Skip to main content

Isolating from external sound; windows; and should I visit Rod Gervais?

I'd like to make a spare room in my house suitable for recording spoken dialogue.

The three problems I see are:

- separating recording equipment/activity from performance
- the room is basically square (23'x23')
- there is a huge sawmill near my house.

A fourth obstacle is that I am poor. I either need to DIY-design and DIY-build, or have someone else who is afforable do the analysis and design, and DIY-build.

The sawmill is actually not that loud, but between the saw running, heavy equipment beeping its way across the lots moving raw and finished materials, and semi trailers loaded with logs and finished lumber passing my house, I'd at least like to investigate the possibility of addressing it. (Currently I just record when there is no activity at the sawmill.)

I attached an image that shows the room layout, and what I was thinking of doing.

What I'm wondering is:

- This space would not be used exclusively for recording. It would be great if I could keep the windows. Maybe replace them with some that have higher insulation, and then also use some acoustic inserts when recording?

- I *think* Rod Gervais lives/works about ten miles from me. Would it be outlandish to approach him about the analysis / design?

My fear is that some error in design will render whatever I build effectively useless. The quote in Rod's book is something along the lines of "the answer is always the same - tear it all down and begin again".
Attached files


komencanto Sun, 01/22/2017 - 22:14
I left a Zoom H4n recording in the room during business hours. Not exactly a scientifically clean test, but it does identify the kinds of sound I need to deal with.
The recorder was placed directly on the step of a step-stool.
I used the onboard stereo mics with the recording level at "100%". No idea how that translates against any absolute standards of measurement.

The "Sawmill activity" clip includes the sound generated by the saw, trucks entering/exiting, and loaders beeping. Truck engine low frequencies resonating a-plenty. I also included a version of this file amplified by 17dB.

The "Saw shuts down" clip captures the sound of the saw being powered down for the day.

"Heat kicks on" records the forced-air heat coming on through the ceiling vents. I think the beginning of this clip is most representative of the room tone when the sawmill is closed and things are "quiet".

I also included "Three-second intervals", which shows some weird wave form stuff going on at 3-second intervals. It's inaudible even when amplified. Not even sure if this is actually sound, or electrical interference of some kind.





Attached files 3-sec-intervals.mp3 (2 MB)  heat-kicks-on.mp3 (398 KB)  saw-shuts-down.mp3 (859.2 KB)  sawmill-activity-plus17dB.mp3 (5.7 MB)  sawmill-activity.mp3 (5.7 MB) 

komencanto Tue, 01/24/2017 - 19:39
I did what analysis I could without a $1500 integrating sound level meter. :D

I recorded for a full day again, this time using the microphone I actually use to record rather than the on-board mics. I also set the recording level about where I normally set it to record dialogue.

Using the rudimentary frequency analysis tool in Audacity, I generated the plot attached. It shows peaks at multiples of 60Hz:

60Hz at -52dB
240Hz at -57dB
480Hz at -79dB

The tone at 240Hz is the clearly recognizable sound of the saw so I'm guessing those frequencies are the speed it spins at and its harmonics.

These peaks are fairly consistent for any stretch of time during the (saw mill business) day.

The room dimensions are:
  • north-south: 21'2+1/8"
  • east-west: 23'3+5/8"
  • height: 98"
class="xf-ul"> There are many diffusive objects against the east wall at the moment so I'm guessing there isn't much chance for resonance to develop in the east-west axis.

I have given more thought to my room usage, and I do not believe any sound generated inside the room will pose a problem. 99% of it will be spoken dialogue at normal conversational volume, and *very* infrequently, raised voices. With microphones blocked off by the 3.5" rock-wool gobos, and reflections dealt with by the rather huge 2" fiberglass panels against the walls, I think I can record multiple people without worrying about room acoustics (it has worked with 2 people no problem).

The problem is the background noise from outside. We frequently edit together dialogue from multiple recordings into a single narrative stream, and if one of those sources has even subtle background noise and the others don't, it doesn't match up.

So I need sound-"proofing" to keep noise out. But if I keep the noise out, and the sound generated from within is not an issue, do I actually need to build out any
of the walls off-parallel?
Attached files

DonnyThompson Wed, 01/25/2017 - 00:18
komencanto, post: 446819, member: 50097 wrote: Rod Gervais , I sent you an email yesterday afternoon - did you get it? No rush, just want to make sure it did not get lost between here and there :)
komencanto, post: 446843, member: 50097 wrote: Oh well... :cry:

Give him a little time. It hasn't even been a week yet since you sent him the email. Rod has always been good about responding to queries, but you're not the only one out there asking him questions.
Be patient, it's not as if you're going to correct all these problems in a 72 hour period, or that you even need to.

OBrien Wed, 01/25/2017 - 04:59
The tone at 240Hz is the clearly recognizable sound of the saw so I'm guessing those frequencies are the speed it spins at and its harmonics.

You will also find, once you do an analysis of the roommodes, that they exist at 120hz and multiples. So the existing room makes these frequencies prominent because it is acoustically sympathetic to them.

komencanto Wed, 01/25/2017 - 12:06
I'm adding an audio sample taken outside the room today along with the Audacity "frequency analysis".
It captures a little of everything - saw, loaders beeping, truck arriving.
This was recorded at the same level as the recording that generated the other frequency analysis a few posts back.
(FWIW microphone is a Rode NT1-A)

Attached files outside-sawmill.mp3 (624.5 KB)