be close to stereo/flat?

Thanks.

Michael

Measure L+Sub and C+Sub. If there's a large dip in the crossover region of one combination but not the other then the polarity of one speaker is probably reversed.
 
#7,063 ·

Quote:
Originally Posted by jkasanic  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/7050#post_23958668



Is the center sitting on a shelf or ledge of any kind? If so, do you have it sitting over the edge?
 
Yes, and yes.
Quote:
Originally Posted by markus767  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/7050#post_23959034



Measure L+Sub and C+Sub. If there's a large dip in the crossover region of one combination but not the other then the polarity of one speaker is probably reversed.
 

Using a legacy system, so can't (easily) do that. But it's my project for the week.

Thanks for the help and suggestions. Graphs (20 to 300, no smoothing 
) to follow.

Michael
 
#7,064 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...ow-to-interpret-graphs/7000_100#post_23959401

Quote:
Originally Posted by markus767  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/7050#post_23959034


Measure L+Sub and C+Sub. If there's a large dip in the crossover region of one combination but not the other then the polarity of one speaker is probably reversed.

Using a legacy system, so can't (easily) do that. But it's my project for the week.

Thanks for the help and suggestions. Graphs (20 to 300, no smoothing 
) to follow.

Michael

The legacy part is irrelevant. When you measure L, C or R + sub you just need to make sure bass management is engaged (I.e. Set speakers to small which they should be already). For center, change the mode to Dolby PLII Cinema with both L and R connected). There has been some recent discussion about measuring all 3 together but I'm not sure if there is consensus on the best means to do so yet or not (e.g. Changing width mode in PLII or using all channel stereo). In any case, for now you're just interested in the three separate responses all overlayed on one FR plot. I would show 15-300 unsmoothed and 15-20k Hz at 1/6.
 
#7,065 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...how-to-interpret-graphs/7060_20#post_23959401


Using a legacy system, so can't (easily) do that. But it's my project for the week.

Thanks for the help and suggestions. Graphs (20 to 300, no smoothing 
) to follow.

Michael

Why would this measurement be difficult on a legacy REW system? Measure center+sub by feeding the REW signal to both left and right channels and set output to PLII Cinema.
 
#7,067 ·
This is the crux of your whole issue. When you measure a mono signal split to the left and right channel of your AVR/AVP, it will route sound to the center channel in PLII mode. If you select stereo then each signal is routed to L and R which gives a combined response measurement.
 
#7,069 ·

I will experiment this week. Thanks for the suggestions.

Michael

 

And, as long as I'm still here, does anyone still have a link to the "limp membrane traps" that were talked about a few pages (eons?) back? The search function here is useless and Google takes me everywhere but anywhere productive.

Thanks.
 
#7,070 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...ow-to-interpret-graphs/7000_100#post_23959561


And, as long as I'm still here, does anyone still have a link to the "limp membrane traps" that were talked about a few pages (eons?) back? The search function here is useless and Google takes me everywhere but anywhere productive.

Thanks.

See here


It was actually the fourth hit on a search of "limp" in this thread. I find the fewer words the better.
 
#7,071 ·

That was it! Thanks.

And a great suggestion, too. I was using "limp membrane traps" and got a million hits on "membrane" and "traps," and the quotes didn't seem to do any good.

Hopefully, we won't get too sidetracked by the "limp" part.



 
#7,072 ·

Quote:
Originally Posted by LastButNotLeast  /t/1449924/simplified-rew-setup-and...how-to-interpret-graphs/7060_20#post_23959480

 

Because PLII Cinema produces sound in the left and right channels also, and I don't really want to disconnect the wires (though I can if I really have to).
 

Sorry, that is not correct.  A mono signal is routed exclusively to the center channel.
 
#7,074 ·
I've been reading this thread for a while now (IIRC I joined it at about page 90, time flies) and thought I'd post some graphs up to see what people think. I have, I think, stuck to the spirit of the house rules but I'm not entirely sure what they are at this point in time given recent posts so hopefully you can bear with me
The one difference to note is that I measure with a loopback (not using USB) hence my IRs/ETCs don't start at 0 because I like to confirm the alignment audyssey comes up with manually via examination of the IRs. I have adjusted time limits on the x axis accordingly though so they should still look the same as yours.


A bit of background info.... My room is my lounge, albeit a lounge that is largely dedicated to the cinema, and is of fairly typical construction for a Victorian house in England (suspended wooden floors, brick walls, 10' ceilings, has an open fireplace). It's not a big room (roughly 14' x 12' x 10'), it has a lot of sofa in it and there is a window behind the MLP... all of which mean that placing treatments is not so easy. Speakers are MK MP150 Mk2 across the front 3 & S150T surrounds. I have a single 15" sub atm using the Fi Q15 driver powered by an inuke, this is powered by a Marantz AV7005/Cinepro 2k6 Mk3 combo. I am currently contemplating a move to a pure (HT)PC solution and replacing audyssey with PC based DRC (have done initial evaluation of acourate), partly because I'm not 100% happy with multeq xt but also because I clearly don't have enough tweakability in my life right now



I know that bass decay could be better but I don't intend to consider that until the next time I do more structural work on the house & need to redecorate the room, at this point I will consider replacing the coving with soffit traps along the lengths of the room (corner traps are impossible because they would block the door or the sofa or the shelves). What I'm interested in are any views on the specular reflection side of things, they're obviously not perfectly given that I have no treatments but I tend to think they don't look too bad compared to other untreated rooms. Ultimately it's really just a Q of whether there is anything in here that anyone sees worthy of particular comment. Any views welcome!

