Find a way to insulate out train noise

Shinzon

Shinzon

New member
Hello everyone,

I'm new the forums but new to recording. I've been doing voice-overs for years and recently bought a house. The good news is, I have lots of room in the house, the bad news is, there is a commuter rail that comes by the (within 500 feet of the house) about every 30-60 minutes. I narrarate audio books and it's gotten to the point that I'm going crazy. Everything I really get into reading, I can hear a train entering the station down the street.

My normal way of insulating (some insulating foam on the walls, padded desk and carpet on the floors is not working. So I've decided I'm going to have to literally strip my office down the studs and insulate from scratch.

Here is some background on the house:

• Very solid, built in 1918
• I rebuilt the kitchen 6 months ago and in doing so found that there is almost no insulation in the exterior walls. They are primarily lath and plaster with horse hair - So I know I'll need to put in R-22 at least in the office walls which two of them are exterior.
• The windows are 5 years old, so I'm not too worried about thier ability to insulate, though I think I'll still do some kind of baffle that I can put over them while recording.
• I use a Neumann U87 which can pretty much hear a pin drop when tuned right.


I would greatly appreciate any advice anyone can give me on what they recommend I insulate my walls, ceiling and floor with.
 
I probably won't be the last, and maybe not the first, to say that foam on the walls is not going to do anything to suppress train noise, ever.

If you go to the studs you might as well consider building a room-within-a-room. A train is going to literally shake your entire house, so you'll need to decouple some of that from your recording environment.

Then make sure to high-pass everything. You can try basic noise removal with something that uses a noise sample, i.e., the train sound w/o anything else, and see what kind of artifacts that leaves. Maybe it will be good enough, or you might need to resort to something like iZotope's RX plugin - expensive but it can do some amazing things.

Good luck.
 
What Keith Said....^^^

Foam suppresses reflections.
You're trying to suppress a *transmission* of sound and that's only done through decoupling and mass.
Two or more layers of sheetrock attached to isolation strips (not directly to the studs) and the same treatment for the ceiling. Yup. Room-in-room thinking. It can hang over the lath, so you don't have to strip anything.
Interestingly, i was involved in something similar near here. After some of the room-in-room measures, including floating on rubber iso pucks, he STILL had some incursion from a heavy freight rail (about that same 500 feet).
Ultimately, the answer came in the form of an LED. We rigged a weather-shielded mic under the roof of his back deck and dialed it in to detect an approaching train. This illuminated an LED in the basement control room. That gave him about 10 seconds notice, usually a little less.
But his intervals were not half-hour. More like once or twice each evening and 3 times each day.
Money or imagination. That's kinda what it comes down to.


Ponder5
 
Ponder5 said:
What Keith Said....^^^

Foam suppresses reflections.
You're trying to suppress a *transmission* of sound and that's only done through decoupling and mass.
Two or more layers of sheetrock attached to isolation strips (not directly to the studs) and the same treatment for the ceiling. Yup. Room-in-room thinking. It can hang over the lath, so you don't have to strip anything.
Interestingly, i was involved in something similar near here. After some of the room-in-room measures, including floating on rubber iso pucks, he STILL had some incursion from a heavy freight rail (about that same 500 feet).
Ultimately, the answer came in the form of an LED. We rigged a weather-shielded mic under the roof of his back deck and dialed it in to detect an approaching train. This illuminated an LED in the basement control room. That gave him about 10 seconds notice, usually a little less.
But his intervals were not half-hour. More like once or twice each evening and 3 times each day.
Money or imagination. That's kinda what it comes down to.


Ponder5
Click to expand...

Right.

Do they make warning systems for mother in laws that just pop up? lol

That was a great idea man! I wonder if there is a way to just see the schedule of the trains coming.

It going to take a huge expense to completely isolate the train noise as it is more a ground vibration thing going on that will be very hard to eliminate. :(
 
jimmys69 said:
I wonder if there is a way to just see the schedule of the trains coming.
Click to expand...
I suppose on passenger trains, maybe. But this was a freight line with traffic (somehow) going both directions on no pattern we could ever really work out.

