|
Post by starandchlorisse on Nov 20, 2016 4:51:20 GMT
Hi I m trying to record a child reading ; I would like to try prevent recording levels being not constant. Should I use a limiter or a compressor and which settings are better for this situation? I m using Logic pro. Many thanks
|
|
|
Post by Hexspa on Nov 20, 2016 23:57:56 GMT
A compressor with moderate settings should be fine. Just make sure the peaks are no more than -10dBFS. You can always compress or automate later.
-m
|
|
|
Post by Ethan Winer on Nov 21, 2016 21:27:09 GMT
I also prefer to record flat and compress later. Or not compress at all and instead use volume automation. If it's a child whose levels might vary too much, Hex's advice to keep stuff below -10 makes sense. Even if you record at -20 the room's own noise will probably be much greater than the "hiss" from 16 bits.
|
|
|
Post by rock on Nov 22, 2016 3:37:50 GMT
I agree, record flat well below FS. -20 leaves plenty of headroom and as Ethan points out, is still very quiet. I typically set levels at - 18 as that's where my DAW meter has a calibration tick mark. When recording a live band, the levels almost always get hotter during tracking after the initial level settings.
Cheers, Rock
|
|
|
Post by Hexspa on Nov 22, 2016 6:27:44 GMT
You could also have one hand on the preamp gain.
I definitely have never recorded into compression though some do.
You could also move the mic back and forth but there could be handling noise, proximity effect etc.
Try to use a quiet room, as Ethan hinted at, so you'll have the greatest flexibility later.
Technically, I think peaking below -10dBFS is too low for a variety of reasons. But if you need that much headroom (doubt it) you can easily add gain later.
I compensate a few dB below -10dBFS to account for what rock mentioned though.
For the record, I don't advise using -10dBFS for distance from the noise floor but because it provides a higher RMS for interfacing with analog and analog modelling plugins.
I also use K system calibration so -14dBFS = 76dB in my room but that's another story.
Like the "clipping detection" thread, using a software limiter isn't going to do much if the peaks get crazy. You'll probably clip at the preamp if levels are too loud.
Unless the child has autism you should be able to just ask it to read again if the take is bad. You could also have them read the passage multiple times and comp the best take.
I really think automation is your answer, perhaps some gentle compression after the fact.
Depending on the purpose of the recording you can use a limiter last to bring up the overall gain.
-m
|
|
|
Post by rock on Nov 22, 2016 12:48:51 GMT
Another aspect is mic placement and pattern. Don't mic too close or use a hypercardioid with too tight of a pattern. The closer the mic is, the more dramatic the level change, as you back the mic off, the less the relative level although overall lower which you can compensate with gain settings. If the child is moving his head side to side, possibly using a second mic, not for stereo, but just for coverage. You'll probably need to compress and or automate but you can do that in post where you don't have to commit to anything specific as when you record with it.
Cheers, Rock
|
|