Acoustic Room Treatment on a Budget
There comes a point in every music lovers' life where, finally, he (or she) can afford the sound system they've been wanting for years.
So, in the early part of 2005, I set about putting together something to do my CD collection justice.
Due to a love of making things, and a desire to save money, I decided to build a set of speakers, based on plans from Wilmslow Audio.
After a couple of months of cutting, drilling, sanding and painting, I had a set of speakers, and was suitably impressed with the sound when I auditioned (and bought) a receiver.
So, great speakers, great amp, and a great (borrowed) CD/DVD player were put in my living room. I fired it up, and... was very disappointed.
My great speakers now sounded odd, confusing, boomy, unclear, and, for complex music, downright unpleasant.
Searching for the solution lead through all the usual audiophile solutions - fancy cables, different CD players... nothing worked.
Finally I found a a group of people over at Audioholics that opened my eyes to the world of acoustic room treatment. Unlike much of the audiophile world of outlandish claims, backed up by no science, acoustic room treatment has a long history, backed up by good, credible science and testing.
I used ETF (www.acousticsoft.com) with a RadioShack digital SPL meter to get a reading of my room's frequency response. Setting it to a full range frequency sweep of five seconds I produced the following graphs (click for larger images):
What these graphs show is that the low frequency response is littered with peaks and nulls - as the sound is being amplified and cancelled out at different frequencies (due to room modes).
The full response shows severe comb filtering, as sound reaching the listening position is being affected by reflections from the walls, floor and ceiling.
The 3D graph shows the sharp peaks and nulls in the low frequency response, and the decay of sound energy.
Finally, the impulse response shows that the listening position is being bombarded with sound impulses after the initial speaker impulse.
After consulting with people at the Audioholics forum I set about building panels to treat the room. To learn about these panels click here, otherwise, read on for the results.
Based on information from RealTraps' Ethan Winer I placed the six thinner HF panels on the side walls and the four thicker HF panels on the floor at the first reflections points. The bass traps were all pushed into the room corners, and I took another set of readings (again, click for larger images).
The bass traps I'd added wouldn't touch the lowest peak (at around 40Hz), but even then, the low frequency response wasn't clearly better.
The full range response was obviously much improved, with reduced comb filtering.
The improvements at the lower frequencies were more apparent from the 3D graph - the peaks are slightly less sharp and the decay is a little more rapid.
The impulse response is slightly improved, though not hugely.
Following more advice over at Audioholics I moved the listening position to 38% of the room's length and took another reading.
This was a great improvement - the sound was much more pleasant, and the graphs showed it was a little better.
The low frequency response is a little smoother, and the full range is a little tighter, but still not perfect at the lower end.
The 3D graph shows some longer decays at the bottom end, but the impulse response is a little better.
I eventually fixed the HF panels that had been on the floor to the ceiling. The results were a great improvement.
The low frequency response is much better. There is more comb filtering in the full range response, though it is arguably a little flatter.
The 3D response shows longer decays, but has less pronounced peaks/nulls.
Finally, the impulse response is a huge improvement, though unsurprisingly, the reflections which were from the ceiling have been removed, but a new reflection just after the initial impulse has appeared. This must be due to the reflection from the newly uncovered wooden floor.
To combat this, I took a duvet (no rugs to hand!) and put it on the floor where the HF panels used to be, then took another reading.
Unsurprisingly, the low frequency response (plus 3D graph) wasn't really affected, but is still a huge improvement from the original readings.
Apart from a null around 4kHz, the full range response is pretty good - not too much comb filtering across the audible range (within 15dB peak to null).
The impulse response is a huge improvement. The duvet obviously isn't as effective as a panel absorber, but still results in a pretty good reading.
So, how does it actually sound?
The difference is huge - music that was OK is now crisp, clear and clean. Music that sounded busy and confusing is now a genuinely enjoyable listening experience.
Comparing the cost of treating your room to buying new audio components, it's clear that room treatment is more cost effective - and much more likely to succeed.
Get yourself some fibreglass today - you won't be disappointed!