icarusfilm
Veteran
We can switch his camera into full auto while he sleeps make all his footage unusable the next day.
FS7: Custom LUT usefulness
- Thread starter JPNola
- Start date
I'm confused by Liam's post on the zebras : "Yeah, they only act on what is fed into the LCD, so you can either have the set for whatever mLUT you are using or, like me and you shoot a lot of HFR, have them set for Slog3".
So - if you set them for Slog 3 - doesn't that just confuse you to turn on the mlut in the vf? Or do you also then turn zebra's off once viewing mlut?
You can have the waveform on the FS7 show you the underlying slog or the mlut - but zebras will always show you or correlate to whatever you're seeing in the viewfinder. In other words, if I set my zebra 1 to 68 - then switch on the mlut to viewfinder - that 68 number will not help me expose for 90% white while viewing the mlut.
So - if you set them for Slog 3 - doesn't that just confuse you to turn on the mlut in the vf? Or do you also then turn zebra's off once viewing mlut?
You can have the waveform on the FS7 show you the underlying slog or the mlut - but zebras will always show you or correlate to whatever you're seeing in the viewfinder. In other words, if I set my zebra 1 to 68 - then switch on the mlut to viewfinder - that 68 number will not help me expose for 90% white while viewing the mlut.
Liam Hall
Veteran
Correct. I just use it to confirm what I see in the WFM. When shooting normal speed, if I have time, I'll toggle the mLUT off and back on to check the zebra - would be great if that was on a user button!
Liam Hall
Veteran
I only use zebras with the slog. Even then, it's a quick on/off.
I am confused how much confusion this camera still creates. Luts and zebra's, slog3 or slog2. I think the real underlying problem is 'camera technology in progress' and lack of knowledge/interest at the client/editor side. A lot editors are not yet used to slog and 4K. And the cameraprocessor isn't capable of doing all the things it should in 4K (but work well in HD) as most of our monitors in editing rooms. So we can use Luts as a temporally solution.
This camera has it's technolocical limitations and a 'Sony overkill' in options to choose from. But I like the FS7 and I have found a workflow that works well for me. (I only use a lut in post (the Amira one that comes with Davinci), shoot mostly slog3 and I edit in Davinci Revolve, use Zebra, and WFM and of course FAM
)
In an ideal world the camera should have higher DR and standard, a viewfinder with HDR that is capable of showing what you shoot without having to think about Luts and the right settings 60% or 62% with or without Lut etc.
This camera has it's technolocical limitations and a 'Sony overkill' in options to choose from. But I like the FS7 and I have found a workflow that works well for me. (I only use a lut in post (the Amira one that comes with Davinci), shoot mostly slog3 and I edit in Davinci Revolve, use Zebra, and WFM and of course FAM
) In an ideal world the camera should have higher DR and standard, a viewfinder with HDR that is capable of showing what you shoot without having to think about Luts and the right settings 60% or 62% with or without Lut etc.
Doug Jensen
Carbonite Member
You took the words right out of my mouth. It astounds me how many crazy ideas and techniques are being kicked around on this thread that are just making it harder for people to work with the camera. I can't beleive there is such a fundamental misunderstanding of how to determine exposure (with or without a MLUT) by some people who's work I respect. Or how to use zebras and even which of the two zebras to use. I think some people don't even know what the difference is between them, or that there even is a difference. I'm surprised what this thread has revealed. The camera is so easy to shoot with, why make it harder than it needs to be?
Not sure what confuses you guys. It’s very simple, most television professionals come from ENG broadcast cameras if they’ve been in this business for any respectable amount of time. These are the cameras they learned in school and grew accustomed to over the years. That means simply 70% skin tones, 100% to make sure you’re not clipping, and WYSIWYG colors. Now more and more clients want this “look” and shooters are trying to adapt. The workflow is completely different and mimics film more than it does any sort of run and gun broadcast format. So now people need to learn what slog is, why they need to expose that in a different way, what LUT stands for let alone what exactly it does.
