View Full Version : In vehicle audio recording


Larrie Easterly
November 9th, 2021, 11:14 PM
I need some suggestions on ways to get better voice audio while riding in a vehicle at highway speeds.

Have been using either a Rode lavalier or a short shotgun microphone to record the voice. As expected, they also pick up the road noise, Have found that both are very placement dependent. It is not possible to dub in the voice track in post.

I currently use several audio tools I post to cleanup the track and would like to spend less time in the cleanup process. So am wondering how others have solved the problem.

Paul R Johnson
November 10th, 2021, 03:32 AM
How pretty does it need to be? An ear loop boom mic would be by far the most consistent positioning and sound, but off course you can see it - the smaller ones like DPAs can be almost unnoticeable .

Rick Reineke
November 10th, 2021, 10:02 AM
Aside from closing the windows and shutting off the interior A/C or heat fan, a high-pass filter can help to attenuate the low frequency road noise.
I usually mount a lavaliere or boundary mic on the person's sun-visor I have also used, hyper-cardioid mics mounted on the center console pointed up towards the talent.

Patrick Tracy
November 10th, 2021, 10:47 AM
Relative proximity is your friend. The closer the mic to the source, the lower the background noise. A headset mic is going to be the optimum, assuming you don't want them talking into a handheld vocal mic with their lips on the grille.

A quieter car would help.

John Nantz
November 10th, 2021, 01:40 PM
This falls under the ‘suggestion’ heading; have not tried this idea.
While downloading the latest iOS update 15.1, there was a feature listed that may have an application for your situation. Since smart phones have been incorporating some capabilities sooner than our kit, who knows?, it may work.

The iOS 15.1 update listed background noise cancelation as a feature, but in trying to find out about it, it appears to have started in iOS 14. The phone has mic input capability so the combination of a better mic (lav?), combined with background noise cancelation, may be something to explore.

One problem may be that it only works with Face Time, in which case, a call to tech support about another way to do it, such as with Voice Memos, may be a work-around. The legal part may be that whoever has the patent for the method that Apple is using may not have addressed a Face Time use so that would be the only method available (for the time being).

Below is what the User Guide has to say.

iPhone User Guide iOS 15
[note: iOS 15.1 may be different (improved?). feature actually started with iOS 14]

Filter out background sounds
When you want your voice to be heard clearly in a FaceTime call and other sounds filtered out, you can turn on Voice Isolation mode (available on supported models). Voice Isolation mode prioritizes your voice in a FaceTime call and blocks out the ambient noise.
During a FaceTime call, open Control Center, tap Mic Mode, then select Voice Isolation.

Change FaceTime audio settings on iPhone:
Spatial audio in the FaceTime app makes it sound like your friends are in the room with you. Their voices are spread out and sound like they’re coming from the direction in which each person is positioned on the screen.
Note: Spatial audio is available on supported models and works with AirPods (3rd generation), AirPods*Pro, and AirPods*Max (sold separately).

https://support.apple.com/en-ph/guide/iphone/iphb54d5dee2/15.0/ios/15.0#iphdf4d7cde3

Bruce Watson
November 10th, 2021, 03:17 PM
I need some suggestions on ways to get better voice audio while riding in a vehicle at highway speeds.

Have been using either a Rode lavalier or a short shotgun microphone to record the voice. As expected, they also pick up the road noise, Have found that both are very placement dependent.

Interesting -- all mics are sensitive to placement. It's what mics do, it's how the laws of physics work. But it's not really complicated. Basically, the closer the mic is to the signal, the higher the signal to noise ratio can be. This is why Broadway shows are using ear-set mics (https://www.bhphotovideo.com/c/product/697575-REG/Countryman_E6IDW6C2LE_E6i_Cardioid_Ear_Set.html) where the capsule is placed just behind the corner of the mouth. Note that ear-set mics come is a large variety of skin tones, so they are difficult to see except in closeups with the talent's head facing the right way. With ear-sets, singers can belt away, and the signal to noise ratio is so high the mics barely pickup the singer 1m away, who is also belting it out.

