Using an 80/85 Filter to Balance

[Brendan H. Banks]

Been reading a good bit about different approaches and techniques used in camera and with lighting and noticed a lot of mentions of using an 85 or 80 filter when shooting to film. Now obviously film stocks are going to be set to their balance and digital can be manipulated on capture digitally, but it did get me thinking. I know that because of the use of bayer patterns, cameras can perform a little bit more cleanly when capturing daylight over tungsten (or I did know this, but please correct me if this is no longer the case). Wondering if there's any advantage to using an 80 filter when shooting in tungsten scenarios while balanced at daylight. Anyone have experience?

The biggest thing that made me begin thinking about this was Deakin's use of solely tungsten stock on No Country For Old Men and correcting back with an 85 filter during daylight scenes. I'm hoping to do some tests but curious if anyone has gone down this road before and cares to share their experience.

 

[Roger Singh]

I use the 80c filter when I have enough light because it takes up a stop of light. It lets you shoot the red cameras at 5000K.

I haven't tested on Helium yet, but on previous cameras (more so on Red One, and Epic MX than Dragon), it was practically mandatory. There'll be less blue noise, but for me, it was more about the skin tones, I always found that I had the best feel for them around the 5000-5500 kelvin.

Although, I felt that the Dragon Sensor handled tungsten lighting much better, so I didn't push for the 80c filter to be used all the time. I also think now with IPP2, tungsten lighting is handled better.

I will always use the filter when shooting green screen though.

I guess to sum it up, even though it's been getting less effective on newer sensors, it's still better to shoot the camera as close as possible to the base kelvin. And yes, I will have to remind myself to do the filter test the next time I have a camera test with a test subject.

 

[Matt W.]

A while back, the M and MX chip in particular were WAY cleaner with a blue filter. You'll notice the Informant was so orange looking because at that point the Red did not behave well at 3200K white balance so the tungsten scenes were processed at 5600K or so.

I don't think it's nearly as true anymore as it once was. The Epic forward seem a lot better but it's still true with it imo. But if you have the light to spare, go for it. I've done this before with the MX and the result was cleaner but MAN did we need a lot of light.

 

[Joe Cage]

Hello,

this is new information for me. Do you have a link to the 80/85 filter?

Thanks!

 

[Nick Morrison]

Or just expose Dragon ar 320 and saturate all your RGB channels that way.

 

[Brendan H. Banks]

Hello,

this is new information for me. Do you have a link to the 80/85 filter?

Thanks!

They're actually two separate filters. The 80 will bring you from tungsten to daylight, and vice versa on the 85. I don't have any good recommendations as I'm just learning about them now but a few pop up with a quick search. Here's a link to what I was reading and made me post: http://cinegleaner.postach.io/post/no-country-for-old-men

That entire site is a gold mine for lighting...especially if you're a Deakins fan.

 

[Matt W.]

Or just expose Dragon ar 320 and saturate all your RGB channels that way.

This seems like a better idea, particularly with the Dragon, which seems to have a nicer color balance with tungsten than the MX did. I would run your own tests... but as I said when I was shooting with an 80C filter it was with the MX. I haven't considered using one (much) since. The best looking Dragon footage I've seen is exposed around 320 ISO and I don't think there's any color filtration there. If you look at the monitors for most of the Taylor Swift videos that are shot on Red the ISO is unusually low and I think that's part of the reason they look so good.

Fwiw the fastest contemporary film stocks (Kodak's 500T) are tungsten balanced. So when you use an 85 filter, which has a filter factor under a stop, you're still working with a faster stock than 250D. So it's not a big issue as regards light loss. But blue filters on the other hand eat up a ton of light. I think an 80A is two stops. Chivo Lubezki I believe used to shoot 500T unfiltered during late day and use the DI to get the color balance so he could get the best low light with film. View camera (4x5) photographers I know will actually skip using all color correction filters and correct digitally because the color filters impart a sort of biased bronzed tint. In a lot of ways, more so than diffusion filters, color filters are a thing of the past. Now we just have white balance and the DI. :/ Despite all logic, color filters do not seem to be neutral and the "tint" observation holds. I find footage shot wth extreme blue filtration (day for night type stuff) will muddy up the shadows with blue and in fact shooting neutrally and adjusting white balance is better (in my opinion, I'm nowhere near capable enough to call myself a DP). So the "bronze" comment, even though I have not shot large format with an 85 filter, makes sense to me, unintuitive as it seems.

