Blown out and dark footage with Mini 3 Pro

[Bunraku]

Hi Guys

How do i stop wasted footage footage like this. I was filming today and used auto settings with 0.0 EV and normal filming mode as i don’t colour grade. The Histogram was on and can’t remember the graph being anywhere, but middle. Some of the footage is usable. however a lot of it was wasted as it either gets blown out or becomes too dark. The first clip is okayish, but as i turn towards the sun the castle becomes too bright and the sky becomes blown out in the sky. The second part it is ok at the start, but as i turn towards the castle which is in shade it goes very dark and loses most of the colours. I know you can Auto Exposure Lock, but not really sure also how to use it. How can i stop this happening?
Much Appreciated

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Meta4]

The second part it is ok at the start, but as i turn towards the castle which is in shade it goes very dark and loses most of the colours. I know you can Auto Exposure Lock, but not really sure also how to use it. How can i stop this happening?

You need to give more consideration to how the scene is lit and how the camera's exposure works.
Unless the sun is covered by clouds, it gives a very directional light.
In this scene the sun is behind the camera and lighting the whole scene evenly.

But when you turn the drone far enough to the right, the scene becomes side-lit and if you turn 180° you end up with the sun in front of the camera, (behind the scene), giving a bright sky and shadows in front of large objects like the castle.

The camera is giving you an average exposure setting for the whole image, but half of the image is quite bright and half is quite dark.
This is a high contrast scene where there are significant differences between brightness and darkness of large parts of the image .
It's not possible for the camera to expose so that both show correct exposure.
If you manually set it to expose correctly for the highlights, or for the shadows, you would end up with darker shadows or more overexposed highlights.
Exposing for the average of the scene isn't ideal either when there are big differences in brightness.
This is just a limitation of the camera technology.

Recognising this and avoiding high contrast scenes and the lighting conditions that contribute is the key to getting better imagery.

 

[waynorth]

@Meta4: Thank you for taking your time to give such a thorough explanation, giving help like you do is what forums like this is all about.

 

[MavicAir2Marc]

The dynamic range of tiny sensors isn't that great in the first place. Your eyes have way more ability to see detail in the shadows and highlights than a small electronic sensor. If you are just taking photos, on some models, HDR may be helpful or bracketing several shots and combining them in your favorite photo editor of choice.

 

[Dogpilot]

Not to start a big debate here. Your eyes actually have very poor dynamic range. What we see is actually a composite image. Our eyes literally take thousand of images at the correct exposure, with the fovea being the center with the sharpest image area and color response, but it is a small part of the image. The eye make tiny movements painting the image with individual properly exposed bits and tiles it all together in our brain. It does this continuously. It does give you the illusion of a high dynamic range image, but it is really a mess-o-images properly exposed for a small area stitched together.

One of the few things I ever felt a loss about film, was its ability to saturate parts of the image, especially colors at sunset. Digital at times is like Commander Data from Star Trek going to a comedy club, a bit too literal. This inability to do a perfect exposure across high contrast scenes is why cameras and processing software have HDR, which is a composite of several images across an exposure range. Eventually, the cameras will be clever enough to do it in one exposure, processing on the fly, individual pixels or zones of pixels, to get the best dynamic range. Not there yet. That would be a big boon for video.

 

[akdrone]

It's been well explained here. I would add that the footage is very normal and to be expected shooting in auto, which is why most pro's rarely shoot in auto or pan from a position where the sun is in back of the camera to in front of the camera because of how dramatically the exposure changes. It's also the reason people shoot in RAW or Cine mode so as to be able to have a higher bit depth and be able to grade the image in post which I realize you don't do but it can help a great deal if you gradually change the exposure of a scene in post to compensate for the changing auto exposure. When filming a scene with dramatically changing light auto exposure can be useful but in a strongly backlit situation you can choose to expose for the foreground or the background but you can't get both. A dimly lit room with a bright window is good example. If you take a photo inside and expose to make the room look good the window will be entirely blown out. If you expose so the are outside the window looks good the inside will be super dark. It's also why, when shooting still images, there is an ability to shoot in AEB, taking 3 or 5 images at differing exposures to blend the HDR final and get good exposures in the dark and bright area which can't be done with video.

 

[Droning on and on...]

The best answer is to shoot 10-bit Cinelike, and color grade. Avoiding this will limit you as you've discovered.

An intermediate choice would be to shoot using the HDR mode of the sensor, which utilizes the dual native ISO feature. There's all sorts of controversy on this forum about this mode, and I'm not going to address any of that.

Rather I'll just say that you WILL get better dynamic range than you get with simple auto exposure.