REW, is it worth using to assess your system/set up?

[AJSki2fly]

I thought some of you may find this interesting and possibly helpful. I am not a sound engineer or an REW expert but have some understanding of how to use it to effect.

Many years ago I worked for the Wolfson Unit Sound and Vibration laboratory at Southampton University, I would set up experiments, take measurements and make changes to determine improvement to various pieces of equipment, generally in lowering sound levels, but also measuring a few loudspeakers. This gave me a basic understanding of how to interpret sound measurements.

Some years ago I tried using REW to assess speaker and listening positions in our previous house, with some success, but limited it to speaker position and listening. We moved into our new house 18 months ago and I now have time to do similar, the lounge is a rather an awkward room, with two large windows along one wall and pocket sliding glass doors on the opposite wall and a glass door in the corner accesing a hallway, so lots of room for reflections.

I am using REW on my MacBook pro (9 years old) and a simple Panasonic TV USB Microphone, the microphone has 2 pairs of mics on each channel and is quite good but nothing special. Before some of your hold your hands up in horror, I do appreciate using a USB mic such as the miniDSP UMIK-1 would give more accurate results as it has a very flat frequency response, however at present I am more interested in comparative measurements so the simple USB mic will suffice for now until I can afford £130 for the miniDSP.

the first thing do in REW was to use Room Simulator, you enter room W*L*H and then you can define where your speakers are in the room, and your sitting position, you also can define the absorption of each wall, floor and ceiling (higher the value the less reflective it is, so carpet is a high value). The software then gives you the anticipated frequency response and modal distribution in a graph, as shown below. What you then do is to adjust the speaker positions, being as practical or not as you wish, and the seating position, you can then compare up to five plans. By taking your time with this you are able to find the best speaker position and listening spot. The aim is to get as smooth response as possible avoiding large peaks and troughs.

Below you can see the room set up I ended up with and the comparison between what I started with and now have. In pic below, top is plan view and below it is elavation. Note my speakers are toed in towards my head which helps minimise reflections from side walls, our room is quite narrow, 10' 6", so awkward, but at least no square.





After the above I took two measurements with REW with the Mic where my head would be and ran a Frequency sweep through my system from REW and recorded the result. The first was the original position and the second with the position from the Room Sim as above. As you can see there is less drop out between 50-80hz and 80-200hz is smoother, and generally at the new position it is smother at the higher frequency as well.


So what does this mean? Well basically it means that there should be less interference to the produced sound at the new listening position. So the proof of the pudding is listening, using some good known tracks at the old set up and then the same at the new. I did this with around 10 tracks and the result was very clear, I had gained an improved and more focused soundstage, more depth and width, the clarity and definition of the mix opened up further and on some tracks I heard aspects of detail more clearly.

Believe me this is not hocus pocus, I trust my ears and I know when it sounds better, give it a go if you can you may be surprised with the results.




We have parque floor and I measure with no rug and with a rug as below, as is easily seen there is less re-enforcement across the mid to upper frequency ranges just my putting a 2m by 2.8m thick pile rug down. So obviously the rug stays. Simpy put the result is more clarity in the sound and less mid range emphasis, so more natural. We used to have the rug down anyway but recently redid the floor and was interested in the difference it caused.




Onto the next stage which is on going. Our lounge, as I have said, has rather a lot of reflective glass, not much I can do about that, but it also has hard reflective walls, behind speakers, rear of sitting position and either side of the sitting position. Based on my limited understanding and experience of how sound behaves in rooms, I want to see what improvements can be gained by placing sound deadening materials at these points, and then decide if some carefully chosen GIK panels would be worthwhile.

The first thing I did was to place a Peruvian decorative rug/wall hanging, that we have had for years, across the rear wall behind the speakers. It is not large only 1.6m by 0.6m and is 400mm down from the ceiling. I followed this by placing an old valet curtain on a clothes horse in front of the wall at mid point, rising to 2m and about 600mm wide. The results you can see below in the before and after frequency sweeps.




