Should I ever raise gain when filming or just do it in post? I'm on an XA-10

oxband

Well-known member
Sometimes, when I shoot with my XA-10, the image is dark. I usually raise the gain by 3 or 5 to make it brighter. However, is there any advantage to doing that as opposed to simply doing it in post? It seems to me the advantage of doing it just in post is it means the raw footage will have less grain from the get go. Thoughts on this? I'd love to hear other ideas.
 
Increasing the gain in-camera has the benefit of amplifying the signal right after the sensor, before it gets encoded. If you do it in post, you're potentially fighting whatever codec the video is compressed with unless you're shooting RAW.
 
Since there's no raw on an XA10, you can't bring back what isn't there, plus the H264 breaks down quite quickly in post. With experience you learn how much gain you can add in any situation before it becomes unacceptable. You can benefit from Neat Video. Also, if you're shooting 50PAL (60NTSC) fps you could shoot 25 (30) fps and depending on what you're shooting drop the shutter speed to 1/25th (1/30th), for overall 4X the amount of light.
 
Noise is not just a symptom of using a higher gain/ISO. It is actually a symptom of underexposure.

If you shoot with high gain/ISO with proper exposure in broad daylight, you'll have very little noise (though it's a silly idea as you'd need to use lots of ND's to counteract the ISO). If you use that same gain and shoot indoors and the image is poorly lit, you'll see a lot of noise.

When dealing with compressed, processed video data (ie: anything other than RAW), it is actually better practice to slightly over-expose in low-light scenarios. Although you start with higher gain to do this, the over-exposure combats noise. You can then crush the shadows in post to bring the image back to how it should look, which further reduces noise.

By doing the opposite - shooting with low gain and fixing it in post, you are getting a lot more noise, it is just hidden in the darkness of the image. As soon as you try to lift the exposure in post, that noise will rear its ugly head. You also have to deal with compression which usually dedicates more data to the properly expose middle range of the image, and less data for the dark parts of the image. Lifting the exposure in post often reveals compression artifacts which would've otherwise remained hidden in the shadows.
 
Back
Top