L+SW and R+SW frequency response, unsmoothed


R+SW waterfall, noise floor in my room is ~45dB


Full Range L and R only ETC


Full Range L and R only ETC, first 3ms


Floor Plan (NB: rears are moved since I made this pic so they're not 1' above ear height and at 90 degrees to the MLP)



Cheers

Matt
 
#7,075 ·

Welcome to the thread, Matt!  And nice job with the measurement graphs.  I can tell you have been following our recent discussions with regards to the posting conventions.

 

I'm sure you will get some valuable comments from the experts who participate in this thread, but I have some observations I would like to share.  First of all, your sub seems to have a very good low end--that is a nice, flat response all the way to below 20Hz. The response in the 60-90Hz range indicates that there may be phase issues at the crossover point.  Have you tried adjusting the sub delay to see if it has any effect on the FR flatness?

 

The waterfall looks quite good, although I would have preferred seeing a 450ms window.  There is a bit of ringing shown by the "mountain" at approximately 90Hz.  the smaller time window might add some perspective.

 

I think the impulse measurement has some areas of concern.  I doesn't surprise me that you have quite a few reflections higher than -20dB, given that you haven't explored room treatments.  The very early reflections are actually quite severe, and are likely impacting the audio in your listening room.  At a minimum, I would recommend that you conduct some of the exercises described in this thread to at least isolate what might be causing the reflections.  This way you might find an easy way to improve the reflections without a lot of effort.  Those reflections must be coming from something quite close to the speakers or your listening point.  The font in the graph is quite small, but it seems as if the reflections colored in blue might be slightly worse.  Consider the "blocking technique" to identify where these are coming from.  Taming the reflections has the potential of significantly improving your listening experience, in my opinion.

 

Looking forward to your progress reports!
 
#7,076 ·

Quote:
Originally Posted by 3ll3d00d  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/7050#post_23961001
I am currently contemplating a move to a pure (HT)PC solution and replacing audyssey with PC based DRC (have done initial evaluation of acourate), partly because I'm not 100% happy with multeq xt but also because I clearly don't have enough tweakability in my life right now
For which I sincerely hope you've seen Equalizer APO:

http://sourceforge.net/p/equalizerapo/wiki/Documentation/

Welcome aboard.

And be careful what you wish for.

 
#7,077 ·

Quote:
Originally Posted by Digione  /t/1449924/simplified-rew-setup-and...nd-how-to-interpret-graphs/6930#post_23951538


The results you obtained are very similar to what I obtained when I added the 3" of Roxsul on the floor except as my ceiling is treated it doesn't introduce any significant reflections.


It is quite amazing how the floor is often neglected as a relevant reflective surface. I suppose it is because it is so difficult to easily treat. It would be great if some company introduced effective floor treatment that you could "walk on".



Thanks for taking the time to post these measurements. Pity you didn't get a chance to listen to the changes.

I finally had a chance to listen. To summarise, I would have to say the imaging became much stronger. I'm quite surprised really, I have all the other early reflection points taken care of, apart from ceiling, and while I had pretty good imaging prior, it's in a whole nother world now. I'm not sure exactly how best to describe it, but it feels like there is more space (depth) from front to rear. Previously it was like the recordings come from a solid wall with some projection as recorded to the MLP, now it's like that wall has been removed, there is more space behind speakers.

I actually went up to the centre speaker at one point to ensure it wasn't producing sound, such is the strength of the imaging now.


I'm within 2.5M of the LR speakers and I had laid some pink fluffy length wise from the speakers towards to MLP, this won't work full time with wife and kids, so the next chance I have I'll try standing some pink fluffy on it's wide end so that it's taking less floor space.


Super keen to dampen the ceiling now
 
#7,078 ·

Quote:
Originally Posted by Audionut11  /t/1449924/simplified-rew-setup-and...how-to-interpret-graphs/7060_20#post_23962198




Super keen to dampen the ceiling now

Of all the treatments I have applied, I was most reluctant to place panels on the ceiling. However, after I did it, the ceiling panels resulted in the most audible improvements I have made to date.
 
#7,080 ·
I'm sorry if this question has been answered. I have mostly kept up with the thread but working 7 days a week as of recent I may have missed. I will post a clip from J's recent comment that I have yet to find a satisfying answer for. Again I have been keeping up mostly with the measurement technique discussion but didn't hear a direct answer.


If you decide to view a full range measurement with L+R+Sub(s) you will most likely be disappointed above about 1khz or so (most of the time and unless you are VERY careful and have very exact speaker and mic placement between the speakers) as the wavelengths start to become so small the slightest movement of the mic will cause phase interactions, comb filtering, and a very poor and inaccurate measurement in the higher frequencies. There really isn't any reason for this. Again, the only measurement necessary for L+R+Sub(s) - combined - should be to look at the overall bass response, below 300hz, unsmoothed.


I did accidentally do this the other day. On my Mac I have to switch audio output to sound flower 64ch as per JPA's wonderful guide to only get one speaker plus LFE channels at a time. I had switch back to my Denon, which allows for both LR to play, to listen to a song for differences during tuning and forgot to change it back. I did notice some not so good looking graph up in the top end.


Why don't we do this? It seems this would be the most important since we listen with both of these speakers (plus a center usually) during a movie or multichannel music. What is the simple answer?
 
Top