Comparing Film to Red , Tests and Example images

[Sebastian_A]

Why are the first 2 frames so noisy ? In general frame grabs from a blu ray movie are pretty clear, or if they are grainy - that grain looks different than what seems to be a fine noise in your frames.

And I'm also curious in what circumstances the grain/noise will increase in film ? The same as with digital ? Low light ?

 

[Bruce Allen]

Another thing I'm noticing is that, despite the increased latitude and highlight handling of film in the first shot, the RED still manages to look essentially the same from a colour/latitude point of view, but again, the motion might make one stand out over the other. Definitely matchable, though.

Can you match the skintones in the first shot? When you punch up the contrast, there is a really big difference in the color look.

Les, is it OK if we attempt to match them or is that not the point of this exercise?

Bruce Allen
www.boacinema.com

 

[Les Dittert]

Film has grain. The film frames of the boy are color corrected to up the shadow details, as the guy was inside of a comparatively dark tent with full sun outside in the background.
My quick color corrections were to get the frames looking fairly close in contrast and look. It was not my goal to match the colors exactly as that would require a 3D lut.
I did the comparison to look at image details and maybe dynamic range.

Why are the first 2 frames so noisy ? In general frame grabs from a blu ray movie are pretty clear, or if they are grainy - that grain looks different than what seems to be a fine noise in your frames.

And I'm also curious in what circumstances the grain/noise will increase in film ? The same as with digital ? Low light ?

 

[Gabe Conroy]

Why are the first 2 frames so noisy ? In general frame grabs from a blu ray movie are pretty clear, or if they are grainy - that grain looks different than what seems to be a fine noise in your frames.

If I'm not mistaken, it's standard practice to degrain film shots and then re-add grain as desired.

 

[Robert Ruffo New]

If I'm not mistaken, it's standard practice to degrain film shots and then re-add grain as desired.

Not standard, just sometimes done (actually rarely)

 

[David Mullen ASC]

Sometimes blu-ray masters of 35mm movies have some noise reduction applied, and often it's one of the things that blu-ray reviewers comment on, the general feeling being that it is better to see some grain if the original had it.

For HD that gets broadcast, sometimes noise reduction is more heavily applied just because a clean image compresses better.

I don't understand why someone thinks that the film frames in this thread look grainy, it looks like a normal amount of grain to me. Often color-correction decisions will hide some of the grain in shadows when they end up crushing the blacks somewhat for more contrast.

It's not standard practice to degrain and then regrain shots. It's done for efx shots sometimes, especially if they have to combine film scans and CGI elements, and it has been done on some restorations of old movies when the original grain was too variable due to elements aging at different rates. But in general it's not done.

 

[Daniel Pearson]

Thanks for the frames Les. I love my Scarlet, but the film grabs just feels richer by comparison.

 

[Sebastian_A]

I don't understand why someone thinks that the film frames in this thread look grainy, it looks like a normal amount of grain to me. Often color-correction decisions will hide some of the grain in shadows when they end up crushing the blacks somewhat for more contrast.

I suspected the reasons mentioned - the subject being slightly in low light and also increasing of the shadows by cc, but I wasn't sure so I asked. And also being used with frame grabs from blu ray movie masters - which are heavily post processed.
Also if I'm not mistaken other film stocks and especially older movies have a different grain type. Perhaps bigger and not so fine. Probably because older film stock was coarser ? Larger grain particles ? I'm not familiar with film stocks so perhaps I said something stupid.

 

[David Mullen ASC]

Modern T-grain is more uniform, tighter, and more even -- older film stocks had more random, coarser grain. But older film stocks were slower so they weren't too grainy unless underexposed and/or push-processed. But many remasters of old movies aren't able to use the original negative, so you also have some grain from the duplicate elements used, whether color interpositives or dupe negatives, or b&w separations. And whether those dupe elements were originally made on an optical printer or contact-printed will effect the graininess. Plus older movies often contain an optical printer dupe element for an effect (fades, dissolves, titles, visual efx) and that shot will be grainier than the rest of the footage.

This test was done using 50D, the slowest and finest-grained stock out there, but it's not grainless.

 

[Les Dittert]

David,
I picked 5203 because of it's fine grain, but was a little surprised when I saw more than I expected!
It is a nice grain though.

 

[Les Dittert]

Thanks for the frames Les. I love my Scarlet, but the film grabs just feels richer by comparison.