I’m one of those people, and yes, I’ve learned the techniques and watched the videos, but I’m also relatively young. So what about the guys who are older, and are now learning all of this from scratch? I say this because where I work this is exactly the case, and I see this happening daily with veteran shooters. Many of you made this transition years ago, but for the guys making it now, as you guys probably remember, it takes time.
I’m one of those people, and yes, I’ve learned the techniques and watched the videos, but I’m also relatively young. So what about the guys who are older, and are now learning all of this from scratch? I say this because where I work this is exactly the case, and I see this happening daily with veteran shooters. Many of you made this transition years ago, but for the guys making it now, as you guys probably remember, it takes time.
icarusfilm
Veteran
I think its fundamentally important for shooters to learn how to deal with the footage in post be it in Resolve or an NLE of any kind.
When you learn basic grading it will help you shoot as well.
When you learn basic grading it will help you shoot as well.
Does the color correction ever involve brightening or darkening the overall image, even slightly? If so, you didn't really have the right exposure.
That's why I say it's impossible to know the ideal exposure at shooting time. It's better to know how much correction was necessary and have measured your exposure latitude than to think your method for setting exposure is always perfect.
Liam Hall
Veteran
Similar, but not the same as in-camera and certainly not the same when dealing with Cine EI. But, I do know what you mean and I'm probably arguing semantics.
Only when I get it wrong, which I try not to do...
It is absolutely possible, it's not that difficult and is easily measurable using a variety of tools, both on the camera and externally.
As one of the so-called "old guys" I'm going to have to take issue with a few things that were just said. Sorry this is a bit of a rant.
1- Many of us veteran shooters started out in film. We know how to shoot film negative and reversal. We remember when the only way to judge exposure was with a light meter and we had film that was balanced for only 1 ISO and 1 color temperature. We lit that way, and shot run and gun that way as well - often with great results. When we moved to video a lot of things - mainly the image quality - were worse but there were a lot of wonderful improvements. Shooting under fluorescents without a white balance button was a royal PITA especially when you had to augment existing light because you never knew for sure what the green levels were (color meters sucked - and still do.) Run and gun shooting without a waveform or zebras was hard to judge and you were constantly pulling out a light meter. A lot of people ( many Hollywood DPs) shot Polaroids with film to help visualize contrast while lighting. - So yes we adopted Zebras and WYSIWYG video because we liked them. One of the advantages of having a monitor was that video exposure and contrast levels were never as reliable as film and the monitor replaced the Polaroids as an aid to visualization. We liked these tools because they helped us work and by the way we never had to look at a negative in the viewfinder while shooting.
- Now along comes Log shooting which is basically trying to emulate film negative , and a lot of young guys seem to think they're the vanguard of the new thinking about how to light and how to expose and the rest of us relics are "old school". Well I'm here to tell you that we were doing "new school film" while a lot of young guys were in grade school. I love Log and the all the choices it opens up in camera and post so there is no problem learning the new methods. But I'm constantly hearing from directors that they get young shooters who seem to think that shooting Log absolves them from having to light carefully because its all done in post. They get a lot of flat garbage they can't do anything with. We didn't shoot film negative that way. I light film, old cameras and new all the same. The lighting tools have changed, the cameras are more sensitive and the video cameras handle contrast better, but its basically all the same. Some of my clients are thrilled that I offer to shoot WYSIWYG because they know its going to look right from the gate.
- Re Zebras Log and waveforms and exposure. Sorry but I think the exposure and color system in the FS7 is lousy. I love the camera, I'm able to shoot great stuff and for the price its incredible but the Sony engineers don't have a clue. They felt they were going to emulate a film camera so they gave us 3 presets for absolutely no good reason when RED and ARRI both have white balance. They gave us Slog and MLUTs but didn't make them easy to use.