Even then, it's my bet they'd not publish such a thing. You know, for 'national security' reasons. Your studio might be a threat ;-)

Ponder5
 
Not only does foam prevent reflection - It prevents reflection of only very high frequency content.

You're talking about low end energy so powerful it can easily be measured by seismometers -- Energy that travels for miles through everything in its path that you can actually feel in the ground.

It's a nearly impossible task.
 
A common mistake many people make is going with absorption solutions (acoustical foam, absorption panels, etc) when trying to block outside noise. The only thing absorption products help with is reducing reverberation in your room, therefore improving sound quality. I recommend checking your room for any sound leaks, such as gaps on the seams of your windows. You can use caulk to seal up any cracks. A good way of finding sound leaks is holding a candle light by your window seams on a breezy day and seeing if the flame bends.
 
Massive Master said:
You're talking about low end energy so powerful it can easily be measured by seismometers -- Energy that travels for miles through everything in its path that you can actually feel in the ground.

It's a nearly impossible task.
Click to expand...

Nearly impossible, true.

But on the good side, what we learned, was that while yes, much of the noise could be classed as seismic, it also planted it firmly beneath any audible frequency range (20Hz).

In fact, we also learned with pretty good analysis gear that *feeling* the place move was not the same as *hearing* it move nor is it the same as any actual artifact in the recording. At first, that sounds counter-intuitive, but it ends up that you can feel far below what you can hear and certainly below what anyone can reasonably record or reproduce.

So what happens is that it's nearly impossible to completely isolate, that's true and we agree, but it's also equally impossible to actually *capture* some of these effects. That is, in lower bandwidths, we couldn't get the train to affect the sound.

You're left with a near paradox. Do you spend the necessary money to isolate out a frequency that you can't even record? Or do you spend the necessary money to record a frequency you can't isolate yourself from?

Obviously, this is in the lowest octave of audible register, and beneath.

Ponder5
 
Unfortunately you need to build a floating room within a room, the only train noise solution. A lot of the train noise is transmitted through the ground, you will eventually get rid of the obvious noise by building insulated internal walls and blocking windows etc, but then you will hear the real problem noise of the vibrations caused by the train. I was involved with a studio build years ago where they had a commuter line 500 mtrs (1500ft) away and they could feel the vibration in the floor. A fully floating studio was built and completely silent.

Do it right from the start and you can beat the problem.

Alan.
 
I fear I have to agree with Witzendoz. Back in the 1990s I had to buid a TV studio that was not too far from a train line and the only solution was the traditional "box within a box" construction with a floating floor. It worked perfectly but I still shudder when I think of the cost. The other consideration is that a second wall inside the other, with them separated by an air gap, eats quite a lot into the size of any room and the floating floor (and inner ceiling) knock about 15 inches off your headroom.
 
Ponder5 said:
Nearly impossible, true.

But on the good side, what we learned, was that while yes, much of the noise could be classed as seismic, it also planted it firmly beneath any audible frequency range (20Hz).

In fact, we also learned with pretty good analysis gear that *feeling* the place move was not the same as *hearing* it move nor is it the same as any actual artifact in the recording. At first, that sounds counter-intuitive, but it ends up that you can feel far below what you can hear and certainly below what anyone can reasonably record or reproduce.

So what happens is that it's nearly impossible to completely isolate, that's true and we agree, but it's also equally impossible to actually *capture* some of these effects. That is, in lower bandwidths, we couldn't get the train to affect the sound.

You're left with a near paradox. Do you spend the necessary money to isolate out a frequency that you can't even record? Or do you spend the necessary money to record a frequency you can't isolate yourself from?

Obviously, this is in the lowest octave of audible register, and beneath.

Ponder5
Click to expand...

But I ponder (sorry for the obvious) how a sensitive condenser mics diaphragm will still be sensitive to the vibrations in a room. You can't cut the vibrations that effect what a mic pics up with EQ. It 'will' pick them up. How it affects the recording is something I have no experience with. I am just questioning myself here. I know that my 5 year old daughter tramps on the floor above, and even if I have a HPF @ 100 Hz, I still hear that shit. It still comes through...

