Film grain simulation vs Film grain scans

[donat]

Hi,

I don't have access to film scans so I'm only using the film grain simulation node. But I'm wondering if the film grain scan node does a better job.

So I'd like to hear some comments from people having used both methods of applying grain.

Regards.

Donat

[Hugh]

If you're trying to match the grain on a plate that you've already got, then I'd imagine that the simulation would work better.... I've not used Nuke enough, though, so I couldn't say...

If the grain scan doesn't match the grain in your actual plate, then I'm not sure how much control you'd have in tweaking it.... One thing that you should get really good at is matching grain in a plate... There are various tricks for this - I usually like to blur my plate and then add grain to that - and then compare my grained blurred plate with the original in various different areas of flat colour.

[5sense]

Hi,

I don't have access to film scans so I'm only using the film grain simulation node. But I'm wondering if the film grain scan node does a better job.

So I'd like to hear some comments from people having used both methods of applying grain.

Regards.

Donat

you can even use a mult node to control your grain, this is only if you have a stock grain to apply to your plate. basically, with each channel, R, G, B, you can control the intensity of the grain. when adding grain either simulating trough a proper node the best way would be in log space. another techniques that I use and which works quite nicely its to have for each channel a grain node. this depends especially of what kind of plate you have.

I use sometimes another trick like this:
at the end of your script use a gamma node and put the value at 2.2 and see the all channel separately and this is really helpful for matching the size and intensity of the grain