Hypothetically maybe there's somewhere to use an 80C or 80A now (tungsten balanced green screen stage?) but yeah... exposing to the right seems like the better strategy now that digital has evolved past its infancy. I'd like to shoot something on 16mm some day soon and in that case I'm renting photo filters, but I have thrown them out.

Interesting, I find Canon's 5600K colors pop more than their 3200K colors so I might consider using an 80A on a C300 except that the noise levels are similar at both white balances and you're not using that camera anyway if you're seeking the absolute most beautiful image.

Weirdly, the dvx100 was cleaner at 3200K than 5600K. Pretty anomalous. Most digital sensors perform best with daylight. Weird.

 

[Brendan H. Banks]

This seems like a better idea, particularly with the Dragon, which seems to have a nicer color balance with tungsten than the MX did. I would run your own tests... but as I said when I was shooting with an 80C filter it was with the MX. I haven't considered using one (much) since. The best looking Dragon footage I've seen is exposed around 320 ISO and I don't think there's any color filtration there. If you look at the monitors for most of the Taylor Swift videos that are shot on Red the ISO is unusually low and I think that's part of the reason they look so good.

Fwiw the fastest contemporary film stocks (Kodak's 500T) are tungsten balanced. So when you use an 85 filter, which has a filter factor under a stop, you're still working with a faster stock than 250D. So it's not a big issue as regards light loss. But blue filters on the other hand eat up a ton of light. I think an 80A is two stops. Chivo Lubezki I believe used to shoot 500T unfiltered during late day and use the DI to get the color balance so he could get the best low light with film. View camera (4x5) photographers I know will actually skip using all color correction filters and correct digitally because the color filters impart a sort of biased bronzed tint. In a lot of ways, more so than diffusion filters, color filters are a thing of the past. Now we just have white balance and the DI. :/ Despite all logic, color filters do not seem to be neutral and the "tint" observation holds. I find footage shot wth extreme blue filtration (day for night type stuff) will muddy up the shadows with blue and in fact shooting neutrally and adjusting white balance is better (in my opinion, I'm nowhere near capable enough to call myself a DP). So the "bronze" comment, even though I have not shot large format with an 85 filter, makes sense to me, unintuitive as it seems.

Hypothetically maybe there's somewhere to use an 80C or 80A now (tungsten balanced green screen stage?) but yeah... exposing to the right seems like the better strategy now that digital has evolved past its infancy. I'd like to shoot something on 16mm some day soon and in that case I'm renting photo filters, but I have thrown them out.

Interesting, I find Canon's 5600K colors pop more than their 3200K colors so I might consider using an 80A on a C300 except that the noise levels are similar at both white balances and you're not using that camera anyway if you're seeking the absolute most beautiful image.

Weirdly, the dvx100 was cleaner at 3200K than 5600K. Pretty anomalous. Most digital sensors perform best with daylight. Weird.

A lot in there.

That "bronze" cast is actually something I desire when it comes to creating a look. I'm lining up a feature right now to shoot and creatively, following that bronze cast is something I really would like.

I didn't know that the 80 filters ate about 2 stops, I thought it was only one...which definitely changes the game a little bit. Now I'm wondering if pulling my ISO and using my WB will result in something more desired than pushing ISO and working a filter. I'm such a pain for doing as much in-camera as possible when it comes to standard manipulation...but some things are best left for the color suite.

Thanks so much Matt!

PS I remember that from my old DVX (which I still have)!

 

[Matt W.]

Check out antique suede filters. Never used them but always liked the look associated with them. You can do most of this in the grade now and the pure unfiltered digital negative will give you the most flexibility. The muddy look to me just looks like saturating the shadows and adding some color in lift. Just muddier shadows with a color bias. But still worth a try using color filters. I think color correction rather than effect filters like 80 or 85 are sort of outmoded though imo now that there's digital that's actually good. But it is a subtly different effect from white balance. Imo not really in a good way. I actually find doing day for night worse when the DP uses a color filter because it muddies up the shadows, but of course doing day for night on a really orange image would be worse. :/ On Zodiac they were using either green or magenta filters to improve the SNR and when I used an 80C filter way back when it was to achieve the same thing with the Red when it was still not doing great things at 3200K. Now the sensors are already really really good, or the processing improved, or something. Just ETTR for low light seems the best option now.

