Wireless Audio with DSLR

[nho]

Buddy of mine is starting a show on NBC Sports and would like to figure out a way to run wireless audio into the few DSLRs they use for certain shots (everything else is shot ENG cam so wireless audio recording into camera isn't an issue). Is it at all possible to have wireless audio record to the camera on DSLRs? Could you not used a juicedlink setup on the camera with a sennheiser G3 wireless setup going into that?

 

[Alex H.]

Are you talking about primary audio from a single source per camera? Or are you talking about sending all cameras the same signal for reference?

The short answer is that, assuming your DSLR cameras have manual audio level controls (or are Canon EOS with Magic Lantern), you can feed a G3 receiver straight into the camera. So, if you're sending out one reference audio signal to all cameras, you can have one transmitter from the mixer and each camera can have a receiver tuned to the same frequency.

The long answer is that audio for DSLR is a tricky beast, and is usually best left as second-system sound. The simple truth is that the audio cirtcuits in a DSLR are pretty shoddy, and while there are tricks and workarounds to minimize the problems you still won't get the best audio recording. Plus, metering on a DSLR is hit-and-miss (again, still don't know which cameras you're talking about) and audio monitoring during recording is hard to come by (Canon + Magic Lantern + a special adapter makes this possible).

 

[Richard Crowley]

Yes, it is almost certainly possible. But any discussion of HOW it could be done or what quality you could expect would require more details about the setup and expectations.

 

[nho]

I'm thinking due to having timecode, the camera to be used would be the Canon 6D, or Mark III (though I know it doesn't have timecode). Only the main cam would have to pick up the audio and it could be sync'd with the other cams. So I suppose only one source of audio to one camera. Though, can the G3 support having two mics per transmitter? or is it a 1:1 ratio thing? 1 transmitter with 1 receiver is all you can do?

 

[Richard Crowley]

1 mic -> 1 transmitter -> any number of receivers. Most wireless mics require matching receivers.

Note that all modern camcorders record "stereo" or two independent audio channels.
You could conceivably record live audio to one track and timecode to the other track.
So these other cameras are for alternate angles of a sync shot (vs. random B-roll, etc.)

Is there a mixer in here somewhere? We are making a lot of guesses here absent any details.

 

[nho]

A small mixer COULD be present, if needed. I really am just wanting someone to tell me what I would need or if it's possible to record good wireless audio straight into the DSLR. Recording into only 1 of the cameras is all that is needed.

 

[John Willett]

I'm thinking due to having timecode, the camera to be used would be the Canon 6D, or Mark III (though I know it doesn't have timecode). Only the main cam would have to pick up the audio and it could be sync'd with the other cams. So I suppose only one source of audio to one camera. Though, can the G3 support having two mics per transmitter? or is it a 1:1 ratio thing? 1 transmitter with 1 receiver is all you can do?

You *have* to have a separate receiver for every transmitter and they have to be on separate intermodulation-free frequencies.

Two transmitters on the smae frequency fight with each other.

But - you can have as many receivers as you like listening to each transmitter.

 

[Steve House]

A small mixer COULD be present, if needed. I really am just wanting someone to tell me what I would need or if it's possible to record good wireless audio straight into the DSLR. Recording into only 1 of the cameras is all that is needed.

It's certainly possible to use a wireless mic with a DSLR but wireless or otherwise, the limiting factor to getting 'good' audio is going to be the camera's audio circuits themselves. Camera manufacturers have treated audio pretty much as an afterthought and the quality of the camera's audio sections are pretty minimal. You need to go double system (separate recorder) for top quality results. As for what you need, if you do want to record direct to camera, the absolute minimum you need is a wireless kit consisting of the mic, transmitter, and receiver plus the appropriate cable to connect the receiver's output to the camera's mic input. That gets you recording - how good it will be reamins to be seen.

 

[Richard Crowley]

I'm thinking due to having timecode, the camera to be used would be the Canon 6D, or Mark III (though I know it doesn't have timecode). Only the main cam would have to pick up the audio and it could be sync'd with the other cams.

My perception from these forums is that almost nobody runs timecode when shooting on DSLR. Timecode implies a well-funded, professional production operation, while shooting on DSLR implies low-budget, almost guerrilla-type informal methods. Modern editing software, most notable including "Plural-Eyes" has made syncing video to audio quite reasonable without timecode.

So I suppose only one source of audio to one camera.

No not at all. You can have any number of audio sources to a single camera or audio track. That is what the mixer is for.

Though, can the G3 support having two mics per transmitter?

Assuming by "G3" that you mean a wireless system. You can feed an audio signal into a transmitter that includes as many microphones as you can get your hands on. Through a mixer, of course.

or is it a 1:1 ratio thing? 1 transmitter with 1 receiver is all you can do?