I don't know Arri that well but i shoot a lot of RED. I have a viewfinder or a nice 5" monitor in which I can flip instantly between Red Color /WYSIWYG and RAW . I have a multitude of exposure tools including false color and zebras - some of which reflect the LUT and some the underlying RAW and the are all available on buttons. I can change the EI and the Kelvin instantly and if i'm under fluorescents I can go to the menu and either hit a white balance button or dial in green and magenta. Likewise when the video goes to Post the Basic user LUT is right there for immediate use. They don't have to use it but its immediately available. There are negatives - Should be quicker to get to the white balance, there is no waveform and of course no internal ND's is criminal.
- With the FS7 I can use a multitude of LUTs but the stupid menu system won't carry the name of the LUT to the EVF so I can remember what it is. I can't flip instantly between Log and MLUT without going in to the menus. I can't use my MLUT on high speed. I can't have my waveform reflect whatever is on the EVF, I have to pick Log or MLUT ahead of time. I have to pick a zebra that is only good for Log or MLUT.
Frankly it would not be hard to fix these issues in an upgrade. Provide a button to switch the EVF from LOG to EVF. Give us one extra zebra to set for Log and one for the LUT and have the waveform reflect whatever is showing. Have the EVF show the name of the MLUT. Its not brain surgery. Apparently changing the color system to allow color balancing under Log is a deeper issue but it was a dumb blunder to design the camera this way in the first place.
- Lastly re: whether those of us who complain simply don't know how to use zebras or expose properly. There's lots of ways to skin a cat. Doug and Liam seem content to use their zebras and waveform in LOG mode and that's fine for them. I want to hear their methods . However many of us don't want to look at LOG in the EVF and have perfectly good reasons for that, neither do we want to jump to Log view like Liam seems to do for zebras. We've all found ways to work and I'm guessing that we're all providing good footage to our clients or we wouldn't be here much longer. So have a little respect for diversity please and use this for an opportunity to share not judge.
sorry for the rant.
i
1- Many of us veteran shooters started out in film. We know how to shoot film negative and reversal. We remember when the only way to judge exposure was with a light meter and we had film that was balanced for only 1 ISO and 1 color temperature. We lit that way, and shot run and gun that way as well - often with great results. When we moved to video a lot of things - mainly the image quality - were worse but there were a lot of wonderful improvements. Shooting under fluorescents without a white balance button was a royal PITA especially when you had to augment existing light because you never knew for sure what the green levels were (color meters sucked - and still do.) Run and gun shooting without a waveform or zebras was hard to judge and you were constantly pulling out a light meter. A lot of people ( many Hollywood DPs) shot Polaroids with film to help visualize contrast while lighting. - So yes we adopted Zebras and WYSIWYG video because we liked them. One of the advantages of having a monitor was that video exposure and contrast levels were never as reliable as film and the monitor replaced the Polaroids as an aid to visualization. We liked these tools because they helped us work and by the way we never had to look at a negative in the viewfinder while shooting.
- Now along comes Log shooting which is basically trying to emulate film negative , and a lot of young guys seem to think they're the vanguard of the new thinking about how to light and how to expose and the rest of us relics are "old school". Well I'm here to tell you that we were doing "new school film" while a lot of young guys were in grade school. I love Log and the all the choices it opens up in camera and post so there is no problem learning the new methods. But I'm constantly hearing from directors that they get young shooters who seem to think that shooting Log absolves them from having to light carefully because its all done in post. They get a lot of flat garbage they can't do anything with. We didn't shoot film negative that way. I light film, old cameras and new all the same. The lighting tools have changed, the cameras are more sensitive and the video cameras handle contrast better, but its basically all the same. Some of my clients are thrilled that I offer to shoot WYSIWYG because they know its going to look right from the gate.
- Re Zebras Log and waveforms and exposure. Sorry but I think the exposure and color system in the FS7 is lousy. I love the camera, I'm able to shoot great stuff and for the price its incredible but the Sony engineers don't have a clue. They felt they were going to emulate a film camera so they gave us 3 presets for absolutely no good reason when RED and ARRI both have white balance. They gave us Slog and MLUTs but didn't make them easy to use.