I think 80C is one stop and 80A is two stops. 80C brings you to 4300K maybe and 80A to 5600K. Anyhow, that's the opposite direction (muddies up the shadows with blue, not orange, again pretty subtle and probably not what you want).

I love real diffusion (nets, classic softs, promist) over post diffusion by far, both for cost and look. But for color stuff I do think the grade and art direction are much more powerful and important. But worth playing around with.

 

[Marc Wielage]

Chivo Lubezki I believe used to shoot 500T unfiltered during late day and use the DI to get the color balance so he could get the best low light with film. View camera (4x5) photographers I know will actually skip using all color correction filters and correct digitally because the color filters impart a sort of biased bronzed tint. In a lot of ways, more so than diffusion filters, color filters are a thing of the past. Now we just have white balance and the DI.

I can tell you in the case of film, the blue emulsion layer does tend to get overloaded when you do this, and while you can sorta/kinda balance it out and match it, it's got an odd look to it. I don't dispute that when they're running out of light, yanking the filters is a last ditch effort to try to allow shooting another couple of takes before the sun is completely gone.

I can't say whether a digital camera reacts better with an 85 or an 85A in place under those conditions, but I can say that mixed lighting feels like more of a problem with digital than it ever did with film. It's particularly a challenge for baked-in files shot at a weird color temperature. Raw files we can generally improve quite a bit, particularly when they're shot at 3200°, provided none of the color channels are clipped.

 

[Matt W.]

I can tell you in the case of film, the blue emulsion layer does tend to get overloaded when you do this, and while you can sorta/kinda balance it out and match it, it's got an odd look to it. I don't dispute that when they're running out of light, yanking the filters is a last ditch effort to try to allow shooting another couple of takes before the sun is completely gone.

I can't say whether a digital camera reacts better with an 85 or an 85A in place under those conditions, but I can say that mixed lighting feels like more of a problem with digital than it ever did with film. It's particularly a challenge for baked-in files shot at a weird color temperature. Raw files we can generally improve quite a bit, particularly when they're shot at 3200°, provided none of the color channels are clipped.

I was in no way involved with Tree of Life or New World, but from what I understand both of those films had some very odd stuff done with intentional overexposure and shooting mismatched stocks without filtration and then correcting in post. But Lubezki is at the forefront of pushing things in the grade and I think his approach is basically what you mentioned, pulling the correction filter when there's not enough light late in the day. Very similar idea to the 4x5 photographer who advised me to skip filtration both for color and reciprocity error and fix the drum scan in Photoshop.

I've had those problems with mixed lighting, too. My friend had something shot on an AF100 graded by Stefan Sonnenfeld and I'm not sure how much I'm supposed to talk about the details of it but he apparently said among other things just what you wrote, color clipping in digital is one of the real bugaboos with lower end digital. But in my limited experience, the Alexa actually handles mixed light pretty well (even if its take on certain sources is different from film), even the prores, and we had some luck making big shifts in color temperature by denoising and using curves on the blue channel rather than just coloring with the gain in the 3 way to correct the overall cast and neutralizing the shadows a bit by adjusting lift in the opposite direction, which was my first approach. But then I had to work with SLOG2 XAVC footage with Sony's old gamut on the F5 and F55 and the chroma clipping with flares, street lights, etc. was simply beyond my skill set to correct, it was hard blocking in chunks of color. The F55 RAW footage I've been working with lately solves this when processed with the new SLOG3 gamut that clamps saturation as the Alexa does.

But I'm curious, when color is hard clipping (traffic lights, flares etc.), what approach do you take? All I can think of is somehow keying that area twice, once targeting the specific chunk of blown out color and pushing it to white, then soft keying the area around it with a mask and blowing it out a bit, too, and then using a global adjustment on the sat v Lum curve. But like.... beyond me... If you have any pointers about this I am curious because it's such an issue.

