Blaise Douros
Forum Replies Created
-
I’m snagging the file from Vimeo and will mess with it a little. I’m wondering if 3D camera tracking combined with a pair of spotlights will do the trick.
-
Your reds are going very slightly past the edge of the 75% usual broadcast safe mark. NTSC has about a 20% safety margin built in, which is what the outer marking is. You could bring saturation back a little to keep it safe. However, I do also recommend looking at your output file on a couple of different monitors to be sure.
-
So, given that the histogram looks correct, you could do one of two things: 1) take your friend’s advice at face value and apply desaturation to the image (not just to the reds, because that would throw off the white balance) to dial back the colors a bit. Or, 2) check with your friend to see if his HDTV is properly calibrated. I suspect that the second option is probably more likely to solve the problem!
It’s super important that you take the time to learn about how to read the waveform and vector scopes before you use them to inform any color correction decisions you make. And notice I say “inform;” the color scopes are a helpful tool, but your eye is far more important. An object that reads identically on the scopes in two different scenes can appear totally different to the eye, so you often have to cheat the color balance a little to make it look visually right.
Larry Jordan has a pretty good overview of using scopes–I recommend taking some time to dig into it.
-
You may also be misreading your scopes, or don’t understand the significance of what you’re seeing. I see absolutely no problem with the image or with the scopes.
As I look at the Chroma scope, which is the circular one in the upper right, the red levels top out right at the broadcast safe mark. Totally acceptable.
Then, I look at the histogram below that in the lower right. This is not read the same way as the luma scope in the upper left; this is like a photo histogram, flipped on its side. It’s showing that you have a lot of reds in the highlights, but none of it is clipping. Since there are lots of bright, saturated reds in the car, I would expect to see that.
Then I look at the luma scopes on the left side, as well as the individual color channels. Notice that the IRE scale is different on the channel scope versus the luma scope. The red channel shows a higher level of red than the other channels, but it’s not clipping. Same with the luma scope–I don’t see anything that suggests there’s a problem with the image.
Which leads me to my final conclusion, which is that you don’t understand how to read a scope. You have an extremely bright red object (and orange cones) that is prominent in the frame, the rest of which is grey pavement. This means that your red channel will show much higher levels than the other channels, because there is more red in the image than any other color.
Your scopes should NEVER show a perfectly balanced amount of color unless you’re shooting a color chart. If you were shooting a scene in the forest, with a bunch of green leaves filling the scene, your green channel would show a spike, and this would be absolutely correct.
Your scopes are useful to look at when you know there’s a problem with the color, and when you need to see exactly what changes you need to make in color grading to get one image to match another, for example. But don’t make the mistake of believing that every image should show a perfectly balanced set of scopes, because any colorful scene will naturally have a unique mixture of colors that will produce a different reading on your scopes.
-
Don’t use 24.00 frames per second. You need to shoot in the 23.976 setting. If you don’t know why, it would take me far too long to explain–there are many, many articles and forum posts out there about the difference.
The short answer is that you can’t shoot 23.976 and still get slow motion. You also can’t get the 23.976 look while shooting 60p (which is really 59.94. Seriously, look it up!).
Good productions plan ahead of time what frame rate they will use for what shots, and set the camera accordingly for each setup. If you’re really a glutton for punishment, do a take of each shot in each framerate. But you’d be better off undertaking some planning.
When you shoot 59.94 (or higher speed, like 120 fps), to get the slow motion effect you have to do a frame rate conversion so that the edit software plays back the frames like a 24p (23.976) clip. Each edit software does this differently, so you’ll need to look up how to do that once you get to post.
-
This is not a problem. The histograms and scopes that you’re showing are correct for an extremely saturated subject in bright sunlight. You could underexpose the scene a bit if you’re worried about the red channel clipping, but white balancing won’t really solve much, given that you’re working with low-bitrate AVCHD. You’d just create an off-balance image which you’d correct in post, probably introducing artifacts from grading a low-bitrate image.
The only way I can think of to keep this from clipping the reds would be to reduce the in-camera saturation control, for a flatter image. Again, though, AVCHD won’t stand up to much grading. If your color is pretty accurate in-camera, and you don’t plan to aggressively grade it in post, I don’t see a problem with the image or the histogram.
-
It really, truly, seriously, for real, for sure, absolutely, positively doesn’t matter how you connect to this projector, or how good the source material is. The projector’s resolution isn’t high enough for you to use it for a large image. The projector has 1280×800 pixels, so roughly a 720P HD display. Imagine taking a 720P TV and blowing it up to the size of a background, at 12 feet wide. You don’t gain any resolution by making it larger; there are not extra pixels between the existing ones that suddenly appear when you project the image on a 12-foot screen. You’re just making the existing pixels larger.
720P video (1280×720) is about a 1 megapixel image. Zoom in on a 1 megapixel image, and you see pixels almost immediately. So a projector that outputs 1280×800 is roughly a 1.1 megapixel projector.
-
Why not just create a light, and add an invisible shadow catcher object to match the plane of the table? You can then control the amount of softness of the shadow in the light’s parameters. And as a bonus, the shadows will behave like real shadows, cast in the correct direction from your virtual light source.
-
I don’t think you’re going to get the kind of results you want out of a cheap projector; you need a lot of resolution and brightness, and those both cost money! When they did this for the movie Oblivion, they used 19 projectors! Granted, it was a really big set, but still. https://www.youtube.com/watch?v=eyHwDizJ-xE
The slide projector would probably be the cheapest option resolution-wise, although it would involve the highest ration of results-to-pain-in-the-ass. It probably isn’t that bright, too, so it might only work for dark scenes.
Have you looked at renting a 4K projector? There’s no way it’s THAT expensive just to rent one.
Some contents or functionalities here are not available due to your cookie preferences!This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.
-
Welcome to the modern era, where there’s no actual standard for the brightness of a camera’s ISO setting. Yep, really. Two different camera models with the same lens and the same ISO setting are going to look different. There’s no physics reason that the full-frame sensor should be brighter.
There’s no magic trick here–the sensors are just different, so you’re going to have to match them using good old fashioned zebras, histograms, and your eyeballs.