fansfe82067d
First Officer
Australia
Offline
|
This is a difficult issue. First of all, what are you recording? If it's just somebody talking, then levels around -20dB may well be correct, because talking isn't a particularly loud sound. If however you are recording a train passing 2 meters away from you, yes, you'd expect that to peak to the top. Cheap cameras tend to use auto gain controls which aggressively boost quiet sounds and aggressively lower loud sounds, but that's not actually very realistic. I would expect some AGC / compression to be used but not aggressively. Good quality audio preserves within reason the original dynamic range of the sound.
As for the sound quality, the sound from the internal mics is, in my view, very good and has the abililty to present a really wide stereo image which, when you see a video of city scenes for instance, makes you feel like you are really there. Traffic rumble really does rumble, but any music from a cafe sounds suitably bright, and voices sound natural. The wireless mic is probably geared more to conveying speech clearly and would be less likely to reproduce sounds of traffic so well. I would expect that. And of course it's mono which changes the whole feel of the sound. One option is to get the best of both worlds by recording on the wirelss mic not linked to the camera, and recording on the camera from the built in mics at the same time, and mixing the two to taste later.
To be honest I have only shot one clip with my Pocket 3 so far, and that was just of me walking round the house I am moving into describing what my plans for it are for some friends. I just used the built in mics. I was impressed with how faithfully it conveyed the very faint but bassy sounds of pop music playing on another floor when I was in the lowest one. Very realistic. The levels sounded as expected. Maybe yours is defective but that would be a pretty weird defect. Lastly, I should mention that 32 bit float audio is likely to sound quiet because the analog part of the audio chain will be set to a low level to ensure that clippng of the audio doesn't happen when the sound is very loud, before it gets digitised to 32 bit float. But then you can increase the gain substantially when editing without any ill effects, that being the nature of the 32 bit float beast.
[I speak as a retired classical music recording engineer - but that was as much art as science, so I don't claim huge knowledge of the technicalities.]
|
|