What do you mean by particularly when they're shot at 3200K though? Isn't the color temperature not baked into raw? This is a discussion that I think is really interesting because hard chroma clipping is the first thing that gives away digital to me but the discussion goes deeper than that into how various sources are rendered or even into how a sensor clips at various color temperatures and what kind of gain is used in each channel. Even color negative and slide film are different in this respect IMO and one reason digital and slide film have richer National Geographic colors than traditional color negative film used in cinema. Imo Arri is actually doing a great job emulating color negative, Red's current color is pretty good, too, but less slavishly color negative film-like. Have you seen Yedlin's infamous Nuke script demo where he shoots film and Alexa footage side-by-side and runs them through a Nuke script with no other correction? It's pretty amazing, the grain/color/halation/tonality is all really superbly matched but the biggest discrepancy is in how they render mercury vapor. It's a correctable difference on the hue v hue line to a perfect match, but that was the one thing the script doesn't procedurally at least get close. There are other discrepancies, including skin tones, but this is the one that stands out.

 

[Björn Benckert]

See the RGB channels as cups filling up with color when you expose. the less color a cup has the more noise will be introduced. Then the cups can also fill upp to the rim and loose color information. So if you shoot in an unbalanced environment that has more of one of the three colors than the other then. A you risk of having a cup empty or not enough filled to stay out of noise, or 2 you might end up overfilling one and loose info in of the cups before the other two are getting enough to not cause noise.


Now if you shoot tungsten your image will be slanted towards the warmer tones, that means that most likely your red channel will fill up and burn before blue and green are properly exposed. So simply shooting tungsten without a 80 filter to balance it out you got less range in your camera. As with range I mean latitude between noise and burn out.

Same thing, when you shot under water or such you need to balance the other way. Push things towards the warmer tones or your blue channel will burn long before the other two channels get proper exposure.


But sure the cameras of today got plenty of latitude and you might say it´s all fine and good and you can just balance in post. And often that works just fine. But then thats a bit like saying I just shoot Compressed HD because that works just fine.... Simply when you dont use all the 3 channels of color in the camera you got less resolution, use your NLE and check your video signal one channel at the time. the Blue channel will always be suffering the most as thats how the Cmos is built so it´s actually worse to shoot tungsten than shoot something that is slanted in the other direction.

Then all filters cut out more or less light. So when shooting low light you need to decide if the balance is more worth than the exposure. I say for a lot of situations it is. So better filter more often than not when not shooting in daylight temped light.

 

[Matt W.]

See the RGB channels as cups filling up with color when you expose. the less color a cup has the more noise will be introduced. Then the cups can also fill upp to the rim and loose color information. So if you shoot in an unbalanced environment that has more of one of the three colors than the other then. A you risk of having a cup empty or not enough filled to stay out of noise, or 2 you might end up overfilling one and loose info in of the cups before the other two are getting enough to not cause noise.


Now if you shoot tungsten your image will be slanted towards the warmer tones, that means that most likely your red channel will fill up and burn before blue and green are properly exposed. So simply shooting tungsten without a 80 filter to balance it out you got less range in your camera. As with range I mean latitude between noise and burn out.

Same thing, when you shot under water or such you need to balance the other way. Push things towards the warmer tones or your blue channel will burn long before the other two channels get proper exposure.


But sure the cameras of today got plenty of latitude and you might say it´s all fine and good and you can just balance in post. And often that works just fine. But then thats a bit like saying I just shoot Compressed HD because that works just fine.... Simply when you dont use all the 3 channels of color in the camera you got less resolution, use your NLE and check your video signal one channel at the time. the Blue channel will always be suffering the most as thats how the Cmos is built so it´s actually worse to shoot tungsten than shoot something that is slanted in the other direction.

Then all filters cut out more or less light. So when shooting low light you need to decide if the balance is more worth than the exposure. I say for a lot of situations it is. So better filter more often than not when not shooting in daylight temped light.

Eloquent description. I'm curious because you're working both in post and on set on some high end content, how often do you see 80A and 80C filters used? Either on Red or Alexa? What ISO do you see the Dragon rated at most often? I recall having very good luck using an 80C filter on a Red MX but figured this practice was outmoded. I have not shot with the Dragon yet, only worked with Dragon footage in post, but I understand that it behaves better with tungsten light.