I don't know Arri that well but i shoot a lot of RED. I have a viewfinder or a nice 5" monitor in which I can flip instantly between Red Color /WYSIWYG and RAW . I have a multitude of exposure tools including false color and zebras - some of which reflect the LUT and some the underlying RAW and the are all available on buttons. I can change the EI and the Kelvin instantly and if i'm under fluorescents I can go to the menu and either hit a white balance button or dial in green and magenta. Likewise when the video goes to Post the Basic user LUT is right there for immediate use. They don't have to use it but its immediately available. There are negatives - Should be quicker to get to the white balance, there is no waveform and of course no internal ND's is criminal.
- With the FS7 I can use a multitude of LUTs but the stupid menu system won't carry the name of the LUT to the EVF so I can remember what it is. I can't flip instantly between Log and MLUT without going in to the menus. I can't use my MLUT on high speed. I can't have my waveform reflect whatever is on the EVF, I have to pick Log or MLUT ahead of time. I have to pick a zebra that is only good for Log or MLUT.
Frankly it would not be hard to fix these issues in an upgrade. Provide a button to switch the EVF from LOG to EVF. Give us one extra zebra to set for Log and one for the LUT and have the waveform reflect whatever is showing. Have the EVF show the name of the MLUT. Its not brain surgery. Apparently changing the color system to allow color balancing under Log is a deeper issue but it was a dumb blunder to design the camera this way in the first place.
- Lastly re: whether those of us who complain simply don't know how to use zebras or expose properly. There's lots of ways to skin a cat. Doug and Liam seem content to use their zebras and waveform in LOG mode and that's fine for them. I want to hear their methods . However many of us don't want to look at LOG in the EVF and have perfectly good reasons for that, neither do we want to jump to Log view like Liam seems to do for zebras. We've all found ways to work and I'm guessing that we're all providing good footage to our clients or we wouldn't be here much longer. So have a little respect for diversity please and use this for an opportunity to share not judge.
sorry for the rant.
i
Last edited:
Doug Jensen
Carbonite Member
Lenny,
Good post. I agree with most of what you have said and I respect your opinion on a lot of the things that I disagree with. But since you invoked my name, I'll add a few comments of my own. I firmly believe, without a shadow of a doubt, or with any misgivings, that zebras (when used appropiately) are by far the best we to nail exposure on a professional video camera and it is the only exposure tool someone should need in 99% of the situations most of us find ourselve shooting in every day. And it makes no difference whether someone chooses to use a MLUT or not. The decision to use an MLUT or not has nothing to do with exposure. And I gurantee anyone who cares to listend to the rantings of an old man that zebras are still the best exposure tool regardless of whether an MLUT is used or not. One thing has nothing to do with the other.
We don't even have to agree on what the best exposure is creatively. There are often legitimate differences in what each of us defines as correct exposure. My target exposure on a given situation might be a stop lower than yours. That is fine. But what is very important is that we are both able to nail that target exposure quickly, accurately, consistently, and with out a lot of hassle. That is what makes zebras so great. Each of us can choose the target level we prefer -- set the meus to match -- and then hit our our targets perfectly almost every time without any extra button-pushing or concious thought at all. And despite what some people have said on this thread, exposure is not some nebulous variable target that doesn't matter. Exposure accuracy is vitally important for S-LOG and even more so when shooting in the Custom Mode with a WYSIWYG workflow. To suggest anything else is wrong. It is true there are many ways ot skin a cat but some ways are much easier, faster, and more efficient that others. If I can skin a cat in 10 seconds and it takes someone else 30 seconds, then I say my way is better. And even though there are many ways to skin a cat does not mean that there aren't even more ways to skin a cat incorrectly. I am not a believer in the modern attitude that anything goes and every opinion or method is credible. Some things are just wrong. And a lot of wrong stuff is being thrown around lately on DVX. That's my opinion, others are free to disagree.
