Dynamic range

P

Paper_bag

Active member
Hi! A complete noob question, but here goes...

When I look at an image and makes guesses about, "What was that shot on?", I find that video "feels flatter". Now this is a matter -- isn't it? -- either of the difference between brightest white and darkest black, or of the number of levels into which that range can be divided?

I've been told that it's possible to create HDR still images by taking multiple shots at diferent exposure brackets, and merging them; for instance, Photoshop CS5 has an upgraded tool for this.

Well... how easy is it to do the same thing for video? Can the same thing be done for video, to extend the range beween highest and lowest brightness points? And what might be involved in the process? If I take multiple shots from a locked off camera, can I get something approaching the same effect with judicious use of blending modes and masking?

Another thing I wanted to ask about -- what exactly limits the dynamic range of video compared to film? Is it sensor sensitivity, codec, display hardware, some mixture of the above? Are there any hard limits here that I'm going to run against if I start experimenting with increasing dynamic range through compositing?

As I said, complete noob here, and completely ignorant of anything to do with this topic, so any thoughts/comments are much appreciated.
 
I'm not an expert on all the technicalities, but there's a number of issues with HDR video. The first one that comes to my mind is that you'd have to shoot the same scene, at the same time, through the same, exact angle/lens at at least three different exposures. There's no camera that can shoot at three different exposures at the same time, that I know of, and if you used three cameras at the same time you'd have to line up the images, which would probably less than perfect. You'd end up with ghosting due to lens differences recording the image from different POV's, although slight, still noticeable.

I believe sensitivity with video chips are a result of the sensor and the math converting those signals over to the desired codec/recording medium. Check out ProLost, he has some great info on all this over there.
 
Paper_bag said:
Hi! A complete noob question, but here goes...

When I look at an image and makes guesses about, "What was that shot on?", I find that video "feels flatter". Now this is a matter -- isn't it? -- either of the difference between brightest white and darkest black, or of the number of levels into which that range can be divided?

I've been told that it's possible to create HDR still images by taking multiple shots at diferent exposure brackets, and merging them; for instance, Photoshop CS5 has an upgraded tool for this.

Well... how easy is it to do the same thing for video? Can the same thing be done for video, to extend the range beween highest and lowest brightness points? And what might be involved in the process? If I take multiple shots from a locked off camera, can I get something approaching the same effect with judicious use of blending modes and masking?

Another thing I wanted to ask about -- what exactly limits the dynamic range of video compared to film? Is it sensor sensitivity, codec, display hardware, some mixture of the above? Are there any hard limits here that I'm going to run against if I start experimenting with increasing dynamic range through compositing?

As I said, complete noob here, and completely ignorant of anything to do with this topic, so any thoughts/comments are much appreciated.
Click to expand...

The sensor has limited dynamic range and using light reduction techniques the range can be extended.

For stereoscopic rig to shoot HDR one simply sets the IO (interocular distance) to 0, sets one camera to clip on back level and the other on white and in the post the left right footage is combine to form HDR footage.

Mathew Orman
 
You must log in or register to reply here.
Back
Top