• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

Boost grain in camera or in post?

superamazing

superamazing

This is more of a hypothetical for me, since I don't have access to post-production software sophisticated enough to intelligently boost grain, but it occurs to me that the capabilities of video camera dB boosting are severely limited. For one, it must happen in real-time, and with only on-board computing power.

My question is, does changing the dB on your camera do anything beyond artificially brightening the image? Couldn't I just do that in post, with more control?

And some related questions (since I'm in the question mood): What are best the options for brightening the image in Post, anyway? Are there other options than just dragging the slider marked "brightness"?

What about raising a DSLR's iso settings? Is that any different? Is there any difference between raising the dB on a video camera or the iso on a DSLR?

I guess the best solution is to seek out more light.
 
Sort by date Sort by votes
Raising ISO and gain are the same thing. You can do it in post, but the results may not be as good. At some point anything that is too dark will just get clipped off at black - raising it in post won't bring the detail back in the shadows, it will just turn the black areas grey. So you want to expose so that any details you want to keep in the shadows are bright enough to be recorded by the camera. The best option is more light (or a fast lens that lets in more light), second best is raising gain/iso, last is raising brightness in post.

Also remember that the recorded signal from your camera is compressed pretty highly. Video compression works by throwing away information that the human eye isn't likely to notice, and one of the areas it can do that in is shadows. When you take that compressed info into your computer and brighten it up you are raising detail from the deep shadows into the visible range - and you'll often find blockyness and other compression artifacts become more visible. If you increase the exposure in the camera those areas are boosted before they are compressed and the detail isn't thrown away.

There are much, much better options than the brightness slider - it effects the entire image and will just make stuff look grey. Depending on your software capabilities I would look for levels or curves adjustments that let you adjust the brightness of parts of the image without changing others, or look at color correction tools like Colorista or Color which give you extensive control over every aspect of your image.
 
Upvote 0 Downvote
Add light, open your iris... expose correctly in camera without adding gain as that will increase the noise in the darks (just like increasing the darks does in post).
 
Upvote 0 Downvote
ItDonnedOnMe said:
Raising ISO and gain are the same thing. You can do it in post, but the results may not be as good. At some point anything that is too dark will just get clipped off at black - raising it in post won't bring the detail back in the shadows, it will just turn the black areas grey. So you want to expose so that any details you want to keep in the shadows are bright enough to be recorded by the camera. The best option is more light (or a fast lens that lets in more light), second best is raising gain/iso, last is raising brightness in post.

Also remember that the recorded signal from your camera is compressed pretty highly. Video compression works by throwing away information that the human eye isn't likely to notice, and one of the areas it can do that in is shadows. When you take that compressed info into your computer and brighten it up you are raising detail from the deep shadows into the visible range - and you'll often find blockyness and other compression artifacts become more visible. If you increase the exposure in the camera those areas are boosted before they are compressed and the detail isn't thrown away.

There are much, much better options than the brightness slider - it effects the entire image and will just make stuff look grey. Depending on your software capabilities I would look for levels or curves adjustments that let you adjust the brightness of parts of the image without changing others, or look at color correction tools like Colorista or Color which give you extensive control over every aspect of your image.
Click to expand...

Exactly. Anything that can be done in camera on set, do it then. There are fixes in post, always, but that's all they are, fixes for not having done it perfect in camera. And hey, often these fixes are great, but they don't compare to the real thing.

At the same time, don't waste half a day on set trying to set the gain perfectly. Gotta be a balance. Good technicians will help with that though.
 
Upvote 0 Downvote
Plays into the "styles" question as well (flat, technicolor cine, etc...) the goal being to capture as much detail as possible in camera because if the data isn't there it isn't there, bringing up a crushed black won't reveal the detail you lost, it isn't there, ditto bringing down a blown out highlight.
 
Upvote 0 Downvote
It depends, in dark areas it (Gain or ISO) will bring in more detail at the cost of noise. I was just pointing out the picture style can be a key to getting more detail to work with in post. It doesn't really expand the dynamic range, but it has the same effect. Highlights don't blow out as fast, blacks don't drop off as fast.
 
Last edited:
Upvote 0 Downvote
superamazing said:
But gain in the camera doesn't give you any new detail, does it? It just brightens the details it already captured, right?
Click to expand...

Once you are changing things in post you can only push around whatever you ended up with in the final file. Using gain and picture styles in camera lets you push a much larger amount of data around before you reduce and compress that data to the final file.

There's a processing pipeline inside your camera that has several steps that all impact what ends up in your final video file. The sensor produces an analog signal in response to light striking it, that signal is then digitized - often at a 12 or 14 bit resolution. Then that signal is processed to extract the individual color channels (in single chip cameras like DSLRs), a gamma curve and misc. color settings are applied (a 'picture style'), the data is reduced down to 8 bits per channel, then the resolution of the chroma channels is reduced (4:2:0, 4:1:1, etc sampling) and finally it is compressed and saved to a file.

So if you apply gain to the analog signal in the first step - before the data is digitized - you can boost up details which might otherwise fall below the digitizer's lowest threshold and be discarded. So that's the only way to 'gain' information that isn't in the digital signal - of course it also boosts the noise at the same time.

At the step where you reduce the data from 14 to 8 bits you have to throw away a lot of detail - and what you throw away is affected by the gamma encoding of the picture profile. Boosting the shadow detail above the bottom of the gamma curve pushes it up into a range where more of that detail will be retained in the bit reduction step. So you don't get new detail, you just keep more of the original detail. You can also change the gamma curve to reduce shadow compression - this is exactly what the new technicolor profile does for the Canon DSLRs. However, whenever you are working with an 8 bit final file using a profile like that means you are trading some midrange details for greater shadow and highlight details.

Infested Films said:
Exactly. Anything that can be done in camera on set, do it then. There are fixes in post, always, but that's all they are, fixes for not having done it perfect in camera. And hey, often these fixes are great, but they don't compare to the real thing.
Click to expand...

I think it's important to make a real distinction between fixing things in post and making full use of post in your overall workflow. The goal shouldn't be to get it perfect in camera or else fix it in post - it should be to get everything in camera that you need to achieve the perfect final image in post. This often means you end up with images coming out of your camera that don't look particularly good at first - but that's done deliberately, and it's much different than finding out your footage looks bad because you didn't know what you were doing and then hoping to fix it later.
 
Upvote 0 Downvote
You must log in or register to reply here.
Back
Top