Question about using the built in limiter.

ironpony

Carbonite Member
A lot of field recorders have them but most experts still say to set the levels in between 12-6. But with a limiter couldn't you turn up it up to 3 or max even and it won't matter cause it's limited? This way you get a real strong signal and it won't matter cause of the limiter, unless of course I am probably wrong.
 
A lot of field recorders have them but most experts still say to set the levels in between 12-6. But with a limiter couldn't you turn up it up to 3 or max even and it won't matter cause it's limited? This way you get a real strong signal and it won't matter cause of the limiter, unless of course I am probably wrong.
Setting the gain too high and relying on a limiter to keep it under control has several drawbacks. One is that limiters are not instantaneous and a few cycles can slip through to produce audible clipping. Another is that doing so squashes the dynamic range of the material into t he top of the scale. Perpetually loud is not a good thing, contrary to the belief of kids who insist on cranking the volume of their iPods all the way up. What you want in levels is a balance - high enough that you are as far above the noise floor as you can be yet not so high that there is a risk of hitting 0dBFS or that you reduce the dynamic range of the material..
 
Excessive limiting will also add distortion as well as increase the amount of BG noise, acoustically borne or otherwise.
"If some is good more is better" is a common mistake.
 
Okay thanks. I have learned some stuff about audio so far, since I have been stuck without a sound person to show up on set sometimes, and have worked the recorder a few times. Sometimes I am told the limiter cuts down on quality but others say no, so which is it? Also I might have already learned this on my own, but what is dynamic range exactly?
 
Sometimes I am told the limiter cuts down on quality but others say no, so which is it?

If used properly, it's merely a safeguard against sudden peaks and will not affect signal quality. All things in moderation, right? If you push it too hard, you'll start to hear problems.

Also I might have already learned this on my own, but what is dynamic range exactly?

It's the difference between the softest sound and the loudest.

Think of sound as a room: it has a ceiling and a floor. Anything above the ceiling will overmodulate (distort). Anything below the floor will not be heard, or is hidden in the noise that sits on the floor (self-noise from the equipment). The actual dynamic range that sits inside the room isn't a set standard. There's the DR that the equipment can realistically generate, and the DR that results from the placement of the mic and the gain staging of the mixer and recorder, both of which have to do with sound during production. There's also DR that is set by the client's standards, be it for theatrical release or for television, and that is created during post-production (though it relies on production sound being clean).

There's a sweet spot between the floor and the ceiling where you want your signal to average. If you record too low, and then try to bring the signal up in post, you also raise the noise ("noise floor") that was recorded. If you record too hot and you push the ceiling, you end up with distortion that cannot be fixed in post. Realistically, you want your dynamic range to go from a little bit above the floor to a bit below the ceiling (headroom).

Limiters keep the signal from going above the ceiling by creating a new ceiling that's below the actual ceiling, but if you drive them too hard they can cause distortion of their own and can have other affects on the signal quality (including frequency response).

Set your dialog record levels to average around -12dB, with peaks (from sudden yelps, claps, coughs, etc.) at -6dB. You can use the limiters to catch anything that slips past your control and starts to push 0dBFS, and you won't have any side-effects from the limiter.
 
Last edited:
what is dynamic range exactly?

dynamic range in the case of a person speaking is basically the range from the softest that a person is speaking to the loudest....no one speaks without dynamics so when you push your signal up to the limiter you lose your dynamics and it sounds very unnatural.


Sometimes I am told the limiter cuts down on quality but others say no, so which is it?

maybe the explanation of dynamic range answered this somewhat but another thing you should know is not all frequencies are equally in power so when the limiter starts squashing your signal the weaker (higher) frequencies are going to suffer first which is where your clarity is... so hitting your limit every once in a while could save you from unwanted clips and you may not notice these things but constant or frequent limiting is going to be a lot more obvious loss in quality...

hopefully that make sense!
 
Set your dialog record levels to average around -12dB, with peaks (from sudden yelps, claps, coughs, etc.) at -6dB. You can use the limiters to catch anything that slips past your control and starts to push 0dBFS, and you won't have any side-effects from the limiter.

Since I work with actors, who are not professional, although they may want to be at some future time... it seems that most do not voice the dialog anywhere near this level unless the mic is about 12" away from their mouth, or the scene such that 'shouting is ok'... which means it is in the shot... move the mic out, and even ignoring any 'reverb' etc. their voice starts to descend to the 'ambient/noise floor'...

My question is... is part of the training of an actor even for 'movie' work to voice 'louder', despite the scene requiring a 'soft' level.

Perhaps a second question is rather than state -12 dB from presumably 0 dBFS... what is the noise floor relative to that dialog 'peak/average'?
 
dynamic range in the case of a person speaking is basically the range from the softest that a person is speaking to the loudest....no one speaks without dynamics so when you push your signal up to the limiter you lose your dynamics and it sounds very unnatural.




maybe the explanation of dynamic range answered this somewhat but another thing you should know is not all frequencies are equally in power so when the limiter starts squashing your signal the weaker (higher) frequencies are going to suffer first which is where your clarity is... so hitting your limit every once in a while could save you from unwanted clips and you may not notice these things but constant or frequent limiting is going to be a lot more obvious loss in quality...

hopefully that make sense!

It sort of makes sense. Costant or frequent limiting is bad, but the limiter is either turned on or off, so wouldn't it being turned on means that it's frequent and and constant, and turned off, means there is no limiting at all? Is there an in between? Or do you mean that hitting above 3 dB is bad, and as long as I stay at 12-6 dB there is no quality loss?
 