I think part of the advantage of color negative film is the non-linear knee. So the cup is maybe shaped more like a cordial glass when you shoot color negative. To me, the highlights are still where film and digital look very different. With film there is a warm halation and color rolloff that is saturated without hard chroma clipping still seems unique to film.

I've long admired the grading in this video:

https://www.youtube.com/watch?v=sxDdEPED0h8

Hard to tell when it's film and when it's digital.

 

[Björn Benckert]

Eloquent description. I'm curious because you're working both in post and on set on some high end content, how often do you see 80A and 80C filters used? Either on Red or Alexa? What ISO do you see the Dragon rated at most often? I recall having very good luck using an 80C filter on a Red MX but figured this practice was outmoded. I have not shot with the Dragon yet, only worked with Dragon footage in post.

I think part of the advantage of color negative film is the non-linear knee. So the cup is maybe shaped more like a cordial glass when you shoot color negative. To me, the highlights are still where film and digital look very different. With film there is a warm halation and color rolloff that is saturated without hard chroma clipping still seems unique to film.

I've long admired the grading in this video:

https://www.youtube.com/watch?v=sxDdEPED0h8

Hard to tell when it's film and when it's digital.

Yes film is a great format in that sense. But I think digital has also quite a few things going that film possibly does not have.

But then the absolute same rules apply to film as well. But shooting film you also got tungsten and daylight filmstocks to play with something we dont have when shooting digital. I guess red could create a more tungsten oriented olpf but thats nothing I seen yet. The H2O olpf is I guess a bit of an opposite of that.

Big productions are generally in total control of their lightning so then color filtering in camera is not done so often. And when looking to get best possible image out of the camera then normally it´s 5600k lamps and good such lamps so then no color filtering.

But I think if people (DP´s) would look at the RGB channels as black and white clips next to each other more often then I think they would bring out the color filters more often. It´s easy to understand that picture quality will not be stellar when you see the separated blue chanel from an underexposed tungsten image... usually they look so bad so you wonder how that could could actually be part of your image.

Now with the in camera luts it´s quite east to make seperate channel luts so you can see your exposure in RGB. Something I think should be implemented as hotkeys, think that would save some from underexposure mistakes.

 

[Matt W.]

Yes I agree digital is far ahead technically, but film looks better in some ways. The highlights in particular–the non-linear knee, the saturation roll off, the halation, the warm bloom and tint, etc. I think another is that colors are more saturated looking with film. I put digital and film color checker charts shot on the same day (it was 5230 and Red MX, I believe) side-by-side and the lines on the vectorscope from the film are more tightly concentrated. Once corrected to equal saturation the colors more vivid looking on film. Red, in particular, ironically, looks very rich and vivid with film but also soft and lush, better with film than digital. Compare the reds in the current Alexa-originating Twin Peaks with the original, which was shot on film, particularly the scenes in the bar at the end. The colors are richer on film.

The red-sensitive layer being at the bottom of the emulsion may have something to do with these differences... at least I always guessed as much.

I like your thinking, but unfortunately see very few people following it. For vfx work in particular that seems helpful. In theory it would maximize dynamic range shooting that way. I'm used to seeing terrible looking blue channels, and have just come to accept it.

Another question, if you don't mind: when you do device comps on green screens, how do you like for the screen to be exposed? I believe the conventional wisdom was 18% gray for green screens but most green screens I see for tvs and iPhones etc. are far far brighter than that.

What do you do when the DP wants to use the screen as a light source? I have seen all sorts of stuff.... shooting live screens to be replaced later (have to paint reflections back in), green screens that are simply much too bright (spill suppressing is a pain), white or gray screens, which seem like a real struggle for keying because even though a luma key works you get these very strong matte lines on the edges of out of focus hair and have to choke the key far too much to get a pleasing result, or even LEDs taped to the screen, which are problematic.

Just wondering. This seems to be a surprisingly common topic of debate. I talked with an old ILM guy (worked on Jedi, etc.) about this and had tried many things but settled on "green screen no tracking markers" simply because it was the easiest and safest request to communicate to production. Other than the poor comps on the Taylor Swift ads, I believe Apple's screens are live, and I think the best looking.

