Well, my quest to get decent sound in a poor room, on a budget and with strict spousal restrictions, continues... Thanks again to everyone who has helped me so far.

Depending on what more experienced people think, this post could be either my way of paying back a little (as much as I can at my current stage of development) or maybe just another opportunity to discover that I have it wrong. I'm happy both ways.

In a nutshell, here's what I've done: I've created an EQ curve that compensates somewhat for my room acoustics and speakers at my listening position and I intend to place it at the very end of my audio chain and mix through it.

Here's how I did it:

I first did some easy and fairly minimal room treatment that cost very little and didn't upset my wife. I got some shiny new monitors -- Focal Solo6's -- (and I know that I'll have to redo this after they've broken in). I love the new Focals, but I felt (by ear) that things were a bit muddy in the midrange and not quite as punchy in the bass as I expected. My ears have recently been "tuned" a bit by the use of a good pair of headphones -- Sennheiser HD800's -- and a marvellous headphone amp -- the SPL Phonitor. (If I felt that I had enough experience and authority to recommend stuff, I would highly recommend that stuff. Certainly I've never experienced sound that came as close to sitting in front of a real orchestra or jazz ensemble.)

So, I loaded a nice reference song with lots of complexity and frequency range into Logic and used Ozone to take a snapshot of the frequency spectrum averaged over time. Then I put a mic at the position my ears are at when I mix and compared the frequency spectrum that it was hearing to the one that was captured in the box. (BTW, I tried two different mics and the differences were small and understandable, given the nature of the mics. Easy to allow for.) My ears were validated: I could see where the spectrum heard through the mic was deficient in the upper lows and between 1900 and 2300 Hz. That made me feel good. Then I fiddled with the settings on the back of my Solo6's until I made the spectrum as close as I could to reference spectrum. The gaps were smaller, but still there.

Just to see what was going on, I put the mic 18" in front of one speaker and fed it the same audio summed to mono. It was very gratifying to see that the speakers themselves, when almost all room acoustics were taken out of the picture, were able to produce a spectrum that came VERY close to the one taken in the box (with different settings at the back, of course, than those used to try to match at my sitting position). So the deviations at my listening position that I couldn't fix by twiddling the speaker settings were due to the room and not because of limitations of my speakers. (I really like these speakers so far. Of course I really haven't much to compare to -- these are the best speakers I've ever owned and I don't have access to a pro studio.)

Back to the second frequency spectrum... the one at my listening position. I used Ozone's matching function to generate an EQ curve that compensated for the differences between it and the one captured in the box. I then applied it to the 2 bus and listened to the reference song through it. To my ears, the sound was much closer to what I hear listening through my HD800's and the Phonitor. I A/B'd a bunch of times with the EQ on and off, through the speakers and through the 'phones. I did the same for a very different song in a totally different genre (1st was jazz, 2nd was alt rock). Things are still not quite as clear as through the 'phones, of course, the room reverberates and smears the sound a bit, but the frequency spectrum sounds more or less as flat as the 'phones to me; I don't hear any big lumps or holes where I did before. (I should disclose, however, that my ears are still fairly new to mixing and are still improving noticeably year by year -- although they are old to the world. Others would no doubt hear things that I can't yet. This is all pretty rough and ready stuff. Real pros and anal retentives should look away.)

So, to conclude, the EQ thing seems to work, as far as I can hear. I'm not suggesting that this a substitute for a properly treated room, but for those of us that need major renovations to attain acoustic nirvana, I wonder if this can be a viable, zero-cost, low effort mitigating strategy while we plan and save. Or maybe even a way for those who do have fairly nicely treated rooms to take the sound from 80% to 90% -- I imagine that that 10% could be very expensive to get for some rooms, using only acoustic treatment.

What do people think? Is this of value? (I'm worried that maybe not, or others would have thought of it and mentioned it somewhere in this forum.)

Don't you hate that? One of my searches just came back with something relevant on the thread titled "Nearfield monitor eq". (I'll figure out how to properly reference threads soon...)

A good response from Ethan Winer where he says:

The point being, I think, that it takes acoustic treatment to make the "sweet spot" large enough to be practical.

I did some informal tests in my listening chair just now and it's certainly true that the frequency spectrum that I'm hearing changes if I move around. Not very noticeably with a few inches, but if I lean over a foot or more to either side. However, the change I hear is symmetrical, right and left. My room is far from symmetrical, in fact the left side is open to a two story living room -- my studio's a loft -- and the right side is closed. Therefore, it seems likely to me that most of this change must be do to the peaks and troughs in the overlapping sound cones coming from the speakers, rather than from nodes in room reflections. No amount of treatment could change this. Ethan? (Of course this is just with my ears. The symmetry would have to be verified by taking actual spectrums and comparing them. An experiment for later...)

