Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Preventive Clipping Tips?

John Gleason

Member
Joined
Feb 24, 2010
Messages
6
Reaction score
0
Points
0
This last weekend I was sound mixer for a short student film. In one scene the 2 actors are lying in bed talking. I got the boom op to get the boom about 2 ft away from the talent's mouths. The test dialogue sounded brilliant barring the occasional car passing by.

As the camera rolled and as the talent said their lines, I kept the peaks to about 12dB, however, they would take long pauses in between sentences. Upon starting the next line they would sometimes speak the next first few words a lot louder causing the audio to clip at 0dB.

The scene involved them waking up early so they were whispering many of the lines. To make matters worse, the director kept them ad libbing in order to give the scene a more natural feel. They would whisper, so I would set the gain high to get a good level, then they would pause. Since the dialogue wasn't consistent take to take, I had no idea what their next volume level would be; I just sort of guessed as best I could. If they started the next line on a whisper and I had lowered the gain for more headroom, the dialogue was recorded too quiet. If I turned the gain turned up for an expected whisper, it would sometimes clip. Luckily for me we did the takes a few times, so all in I believe we got the lines recorded well, but I never felt 100% certain the recording wasn't going to clip at any given moment.

Thinking back now, I think I would to have liked adding another boom and then recording that to a separate track with split gains so I could get the dialogue's full dynamic range, but at the time I had only had 1 mic available.

What things can one do in a situation like this to prevent clipping? At what dB is it acceptable to capture whispers? Could a limiter have helped (is that what they are for)? Is it a matter of experience controlling the gain nob? Thoughts? Thanks.
 
Engage the limiter on both the input channel and the mixer outputs.
Print to 2 channels instead of 1, driving the second channel with less gain to keep a safety net.
 
send the actors to dialog coach , send your booman to work as grip if doesn't feel the dialog going up and not rise the boom .
prevent any noise outside and inside the room and lower the input gain , get better preamp with less self noise and keep the dialog so the picks will not go above input clipping level .
never work with limiter if you don't have 2 , run 24bit to get able rise the level in the post for 20 db without problem .
by the way, 12 db or 0d have nothing with -12 dbfs or 0dbfs and have nothing with if it clean or if you distort or not the signal in the first stage (the preamp)
 
Engage the limiter on both the input channel and the mixer outputs.
Print to 2 channels instead of 1, driving the second channel with less gain to keep a safety net.
to make it right you need split the mic before the preamp :-)
that's only work if you don distort in first place , don't really works in real life when you use more then 1 mic :-)
will not work with most of the existing field recorders , will work n field mixer which has pan pot .
for people who need 2 tracks to record 1 mic , Tascam came with drd-2 recorder which has dual preamp stage :-)
 
a limiter will CHOP off anything above the threshold which will throw away info.
You'd be better off using a compressor with a super soft knee/attack with a long release to keep as much info as possible "compressed" into your range. very light compression ratio tho just enough to keep it under without clipping.
Also the clean split before pre-amp with a lower safety-net gain on one is a great idea.

Jason
 
Thanks for the replies. So the suggestions so far are to:


  • Engage the limiter on both the input channel and the mixer outputs (.
  • Print to 2 channels instead of 1, driving the second channel with less gain to keep a safety net.
  • Lower the actual room noise, then I can set a lower the input gain to get more head room (makes sense, in this case though the room was actually very quiet (except for occasional cars) even without much treating).
  • Get better preamp with less self noise and keep the dialog so the picks will not go above input clipping level .
  • Never work with limiter if you don't have 2.
  • Run highest bit depth (in this case I was already at 24)
  • Split the mic before the preamp (although it will not work with most of the existing field recorders), will work n field mixer which has pan pot
It seems like a compressor would be my friend in this situation. Also a mixer which I could split the single boom input into two channels with separate gain control for each (is this a standard feature for pro field mixers?)

What then is the precise difference between a limiter and a compressor? Is it that the limiter just has a faster attack and a much higher ratio?
 
The scene involved them waking up early so they were whispering many of the lines. To make matters worse, the director kept them ad libbing in order to give the scene a more natural feel. They would whisper, so I would set the gain high to get a good level, then they would pause. Since the dialogue wasn't consistent take to take, I had no idea what their next volume level would be; I just sort of guessed as best I could.