If you can do this, it's about as good as you can get for this kind of work. Much depends on how sensitive you are to having the mic seen in the shot. Even for this there are ways around it. You can use the far ear as long as the talent never turns too far into the shot. If that won't work for you, you can mount the mic on the bill of a ball cap. Or just back from the hair line. Or on the frame of a pair of glasses (or sun glasses). Etc.

The key here is to get the mic as close to the mouth as possible, without being directly in front (need to avoid plosives). Then, embrace what road noise you can hear as a feature -- everyone expects to hear some road noise in a car. If you take it all out, you'd find yourself putting it back in later.

If this is still too much noise for you, the next step us is to go "full Hollywood" and put the car on a process trailer (https://www.midwestcameracars.com/proces-trailer/) and tow it around the route. This eliminates the motor sounds and tire noise, and a lot of the bumps. Probably more time and money than you want to put into it, but you aren't going to get much less noise than a process trailer.

Paul R Johnson
November 11th, 2021, 02:41 AM
We have loads of cop documentaries here and sun visor to a zoom, plus the on camera mics seem to work pretty good even when they're chasing people at silly speeds! Have you a short clip of your problem?

Larrie Easterly
November 11th, 2021, 10:54 PM
Thanks for all the information. Having the mic’s visible is not a problem although I do hide them the best I can.

By coincidence our local news channel did a piece on all the rain we have been having here in the Pacific Northwest. In one segment they were narrating the story in a moving vehicle on the freeway. I was surprised at how much it sounded like what I am getting. Maybe things are not as bad as I thought.

Will experiment with different mic’s and their placement over the next few weeks. Am also planning on driving on different road surfaces. I like the idea of a mic with a high-pass filter. Had not thought of that. The local photo rental shop has several to choose from. Need to get this figured out before heading to Baja after the first of the year. Will be shooting another multi part backcountry travel video to post on my YouTube channel, https://www.youtube.com/channel/UCONQUj93Tn3IXQcTM2LWovA.

Jim Michael
November 14th, 2021, 02:23 PM
You might investigate how aviation headsets are designed since the same issues apply, the need for clear communication in a high noise environment. E.g. here are a selection of dynamic and electret mics. https://www.pilot-usa.com/headset-accessories/microphones.html Impedance is worth noting.

Allan Black
November 14th, 2021, 08:56 PM
Different road surfaces will generate various levels of road noise in the car, the louder it is the louder the talent should speak. With some experience you should easily be able to correctly judge the right balance on your headphones.

How loud, clear and decisive the talent delivers has a lot to do with satisfactorily recording voices in noisy environments. Too fast, indistinct and mumbling will cause problems in post and new talent is the worst. Even though they’re alerted to start, it’s easy for nervous newbies to become generally inaudible and the director should keep reminding them based on headphone results from the audio guy.

If you do a search, there’s a few good UTube suggestions including hiding lavalier mics under clothes. Don’t forget to record some car interior ‘road noise’ as possible back up in post.
Cheers.

Greg Miller
November 17th, 2021, 02:14 PM
Lots of good suggestions above.

The low-cut filter doesn't need to be integral to the mic. It just needs to be somewhere in the audio chain. Probably the earlier the better, provided you don't filter out too much of the voice and then want to restore it later.

Several years ago I looked at a few headsets designed for loud environments (aircraft, sports arenas, etc.) The mics were intended for very close talking. IIRC the mic elements were open front and back. Since LF is largely non-directional, the theory was that any LF noise in the vicinity would reach both sides of the diaphragm at nearly the same time, so the pressures would be equal and the output would be minimal. Of course there would be less rejection at higher frequencies. I suspect that these would suffer badly from sibilants and plosives, given that the person was speaking directly into one side of the diaphragm, from probably less than an inch away. Of course you'd lose a lot of LF from the voice, too. Bottom line: you'd sacrifice "pleasantness" in the voice in order to achieve maximum noise reduction.

Cellphones, of course, are good at noise rejection, given their multiple mics and sophisticated signal processing. But the result is often very robotic sounding speech.

You're up against a trade-off. The importance of good natural speech reproduction, vs amount of noise reduction.

I'm inclined to like the suggestion of mini head-worn boom mic, positioned beside the cheek, just slightly behind the corner of the mouth (to avoid sibilants and plosives). That seems like the best alternative.