 

[Björn Benckert]

Yes I agree digital is far ahead technically, but film looks better in some ways. The highlights in particular–the non-linear knee, the saturation roll off, the halation, the warm bloom and tint, etc. I think another is that colors are more saturated looking with film. I put digital and film color checker charts shot on the same day (it was 5230 and Red MX, I believe) side-by-side and the lines on the vectorscope from the film are more tightly concentrated. Once corrected to equal saturation the colors more vivid looking on film. Red, in particular, ironically, looks very rich and vivid with film but also soft and lush, better with film than digital. Compare the reds in the current Alexa-originating Twin Peaks with the original, which was shot on film, particularly the scenes in the bar at the end. The colors are richer on film.

The red-sensitive layer being at the bottom of the emulsion may have something to do with these differences... at least I always guessed as much.

I like your thinking, but unfortunately see very few people following it. For vfx work in particular that seems helpful. In theory it would maximize dynamic range shooting that way. I'm used to seeing terrible looking blue channels, and have just come to accept it.

Another question, if you don't mind: when you do device comps on green screens, how do you like for the screen to be exposed? I believe the conventional wisdom was 18% gray for green screens but most green screens I see for tvs and iPhones etc. are far far brighter than that.

What do you do when the DP wants to use the screen as a light source? I have seen all sorts of stuff.... shooting live screens to be replaced later (have to paint reflections back in), green screens that are simply much too bright (spill suppressing is a pain), white or gray screens, which seem like a real struggle for keying because even though a luma key works you get these very strong matte lines on the edges of out of focus hair and have to choke the key far too much to get a pleasing result, or even LEDs taped to the screen, which are problematic.

Just wondering. This seems to be a surprisingly common topic of debate. I talked with an old ILM guy (worked on Jedi, etc.) about this and had tried many things but settled on "green screen no tracking markers" simply because it was the easiest and safest request to communicate to production. Other than the poor comps on the Taylor Swift ads, I believe Apple's screens are live, and I think the best looking.

To me comparing different formats and their qualities is difficult. Saying film has more color than digital? How do you compare the two? the digital image need debayering and grading to even be visable, same goes for film it needs to be developed and you ned to strike a print to view it, now did you do a lowcon or hi con print or what setting did you have when transcoding your digital file. Sure one can say when all things are optimised one might look better than the other but as they are different they simply dont perform their best in the same setups. You can even narrow it down and say you can not compare two film stocks and say that one is better than the other if you dont also add for what that they are better. And when you compare twin peaks with twin peaks it´s also different teams making different decisions when they shoot so basically the test samples are more than a little contaminated, to me to a degree so you can not really judge anything out of it.

For replacing screens it´s all depending on what you do and what the scene is, today you can get pretty much any clip rooted for about 2.5USD per frame overnight. So it makes a lot of sense not to use blue screens etc. Same with taking points planar tracker got better and you can 3D track much easier now than before but still for sure if you can have small tracking dots that is easily removable then that gives a better track than when you have nothing to track on. So again depends what you shot, there is no general solution for this things. It´s a bit like asking what lens is the best to use... well it depends what you want to shoot.

 

[Blair S. Paulsen]

IMO, with Dragon and Helium, especially via the IPP2 pipeline, color filtering on camera is not necessary in most situations. M/MX is a different story.

When the color temp of the scene is particularly low, under 3000° with funky CRI for example, I mount a Schnieder 1/4 CTB 4x5.65 filter. It eats roughly 1/2 stop, but per Björn's analogy, it fills the cups more equally which I find is a worthwhile trade-off in terms of what I can pull from the image in grading. I should also mention that traditional filters for film like the 80A/85/etc have a MIRED shift specifically targeted to the properties of film. For digital, I find the CTB filtration manages to push me up the Kelvin scale enough to improve channel balance while eating less light. YMMV.

Cheers - #19

 

[Björn Benckert]

IMO, with Dragon and Helium, especially via the IPP2 pipeline, color filtering on camera is not necessary in most situations. M/MX is a different story.

When the color temp of the scene is particularly low, under 3000° with funky CRI for example, I mount a Schnieder 1/4 CTB 4x5.65 filter. It eats roughly 1/2 stop, but per Björn's analogy, it fills the cups more equally which I find is a worthwhile trade-off in terms of what I can pull from the image in grading. I should also mention that traditional filters for film like the 80A/85/etc have a MIRED shift specifically targeted to the properties of film. For digital, I find the CTB filtration manages to push me up the Kelvin scale enough to improve channel balance while eating less light. YMMV.