It totally sucks to be next to a commuter train. The obvious solution is not the easiest. "move". :(

Hopefully you can find a way to make it work for you man.
 
Ponder5 said:
Nearly impossible, true.

But on the good side, what we learned, was that while yes, much of the noise could be classed as seismic, it also planted it firmly beneath any audible frequency range (20Hz).

In fact, we also learned with pretty good analysis gear that *feeling* the place move was not the same as *hearing* it move nor is it the same as any actual artifact in the recording. At first, that sounds counter-intuitive, but it ends up that you can feel far below what you can hear and certainly below what anyone can reasonably record or reproduce.

So what happens is that it's nearly impossible to completely isolate, that's true and we agree, but it's also equally impossible to actually *capture* some of these effects. That is, in lower bandwidths, we couldn't get the train to affect the sound.

You're left with a near paradox. Do you spend the necessary money to isolate out a frequency that you can't even record? Or do you spend the necessary money to record a frequency you can't isolate yourself from?

Obviously, this is in the lowest octave of audible register, and beneath.

Ponder5
Click to expand...

Not true, a condenser will record the frequencies around 20hz, you don't hear them but if you run the recording through a sub you will feel them. You should hear the drag racing video I filmed many years ago with just a camera mic, when played back through the surround system with the sub you could feel the earth move LOL, I was standing on the start line when filming. Even in my studio which is very heavily sound proofed but no floating floor, I can still hear a noisy car faintly if recording spoken word, more as a vibration, so I get around it by Hi Pass Filters.

However, if most of the recordings are done with gear that is just plugged in, and the vocals are the only mic type recordings use a suspension mount and place the mic stand feet on blocks of foam.

I get the impression there is a fair bit of train noise involved here?

Alan
 
To the OP Shinzon who came here once and not yet responded to our brilliant team of sound deadeners...

You didn't mention what kind of a budget you are on ..or if this is a full time gig or a hobby.

How you deal with it really depends on those two factors as much as anything else.

Some really good insight has been provided by the previous posters....

If you have the $ the room within a room, well planned out will provide you a place where you should be able to record voiceover tracks without "noticeable" audio interference from the trains.

Another suggestion was don't fight it, sell the shack and next time you buy a house do a little due diligence before you sign the papers. Stupid ass barking dogs and Crowing Roosters can ruin your day just as well as a train or plane or other common noise pollution problems home recording folks encounter...

One other suggestion is you may consider sub letting or renting a small room in a quiet part of town. There are a ton of businesses at least in my area that have extra space that they rent out pretty cheap....under a few hundred a month that include the electricity.
 
The bottom line is that for the microphone not to pick up the train, you need to be in a room where you can not hear the train.
It's going to take a hell of a lot of isolation to achieve that.
Thinking back to college, they had double skin insulated everything. Even the doors were doubled up with a 1' gap in between. Heavy fire doors, double glazed with much thicker glass than normal....
A lot of time an money went into that.

You could still hear everything between rooms. Sure, it was diminished greatly, but you'd be aware of someone playing an instrument for sure.
The carpark was outside and you could certainly hear mopeds coming and going.

Lots of good advice coming in with regard to that isolation approach but, personally, I'd rather move house, rent something ^^, or pause every 30 minutes.
Easy to say, sure, but I think sound proofing against a nearby train is going to be a mammoth task.
 
I would first see how much noise you can get rid of without doing any construction.

First, you need to be as close to the mike as possible. This will allow you to reduce the gain of the preamp and boost your signal-to-noise ratio. A close-working dynamic like a Shure SM-7B or an EV RE-20 would be a good choice.

Second, Make sure your mike stand is sitting on some kind of rubber padding so it isn't picking up train noise through the building structure. Your mike should also be hung using a good shock mount.

Third, use the high pass filter switch if your mike has one to get rid of as much of the low-frequency noise as possible.

Finally, process the audio to remove any low frequencies below about 100 Hz.

