Jump to content

The secret to Red cams smooth highlights...


Recommended Posts

Could it be the Red Olpf contains a polarizer?  

After watching the below linked video... It made me wonder are those extra highlight details just down to polarization? Cause a polarizer would fix that on any cam and on another comparison, red was 1 stop darker than he Ursa mini pro. So it would make sense? Or am I crazy? 

 

Link to post
Share on other sites

I don’t think it makes sense.

There are many reasons why one camera would appear darker or have shifted colors in a camera test, the most common being user error. Here, I suspect that using a Variable ND filter caused some matching issues - you can get color shift and other effects when rotating it.

Usually for a side-by-side camera color/exposure test like this, it’s better to use high quality single strength ND sets like Mitomos or Rhodiums, or better yet avoid ND entirely and control exposure with the shutter and T-stop. We also have no data on how each camera was exposed and processed for viewing, which is usually the other point when large variables appear.

Red cameras have had documented OLPF issues, like the ‘red dot’ flares (eventually solved with the Skintone-Highlight OLPF and mitigated by the Standard OLPF). But polarization is not one of those issues. If it were, then you could not use a standard polarizer filter on the lens without creating a variable ND effect.

Link to post
Share on other sites

I'm not sure I was being too clear. What I'm speculating is, pherhaps red has a polarizer built into the Olpf. In that comparison it's the same lense and filter but the specular highlights on red seem very diffused Vs the Ursa. I think a polarizer would soften speculars in the same way no? Maybe it's just the sensor... 

Link to post
Share on other sites
6 minutes ago, David Pritchard said:

I'm not sure I was being too clear. What I'm speculating is, pherhaps red has a polarizer built into the Olpf. In that comparison it's the same lense and filter but the specular highlights on red seem very diffused Vs the Ursa. I think a polarizer would soften speculars in the same way no? Maybe it's just the sensor... 

I think I understood you correctly. As I said before, if there was a polarizer in the OLPF filter stack, then you would not be able to use a normal polarizer filter on the lens without getting a variable ND effect. Since that doesn’t happen on Red cameras, one must conclude that this isn’t the case...

Re: speculars

No, polarizers have no effect on specular reflections. Polarizers are used for car commercials all the time, and the speculars you get from the backlit metal is still there. 

If the filter is dirty, then it may have a diffusion effect, but that is the same for all filters. Note that the test has also listed a Black Promist filter in the filter stack.

Also, most variable NDs (being two polarizers stacked together), will soften your footage and can create multiple reflections. Here, there is essentially a triple stack of filters in front of the lens. This is generally why you avoid stacking filters for a camera test unless you’re testing the filters themselves. 

 

  • Upvote 1
Link to post
Share on other sites

Ohhh I see so now I understand you clearly. Well I guess I'm no closer to figuring out why the red highlights look much nicer there! Maybe the sensor simply has much more stops in the highlight range and/or the angle of light hitting the face was less harsh. 

Link to post
Share on other sites

Well, I have no idea. But I can tell you that in scene two, the Scarlet definitely looked nicer. No question about it. The Ursa was a bit 'hot', although the colours were a bit richer. I would have preferred that no filters were used, though.

Link to post
Share on other sites
6 hours ago, David Pritchard said:

Maybe the sensor simply has much more stops in the highlight range

I think this is the answer, it’s just a better sensor (as reflected in the higher price of the camera). 

Personally, I always thought the MX sensor was really good, even with its color issues. I shot some of my best work on it: 

https://satsukimurashige.com/sigur-ros

https://satsukimurashige.com/abbadon

 

 

Link to post
Share on other sites
57 minutes ago, David Pritchard said:

That's some beautiful work Satsuki! Yes I guess I'm clutching at straws for getting the same look on my Ursa Mini Pro lol. 

It might be nothing more complicated than what LUT was used for the transform from Log to Rec709. I’ve seen Alexa footage with atrocious highlights because of a poorly executed transform.

Link to post
Share on other sites
4 hours ago, David Pritchard said:

That's some beautiful work Satsuki! Yes I guess I'm clutching at straws for getting the same look on my Ursa Mini Pro lol. 

Thanks!

Link to post
Share on other sites
3 hours ago, Stuart Brereton said:

It might be nothing more complicated than what LUT was used for the transform from Log to Rec709. I’ve seen Alexa footage with atrocious highlights because of a poorly executed transform.

