Blaise Douros
Forum Replies Created
-
Re-read Walter’s comment, but slowly this time. He tells you the exact circumstance under which you would use this.
-
Trying to do this is your worst possible option for integrating your CGI footage into this scene. Why not go the other route, and work on masking off your footage so that the tree appears to be in the foreground of your footage? That way it will look more integrated with your environment, and that is MUCH easier than removing the object entirely. Unless that tree is blocking the number one most important feature of your CGI, I’d do it this way and move on.
You can do this easily by masking off a garbage matte that surrounds the tree, and then using the color of the leaves as a key to refine the matte.
I’ve done this a bunch in videos for a utility company, showing the trajectory of underground pipelines; in areas where trees are present, you can make the object appear to be under the tree canopy.
-
Looks like garden-variety moire to me.
-
I use a Syrp system for motion control for our B cam, which is an a7III. It’s a shade less precise on slider moves than other systems, but its flexibility in setup makes up for that.
-
This is such a big topic, and requires so much background knowledge, that I pretty much have to tell you that you need to go research the history of film sizes, sensor technology, noise reduction, codecs, and processing for a while before we can even begin to have an intelligent discussion on it.
For example: the a7sII has such a great reputation for low-light performance because it processes the video like crazy, and uses slightly different back-illuminated sensor technology than RED and ARRI. So while RED and ARRI may not get super clean video at super-high ISOs, they will get much BETTER results at normal ISOs, and really, the a7sII is just applying tons of noise reduction at high ISOs anyway, which doesn’t depend on sensor size, plus the codecs capture differing levels of detail, so you could argue that there is room in RED and ARRI footage for post-processing of low-light footage that the Sony doesn’t offer, and…and…and…
You see what I mean. There’s a lot to learn, and nobody here realistically has the time to take you through the (quite extensive) amount of knowledge that it will take for you to understand all the whys behind the answer to your question. So for now, you’ll need to accept the short answer, which is that low-light performance does not depend on sensor size.
-
The answer is dead simple: noise doesn’t depend on sensor size. Noise depends on noise-reduction algorithms and the camera’s overall high-ISO performance. So some cameras are better at it, and some are worse.
-
Apply for a waiver from the FAA for Section 107.25 – Operation from a Moving Vehicle or Aircraft. If you can prove you can do this safely, and get the FAA to sign off on your waiver, just follow the drone in a car and do it in one take.
Otherwise, forget it.
-
Actually, vintage stills lenses work great for video–many of them are great quality, they’re available for low prices, and they can be modded so that the aperture changes smoothly. The focus breathing isn’t always well controlled, but neither is that true on a $2000 Canon L-series lens, so…
Anyway, don’t knock ’em until you’ve tried ’em. They can be a great solution. There are even Hollywood DPs that use them, in certain projects.
Peter, the M/MD just means Minolta MD mount. Any MD-m43 mount adapter should do the trick, and if you can get one with a Speedbooster optic included, so much the better.
-
Whatever you want it to be for the scene. Narrative is different from documentary/eng/beauty/commercial. Lighting can greatly effect how the scene is perceived, so if you’re trying to keep someone’s identity hidden, or hide the backdrop as they’re lit by the light spilling in from wood slats, you’re going to make exposure decisions based on THE WHOLE SCENE, not just the actor’s skin IRE level.
Put it this way–if the scene calls for you to shoot the actor in silhouette, are you going to expose for skin tones and change the exposure in post? No way!
-
Took about eight minutes. Precomped it, and lightened the color up, increased contrast so that the plain had lots of easily-trackable black spots. Used 3D Camera Track, and used Detailed Analysis, selected Typical from the scene type, and that was it.
Here’s the AEP file: 13158_dimitrovlandscapetrack.aep.zip