Costant or frequent limiting is bad, but the limiter is either turned on or off, so wouldn't it being turned on means that it's frequent and and constant, and turned off, means there is no limiting at all? Is there an in between? Or do you mean that hitting above 3 dB is bad, and as long as I stay at 12-6 dB there is no quality loss?

The limiter has a threshold, or, the point at which is starts to take effect. Below that threshold, it is inactive.
 
Okay thanks. Anyway of telling what the threshold is? I am doing tests but just want to be sure. I am using the FR2-LE and I noticed that the headphone jack has more hissing coming out of it, than there actually is when done. I was told that the jack does not have as good of monitoring quality if that's true. But it's more difficult to tell on set than later when it sounds more clear, even through the same headphones which I use for both.
 
Okay thanks. Anyway of telling what the threshold is? I am doing tests but just want to be sure. I am using the FR2-LE and I noticed that the headphone jack has more hissing coming out of it, than there actually is when done. I was told that the jack does not have as good of monitoring quality if that's true. But it's more difficult to tell on set than later when it sounds more clear, even through the same headphones which I use for both.
Better professional recorders and mixers have adjustable limiters so with them the threshold is whatever you choose to set it at. If it's not adjustable, RTFM or send an email to the manufacturers tech support and see what the manufacturer says it is in the specs. I've tried to find the spec for the FR2LE and can't located it so it's anybody's guess at this point.

Correction - I meant to say I found the specs for the FR3LE but they don't list the limiter threshold in them.
 
Last edited:
Since I work with actors, who are not professional, although they may want to be at some future time... it seems that most do not voice the dialog anywhere near this level unless the mic is about 12" away from their mouth, or the scene such that 'shouting is ok'... which means it is in the shot... move the mic out, and even ignoring any 'reverb' etc. their voice starts to descend to the 'ambient/noise floor'...

My question is... is part of the training of an actor even for 'movie' work to voice 'louder', despite the scene requiring a 'soft' level.

Perhaps a second question is rather than state -12 dB from presumably 0 dBFS... what is the noise floor relative to that dialog 'peak/average'?


I found this to be quite true last feature I shot, so this time a key factor in the auditions was voice projection. It's a fine line because you don't want the kind of projection done in the theater, just a good strong voice that comes through on mic. As a result of auditoning this way I'm having fewer problems with weak actor voices making higher recording levels necessary. Another factor was the mic I chose (for outside use), an AT 4073. It's one of the most sensitive mics available and often the level control (mic is plugged straight into the camera) is at the 3 or 4 level, where previous mics would have been around 6 or 7, raising the noise level. The only drawback is with a mic this sensitive,over peaking is a bit more likely. But it hasn't been much of a problem for me.
 
Hi

Limiters can be configured in higher end gear as mentioned above. Also on the more economically oriented gear they tend to be either configured for voice or music. I haven't any experience with your recorder and can't help with how it may be configured. Additionally limiters tend to also be sorted between "brick wall" and "soft" limiters. Brick wall kick in all at once at their setting to prevent an over. Soft limiters (at some point becoming a compressor) come in gradually bringing the rising level to a more gradual stop. I have heard that most economically oriented recorders are purchased for recording music and have limiters oriented for that. The other big configuration are attack and release times. I found this to be very well written: http://en.wikipedia.org/wiki/Dynamic_range_compression. Read selectively as there is definately TMI (too much information as my kids say). A limiter is an important safeguard but as mentioned in other posts it is best used that way for all the reasons mentioned. Overages in the old analogue world failed gracefully and rock n roll made distortion an art. Digital overages are nasty sounding creatures that are best strongly avoided. Any limiter used as a rare safeguard to catch a mistake is usually better than letting a digital over occur. A digital over can't really be fixed after the fact. I reccomend running little fifteen second tests deliberatly making overages and testing a limiter and learning the sound and what you can and can not do about it. Also try recording too low and listen for the noise that is revealed on the too soft end.

Best Regards
 
Hi
Meters used in most equipment aren't fast enough to catch transient overs and aren't slow enough to give a RMS (root mean square) reading. I am amazed that on my Digicheck program which features both configurable peak meters and RMS meters (both run in hardware on RME Fireface 800) that with well recorded speech there can be a 20 dB difference between the highest peaks and the RMS. Fifteen dB spread is common. I have taken to running my inputs softer than I used to. The performer makes a big difference as mentioned previously. But I find the difference between true peaks and a RMS level is still more than I had previously expected. My lesson is play it safe with the overages. very quick transients can make it past many limiters if the attack time is not quick enough combined with the what specific level it kicks in. So even a limiter is not in all cases a a total solution.

Best Regards
 
I am using the FR2-LE and I noticed that the headphone jack has more hissing coming out of it, than there actually is when done. I was told that the jack does not have as good of monitoring quality if that's true. But it's more difficult to tell on set than later when it sounds more clear, even through the same headphones which I use for both.

The FR2-LE has a separate headphone volume knob that may be turned up higher than your actual recording volume.
 
Yep I know, but there is still more hissing when the volume is increased, but that hissing is not actually there. It's just coming from the jack.
 
Yes, the FR-2LE has been known to have a noisy headphone amp. None of that noise is recorded; it's just in the headphones.

However, I have an FR-2LE and the headphone amp is pretty darned clean.
 
Back
Top