Cheers - #19

Yes, better cameras with better DR has more tolerance. But again you toss away resolution when not having properly exposed photo sites and letting a tinted image hit the sensor and then balance it in post is easily narrowing down the resolution by a third.
For that reason, even stuff that should be heavily red, blue or green looking benefit from being shot more balanced and then tinted in post.... I done a fair load of ketchup commercials and if you want that ketchup to really pop you should not shoot balanced but very blue greenish tinted and then balance back in post, then you get all the details in the red ketchup preserved... most DP´s dont grasp the logic of that, but many have asked me why the picutre looks so low res when they shoot in coloured light...

 

[Brendan H. Banks]

Yes, better cameras with better DR has more tolerance. But again you toss away resolution when not having properly exposed photo sites and letting a tinted image hit the sensor and then balance it in post is easily narrowing down the resolution by a third.
For that reason, even stuff that should be heavily red, blue or green looking benefit from being shot more balanced and then tinted in post.... I done a fair load of ketchup commercials and if you want that ketchup to really pop you should not shoot balanced but very blue greenish tinted and then balance back in post, then you get all the details in the red ketchup preserved... most DP´s dont grasp the logic of that, but many have asked me why the picutre looks so low res when they shoot in coloured light...

HA! The ketchup commercial is perfect example of that approach, I had been having some trouble understanding prior to that. And I love the cups metaphor...I'll be using that in the future

So I totally did not think that this thread would get any replies...and now it's an incredible discussion of approaches, tech, and artistry. Thanks everyone, this is my new favorite thread! But please don't let this thanks derail....

 

[Matt W.]

To me comparing different formats and their qualities is difficult. Saying film has more color than digital? How do you compare the two? the digital image need debayering and grading to even be visable, same goes for film it needs to be developed and you ned to strike a print to view it, now did you do a lowcon or hi con print or what setting did you have when transcoding your digital file. Sure one can say when all things are optimised one might look better than the other but as they are different they simply dont perform their best in the same setups. You can even narrow it down and say you can not compare two film stocks and say that one is better than the other if you dont also add for what that they are better. And when you compare twin peaks with twin peaks it´s also different teams making different decisions when they shoot so basically the test samples are more than a little contaminated, to me to a degree so you can not really judge anything out of it.

For replacing screens it´s all depending on what you do and what the scene is, today you can get pretty much any clip rooted for about 2.5USD per frame overnight. So it makes a lot of sense not to use blue screens etc. Same with taking points planar tracker got better and you can 3D track much easier now than before but still for sure if you can have small tracking dots that is easily removable then that gives a better track than when you have nothing to track on. So again depends what you shot, there is no general solution for this things. It´s a bit like asking what lens is the best to use... well it depends what you want to shoot.

That's all totally fair and true. I still think color generally appears more subjectively saturated with film stocks than with digital for a given scene, and the only technical correlation I've found is a tighter point cloud on the vectorscope. Then again, of course, this varies tremendously. But trying to get digital to have that same rich saturation is more difficult for me, particularly with the color red (the ketchup example might be the trick though). I started shooting stills with a Foveon camera and am a bit more satisfied with those colors and how the highlights clip.

Can you PM me the name of the place you to for roto? It might be the same place I know of but that's a good price. Would definitely appreciate it. Do they send back silhouette files or hicon image sequences? I think I'm usually paying more than that.

And yes, I think black screens are generally most artists' preference, but what do you do if the DP wants the light from the screen to play as a source? And when you do use green screens, what exposure do you usually prefer for device comps? I know this is all a matter of horses for courses, but you can't always be on set for this stuff and sometimes you have to hand down whatever guidelines you can, especially on lower budgeted stuff like network tv and Netflix and lower budgeted ads the best you can do is hand down guidelines if they can't afford a sup on set. For instance it was a very highly rated network tv show with a rather larger vfx budget where the sup I talked with wasn't able to be on set so he just went asked for blank green screens and was done with it. Not all of us get to work on high end ads all the time, even if we'd like to.

In the old days 18% gray was the starting point for green screens from what I've seen, so I'm just wondering what the guideline would be for device comps, in your opinion.

That ketchup trick might be brilliant. I will remember that. Very very interesting.