Larrie Easterly
November 18th, 2021, 10:49 PM
Lots of good suggestions above.

The low-cut filter doesn't need to be integral to the mic. It just needs to be somewhere in the audio chain. Probably the earlier the better, provided you don't filter out too much of the voice and then want to restore it later.


If I understand this correctly, you are saying get the best audio I can based on mic position and then filter out the noise as the first step in post rather than trying to filter the noise with filters, low cut/high-pass on the mic. I like this solution.

I also like the idea of a mini head-worn boom mic. Will look into that. Thanks.

Rick Reineke
November 19th, 2021, 09:46 AM
Most location preamps/recorders have a HPF (high-pass filter) option(s). Additional LF filtering (EQ) can be applied in post. There are in-line HPF filters available (like the Shure A15HP) if the mic and preamp does not have the option, I would recommend an omni headset/earset mic. Despite being directional, cardioid mics inherently pick up more extraneous LF noise and are subject to air turbulence (wind) than a omni.

Bruce Watson
November 19th, 2021, 10:01 AM
If I understand this correctly, you are saying get the best audio I can based on mic position and then filter out the noise as the first step in post rather than trying to filter the noise with filters, low cut/high-pass on the mic.

I don't think that's exactly what he's saying. Many mics have low cut filters on them because using them on the mic has some advantages. The main one is to lessen the chance of overloading some component down stream from your mic. Say, a microphone pre-amp. Something that, if it does overload, is very difficult to fix in post.

Just know that a lot of low frequency (below human voice frequencies) energy is large. Cutting that out of the signal path early is usually better.

Also probably a good idea to wrap your head around limiters. They can save you from the same basic problems, but across the entire frequency band. I had a woman drop a glass mixing bowl onto a stone counter top during a cooking demonstration. The limiter clamped down on that way before I even understood that it had slipped from her hands. The sound was actually quite usable (no clipping at all), but there was no way we'd use that scene in the final edit -- real cooks don't shatter mixing bowls (on camera) -- something about not wanting the crunch of glass in the brownies. ;-)

Given that you're going to be recording in a moving car, you might really want to use a limiter. Works great when "talent" yells or screams or otherwise goes up in level higher than you'd planned. A good limiter, set properly (yes, there's always a catch), can definitely save a recordist's tail on occasion.

Greg Miller
November 19th, 2021, 10:51 PM
I may have been unclear when I said "early in the audio chain" because in fact the mic is the earliest part of the chain.

To elaborate, there is a tradeoff (as pointed out by others). If you don't filter, or don't filter enough, at the mic, then there is a risk of overloading the system with LF signal. (You can later filter out the LF, but can't remove the distortion products.) But of course you have a recording gain control and meters, so hopefully you can prevent overloading even if there is a lot of LF.

The other side of the coin is that if you filter too much before the recorder (either the mic's integral filter, or a plug-in filter between mic and recorder) then you run the risk of removing some desired LF from the voice. In that case the recording might be less noisy, but the voice might be less natural. A gradual LF rolloff might be better than a sharp low-cut filter, because the former gives you a better chance of restoring lows compared to the latter.

Ultimately you need to monitor closely (using headphones with GOOD isolation, since the car interior is so noisy) and watch levels, to be sure you're not overloading or clipping. You need to use "the right amount" of LF rolloff to reach the best compromise that avoids the above problems.

One question that's important IMHO is this: which is more important, intelligibility or naturalness? (e.g. an old telephone, with a carbon mic element and limited bandwidth earpiece, is designed to be intelligible ... sure it surely doesn't sound natural.) Of course that's not a binary yes / no answer ... it's which point you pick along a continuum.

Remember, too, that there are a lot of near reflections and resulting response peaks, caused by nearby surfaces like windshield. This will greatly affect naturalness, but will probably have less effect on intelligibility. (Again, the closer the mic is to the mouth, the less this will be a problem.)

Larrie Easterly
November 19th, 2021, 11:18 PM
Thanks for the clarification. I realize I will not be getting pristine sound. With the information and ideas you all have suggested I should be able to get what I need. Will be running tests with different microphones and their placement the beginning of December.

Greg Miller
November 21st, 2021, 10:41 PM
Also be sure to record some road noise immediately before and after each take. That will give you noise spectrum samples, in the event that you want to use some spectral noise reduction software.

Bob Hart
November 24th, 2021, 12:18 AM
You might try a Rode NT2a studio mike switched to figure 8 inside of a good shockmount or piece of foam with a hole in it placed on the console between the seats just below image frame edge. Also stuff as much old clothes in rubbish bags and blankets on the floor and rear seats as you can to suck up in-cab reverb and road noise. The sound was not pretty but of sort of worked.

Greg Miller
November 26th, 2021, 12:16 PM
Larrie,

I quickly skimmed through a couple of your videos. I did not hear any objectionable road noise, with the exception of one short stretch that was obviously highway driving. I didn't try to extract the audio from the video, and I don't recall exactly where that was. But the off-road footage IMHO did *not* have objectionable noise. In fact if it had sounded like a quiet studio it would have seemed hokey.

PS: Gawd, that's a rocky road. I would have hated to make that trip in my '71 Land Rover.

Larrie Easterly
November 27th, 2021, 11:07 PM
Thanks Greg.

Yes, I do get to pretty out of the way places.

Greg Miller
November 28th, 2021, 10:28 PM
If you want to give me a link to an existing file (and approximate time within the file) that concerns you, I'd be interested in hearing what sort of issues you've having.

I recorded a really informal test earlier today, but haven't had time to analyze the tracks yet. I'll post a brief report on what I find out.

Greg Miller
December 1st, 2021, 11:07 AM
As promised, here's a brief summary of some non-rigorous testing. The aim of the test was not to evaluate a specific mic, it's to compare pickup of [essentially] the same mic in two different locations.

I used two off-brand lav mics (the kind marketed to amateur videographers), both small diaphragm cardioids, both the same model, but not claiming to be a matched pair. A side-by-side test shows that their performance is similar.

I mounted one mic on the sun visor (in its usual "up" position) sandwiched inside a larger SM-58 size foam windscreen, to isolate the mic from the car's vibration. I clipped the other mic to a headset boom arm (with no other mic attached).

Results:

With gain on both mics the same, the visor mic picked up much more low frequency rumble from engine & exhaust, even with the car stationary. This makes sense when I remember that LF from a loudspeaker is boosted when the cabinet is located on one room boundary (wall or floor), boosted more at the intersection of two boundaries, boosted even more at the intersection of three boundaries. The visor mic mounted roughly an inch from the roof couples more LF from the car, compared to the boom mic which was roughly 10" from the roof.

With gain on both mics the same, and the vehicle in motion, the visor mic picked up at least 10-12 dB more overall noise, sometimes much more than that.

With the engine off and the car stationary, the visor mic (~ 10" from my mouth) picked up my voice ~14dB lower level than the boom mic (~ 1.5" from my mouth). Of course a difference was expected, and this test gives us an approximate numerical amount.

Conclusion:

If I raise gain on the visor mic channel, so the voice level is similar to the boom mic channel, then noise level on the visor mic is ~24 dB or more louder than on the boom mic.

Of course I expected the boom mic to be a better choice, but I didn't expect the difference to be this drastic, so this was a worthwhile experiment. It also opens up some other questions for possible future tests. I hope this is helpful to folks like Larrie who might be planning to record in a moving vehicle.

David Peterson
September 5th, 2022, 03:33 AM
I've shared lots of videos demonstrating car rigs on my YouTube channel:

https://www.youtube.com/c/SoundSpeeding/search?query=car

Andrew Smith
September 5th, 2022, 07:37 AM
I've shared lots of videos demonstrating car rigs on my YouTube channel

David, mate, it's so good to see you here. Love the videos on your channel when I come across them.

Andrew

PS. Is it true you were a stand-in on LOTR?

David Peterson
February 21st, 2023, 03:06 AM
PS. Is it true you were a stand-in on LOTR?

ha! No.

But I did get rung up by a line producer on Amazon's LotR to be a Production Sound Mixer for them on one of their units. Sadly I could not :-( As I was already busy working on Cowboy Bebop. That's ok though, I could always end up working on them next season instead?

oops

LotR left NZ :-( :-/ No more filming here!