-
Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.
color correction tips.
- Thread starter peter917
- Start date
Are you looking for tips on the correction? The digital sun is a little obvious at the corner of the screen for me personally... you may want to either move that farther off screen to enlarge the point of impact, or just make it a little larger to spread it out a little bit.
One of the drawbacks to shooting with the wrong WB setting is that you will lose quite a bit of saturation in all but the color channel of the color cast (in this case, the blue channel has more information than the R or G channels)... This can make it difficult to color correct back to normal as the red is flesh tones and the green is the detail in those tones. Since we generally try to correct for flesh tones to make our subjects look healthy, give the grade less information is difficult.
If you start with more information (shooting a slightly flat image and having it white balanced to the dominant light source -- I only use the presets for indoor and outdoor), you'll be able to do much more in the color grade with your image as it'll have more data to work with which will help keep it from breaking down as you push/pull it around.
One of the drawbacks to shooting with the wrong WB setting is that you will lose quite a bit of saturation in all but the color channel of the color cast (in this case, the blue channel has more information than the R or G channels)... This can make it difficult to color correct back to normal as the red is flesh tones and the green is the detail in those tones. Since we generally try to correct for flesh tones to make our subjects look healthy, give the grade less information is difficult.
If you start with more information (shooting a slightly flat image and having it white balanced to the dominant light source -- I only use the presets for indoor and outdoor), you'll be able to do much more in the color grade with your image as it'll have more data to work with which will help keep it from breaking down as you push/pull it around.
My opinion: Your correction looks much better, but some of the colors (skin and shirt) are a little weird. It's probably because your original image was so far from color realistic.
I'm a huge proponent of post color correction and grading, but in the future, take a little extra time getting the right balance from the camera first.
Now that you're in post, if you want a more cinematic look, desaturate, and add tealish blues to your blacks, and yellowish oranges to your highlights.
I'm a huge proponent of post color correction and grading, but in the future, take a little extra time getting the right balance from the camera first.
Now that you're in post, if you want a more cinematic look, desaturate, and add tealish blues to your blacks, and yellowish oranges to your highlights.
I'd like to say that I can't, for the life of me, understand why people think the "film look" means "desaturation" or lack of contrast... but, I know why people do this.
Trying to mimic the range of Film by desaturating an image and lifting shadows does not make your stuff like like film, it just makes it look like you don't understand color theory and what contrast is.
Trying to mimic the range of Film by desaturating an image and lifting shadows does not make your stuff like like film, it just makes it look like you don't understand color theory and what contrast is.
Neutral is probably the best way to roll on any Canon camera, but do realize that no amount of tweaking the camera will allow you to move the image every which way but loose. It's still just DSLR footage.
The closer you get to your look in camera, the more people look like people, the better your lighting ratios, the better off you'll be.
Of course, take it or leave it! Int he end, if we all did everything the same then everything would look the same.
The goal of limiting contrast is to make sure you capture all of the black values and all of the light values in your image that are available to you (prevent the camera from clipping either on the way in)... this process assumes you will be CC/Grading EVERY SHOT you take to get your contrast back with blacks all the way down at 0 and the hilights all the way at 100.
In camera sharpening off (or down as far as it'll go)... and saturation down slightly to protect the individual color channels from clipping (this will serve to flatten the image slightly) ( http://prolost.com/flat ). If you read the posts I made years ago, this is fundamentally the same setup I used on my XL1s. You're protecting the data so that the camera with its limited data capacity (example: in an 8 bit image, each color channel will only have 256 possible values of gray in each color channel -- not much information when it comes down to it, spend it wisely).
This way, in post, you can take the pixels with their R, G and B values and move them where you want them... if they are clipped, values that are darker than 0 become 0... and brighter than 255 become 255, making big oceans of black and white in your image that looks like crap (video look). Alleviating these clipped areas is a bigger solution to the "Film look" than 24p or anything else you can do... and is fundamentally a lighting challenge... expose for the hilights and light the darks into range -- coincidentally, just like you would with actual negative film (although it works more like reversal film where you have to protect the highlights more than the shadows).
I don't let actors wear pure white, pure black or pure bright red, green or blue on set ever, and I make sure the set folks take care of things on set with these colors as well... moving, painting, scrimming, lighting, whatever it takes.
In camera sharpening off (or down as far as it'll go)... and saturation down slightly to protect the individual color channels from clipping (this will serve to flatten the image slightly) ( http://prolost.com/flat ). If you read the posts I made years ago, this is fundamentally the same setup I used on my XL1s. You're protecting the data so that the camera with its limited data capacity (example: in an 8 bit image, each color channel will only have 256 possible values of gray in each color channel -- not much information when it comes down to it, spend it wisely).
This way, in post, you can take the pixels with their R, G and B values and move them where you want them... if they are clipped, values that are darker than 0 become 0... and brighter than 255 become 255, making big oceans of black and white in your image that looks like crap (video look). Alleviating these clipped areas is a bigger solution to the "Film look" than 24p or anything else you can do... and is fundamentally a lighting challenge... expose for the hilights and light the darks into range -- coincidentally, just like you would with actual negative film (although it works more like reversal film where you have to protect the highlights more than the shadows).
I don't let actors wear pure white, pure black or pure bright red, green or blue on set ever, and I make sure the set folks take care of things on set with these colors as well... moving, painting, scrimming, lighting, whatever it takes.
Well, I guess take this or leave it: use Neutral at most on a Canon cam.
When you are working with cameras that constrain all of their information within a certain set of values, that gets compressed, you have to play by the camera's rules.
By flattening out an image, not white balancing to a card (dialing kevlin, using presets) all of this stuff, you're just cutting the 8-bits of usable information down bit by bit.
If you want a flatter image, lift it in post. You aren't doing yourself any favors (if you like to grade) by doing it in camera.
Want a simple test to try? Shoot Cinestyle or a flat image. Shoot the same exact image with the contrast dialed back up to a level that that's slightly darker than you'd like.
Balance both to your liking, see which one is the better overall image concerning noise floor, color, etc.
But, yeah, even knowing that... at the endof the day, people should just work how they feel comfy! xD
Edit: And what Knightly said.
When you are working with cameras that constrain all of their information within a certain set of values, that gets compressed, you have to play by the camera's rules.
By flattening out an image, not white balancing to a card (dialing kevlin, using presets) all of this stuff, you're just cutting the 8-bits of usable information down bit by bit.
If you want a flatter image, lift it in post. You aren't doing yourself any favors (if you like to grade) by doing it in camera.
Want a simple test to try? Shoot Cinestyle or a flat image. Shoot the same exact image with the contrast dialed back up to a level that that's slightly darker than you'd like.
Balance both to your liking, see which one is the better overall image concerning noise floor, color, etc.
But, yeah, even knowing that... at the endof the day, people should just work how they feel comfy! xD
Edit: And what Knightly said.
Last edited: