As some of you know getting your system to sound just right can be a pain in the butt. No bass, too much bass, echo, thin sounding and shrill, muddy, poopy diaper. etc.
I'm hoping I can start a discussion that will let everyone ask questions and chime in with how they solved issues in their room which may help others. I would like your help in interpreting results and making decisions that will impact the acoustics of my HT room eventually. I don't know what I am doing in practice, but have read enough to guess the theory behind what is happening, why, and what to try first.
I can start by recommending first off if you wanna get consistent feedback with what your tweaks are doing to your system you really need to get over that hurdle of measuring your room with software. I know right. Lame, boring, pedantic, tedious etc. But in the end very fast, repeatable and rewarding.
I use the Audio Tools App by Studio 6 Digital. It is inexpensive and works well on my ipad. It has pre-calibrated mic corrections for i devices and bypasses the onboard audio filters in the OS. I get most use out of the included measurement tools. The SPL, RTA plots, FFT plots, RT60 measures and ETC plots are very handy. The ETC plot requires the purchase of the LARSA plugin for an additional fee and is not essential to see trends in frequency response in a room. There are also many included audio calculators and a handy curve overlay function that lets you quickly guestimate NCB levels, STC, Transmission loss etc. Dont worry it's easier than it seems.
For test material I have found using the included pink noise generator helpful, but that tethers me to the equipment with a signal cable. I invested in the Stereophile Editors Choice disc for $10 and it has many helpful test tracks and reference material including a very nice piano piece. Of all the test tracks the dual mono pink noise track is the most used. Any full bandwidth dual mono pink noise test track will do and there are many available free for download.
I'm just cutting my teeth at all of this, but a couple of beers with my father in law and the wife being away got my courage up enough to unbox my 2 sealed subs and tinker in my listening room. It helps when someone is urging you on with a grin on their face. The goal was to determine where my 2ch system was performing compared to a desktop powered M2 setup we both had some experience with. My system was coming up really short in the bass dept. as "Peg" by Steely Dan demonstrated.
The room is setup width wise in a 11' x 18.5' x 8' room. The speakers are setup symmetrically from the sidewalls\centerline of the room and the listening chair puts my listening position about 1' off the back wall. The room has been modestly treated with broadband 2x4 absorbers on the front wall and a hanging tapestry behind the listening chair on a tapestry rod 3" or so off the wall. These treatments were deployed to tackle the flutter issues the room had and reduce the overall RT60 to a measured overall 0.4s. The low frequencies falling off faster at 0.35s and the upper mids at 0.45s. The result is a room that sounds live but still clear and focused. I'm not allowed to treat further.
Here is the basic setup.
Using the pink noise test we measured a response from both speakers at the listening position. This was the before subs measurement. We used RTA at 1/3 octaves because it is easier to read from far away, is easier to spot trends and I knew this was more an exercise in experimentation rather than a lasting setup. A 1/12th octave FFT plot is a much more accurate way of measuring response and finding high Q issues. However, these are often red herrings that lead to tweaking problems that do not matter outside of sound tests.
Before subwoofers added response at MLP. This measurement was taken later with the subs turned off after all the levels were set.
We can see the response rolls off sharply just before 60hz. The peak centered on 50ish hz is due to the proximity of the back wall to the listening position. These are modal issues I really can't do anything about without some serious bass trapping. The effect seems to be a heavier thwack on bass drums and upright bass tracks. Luckily it doesn't seem to be an issue for muddying sound in my room.
We added the subs. One at first at the right side and then the other. They are just inside of the 1/3 mark from the sidewalls my speaker stands occupy. I had been talking them up quite a bit as we were building the basement together and he was disappointed when the first one was set up. It took a bit of messing around with dials, but I thought I had it. We listened to a few tracks and he was still convinced his powered M2s provided an overall more rounded and smooth listening experience.... I was a little miffed and quickly flew down the stairs to retrieve the other sub (afterall, the M2s are under $600!). I had it set up and matched the dials to the other in the hope the low frequencies would now blend better into the minimonitors. The result was a little fatter bassline than before, so I turned each sub's gain down about an "hour" or so on the dial.
With the subs integrated the front of the room took on a different feel altogether. Why is WAF an issue I wonder?
We re-listened to the material we were juggling around and decided it wasn't right at all. The bass was thick and slow. It was not boomy, but seemed to hang in the room too long and hide the detail in the recordings. This was all setup by ear at this point. I decided to break out the ipad and was prepared for the "jeez, really?!" But he was eager to get the system up to snuff. He was still riding high on his M2 performance afterall, and wanted to see just where this system was "better" -as I talk about audio all too frequently....
We returned to the pink noise test and set the levels of the sub gain to closely blend into the monitors. But something was wrong. The bass was now thicker and louder than before. "Peg" was now "PEG!" It dawned on me I had been testing with a graph response readout in dBA instead of dBC.... dBA is a better measure for frequencies in the speech range and is often associated with general SPL readings in industry guidelines. Its range is not appropriate to set subwoofer frequencies. Doh! I quickly changed the program settings, re-ran the pink noise test and we returned to "Peg." Much, much better.
Here is the response at the MLP after tweaking the low pass and gain settings with the help of real time meaurements.
The bass still measures ~5db fat, but in the room it blends perfectly with the mini monitors and provides a pleasing warm sound. We finished up our beers and he went home to leave me to listen to the Pink Floyd's "The Wall" for the first time ever. I know, I know. I'm not old like you guys.
I don't need to hear "Peg" for a while now...
Things I learned from this experiment.
1. Setting things by ear is a great way to start, but you will only ever reach ideal settings by fluke or by using a lot of time.
2. Knowing where your system starts is a better way of knowing where improvements can be made.
3. Having a second listener is an excellent way to prove your perceived results are real, and to keep you humble when you muck up.
4. Axiom makes good speakers.... Now I have someone to remind me of that constantly.
I will follow up with a plot from the M2 desktop system for a point of reference later. It will be interesting to see just how his setup has influenced his listening preferences. He loves his speakers.