I guess the real question here is: "Am I (and maybe others) better off by doing this quick, zero cost, partial fix or not?" (So, just to be real clear -- the question is not "Would I be better off with a treated room?" I think even I know the answer to that one.)

It seems to make a real difference, to me. But then I'm easily fooled.

I'm sure that there's lot's of material on this elsewhere, but I had to try to figure things out for myself. I've attached three snapshots of a model of my monitor setup.

Certainly it's true that moving around, out of my normal listening position changes the sound. Because the sound seems to change symmetrically close to the sweet spot, and my room is not symmetrical acoustically, my suspicions were that most of that change comes not from nodes caused by reflections (which could potentially be treated) but from from the interactions of the direct sound coming from the monitors, reinforcing and cancelling each other differently at different frequencies. I knew there would be areas of change comparable in size to the wavelength of the sound, but what would the pattern look like?

The model I built is very simplistic -- it looks at speakers as point sources producing the same sound. Speakers are not point sources and, hopefully, the audio is stereo. The model also doesn't look at attenuation due to distance or air absorption (which is frequency dependent). But I think I can still learn what I wanted to from the model.

What the snapshots show is the sum of two sinusoids coming from my monitors at three frequencies: 80Hz; 200Hz and 2500Hz. The grid resolution is 10cm.

It looks like (if I've got my scale correct), at 2500Hz the granularity of the peaks and troughs begins to overwhelm the resolution of my grid. Moving my head 5 cm to the left or right will dramatically change what I hear at that frequency. This kind of change to the frequency spectrum of a sound cannot, as far as I can see, be mitigated by room treatment. (It is mitigated to some extent, I imagine, by the fact that speaker cones have area - they are not point sources -- that smears the effect. And also the frequencies we are hearing are a complex, changing blend where cancellation kind of averages out over the spectrum and over a few milliseconds. So what I hear isn't nearly as bad as the model shows.)

Most of the sound energy arriving at my ears has got to be from the direct sound of my nearfields. I'm not sure of the %, but looking at the additional travel distance of first reflections from the sides of the room, where the shortest path to my ears is double the direct distance, the amplitude of the sound from that reflection would be, at most, 25% of the direct signal (since the amplitude decreases with R squared), given perfect reflection. And I can't believe that even the most reflective surface in the room reflects more than 50% of the sound energy. The sound pressure levels coming out of the back of the speakers have got to be less than 25% of those coming out of the front (haven't measured yet, though). So I'm thinking that more than 75% of the sound reaching my ears (maybe more than 85%) is direct. If this reasoning is correct, then the effect that reflection nodes can have on the frequency spectrum is small compared to the effect of interactions of the direct sound.

So, it's quite possible that the deficiencies in the frequency spectrum that I recorded at my listening position come largely from interactions of the direct sound (and from the absorbtion of high frequencies with distance, even with nearfields -- part of the EQ that I derived with Ozone's matching function boosts the highs a bit) and that that is what my EQ is correcting. The first snapshot might even show why I felt that the bass wasn't punchy enough (confirmed by comparing mic'd and in the box EQ's): I seem to be sitting in a big trough at 80Hz!

Does any of this make sense? (Or perhaps just not interesting. I don't see any replies...) I know I'm doubtlessly duplicating stages of thinking (and mistakes) that many people must go through.

You are making the mistake of assuming that all of the frequencies/wavelengths behave identically in a small bounded acoustical space.

They do not.

Wavelengths larger than the dimensions are superposed into modal standing waves.
While frequencies/wavelengths smaller than boundary dimensions behave specularly.

You might want to read the tutorials on the RealTraps and GIK websites and investigate the characteristics of both modal versus specular behavior and how they vary.

Thanks for your reply SAC. I've read a little on the RealTraps site. I'll read the rest and on GIK.

I guess what you are saying (???) is that modal standing waves will accumulate energy over a number of vibrations and that this could skew the proportion of sound energy coming from reflected vs. direct sound. Am I interpreting you correctly? I'll look into this.

The fact remains that my monitors sound better, flatter, less muddy at my listening position with my EQ on. They sound a lot like the spectrum I hear through my HD800s and my Phonitor. I'd like to explain that somehow. In the end it's the ears that count, no? Of course my ears are just babies, I do understand that. It will interesting to hear what I hear in a few years.

eq is unable to do anything regarding the time domain (decay times).

and ears are easily fooled...hence why most statements require measurements

Yes. Understood, I think. The EQ solution is just about static frequency spectrum. I'm just wondering if this can still be a significant proportion of the sound deficiencies in many rooms.