As can be seen from the above whilst all the frequency sweeps look similar there are variations but the most noticeable thing is using makeshift sound absorption at each end of the room the general sound re-enforcement across the frequency range is lowered, my investigations will continue with side walls as well. Lowering frequency re-enforcement is something that we should all aim to achieve, especially at the lower frequencies, but do not miss-understand me on this, if you go out and buy a myriad of sound deadening foam and bass traps and stick it all over the room you will start to approach a room that sounds like an anechoic chamber and your music will sound completely dead and lack lustre. The key is to smooth things out and to try lessen re-enforcement at frequencies that are being enhanced.

Summing Up

You may be wondering how my system sounds compared to when I started on this three days ago, using my ears and listening at each stage I could detect decernable improvements in soundstage, clarity and perceived detail. I would not be wrong in saying that as a whole the sound is now more enjoyable overall, and the exercise has been definitely worthwhile as far as I am concerned.

In answer to the title of this thread I would say that REW is a very worthwhile tool to use to help you adjust your listening room to get the most from your system. What's best of all it is free, runs on a Windows or Mac machine, and with just a simple USB Mic it can be used to reasonable effect for comparative measurements, or for a £120 USB Mic get very accurate results.

From what I have read and listened to, one of the most overlooked aspects of an audio system is setting it up and the room. There is no point spending large amounts on HiFi and just placing it in a room where it looks best cosmetically for what ever reason, it will more than likely sound poor and be a real disappointment, also if the room has few furnishings or has lots of reflective surfaces this will also have a bad impact on the sound. This does not just apply to more expensive equipment, any carefully set up system in a well adjusted room can give very good results, at least the best that can be got from it. IMO making the most of your system and spending time setting it up and tweaking the room to get this best from it is very important, it may even stop you changing speakers or amplifiers unnecessarily if you realise the full potential of your system and the room it is in.


PS If you do change speakers or amplifier because you want to, I would strongly advise using REW and taking measurements again to get the best Speaker/Listening position, this is because all speakers have different frequency characteristics and have a different synergy with amps they are used with. So changing either is likely to result in a different listening experience from subtle to large. Obvious really but I thought I would mention it.

[Macca]

interesting stuff. I keep meaning to buy a mic and measure the response in my room. I have applied some treatments but only assessed them by ear.

Looks like you have almost a 10dB reduction between no wall coverings and all walls covered - that's a lot!

Is your room naturally well damped (lots of padded furniture, booshelves, clutter) or is it quite barren?

[Qwin]

Measurements have their uses, I use REW when working on my crossovers.
A little knowledge though can be a dangerous thing, I'm no expert either, so I bear this in mind when taking measurements, as its a very complex subject.

What you have to consider is that measuring gear does not hear like we do!
A 3dB rise equates to twice as loud when measured, but it takes nearly a 10dB rise for our non linear hearing to perceive things as twice as loud.
We also hear high frequencies as louder than low frequencies, when fed two signals of the same SPL.
So there are elements at work, that make the ideal flat measured response invalid.

I stopped aiming for a flat measured frequency response some time ago, to me it sounds dull and lifeless.
I aim for a gentle 3 to 4dB slope, from the bass roll off point down to the HF roll off point.
This sounds infinitely better to me and instruments like Piano are more realistic across their range.
It's no coincidence that many well regarded speakers measure similar.

So ultimately, let your ears tell you what is working best in your set up.
Just my take on it, YMMV.

[Macca]

I don't think anyone aims for a flat response in room. As you say it would sound very toppy.

Regrettably this is often confused with a flat response in an anechoic chamber. If you put a speaker that measures flat anechoically in a room you will get room gain boosting the bass and attenuation from furnishings reducing treble. So you get a curve downwards in FR from 20Hz to 20Khz and it will sound natural.

Harman curve is 10db down 20Hz to 20KHz, according to their research that's what most people prefer.

[Qwin]

You'd be amazed how many people don't get that Martin.
I think 10dB is a bit much and prefer up to 5 dB slope.
When I had my DSP unit, I played around a bit with this, but its really down to personal preference.

[AJSki2fly]

Following on from my earlier REW and room treatment investigations after some thought I decided to try out some simple side wall treatment to see what impact it would have, focusing on 1st and 2nd side wall reflection points. I have been told these points are essential to treat to avoid issues with stereo imaging and comb filtering. Basically any reflections reaching your ears very shortly after the direct sound can generally create an inaccurate listening environment.