Also this. How you process and display the footage is as important as how you expose it. Grading can make a world of difference. 

Link to post
Share on other sites

Well after that above video and a discussion with another cinematographer who was slating the highlight handling of the Ursa I did a little test... Similar lighting setup to that lovely lady video above and I used a variable polarizer. Without the polarizer, specular highlights are pretty nasty with side lighting like that but a circular polarizer and blackmagic pro-mist 1/8th I personally think looks pretty nice now apart from the fat model/cameraman/me. Not sure if my grade is a little funky, playing with transforming the footage to both arri log and red log before grading.

Just to add, specular highlights are quite ugly and blown out without the polarizer, I didn't fully engage the polarizer however to leave a slight glow... I think it's a nice combo?  

 

Edited by David Pritchard
Link to post
Share on other sites

If you’ve found a recipe that works for you, then that’s fine. I would just be careful about conflating the results with the process. Correlation is not causation. 

Polarizers remove reflections from some surfaces at certain angles. If you have soft reflections on the cheek, then a polarizer can reduce those. It’s highly dependent on angle and won’t always work.

Also, a pola cannot remove specular reflections, which are hard reflections that you get off of highly reflective surfaces like water, glass, and metal. A soft surface like skin will scatter light and cannot reflect hard speculars unless there’s also sweat or oil.

A diffusion filter like Black Promist spreads the light around highlights and brighter midtones, creating a localized glow. This halation will reduce the contrast in bright areas, so when it’s combined with reducing bright areas in the frame, your highlights will appear less harsh. Camerafolk have been using lens diffusion to tame the highlights on video cameras for many decades.

I would suggest that you test each filter individually at different exposure levels, so that you can see what effect you’re getting from each. You could also try a color grade where you lower the highlights and upper midtones as a starting point and see if gets you to a similar place. That’s invariably one of the first things I always do when grading my own footage.

Link to post
Share on other sites

Great advice as usual! Ah I see I'm confusing the term specular highlights then. I just meant any hard ugly bright spec of highlight that has exploded way out of control. On skin, polarizers do a great job with them when placed 90 degrees approx from the camera. Actually it's too many filters stacked together though, I'd need to use a mattebox or slimmer profile filters to do this realistically. 

Link to post
Share on other sites

I think this has to do more with the color science employed by Red VS Blackmagic and that you'd have a better experience customizing whatever color transform you use to covert your log footage to working color space rather than using filtration to emulate the look.

If I am correct, this is essentially the same advice Stuart is giving.  By using filtration to compensate, you have to fight the transform to obtain your desired look and that can be less predictable than "fixing it up front".

 

Link to post
Share on other sites
29 minutes ago, Frank Wylie said:

I think this has to do more with the color science employed by Red VS Blackmagic and that you'd have a better experience customizing whatever color transform you use to covert your log footage to working color space rather than using filtration to emulate the look.

If I am correct, this is essentially the same advice Stuart is giving. 

Yeah, that is what I'm saying. The vimeo clip doesn't say what format was originally shot. The RED material was obviously some flavor of RED Raw, but there's no mention which 709 LUT was used. The BM footage might have been BRAW, or could have been ProRes, again no mention of what LUT was used.  Not all LUTs are created equal, and all this video really shows is that some LUTs work better with some cameras than others. I don't have a lot of experience with BM cameras, but I do remember thinking that their own 709 LUT wasn't that great.

With regards to using polarizers, as Satsuki has already pointed out, they only work at certain angles, which might be fine for a still frame, but as soon as the actor moves, the highlight will return.

Link to post
Share on other sites
4 hours ago, Stuart Brereton said:

The BM footage might have been BRAW, or could have been ProRes, again no mention of what LUT was used.  Not all LUTs are created equal, and all this video really shows is that some LUTs work better with some cameras than others. I don't have a lot of experience with BM cameras, but I do remember thinking that their own 709 LUT wasn't that great.

I think so too. I’m sure they’ve gotten better in the more recent models, but the Rec.709 LUTs on the Mini 4K and v1 4.6K were very clippy compared to their contemporaries. Kinda reminded me of how SmallHD monitors used to look in their first few generations.

Link to post
Share on other sites

The latest gen 5 out for bmraw is really quite low contrast. In the past it really killed a lot of the detail and often looked pretty ugly. I would be confident in saying the above was at best gen 3 color cause or the age but likely gen 1 or 2. 

Anyway thanks guys a lot of good info here 🙂

 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...