OK Daniel, here is something to make you feel better :

First, the film image of the boy, then the RED image of the boy, but degraded ( softer,grain, halation simulation ):

RED image degraded to look like film at 4K image :

 

[Marc Wielage]

I started working at ILM in 1986, to take over the Pixar laser film recorder/scanner when Pixar co. was sold by George. That's where I was. We also scanned before the Cineon scanner was done. If memory serves, it was an Ikonix line array camera that Scott S. bolted to an optical table behind the laser machine.

Yes, we did tests at Cinesite in LA on THX 1138 in 2004 with Fred Meyer from ILM and compared the Cineon scanner, a Spirit 2K scanner, and the ILM scanner, and everyone was appalled how bad the ILM scanner was. There were some huge issues with 2-perf Techniscope, which is why Lucasfilm opted to do the scans outside their own facility. It was clear that ILM had never really updated the scanner from 20 years earlier. My memory is that there were "patchy" issues with horizontal lines in the middle of the screen, where the CCD segments did not match in terms of density.

And I was the digital opticals supervisor on Back to The Future 2, doing the first ever fully digital composite. Before that ILM was filming out mattes on hicon and using the optical printer to comp. the shots.

An incredible job. I'm always quick to credit ILM for forging the groundwork for scanning 25 years ago. I can recall reading the American Cinematographer article on BTTF2 and they threw in a little mention of "wire removal," and I just about fell off my chair. This was very heady stuff back then. The line in the story that stood out to me was, "we got so good at wire removal, we could almost literally use a 1-inch rope on Michael J. Fox, and we could easily remove it." A huge achievement for that time.

Again, can you tell us what you used to scan this film test in 2013? Also, if you're in LA, you really should check out the ASC StEM test, which does a fairly exhaustive comparison between film and digital. (We did similar tests when I was at Cinesite and at Technicolor over a 10-year period.) I'd love to see the ASC update this with a Red / Alexa / Sony / Film demo, which I bet would be very interesting... but my guess is that the differences would not be as great as people think.

 

[Les Dittert]

Marc,
The scans you see were from a custom area array scanner, using a 4.8K Kodak sensor. It was also scanned on a Northlight at 6k, but they looked pretty much the same.... no further detail was showing up.

I'd like to see what's up with the ASC test. They charge for seeing the test results don't they ?

 

[Gabriele Turchi]

halation simulation

Hi les ,
how did you simulate the halation ?
thanks
g

 

[Martin Stevens]

Hello Les,

Can you tell us what program you used etc. to add grain the the RED image?
Did you use "Film Convert" at all?

I too am interested in how you simulated the halation.

 

[Rudi Herbert]

I think the grain and general "feel" of film can be matched fairly well on the digital file, the problem is with the actual color rendition, with all the green/yellow splotches on the skin and red/magenta ones on the sweater, pretty nasty stuff that is, still, the one downfall of digital.

 

[BRANDON JAMESON]

I think the grain and general "feel" of film can be matched fairly well on the digital file, the problem is with the actual color rendition, with all the green/yellow splotches on the skin and red/magenta ones on the sweater, pretty nasty stuff that is, still, the one downfall of digital.

We don't see many of these artifacts from images produced with digital sensor of the F65.

 

[Les Dittert]

Halation sim:

Basically you want to simulate the 'blooming' of bright things, where light tends to go sideways in the emulsion.
What I did was fast and a partial simulation. Look what it does to the hair against bright BG, compared to the film version. I think it gets 80% there.

Pull a luma matte on the brightest parts of image.
Blur and dilate this matte a little ( a few pixels )
Use matte to color correct the bright areas, lifting the mid-brights up in intensity.
That's it. Mind you I also blurred the starting RED image a bit to kill the high frequency details as well.
This was all done in Fusion. Fusion grain as well.

I made a quick attempt to fix his skin color as well. Keyed it and CCed a little to get closer to film.
Just like many of the colorists are doing in RED feature production

Hello Les,

Can you tell us what program you used etc. to add grain the the RED image?
Did you use "Film Convert" at all?

I too am interested in how you simulated the halation.

 

[Daniel Pearson]

OK Daniel, here is something to make you feel better :

First, the film image of the boy, then the RED image of the boy, but degraded ( softer,grain, halation simulation ):

RED image degraded to look like film at 4K image :

Hi Les,

Thanks, much closer. I do think film still has an edge, but with grading they will both produce amazing results.

 

[Vadim Bobkovsky]

That's just MX's noise in the blue channel, I'm sure more sensitive chips like Dragon will perform better and be a nicer starting point to emulate film look in post with all the highlights blooming, grain and slight color shift. It's always better having the source as clean as possible.