How often do you color grading ?

You don't actually gain anything from a flat profile. With DSLRs and other low bit rate cameras its best to dial in as much as you can inside your cameras settings. I hardly grade anymore unless I'm fucking around. I usually only adjust blacks/mids/highlights if I even do anything.

What about in camera sharpness? I heard that adding it in post is a lot better than doing it in camera (could just be a myth)

My last color grading was an A+. Usually I get B's or C's but ever since my dad had a sleep over with my teacher, I've been getting pretty good grades on my coloring. But the best way to color correct is with whiteout. White is the best color ever.

You get so much more dynamic range in a flat image where blacks arn't "blocked". Basically, to efficiently color grade, you just the the DR to it's max, for that, you must get very low contrast, low sharpness, and highlights must be low too. That's what I was taught at shcool and what I see on many shoots if your workflow requires solid color grading changes.

You don't actually create more DR and when you use a flat profile you're just causing yourself to have to grad a file that wasn't made to be graded. Unless you're shooting RAW, shooting flat is a waste of time. Whoever at your school taught you that needs to be fired.

I really don't know much about color grading but (I'm shooting on a t2i) should i be adjusting anything else besides my iso, shutter, WB and f-stop? I don't even know what shooting flat is. And I know that shooting raw pretty much gives you full control over your image in terms of color grading, but does say the fs700 over the t2i? It does right? Sorry I'm sounding like a complete noob here. But someone please elaborate on this.

Not that I'm saying you're wrong (ehv), I've heard a bunch of other people talking about how flat profiles on canon dslrs are pointless in different threads here on ns, but it'd be nice to see an example video or something showing examples of both and proving it. I feel like so many people use magic lantern + flat picture style or technicolor cinestyle that they probably aren't all just completely wrong.

It's really just my opinion, if you wanna shoot flat and grade the fuck out of your h264 file to the point where it breaks down the pixels, thats fine.

I just don't want people posting in here that it creates more DR or creates a higher IQ. Thats all.

The problem is when people shoot footage with stock profiles that look rich and contrasty in camera, and then end up with blown out highlights and crushed blacks when you look at it on a monitor. I've sat in way too many color sessions with a shitty cinematographer who's saying "uh, all the information looked like it was there in camera... can we just roto in some blue and some clouds into the sky?"

Just don't shoot with a damn stock profile. It's like shooting with auto settings. I'm not much of a camera op, so I prefer shooting with Technicolor Cinestyle or Canon Neutral, which are safe, but not too flat.

I used to always look at the screen on my t2i (which is only like 480p I believe) and see that it looked really good and I was stoked to make a sick edit with it. But then I pull it up on the big monitor and realize it looks like shit. I'm probably going to do some test footage to see if it actually does make a difference (once graded)

Il try to break it down in the most comprehensible way possible. It might sound dumb at parts but Im trying to explain so everyone understands.

There are basically two parts to a video process. Input and output. The input is basically what your camera is receiving as information (video recorded in cam) and your output is the final video that you will export to share with the world.

When working on a project with a certain output(look)in mind no matter how you shoot it, you want the final product to look a certain way. You will do this by manipulating the video that you originally record. Weather your input is high contrast or low contrast can generally affect the ability of obtaining that final result. With video dslr's that margin is very narrow and the negative effects of stretching your colors can cause very serious video bugs called artifacts.

The idea of having a super flat input comes from film and raw video. The idea is to be able to underexpose your image so that you can later bring back the lowlights to higher levels while leaving the highlights in the sky more detailed as opposed to being blown out.

The problem with the application of this technique with video dslr's is the codec compression. Raw video is recorded in higher color spaces and bitrates which basically means it records more subtle differences in color over a larger spectrum of light.(ex:4:2:2 12 bit) This helps when we are pushing the saturation up.

In video dslr's the video is recorded with less color variance.(Ex 4:2:0 8 bit color space) When we push the saturation up on this file, the colors "block." Blocking means that similar colors are grouped into one solid color rather than a smooth gradient. This is due to the bit depth: 8 bit instead of 12 for example. This effect more importantly happens in the blacks when they are shot low contrast (grey) they will group together in blocks when contrast will be added. This is due to the color space 4:2:0 as opposed to the optimal 4:4:4 color space in really really high end cameras.

All of the above basically means that your file format lacks the proper quality to be able to really get the benifits of shooting flat. Your sky now has details like clouds but it is gray and needs grading. You simply cant stretch the extra info enough for it to be good without causing serious blocking. Result, most people grade the footage to look like it would have in camera because any further and you are plagued with artifacts.

So with what was said above, we have determined that people use it and usually acheive the same output weather they shoot flat or if they shoot in cam.

Here is the reason why you shouldn't shoot flat, not only because its the same now but because its actually worse.

In order for your camera to receive "additional dynamic range" your processor has to do a few tricks when processing that information. In short what this profile does is boost your low lights and lower your highlights in order to gain more dynamic range. This does work, but with a cost. The lowlights have a gain added and the highlights receive an iso pulldown. This causes noise, lots and lots of grain. On a high end video camera, your sensor has that much dynamic range, it doesnt need any tricks to get there. The result is that good cameras dont have more grain in their image but yours does and your going to boost that grain in order to shape your image but you cant because your codec is not flexible enough.