That's a worst-case scenario. If they're not sticking to the script, then there's no way you can anticipate level changes and ride gain. All bets are off.

I think the idea of a mono mix split to two tracks, one normal and one set -10dB lower, will work. This has been a common technique used on documentaries, going back to the DAT days of the 1990s.

Warn the director and editor that they're going to need a good dialog editor to chase the levels in post. And I hope the backgrounds are clean enough that they can bring up the levels for the whispering. If the audience can't hear the dialog clearly, then the characters and story are lost.

Limiters (in theory) affect only peak level; compressors bring up quiet sounds and prevent loud sounds from getting too loud. The problem is, you can hear the effect of both, and if you record with either in a master track, you can never undo it later on. Some mixers use a little bit of peak limiting, but I haven't seen many people use compression.
 
Thanks for the replies. So the suggestions so far are to:


  • Engage the limiter on both the input channel and the mixer outputs (.
  • Print to 2 channels instead of 1, driving the second channel with less gain to keep a safety net.
  • Lower the actual room noise, then I can set a lower the input gain to get more head room (makes sense, in this case though the room was actually very quiet (except for occasional cars) even without much treating).
  • Get better preamp with less self noise and keep the dialog so the picks will not go above input clipping level .
  • Never work with limiter if you don't have 2.
  • Run highest bit depth (in this case I was already at 24)
  • Split the mic before the preamp (although it will not work with most of the existing field recorders), will work n field mixer which has pan pot
It seems like a compressor would be my friend in this situation. Also a mixer which I could split the single boom input into two channels with separate gain control for each (is this a standard feature for pro field mixers?)

What then is the precise difference between a limiter and a compressor? Is it that the limiter just has a faster attack and a much higher ratio?

It would really help if you had listed the equipment you used. You probably have a really awesome and sweet setup. Maybe your boomer could just scoop the sound from below and get the chest voice with the whispers as opposed to overhead booming. Also I assume you were picking up loud ambient noise from outside so you had to mix and compensate, a very common problem we all face.
 
It would really help if you had listed the equipment you used. You probably have a really awesome and sweet setup.

Ha, well part of my problem was that my setup was minimal, so I didn't feel I had that many options. I was recording on a Marantz 660 with the boom feeding directly in. I forget the exact name, but the mic was a ~$1k super cardioid condenser Sennheiser (internal battery). Now here's the key, I didn't/don't have a mixer.

My reason for this post was to deduce whether I could get great un-clipped sound with that same setup (are the tools not enough, or is it that I am not using them to their greatest potential?). If the problem isn't me, what other tools do I need? I think I have these two questions pretty much solved, but I think I need to get a hold of a decent field mixer and do some tests first.

The other thing was I was hoping this thread could turn into an advice center for mixers who want to keep their signal hot without jeopardizing their sound.
 
The 660 has no limiter, so that's your greatest problem there. It also has really crappy pre-amps (unless it's been modded,) so you'd do quite well to add a mixer with good pre-amps and a limiter between the mic and the 660.

If you only need one channel, a Sound Devices MM-1 would do the job. If you need two, a Sound Devices MixPre. If you wanted three channels, plus getting a much better meter to make mixing easier, a Sound Devices 302.

I ran a number of tests years ago, between running a mic directly into various cameras, a Zoom recorder, or a Sennheiser wireless transmitter, vs. first running the mic through an MM-1 and using the limiter, low-cut filter, and better-quality pre-amp.
In every case, the audio had noticeable less noise and hiss by virtue of the pre-amp, was free from low-end wind and movement rumble by virtue of the high-pass (aka low-cut) filter, and was virtually impossible to clip even when the signal was run really hot and hard thanks to the limiter, even for very short bursts of loudness.
It was such a fantastic difference that I started using the MM-1 even for crappy real-time audio streaming over the internet of corporate meetings.
 
I am not familiar with Marantz products.

But like Stephen said, if you plan on doing the audio post yourself it is highly advisable to get a mixer with a good preamp. You will off course loose some frequencies if you go the lowpass limiter route but you will get great sound. If you work with an experienced post sound engineer, you would most likely be told to record 3 channels of full bodied sound just slightly pre-amped but with lots of headroom. Recommended mixers for the money are sound devices or Wendt line give you bang for buck. in a situation where there is whispering try renting DPA 4060' s 4061's. They are simple full bodied omni lavs that deliver.
 
Back
Top