Good post. I agree with most of what you have said and I respect your opinion on a lot of the things that I disagree with. But since you invoked my name, I'll add a few comments of my own. I firmly believe, without a shadow of a doubt, or with any misgivings, that zebras (when used appropiately) are by far the best we to nail exposure on a professional video camera and it is the only exposure tool someone should need in 99% of the situations most of us find ourselve shooting in every day. And it makes no difference whether someone chooses to use a MLUT or not. The decision to use an MLUT or not has nothing to do with exposure. And I gurantee anyone who cares to listend to the rantings of an old man that zebras are still the best exposure tool regardless of whether an MLUT is used or not. One thing has nothing to do with the other.
We don't even have to agree on what the best exposure is creatively. There are often legitimate differences in what each of us defines as correct exposure. My target exposure on a given situation might be a stop lower than yours. That is fine. But what is very important is that we are both able to nail that target exposure quickly, accurately, consistently, and with out a lot of hassle. That is what makes zebras so great. Each of us can choose the target level we prefer -- set the meus to match -- and then hit our our targets perfectly almost every time without any extra button-pushing or concious thought at all. And despite what some people have said on this thread, exposure is not some nebulous variable target that doesn't matter. Exposure accuracy is vitally important for S-LOG and even more so when shooting in the Custom Mode with a WYSIWYG workflow. To suggest anything else is wrong. It is true there are many ways ot skin a cat but some ways are much easier, faster, and more efficient that others. If I can skin a cat in 10 seconds and it takes someone else 30 seconds, then I say my way is better. And even though there are many ways to skin a cat does not mean that there aren't even more ways to skin a cat incorrectly. I am not a believer in the modern attitude that anything goes and every opinion or method is credible. Some things are just wrong. And a lot of wrong stuff is being thrown around lately on DVX. That's my opinion, others are free to disagree.
I'd be interested in hearing people's different methods for setting exposure, and especially how you decide to set exposure when there are conflicts between the things you're measuring, like some people in sunlight and some people in shadows, people with very different skin tones, or different light meter readings in different parts of the scene.
Peter James
Member
Doug you will be upsetting a lot of cat lovers with all this talk of skinning them!
I find using a MLUT helps with focusing especially when using peaking. Peter
I find using a MLUT helps with focusing especially when using peaking. Peter
cameramanben
Well-known member
balazer,
for me, it depends on the setup.
In the case of your suggestion with figures in both sun and shade and no control of light I will aim to avoid clipping and then think, is the shady one interesting as is? I'll try to frame up to work with the light provided, and I tend to make extensive use of ND grads in all kinds of orientations to control contrast to where I want. Also a nut for Polas to control to control backlight. Mind you, if silhouette looks cool and makes sense, great. Shooting log though, I know that handled properly and with clipping avoided, anything beyond really noisy underexposure is inherently fixable, so if needed, I go with just avoiding clipping.
As for more controlled light, I'll generally light and expose to place subject skin tones appropriately in the midrange. I use a combination of waveform and zebras (set to 3% aperture @ 65% which with my preferred MLUT is equivalent to 70% on conventional Rec709, around a stop and a bit over mid gray). I use that 65% as a datum, but clearly I don't end up placing everyone's face at the same level. I just know how to play off that, in the same way that many people reference off 90% matte reference white. I'd only use 18% gray with a light meter - so not very often ;-).
I did used to use 90% ref white on previous cameras, but find the lower datum easier to work off considering the headroom now usable above reference white and consequent ambiguity over matte white and reflected highlights. The 3% zebra aperture means I don't have zebras all over the interesting bits annoying me, and I can feel like I'm being precise!
Within a sequence, I will expose for consistency in the subject. Going from the wide to the close, I don't want the edit to have to compensate for jarring jumps. 2.8 on the wide should still be 2.8 for the closeups, though I may well open up on someone who is under in the wide and doesn't work on the close; along with finessing lighting, whatever works to make shot and reverse cuts smooth in the edit. If the lighting is under my control, I'm not going to set anything up which doesn't work, though I do like to make use of the dynamic range of log.