SPARKNOTES:
Dont shoot flat: It adds grain to get more dynamic range which you end up losing because your cameras files cant be stretched enough to see the benifits. your making your image worse in a very subtle way but mostly you are waisting your time.

Nicely explained, I didn't really know about DSLR's. At school we were taught (usually on 2K and 4K shots)to make it as flat as possible so that at the end of the chain, the editors will be able to color grade without killing the colors too much. So i just thought, "great, i'll do the same with my DSLR" but with what you are saying, I might do it less.

I looked into it for a really long time and I never was satisfied with hearing peoples opinions.

Personally I need to understand something in order to believe it and so I did a lot of tests and research combined with the knowledge I already had at the time of making my decision and after all of that I have come to the conclusion that its a gimmick.

Understandable that people use it when they know the surface information, there are a lot of advantages in theory. However, when looking into the technical aspects of those theories I realised that there are no benefits, only downsides.

Why it exists if there are no benifits? Well its complex enough that most people wont realize that its pointless. They got it supported by popular Indy filmmakers and got people to use it and feel like they are in full control of their production. Now everyone knows the name technicolor.

exception: I think this will be a good picture profile to use when the 5D is fully functional with the hacked raw capabilities.

This is like saying that spreading food around your plate creates more food, when all you've done is taken the same amount of matter and stretched it out. Think of it as conservation of matter, only with data.

Shooting dedicated flat profiles makes sense with solid bit depths (12-bit), but with low-end bit depths (8-bit) it only harms your footage.

The Cinestyle fad is nothing more than armchair DSLR-cats with little to no practical understanding of color processes who regurgitate the same handful of blog posts.

If you want a good flat profile, use neutral with everything turned down. Or better yet, learn how light works. 80% of the image is made before light hits the sensor.

Is this the reason programmes on tv will have extremely high saturation and contrast, but won't look like shit, presumably if you did the same with 8 bit it would?

it depends which shows, a lot of them are very heavily post processed like true blood, walking dead and shows in that HBO production style. They shoot on high end digital or film cameras and tweak the image to no end so they can get really crazy looking shots and they do this very well.

Sitcoms for example will usually have a structure where they need to shoot fast and edit fast. They dont really have much post processing done so they could get away with dslr's. Most of them already have the cameras but some shows use dslr's. I know House used some 5D's for at least a show, maybee a season. It looked great because they dont have to manipulate the image as much.

If walking dead was shot on dslr it would look absolutely amateur because the sensor cant handle the dynamic range needed for their eery sky and crazy detail. A flat profile would be stretched to the point of blocking because it doesn't have enough colors to work with ( low bit depth 8 bit.) It is shot on 16mm film because they have a real wide dynamic range and tonnes of detail.

But isn't one of the main points of shooting flat so that you can adjust the contrast/sharpness/everything later (to an extent obv. since its just a dslr with h.264)? Much better to shoot flat than to shoot with tons of contrast in camera and have it clip the blacks or highlites, or over sharpen it (have had this happen, my friend set it to max sharpness since it looked better on the LCD of the camera, but looked like ass on the computer...) Even if it doesn't give more dynamic range its still better than losing info in camera, if that made any sense

your better off shooting in a slightly low contrast with the sharpness to the lowest setting.

The information that you gain by shooting flat is only gained on the screen. It is far harder to clip your image than you would think. Your darks will only be clipped when severely under exposed. Even though they appear pure black, up your blacks and you will see new details come out. By doing this your doing exactly what technicolor is doing.

The one thing I agree on doing 100% is the sharpness being lowered to basically off. Your sensor is 18mpx and has to down scale to roughly 2-3mpx. It does this by essentially skipping lines on your sensor. Sharp details and lines are skewed or lost when these lines are removed. This can cause aliasing which is when a detail doesnt align properly with the next pixel due to the process and causes weird waves in your image.

Best solution is to take it off so that this effect is limmited as much as possible and add sharpening as needed in post.

Secondly for sharpening: The camera's sharpening is automatic and can not be fine tuned. It tends to be harsh and un precise. In a good program you can control not only the itensity but also the radius and also have specific tools like unsharp masks and precise sharpening.

I know you might laugh now with this question but... I have a GoPro hero3 black edition. But thats what i use right now until i feel upgrading to a DSLR(which is soon). But GoPro came out with Protune if you know that and it has a bitrate of 45 Mb/s and shoots with low contrast and sharpness. Now would this be the same as shooting with the Cinestyle preset on DSLRs or would you actually benefit from it ex. more information, easier to color correct and so on? Protune has higher bitrate than the normal recording modes.

Well if it has higher bitrate than you're getting more information so yes it would be better.

I'm pretty sure that is done by putting a transparent tan layer over the video. Can anyone confirm?

How would it be possible to make footage sharper after shooting it apposed to in camera?

In post. Unsharpen mask in ae/prem

I dont know the specifics of that format but a general rule is that a higher bitrate is better. Your essentially compressing your image less.