Look: In-Camera or Post-Production?

While a histogram would be a wonderful feature, I find that you can achieve optimum contrast range by a number of different methods.

One would be to use a product like DV-Rack this allows you to quickly analyze the min / max illumination levels in your scene. You can also use your NLE to create frame grabs that you can import into Photoshop.

Another method (one that I use most often) is to have advance knowledge of the camera (based upon prior testing) and then use either an incident meter (and / or spot meter) to set your highlights and shadow levels. With a bit of experience, you can fully exercise the dynamic range of the system with very predictable results. This method ends up being nearly identical in practice to how it would be done if your are shooting film (though film stocks have wider dynamic ranges with a bit more curve in both the heal and shoulder areas). The spot mode available on most mid range and up cameras along with a calibrated reflectivity object (gray card, chip chart or similar) are also equally useful for accomplishing this task.

I fully disagree with the comment

"...shooting with a higher than -6 MP record any additional detail? I don't believe so."

It is entirely dependent upon the dynamic range of the scene that you are shooting. A wholesale statement such as this does not take into account the real world conditions encountered.

For anyone interested in this subject, a bit of simple testing will easily reveal what is optimum for any given set of conditions. No speculation is required, this is all very basic concepts.
 
khmuse said:
While a histogram would be a wonderful feature, I find that you can achieve optimum contrast range by a number of different methods.

One would be to use a product like DV-Rack this allows you to quickly analyze the min / max illumination levels in your scene. You can also use your NLE to create frame grabs that you can import into Photoshop.

Another method (one that I use most often) is to have advance knowledge of the camera (based upon prior testing) and then use either an incident meter (and / or spot meter) to set your highlights and shadow levels.
Click to expand...

Totally agreed, there are other ways, but a histogram would rock. Some of the new HD cams have them, but not Pana.

I fully disagree with the comment

"...shooting with a higher than -6 MP record any additional detail? I don't believe so."

It is entirely dependent upon the dynamic range of the scene that you are shooting. A wholesale statement such as this does not take into account the real world conditions encountered.
Click to expand...

I'm trying to find the link to the test results, but consider the scenario again. The camera has the lens cap on. ND 1/64 is on. Iris is fully closed. It doesn't get any darker than that. At -6 MP they got a 0 digital value. At -5 MP they got a positive digital value (probably 1), and at higher -4 and up MP values they got increasingly higher digital values... even though there was no light to register. What was registering? Artificial digital noise and processing errors from the DSP's. Amplification based on a certain MP value. No actual visual information, even though some was registering.

Mind you, I'm not sure the control params in this test (gamma, matrix, and other settings), but it goes to show that under any given circumstance there is an appropriate MP setting that sets BLACK to BLACK. Not BLACK to DARK GRAY or vice versa. Increasing the MP won't captuer additional detail, it'll just waste bits.

It's all about setting the MP accurately.

Check out three informative posts by Barry here. Again, no word about what the other control settings were. You'll note that the default scene files for 24p have -5 MP set by default.

As with many things, this may be a camera-specific tuning function. You may need -5 to get 0 digital values in the dark, my cam may need -7. The point is that there is a "true" black calibration that can be done using MP, and that it is fairly easy to determine. Set all your other scene file settings, then close it down and calibrate your MP to get a true black. Then, expose as appropriate.
 
While -6 might be a good starting point, it isn't an absolute (perhaps except for the conditions of near zero light striking the sensors, but when does this ever occur in the real world?).

Keep in mind that many of the other parameters will effect the transition to absolute black. Certainly a higher scene illumination level is one very obvious one, one might have a scene that has a section illuminated above the absolute black condition that you describe, but you still want some aspect of this section to reach the bottom of the dynamic range, here you would want to raise the threshold (black level offset) to accommodate your scenes characteristics.

My underlying point is that in order to extract the maximum dynamic range, you need to test under the conditions (lighting and other camera settings) that you are employing in a scene.

Now for the concept of a histogram, I would be one to vote for that feature in a heart beat.
 
Forget that in-camera appraoch; once yougo there you cant go back

I prefer flat & true no matter what look I have in mine from the beginning
 
If there is where you want, why would you want to go back? I shoot and light for my end goal.

Oh well, to each his (or her) own.
 
khmuse said:
While -6 might be a good starting point, it isn't an absolute (perhaps except for the conditions of near zero light striking the sensors, but when does this ever occur in the real world?).