In terms of the value of measurement, I'm totally with you. Unfortunately, it's sometimes hard. How about this for way of measuring the portion of the sound energy at my listening position that comes from reflected sound:

1) I know that the sound pressure levels measured quite close to a monitor are almost totally due to the direct sound. I can measure these with my sound level meter, calibrating Logic's Test Oscillator and my speakers at say 20 different frequencies.
2) With a little thought, I can probably figure out the theoretical curve for the reduction of sound pressure levels due only to distance.
3) Any differences from the R squared attenuation have to be due to either: energy contained in reflections (these will make the curve deviate upwards with distance); or absorption due to heating the air (these will make the curve deviate downwards and will effect mostly the higher frequencies).

I'm thinking that, if the absorption of sound energy travelling through air is significant at nearfield distances -- and I suspect it might be -- I will be able to see it in a gradual difference as the frequencies get higher and be able to then allow for it. Reflection energy should be a lot lumpier, I'm thinking, especially where there are modes.

Once I've subtracted (if necessary at the level of accuracy I'll be working at) the absorption factor I should be able to see, at those 20 frequencies, how much of the sound energy at my listening position is due to reflected sound. (It would help if I can find the theoretical air absorption factor by frequency and distance.)

Does that sound like proper measurement design? A lot of work. :(

If your room (or tracks) have too much energy in the tubby bass range around 150 to 200 Hz, then cutting that range will seem to sound better. Sometimes even a good mix will seem better when reducing the tubby stuff with EQ. But in the end, bass traps are the only viable solution. What you need when mixing is accuracy, not "sounds better."

--Ethan

________________
The Acoustic Treatment Experts

Agreed. Accuracy is the thing. When I say "sounds better" it's short form for "sounds like it will make for more accurate mixes". Better = better mixes.

In terms of frequency spectrum (stereo field is another thing, as is visceral bass perception), I have been taking my HD800 headphones with my SPL Phonitor to be the best reference that I have access to at home. When I add EQ and it sounds more like that reference, I'm thinking that that's "better". No?

You're there and I'm not so I can only guess. I can tell you that the best way to know what's "better" is to measure your room with and without the EQ.

--Ethan

________________
The Acoustic Treatment Experts

Yes. Thanks. That makes perfect sense.

I've done that partially, I think, by comparing the frequency spectrum of complex songs mic'd at my listening position with the spectrum captured in the box. That's how I came up with the EQ curve in the first place. (My first post in this thread.)

I understand that proper measurement software will show me all kinds of stuff and I've been meaning to do some real measurement for weeks now. Unfortunately, I don't have an omni mic and can't find one in my limited circle of musical friends. Not sure yet if i want to buy one just for this. Maybe. Or maybe I'll give the measurement a try with a cardioid.

Back to practical considerations inside a small room.

Effective room equalization starts with a good pair of quality speaker stands (ear height 1.2m or 4ft) and moving speaker around by iteration in the room (according to floor marked trial pattern). Toward more satisfactory low frequency response, overall in-room flat response and precision stereo image (minimizing unwanted reflections).

Once you have found few best spots (maybe 1, 2, 3 pair of positions provide flatter response) in comparison to reference headphones, you may reconsider your need for any electronic EQ.

DIY speaker in room integration optimization. Many good readings (from simple to more complex):

• www.genelec.com/documents/qsg/QS2-way.pdf
• www.hhb.co.uk/hhb/uk/source/brands/downloads/Step_by_Step_Guide_for_Louspeaker_Placement.pdf
• www.genelec.com/documents/publications/PlacementandRoomAcousticsInteraction.pdf
• www.rpginc.com/news/library/ropt_wp.pdf

Thanks Unique. That sounds very sensible too, to me. I have tried a couple of different spots for my monitors, but haven't done any comparison of the frequency spectrums. I will do so. Also buy some measurement software. (My trial version of FuzzMeasure Pro has expired...)

BTW, have you folks seen this: DECWARE - Article about Setting up a Listening Room without Treatments

The article is interesting... It suggests that one way to make up for a poor room is to try a diagonal speaker setup, rather than one that conforms squarely to the sides of the room. I think the idea is that with a speaker set up that is rectilinear, there are more opportunities for early reflections to end up back at your ears. His idea is to find a set up that works best in the untreated room, and that that set might very well be diagonal, not straight on, and then treat the worst reflection points.