A wireless link is just a way of getting audio from one place to the other. You can use a wire, or you can use a transmitter and a receiver.

Now what you feed into the transmitter is some audio signal. It could be just a single microphone clipped to the talent. Or it could be the complete field-mix from a mixer operated by the production audio crew member.

You can NOT have more than ONE transmitter on any single channel. However you can have any number of receivers. The only difference between a wireless audio link and your favorite radio station is the radiated power.

It wouldn't be unusual to use a wireless setup to take the mix (from however many microphones) and send it to all the cameras, each with their own receiver. However, most people shooting with DSLRs just use the audio from the built-in mic to sync the video with the proper audio track. Then, of course, the audio from the DSLR is discarded after it is used for sync.

 

[nho]

Thanks for all the help guys. Definitely gave me some useful information

 

[TheDingo]

I really am just wanting someone to tell me what I would need or if it's possible to record good wireless audio straight into the DSLR

Yes, but it will depend on the camera.

The new Panasonic GH3 has a noise-floor of -66 dB RMS ( this is equivalent to a Zoom H4n recorder, and better than most other DSLR cameras ), it also offers timecode ( though currently there doesn't appear to be any way to sync it with other devices ), and 1080 HD 50 Mbits IBP recording with uncompressed LPCM audio recording. It works really well with the Sennheiser G3 wireless systems ( set the receiver to 0 dB output, set the GH3 camera's audio level to 1, if you are using a Sennheiser lav mic then set the transmitter to -18 dB ) that you can plug directly into the camera's mic input jack. The GH3 also offers a headphone jack that you can use to monitor your recording.

I don't have any samples of my GH3 audio online, but here's a sample I shot with the Panasonic GH2 using a Sennheiser wireless system plugged directly into the camera : Interview with Ronnie Burket

 

[OzGizmo]

Yes, but it will depend on the camera.

The new Panasonic GH3 has a noise-floor of -66 dB RMS ( this is equivalent to a Zoom H4n recorder, and better than most other DSLR cameras )

????? Are you kidding.
Interesting specs you quote for the H4n...... If that was for real no one would be using the zoom.

 

[TheDingo]

Interesting specs you quote for the H4n...... If that was for real no one would be using the zoom.

The Zoom H4N is not my favorite recorder for several reasons, one of them being it's noise-floor performance.

A noise-floor of -66 dB RMS is very good for a DSLR, but not so hot for a $250 audio recorder. Several of the competing brands of recorders have noise floors that are 10+ dB lower.

The Panasonic GH2 has a noise-floor that starts at -63 dB RMS ( the same as my HMC-150 camera ), and the GH3 improves on this by 3 dB.

Olympus LS-11 / LS-12 / LS-14 recorders : All have a noise-floor that starts at -75 dB RMS

Sony PCM-M10 recorder : noise-floor starts at -75 dB RMS

Tascam DR-40 recorder : noise-floor starts at -79 dB RMS

Tascam DR-100 MK2 recorder : noise-floor starts at -82 dB RMS

Zoom H4N recorder : noise-floor starts at -66 dB RMS

With most low-cost recorders the noise-floor will rise as you apply more audio gain, so some recorders may start with a -75 dB noise-floor, but this can rise to as high as -50 dB at maximum mic gain.

The key to getting good results with low-cost gear is to feed the camera or recorder a strong audio signal, so the internal pre-amps never reach their noisy setting. For the Panasonic GH3 the optimal audio signal seems to be -15 dB, which means you have to use an external pre-amp ( like a JuicedLink unit or a good mixer ) or a good quality UHF wireless system. ( i.e. Sennheiser G3 system )

 

[John Willett]

The Zoom H4N is not my favorite recorder for several reasons, one of them being it's noise-floor performance.

A noise-floor of -66 dB RMS is very good for a DSLR, but not so hot for a $250 audio recorder. Several of the competing brands of recorders have noise floors that are 10+ dB lower.

The Panasonic GH2 has a noise-floor that starts at -63 dB RMS ( the same as my HMC-150 camera ), and the GH3 improves on this by 3 dB.

Olympus LS-11 / LS-12 / LS-14 recorders : All have a noise-floor that starts at -75 dB RMS

Sony PCM-M10 recorder : noise-floor starts at -75 dB RMS

Tascam DR-40 recorder : noise-floor starts at -79 dB RMS

Tascam DR-100 MK2 recorder : noise-floor starts at -82 dB RMS

Zoom H4N recorder : noise-floor starts at -66 dB RMS

With most low-cost recorders the noise-floor will rise as you apply more audio gain, so some recorders may start with a -75 dB noise-floor, but this can rise to as high as -50 dB at maximum mic gain.

The key to getting good results with low-cost gear is to feed the camera or recorder a strong audio signal, so the internal pre-amps never reach their noisy setting. For the Panasonic GH3 the optimal audio signal seems to be -15 dB, which means you have to use an external pre-amp ( like a JuicedLink unit or a good mixer ) or a good quality UHF wireless system. ( i.e. Sennheiser G3 system )

These are all pretty horrible figures (with the possible exception of the Tascam).

My pro. recorder quotes a signal to noise ratio of <114dB and the mic. inputs on my other one quote -128dB.

My baby pocket machine (Nagra SD) quotes -80dB.

 

[TheDingo]

These are all pretty horrible figures (with the possible exception of the Tascam)

Yes, compared to professional audio recorders these low-cost recorders are pretty noisy.

Keep in mind that I am listing RMS noise values from my own audio tests not the A-weighted noise-values that most manufacturers use. ( I prefer to run my own noise-floor tests, as published specs don't always hold up )

I haven't had a chance to test a Sound Devices recorder, but I expect it will be -100 dB RMS or better when it comes to measuring the noise-floor.

If anyone wants to confirm my results, here's how I test the noise-floor of a recoding device...

A Simple Method to Test the Audio Noise Floor of a Device

You will need a professional quality mixer that has a 1 kHz test-tone built-in. I use the Sound Devices MixPre-D mixer for this, which has a dynamic range of 110 dB or better.

1- Connect your recording device to the output of the mixer via an XLR cable. ( I use a 10 foot Canare Star-Quad cable with Neutrik gold connectors ) For devices with a 3.5mm audio inputs you will need to use a good quality passive XLR to 3.5mm adapter. ( I use a BeachTek DXA-2s passive adapter )

2- Set your audio recording device to record at 48 kHz 24-bit ( most consumer cameras will record at 48 kHz 16-bit )

3- Turn on the test tone of your mixer and adjust the recording level of your device to -12 dB. ( you may have to guess with your camera, as many consumer cameras don't have -12 dB markings )

4- Start recording and after 10 seconds turn off the mixer test-tone, but keep recording for another 20 seconds.

5- Bring your recording into your audio editor ( I use Sony Sound Forge ), and check to see what the test-tone recording is showing in your editor's audio meters. ( Sony Sound Forge meters are calibrated in RMS, which is one reason why I use the RMS scale )

6- Tweak the volume of the audio recording until the test-tone recording is displaying exactly at -12 dB. ( it may be off by up to 2 dB depending on the recorder )

7- Now check to see where the audio meters are sitting when the test-tone is switched off, and this level is the noise-floor ( i.e. self-noise ) of the recording device you are using.

8- To test the audio device pre-amp performance, add a passive mixer to the audio chain between the mixer and the device, this will allow you to lower the audio signal strength of the test-tone without introducing additional noise. ( assuming the passive mixer has a good dynamic range ) To test a device's pre-amps, simply set the recording level of the device ( i.e. set your recorder to gain level x, where x will vary depending on the device ) then repeat steps 3 and 4 for all of the gain settings that you want to test, and use the passive mixer to adjust the audio signal strength to keep the test-tone at the -12 dB level on the device you are recording with.

Here is how the Sony PCM-M10 recorder performs when tested like this:

Sony PCM-M10

RMS MIC NOISE FLOOR
----------------------
MIC LEVEL 2 : -75 dB
MIC LEVEL 4 : -72 dB
MIC LEVEL 6 : -60 dB
MIC LEVEL 8 : -51 dB
MIC LEVEL 10 : -49 dB

RMS LINE NOISE FLOOR
----------------------
LINE LEVEL 2 : -76 dB
LINE LEVEL 4 : -75 dB
LINE LEVEL 6 : -76 dB
LINE LEVEL 8 : -76 dB
LINE LEVEL 10 : -75 dB

...As you can see the mic pre-amps on this recorder aren't very good, but for LINE level audio recording this recorder is fine.

I consider -60 dB to be the threshold for what I would consider a "noise-free" recording, as most people won't be able to hear the noise-floor at -60 dB under typical listening conditions.

But this means that any volume increases in post may push the noise-floor into the audible range. So if your device's noise-floor is -66 dB like the Panasonic GH3 and you apply +12 dB of volume in your audio editor, your noise-floor is now -54 dB and you will probably be able to hear a faint "hiss" in the background. ( noise-reduction filters can help to eliminate this "hiss", but it's better to not have to resort to NR filters )

 

[OzGizmo]

Interesting that your noise figures are 20+db different (worse) than other tests made by other people ....... what reason might you give?
http://www.kenrockwell.com/audio/zoom/h4n.htm#meas

And as an example how do the current trent of digital recorders compare to the Nagra analogue recorders used film production for many years?