Keep in mind that many of the other parameters will effect the transition to absolute black. Certainly a higher scene illumination level is one very obvious one, one might have a scene that has a section illuminated above the absolute black condition that you describe, but you still want some aspect of this section to reach the bottom of the dynamic range, here you would want to raise the threshold (black level offset) to accommodate your scenes characteristics.
Click to expand...

The test at least calibrates your camera for absolute black. In this test, -6 MP. Anything -5 to 0 and in the + side is only using valuable bits to record noise and error that will need to be corrected back out in post. This is basic calibration of the black point of the camera.

As you point out, it is unlikely anything in your real world scene will be that black. You will control your blacks through exposure and lighting, and if necessary you can lower the MP below -6, all the way down to -15. Raising it to a point where it is known to record only noise and error has no advantage.

Even shooting flat, setting an accurate black point is fundamental to capturing a high-quality image.

My underlying point is that in order to extract the maximum dynamic range, you need to test under the conditions (lighting and other camera settings) that you are employing in a scene.
Click to expand...

I totally agree, but the usable range of the MP setting can be determined in advance, for scene file F6 presumably (CINE GAMMA CINE MATRIX) it is -6 (flat) down to -15 (crushing or forcing the blacks down).

This, really, doesn't necessarily have to do with setting a look in-camera, it is basic calibration of the camera. However, it can also be used creatively.
 
I used to be of the mind that you fix your look in post. That was when I didn't know how to get the look on the set. I still don't but I've seen by a cinematographer that knows what he's doing. After getting his shots in camera, there is only the need for minor touch up in post. Of course you have to be able to communicate to your cinematographer the look you're going for. And if you don't know the look you're going for, perhaps you need to spend more time in pre-production and figure it out.
 
John_Hudson said:
Forget that in-camera appraoch; once yougo there you cant go back

I prefer flat & true no matter what look I have in mine from the beginning
Click to expand...

This gives you freedom for post. By locking your look in camera you're pretty much slaved to it.
 
Well just have to give my two cents on the subject :)

I really like getting the final look of the image in camera, or at least get as close as possible. But with me its not that much a question of preserving the quality of the image, its more to just get a good image.

Contrast for example heavily affects the composition of an image as well as the mood. Operating the camera myself while shooting my shorts, i really want to see it through the viewfinder as it will appear in the movie, because it lets me judge the image i am getting right away. I know when i got a shot and when not.
 
Shooting completely flat and leaving everything to post to me smacks of lack of a clear creative vision.

On a major film, there are numerous choices made prior to the shoot that will effect the look: film stock (can have a huge impact on the look), over/underexpose with push/pull processing, other processing options, high/low key lighting, production design. Not to mention camera placement, focal lengths, etc. Every choice impacts the image, and most are irreversible. You make a choice because it gets you what you know you want.

I'm not saying one way is better or worse, but c'mon, do you know what you want your film to look like or not? The look is part of the story, part of the emotion and characters. You need to know, you need to communicate it, and wrangle everything aspect of the production to deliver that. After that, you "correct" in post, meaning the image didn't quite hit the intent and you need to finesse it.

If you haven't done anything to craft a look waaay before you get to color correction, then you won't have a very good look.
 
Joshua Provost said:
Shooting completely flat and leaving everything to post to me smacks of lack of a clear creative vision.

On a major film, there are numerous choices made prior to the shoot that will effect the look: film stock (can have a huge impact on the look), over/underexpose with push/pull processing, other processing options, high/low key lighting, production design. Not to mention camera placement, focal lengths, etc. Every choice impacts the image, and most are irreversible. You make a choice because it gets you what you know you want.

I'm not saying one way is better or worse, but c'mon, do you know what you want your film to look like or not? The look is part of the story, part of the emotion and characters. You need to know, you need to communicate it, and wrangle everything aspect of the production to deliver that. After that, you "correct" in post, meaning the image didn't quite hit the intent and you need to finesse it.

If you haven't done anything to craft a look waaay before you get to color correction, then you won't have a very good look.
Click to expand...

I see your point about having the vision. I agree that you should know what it should look like before you shoot. Light, framing, testing and editing are all parts of the puzzle. There are certain looks that simply cannot be achieved in camera and therefore have to be done in post. I would say that if you're fairly consistent about your choices once you have made them and seem to be on the ball most of the time then affecting the camera look prior to post is a really good advice. If you're testing your Color Correction/Color Timing skills and want to go beyond the exposure and saturation levels then it would be wise to see what you can do with a flat picture. It really comes down to your preference and workflow.
Jgastelb
 
You must log in or register to reply here.
Back
Top