In case others want to try this out, sit in your listening position and get someone to move a mirror along the left side, wall starting with the mirror at the side of the speaker, gradually ask them to move the mirror along the wall until you can see the left hand speaker, mark that point, it is the 1st reflection point. Continue moving the mirror along the left wall until you can see the right hand speaker in it, this is the 2nd reflection point. If your seating position is central to the room and speakers then the reflection points should be the same on the right wall, if not repeat for the other wall but will looking for right speaker 1st and then left speaker, mark those points. 

At the 1st and 2nd reflection points on each wall you need to place some room treatment to minimise reflections. As you can see in my case below I so far have tried out just hanging curtains using simple metal poles placed over hooks, which go across both 1st and 2nd reflection points.




The result is an immediate improvement, the detail, timing, attack and decay of the music is much improved, as a result instruments are better defined and overall clarity is better. In fact I would go as far to say that it is no brainer to do this, the overall improvement definition is superb, and soundstage is more 3D. I think if I did similar to the ceiling then there would be further improvement. I will try out other materials, such as blankets and sound absorbing foam to see what work best and which is most practical. The beauty of the velvet curtains is I can easily pop them up for a listening session and take down when finished, quite SWMBO friendly IMO.

So all this cost me was about an hour of my time and some spare velvet curtains. I would thoroughly recommend that anyone tries at least this out. 

The graph below shows the change in the sweep, you can see high frequency improvement, and 3.5-4khz uplift(I do not know why this would occur at present), and the sweep is a little smoother.  

[Macca]

magnitude of the effect will depend on how narrow the directivity of the speakers is. The narrower the directivity the less improvement to be had.

I found treating first refection points to be pretty much unnoticeable, but treating the areas behind and to the sides of listening position made obvious improvements.

[Gazjam]

Always good to hear 'real world' experience over Theory.

Where the smart stuff lives.

[Magna Audio]

I've used Holmimpulse a huge amount in setting up, modifying and tuning my system.

I don't think Holmimpulse is still available but it's similar to REW (was free) and has all the func. I need.

Measuring crossover points, time alignment, dB level settings, controlling room modes (29Hz and lesser at 56Hz), very helpful.

My system does not require any correction above those frequencies.

I do fine tuning of the bass cuts by ear.
Never add, always cut dB.

Time alignment by measuring first peak positive pulse is very beneficial.

These are not the sort of things an off the shelf speaker user can do anything about.

Once set I don't tinker unless I change something, like go from Tractrix to Le Cleach horn profile, change a driver pair or move house

[AJSki2fly]

Today my audio room tweaking extravaganza continues. I decided to try a somewhat larger damping/diffusion tactic at the wall behind me. This is with a thickish fleece blanket simply draped over two clothes dryers.

I took more REW measurements and these are quite interesting. As you can see below, with the most recent, Pink a dB average of 5 sweeps, the response has been raised overall, and most interestingly the response from 20-60hz has been smoothed, quite dramatic is that frequency response above 10khz has been raised.

The improvement down at 20-60hz is good I think, and the improved higher frequency is good, although above 13khz it will be lost on my ears.

Listening to the system now is subtly different, well maybe not so subtle in fact. The immediate impression is of a slightly drier sound, or a better description might be it is as if the bass and low-mids have more impact and depth on the sound overall, maybe better tonal balance.

I think what has happened is a softening of mid-high end frequencies due to less reflections/echo, and a subtle improvement in low bass. It would be interesting to see how adding bass traps would change things. I need to come up with a way of trying this out simply and cheaply, maybe these for a small outlay initially these bass traps could work.

If they work well then simply covering with something like white speaker fabric would be better.

Wall treatment behind me could be achieved with something like these wall treatment mounted on a ply panels and then covered with the same cloth, I could hang them on hooks, so are easily taken down if I want.

I have listened to quite a few of my favourite listening tracks and I am coming to the conclusion that there is a definite improvement in the overall presentation. My initial thought was it sounded a little dry, but after listening to several tracks it has become clear that generally there is better clarity to the music, and on certain tracks with big dynamics that elusive ‘darkness’ in between notes is more pronounced. This is particularly evident on ‘Moon Light On Spring River’ - Zhao Cong, “Sounds Of China”, a superb test track with huge dynamics, lots of very deep natural bass from a temple drum, and a variety of percussion instrument and bells.

I would be interested in any of your thoughts, suggestion or experience with room sound treatment.