Take the 2-minute tour ×
Video Production Stack Exchange is a question and answer site for engineers, producers, editors, and enthusiasts spanning the fields of video, and media creation. It's 100% free, no registration required.

I've read a bit about how it can be useful to use a "flat" profile for shooting video on a DSLR which produces only compressed video. For example as described in this article. The reasoning is that most lossy compressors (like H.264) will do more compression in the darks, removing that data and making it impossible to adjust in post. (For example, if you brighten it, you'll get bright gray squares instead of whatever data was there before compression.)

The author of the above seems reasonable. He worked at LucasFilm and has directed many music videos and feature film VFX shots. It doesn't seem like he's suggesting some fad, but is talking from experience.

In the answer to [Post-production workflow: VFX before color correction? workflow: VFX before color correction), this was included in the accepted answer:

As far as low contrast and low saturation, I would disagree rather strongly with shooting for low contrast/low saturation. This is how you get noisy, low quality color. Your camera has a limited number of colors it can describe, if you ignore a large portion of that color space, then you can't describe as many colors and the resolution of your color after boosting the saturation and contrast will be very greatly reduced.

These sound contradictory to me, but maybe I'm misunderstanding. Is the advice to shoot flat good advice for DSLRs with lossy compression or not?

share|improve this question

2 Answers 2

up vote 1 down vote accepted

Neither viewpoint is wrong, they rather have different advantages. It is a question of dynamic range vs compression noise. My answer on the previous question may have been a little too strongly from my particular camp of thought (I was not aware of the compression related dark color issues at the time as I shoot high bit rate video primarily, which makes it much less of an issue.)

If you are shooting non-raw and using a flat color space, you have fewer values to represent the range of colors in your shot. You can get more detail on the brightest and darkest spots since they are exposed as mid-range colors, but you lose the depth of color since you have to artificially expand it later in post. I'm not sure what I was thinking with the noisy bit, probably thinking about the loss of color detail from reducing your color space based on the surrounding explanation, but noise was a poor choice of term for that. (I have updated accordingly.)

The flip side of that is to get it mostly right in camera. You have to be careful about your black point and make sure that you actually have sufficient detail in the shadows, but utilizing the full color space allows for far more colors to be described, even if it offers less ability to adjust after the fact.

It is all a question of recovery. When you deal with RAW, you have detail going all the way down to the darkest and all the way up to the brightest point from which you can work to pull out any details that the camera captured. When you are dealing with an encoded video though, you are limited to whatever is displayed. Using flat allows more recovery since you capture beyond the black and white point of the final video (and does reduce sensor noise in the darks since they are actually exposed as mid-tones) but it means much less color detail to work with.

I prefer making full use of my color space and controlling shooting, others prefer to have extra control at the cost of color depth. Neither is a wrong choice.

share|improve this answer
    
Thank you for the clarification! This makes more sense to me now that I see what the trade-offs are. –  user1118321 Jul 18 at 16:04
    
I'm confused about the "fewer values to represent the range of colors" bit. Is there a noticeable difference between applying in-camera saturation as opposed to post saturation? I didn't think changing in camera contrast and saturation (color profiles) changed the "color space" of the camera. It's still sRGB, Rec 709, NTSC, etc. no matter what your settings. Am I misunderstanding? –  Scott James Walter Jul 18 at 16:34
    
@ScottJamesWalter - it is the depth of the color space that changes. It still captures the same min and max color (actually, technically, it gets a wider min and max arguably), however there are fewer actual values in between. You have a fixed color depth (number of distinct colors). If you use a bunch of those as buffer, you lose color depth. Taken to and extreme, lets say you almost made the image black and white. You could expand the limited color you had so that they are strong colors again, but it's gonna look something like early color footage without a wide selection of colors. –  AJ Henderson Jul 18 at 16:58
    
In the finished product, for a "flat" color file, you are going to make a lot of greys black and possibly some light greys white. Those are all possible color values you aren't making use of. –  AJ Henderson Jul 18 at 16:59

There are two schools of thought on this topic but nothing I know has suggested that shooting with a flat profile is a bad thing. Here is what I understand:

1) When shooting with Canon DSLRs, it's generally advised that you shoot with a flat profile (Technicolor cinestyle, Prolost, etc.) and over expose (but don't clip!!) and you get a picture with much more detail, specifically in the shadows. Here is a video showing this: https://vimeo.com/73909925

It's rather long, but the entire series is good. In there he specifically shows the effects of shooting standard and exposing correctly v.s. shooting flat and over exposing. Lots of detail in the shadow is preserved when performing the latter.

2) A lot of more professional film makers (Shane Hurlbut, and Stillmotion) suggest that you should use a color profile that best matches your final color result. Less work is needed in post, and you have a better feel for your film. I think I see this school of thought from professionals more often because the gear they own has high dynamic range and retains so much information anyways, so detail isn't something to worry about.

Personally, I'm not a professional and I record with a Canon 550D but I use the Standard Profile when exposing my shot (aka, over exposing) and switch to technicolor cinestyle when I hit record to give me more color grading flexibility in post.

share|improve this answer
    
+1 on the observation about dynamic range. I suppose that really is kind of key that if you have good dynamic range in to shadow, then it is worth preserving color depth, but if you don't, you are really just adjusting the format to use the best part of the dynamic range available and discarding part deemed unusable. –  AJ Henderson Jul 18 at 16:00

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.