 

[TheDingo]

Interesting that your noise figures are 20+db different (worse) than other tests made by other people ....... what reason might you give?

I doubt that we are performing the noise test using the same methods, so the values from the tests will be different. Just like A-weighted tests will show absurdly low noise-levels. ( i.e. -115 dB for an Olympus LS-10 recorder )

...I like to go by empirical evidence, where I can make my own tests, and confirm with my own ears what the audio performance of a device actually is. This is why I posted the basic instructions for anyone who wants to confirm the results I am getting. Anyone can perform this test. It only takes a few minutes to complete.

I was very surprised when I tested the noise-floor of the Sony PCM-M10 recorder, as this is a very popular recorder, especially in the indie-audio world, where it has a reputation for being a really good low cost recorder. But after testing, I would never use this recorder for anything other than LINE level audio inputs. The pre-amps are just too noisy otherwise. I kind of feel the same way with the Zoom H4n, which like the Sony PCM-M10 recorder, the H4n does well with strong audio signals, but becomes very noisy if you have to rely on it's built-in mic pre-amps. From my own tests, the Tascam DR-40 is a better recorder, and it's almost half the price of the H4n.

When testing the Zoom H4n, I borrowed two different recorders from friends. The first recorder was a bit older and showed a noise-floor of -63 dB RMS, and the second newer recorder had a noise-floor of -66 dB RMS. I figured that Zoom might have improved their manufacture of the H4n, so I kept the results from the newer recorder as my reference for the H4n model.

Another issue I have with the Zoom H4n is that it does not have a MIC / LINE level switch, so you have to add a 10 or 20 dB inline pad in order to record from a LINE level audio source. I have no idea why Zoom did this, as almost every other recorder on the market offers the ability to take a LINE level audio signal directly.

And as an example how do the current trenf of digital recorders compare to the Nagra analogue recorders used film production for many years?

The Nagra analog recorders were famous for having very good sound, but todays professional digital recorders have noise-floors that are below -100 dB, something I don't think any analog recorder can achieve.

 

[Noiz2]

There are actualy a few issues with talking about self noise levels. And really that is what we are talking about. First noise isn't a fixed thing. This shows up in mics all the tjme. Some "noisey" mics sound quieter than "quiet" mics because the quality of ghe noise is more noticable. And as The Dingo sais the specks quoted are pretty hard to relate to real world experience.

Nagra used a special biassing that made for very quiet sounding recordings. I'm not sure
e how you would compare that abd say Dolby Sr wich is very quiet sounding.

I do think his tests are good for evalhating, providing you only compare numberes generated by the same tests. His don'tnec. conflict with manufacturers but his are good as a comparison between devices since nobody uses the same system so comparing specks is near usless.

It would be Acool project to get people to do this test on their equipment and submit to a central database so there wss a real place to compare apples to apples.

 

[OzGizmo]

Another issue I have with the Zoom H4n is that it does not have a MIC / LINE level switch, so you have to add a 10 or 20 dB inline pad in order to record from a LINE level audio source. I have no idea why Zoom did this, as almost every other recorder on the market offers the ability to take a LINE level audio signal directly.

The H4n was NEVER designed as a location recorder it was designed as a musicians recording tool, the 'Line' input is designed to take a guitar directly (how many other recorders offer that function?) and not line level. Many of the functions of the H4n are music orientated...
The reason it sold so well is it was the first unit on the market that fitted a 'price point' and 'usability point' for a consumer market it was NEVER designed as a location recorder as such...... It also came out around the same time as the DSLR trend took place. People have just worked around many of the negatives of the H4n. Perhaps a bit like DSLR cameras... They were never designed as a video camera but some people do shoot video on them and work around the shortcomings of the format.

The H4n is actually getting quite old in a technology point of view and a lots of other recorders have been released onto the market more suited to location sound recording, for example how long ago was the 5D released and how many camera model changed has occurred since then.

 

[TheDingo]

The H4n was NEVER designed as a location recorder it was designed as a musicians recording tool

My first portable recorder was the Zoom H4, which I owned for 1 year and then upgraded to a Sony PCM-D50. The Zoom H4 had many different technical problems, and there wasn't much competition at the time, so you had to live with these problems.

Despite being designed for musicians, I think Zoom still compromised the sound quality of the H4n recorder in order to make it low-cost, where their competition was able to beat them on price and still produce a better recorder. Perhaps Zoom will get it right the next time, but there is a lot of competition in the market, so they really have their work cut out for them.

The recorder I am most curious about is Sony's replacement for the PCM-D50, which should be outstanding in the $500+ price range. The PCM-D50 is very good ( the built-in mics in the PCM-D50 are the best I've heard for a recorder in this price range ), my only issue with it was Sony's poor support for standard memory card formats. I hope they do a better job with the next model.