With your 'one figure in light, one in shade' example and no control of lighting, I'd be much more flexible in exposing the close shots to just look 'right', having just exposed the high contrast wide to protect highlights and 'fix it in post' ;-).
Under CineEI I tend to rate the camera 2/3 stop pulled as my base. That's like placing 90% white around 68% IRE on S-Log2 or S-Log3, where Sony defines it at 'Native ISO' as around 60%. That pull knocks back noise a bit, making the shadow range a bit more comfortable, and still leaves me a little over 5 stops of headroom. I make use of zebras and an MLUT so CineEI is trivial; my 65% datum is 65% whatever the CineEI and my look means that it is pretty easy to judge if anything is close to white clip on the recording, though I am very conservative about bouncing around CineEIs within a project to keep things simple for post.
Obviously log is really forgiving if post has time, knowledge and inclination to work on it, but however I set up, I aim to remain consistent in my approach. That way, the director can work his edit with a 'Ben' LUT deuglifying the log (maybe a couple of CineEI corrected versions if we have bounced around CineEI), and post can work off a base 'Ben' setting - whether a nondestructive LUT or bunch of control settings - and polish things to a finished product reasonably painlessly. Then, when I send in the horribly contrasty shot with important stuff at either extreme, they hopefully won't begrudge me the extra work!
Ben
for me, it depends on the setup.
In the case of your suggestion with figures in both sun and shade and no control of light I will aim to avoid clipping and then think, is the shady one interesting as is? I'll try to frame up to work with the light provided, and I tend to make extensive use of ND grads in all kinds of orientations to control contrast to where I want. Also a nut for Polas to control to control backlight. Mind you, if silhouette looks cool and makes sense, great. Shooting log though, I know that handled properly and with clipping avoided, anything beyond really noisy underexposure is inherently fixable, so if needed, I go with just avoiding clipping.
As for more controlled light, I'll generally light and expose to place subject skin tones appropriately in the midrange. I use a combination of waveform and zebras (set to 3% aperture @ 65% which with my preferred MLUT is equivalent to 70% on conventional Rec709, around a stop and a bit over mid gray). I use that 65% as a datum, but clearly I don't end up placing everyone's face at the same level. I just know how to play off that, in the same way that many people reference off 90% matte reference white. I'd only use 18% gray with a light meter - so not very often ;-).
I did used to use 90% ref white on previous cameras, but find the lower datum easier to work off considering the headroom now usable above reference white and consequent ambiguity over matte white and reflected highlights. The 3% zebra aperture means I don't have zebras all over the interesting bits annoying me, and I can feel like I'm being precise!
Within a sequence, I will expose for consistency in the subject. Going from the wide to the close, I don't want the edit to have to compensate for jarring jumps. 2.8 on the wide should still be 2.8 for the closeups, though I may well open up on someone who is under in the wide and doesn't work on the close; along with finessing lighting, whatever works to make shot and reverse cuts smooth in the edit. If the lighting is under my control, I'm not going to set anything up which doesn't work, though I do like to make use of the dynamic range of log.
With your 'one figure in light, one in shade' example and no control of lighting, I'd be much more flexible in exposing the close shots to just look 'right', having just exposed the high contrast wide to protect highlights and 'fix it in post' ;-).
Under CineEI I tend to rate the camera 2/3 stop pulled as my base. That's like placing 90% white around 68% IRE on S-Log2 or S-Log3, where Sony defines it at 'Native ISO' as around 60%. That pull knocks back noise a bit, making the shadow range a bit more comfortable, and still leaves me a little over 5 stops of headroom. I make use of zebras and an MLUT so CineEI is trivial; my 65% datum is 65% whatever the CineEI and my look means that it is pretty easy to judge if anything is close to white clip on the recording, though I am very conservative about bouncing around CineEIs within a project to keep things simple for post.