This might solve the problem. If it doesn't, does the house have a basement? Being in the basement should block most of the noise that is travelling through the air and all you'll have to deal with are the low-frequencies being conducted through the ground. I ran a commercial studio for 40+ years that was in a basement and 400 feet from a busy commuter and freight line. The train noise was rarely a problem as long as the narrators were close to the mike and I used a good shockmount.

Good Luck!

Mark
 
Shinzon said:
I would greatly appreciate any advice anyone can give me on what they recommend I insulate my walls, ceiling and floor with.
Click to expand...

There is really nothing you can do as far as your question above without doing a major studio build. Your best bet may be to use an ISO booth. You can buy them for around $3,500 or build your own. Like the AC specialist say in their commercials: "It's Hard To Stop A Trane". Can you upload an audio sample to show the issue?
 
Last edited:
The problem with giving advice is we don't know how bad it is.

We could say insulate the walls, put ion double doors and seals, block windows etc, and at the end of the day you spend all this and it still does not work. On the other hand you could do it and it works well enough?

Have a read of this link, you don't have to buy the products, but the advice is good regarding insulating sound transmission.

Alan.
 
witzendoz said:
The problem with giving advice is we don't know how bad it is.
Click to expand...

The other problem, as i've learned in this thread, is that the people that are not there know more than the people that are.

In the annuls of sound treatment, many have tried and many hucksters have hawked, but it still comes down to keeping sound you don't want away from sound you do - and those techniques are well documented and understood. As there is no such thing as total isolation, no matter how much construction is involved, then the only question is "how much for how much?"

StudioMicZone said:
I would first see how much noise you can get rid of without doing any construction.
Click to expand...

The answer is and has always been Mass. Decoupled mass. And that's really hard to do "without doing any construction". The momentary illusion of "quieter" can be had by controlling reflection. After all, once the train noise enters it does, yes, usually wander around a bit. But it's still there.

I'm with the rest: Best of luck.


Ponder 5
 
jimmys69 said:
But I ponder (sorry for the obvious) how a sensitive condenser mics diaphragm will still be sensitive to the vibrations in a room. You can't cut the vibrations that effect what a mic pics up with EQ. It 'will' pick them up. How it affects the recording is something I have no experience with.
Click to expand...

Consider this: good gear can record down to 1Hz. Other good gear can play it back. But how good are your ears? (or anyone you've ever met?) My trouble, probably big trouble, is that I only use instrumentation and I trust it implicitly. I don't use nor trust 'golden ears' or any voodoo-borne reasoning that is subject to a psychosomatic template. If the screen says we got a notch at 22.2Hz, then that's what we got. If the screen says we got a spike at 16.6Hz, then that's what we got - whether anyone hears it (or agrees with it) or not.

But now procedural, and not to bore, if we take multiple passes of the (more or less) same train and record tone bursts against intermittent silence in each pass, then we should have a pretty darned good baseline of what's there, easily visible on the screen. After all, it's true that if you can hear it, your mic can hear it, but the converse is not true. Not for anyone.

That's how the decision was made to go forward with the isolation he had (it was actually not bad) and simply be inconvenienced with the warning LED. Personally, I thought it was a wonderfully practical compromise. I only wish I coulda taken credit for the whole thing :-)

jimmys69 said:
I am just questioning myself here. I know that my 5 year old daughter tramps on the floor above, and even if I have a HPF @ 100 Hz, I still hear that shit. It still comes through...
Click to expand...
See that's the stuff this installation already had taken care of. Double sheets of 5/8" sheetrock on wire iso with rubber plug attachments... or something like that. It was that 'seismic' stuff that was worrying him. That's actually the only reason I keyed-in on this thread.

You know, now that i think about it, on my case he wasn't even complaining about bleed or interference because he wasn't actually doing any serious recording just yet. He just noticed he could detect the train from the studio and that was enough to send up a flare and bring us in.

We tried a bunch of test recordings and it basically came down to, "you got about 5 seconds after the LED comes on before audible freqs happen. Whether or not the program material masks it is another question. But it's detectable."

Ponder5
 
You must log in or register to reply here.
Back
Top