Obviously log is really forgiving if post has time, knowledge and inclination to work on it, but however I set up, I aim to remain consistent in my approach. That way, the director can work his edit with a 'Ben' LUT deuglifying the log (maybe a couple of CineEI corrected versions if we have bounced around CineEI), and post can work off a base 'Ben' setting - whether a nondestructive LUT or bunch of control settings - and polish things to a finished product reasonably painlessly. Then, when I send in the horribly contrasty shot with important stuff at either extreme, they hopefully won't begrudge me the extra work!
Ben
Last edited:
Liam Hall
Veteran
Good post Lenny. I agree completely. And, as an old tart myself who started in film I tend to also think of shooting in CineEI as if I were exposing ISO1250 film. It need be nothing more complicated than that.
I skin my cat the way I do, with WFM and zebras simply because it works for me, as I tend to flip around frame rates all the time. Certainly, there are many other methods that work just as well. Today, I am tethered to my Odyssey O7Q+ which makes it so easy to get everything right, not so much fun lugging it around though!
Last edited:
Liam Hall
Veteran
balazer,
for me, it depends on the setup.
In the case of your suggestion with figures in both sun and shade and no control of light I will aim to avoid clipping and then think, is the shady one interesting as is? I'll try to frame up to work with the light provided, and I tend to make extensive use of ND grads in all kinds of orientations to control contrast to where I want. Also a nut for Polas to control to control backlight. Mind you, if silhouette looks cool and makes sense, great. Shooting log though, I know that handled properly and with clipping avoided, anything beyond really noisy underexposure is inherently fixable, so if needed, I go with just avoiding clipping.
As for more controlled light, I'll generally light and expose to place subject skin tones appropriately in the midrange. I use a combination of waveform and zebras (set to 3% aperture @ 65% which with my preferred MLUT is equivalent to 70% on conventional Rec709, around a stop and a bit over mid gray). I use that 65% as a datum, but clearly I don't end up placing everyone's face at the same level. I just know how to play off that, in the same way that many people reference off 90% matte reference white. I'd only use 18% gray with a light meter - so not very often ;-).
I did used to use 90% ref white on previous cameras, but find the lower datum easier to work off considering the headroom now usable above reference white and consequent ambiguity over matte white and reflected highlights. The 3% zebra aperture means I don't have zebras all over the interesting bits annoying me, and I can feel like I'm being precise!
Within a sequence, I will expose for consistency in the subject. Going from the wide to the close, I don't want the edit to have to compensate for jarring jumps. 2.8 on the wide should still be 2.8 for the closeups, though I may well open up on someone who is under in the wide and doesn't work on the close; along with finessing lighting, whatever works to make shot and reverse cuts smooth in the edit. If the lighting is under my control, I'm not going to set anything up which doesn't work, though I do like to make use of the dynamic range of log.
With your 'one figure in light, one in shade' example and no control of lighting, I'd be much more flexible in exposing the close shots to just look 'right', having just exposed the high contrast wide to protect highlights and 'fix it in post' ;-).
Under CineEI I tend to rate the camera 2/3 stop pulled as my base. That's like placing 90% white around 68% IRE on S-Log2 or S-Log3, where Sony defines it at 'Native ISO' as around 60%. That pull knocks back noise a bit, making the shadow range a bit more comfortable, and still leaves me a little over 5 stops of headroom. I make use of zebras and an MLUT so CineEI is trivial; my 65% datum is 65% whatever the CineEI and my look means that it is pretty easy to judge if anything is close to white clip on the recording, though I am very conservative about bouncing around CineEIs within a project to keep things simple for post.
Obviously log is really forgiving if post has time, knowledge and inclination to work on it, but however I set up, I aim to remain consistent in my approach. That way, the director can work his edit with a 'Ben' LUT deuglifying the log (maybe a couple of CineEI corrected versions if we have bounced around CineEI), and post can work off a base 'Ben' setting - whether a nondestructive LUT or bunch of control settings - and polish things to a finished product reasonably painlessly. Then, when I send in the horribly contrasty shot with important stuff at either extreme, they hopefully won't begrudge me the extra work!
Ben
Excellent post as ever Ben, you write everything I like to say, but in a more eloquent way.