FS7: ND Filters when wide open in bright light

S

Soldier

Well-known member
Has anyone run into the issue of trying to shoot wide open in CINE EI mode in bright sunlight? The gain is locked at the camera's native sensitivity which, being the equivalent of 2000 iso, means the ND needs to be about as opaque as a welding visor.

Even with the strongest internal ND filter, it's pretty hard to stop down any lower than f/8 (without resorting to the shutter) I'd love to hear how you deal with the situation. Slap an extra variable ND on the lens?

Glen
 
I have a 3 and a 4 stop ND to use in extreme circumstances.

Vari-NDs are the devil and do horrid things to your image, although occasionally are a lifesaver.
 
Soldier said:
Has anyone run into the issue of trying to shoot wide open in CINE EI mode in bright sunlight? The gain is locked at the camera's native sensitivity which, being the equivalent of 2000 iso, means the ND needs to be about as opaque as a welding visor.

Even with the strongest internal ND filter, it's pretty hard to stop down any lower than f/8 (without resorting to the shutter) I'd love to hear how you deal with the situation. Slap an extra variable ND on the lens?

Glen
Click to expand...

I keep a .6 (2 stops) and .9 (3 stops) ND filters handy and usually use the .6 filter on top of the strongest internal filter. Because you're stacking filers don't use cheap filters to avoid color issues or distortions.
 
Last edited:
I tend to use the 4 stop Hoya PRO NDs on my lenses. Combined with the internal ND on the FS7 they work really well.
 
Soldier said:
Has anyone run into the issue of trying to shoot wide open in CINE EI mode in bright sunlight? The gain is locked at the camera's native sensitivity which, being the equivalent of 2000 iso, means the ND needs to be about as opaque as a welding visor.

Even with the strongest internal ND filter, it's pretty hard to stop down any lower than f/8 (without resorting to the shutter) I'd love to hear how you deal with the situation. Slap an extra variable ND on the lens?

Glen
Click to expand...
An issue for sure but that's way they make Matt boxes and external filters.
That said, my Matt box only fits a few of my lenses but then that's why I also have a selection of screw on polas and NDs. Even with my XDCAM I needed front end glass NDs to keep the stop where I wanted it, which is about f-4 to f 5.6.
 
Soldier,

I think that there are two sides to this. First off is the way that CineEI is designed to be used. Whilst 'gain' is not available in CineEI, this is because the camera is capturing the entire dynamic range that the camera is capable of. Baking in gain - for example as the C300 does to get around an 8-bit codec - means that the full dynamic range would only be available at native ISO.

Instead of gain which is baked in, the idea is that you change the ISO rating, the camera records it as a piece of metadata and the post software later does the exposure compensation required. In the real world software currently often does not read the ISO metadata, so in that instance exposure compensation LUTs will do it properly (its not just conventional gain). I wrote a web app to make them called 'LUTCalc'. The big advantage of CineEI is that you always have 14 stops, it's just that at low CineEI ISO you have extra shadow detail with reduced headroom, at high ISO extra highlight headroom with reduced detail in deep shadows.

This is where monitor LUTs (MLUTs) become pretty vital. Set the camera to show MLUTs and the waveform and zebras will shift with the ISO set. Provided you don't clip white, you can drop the ISO to open the lens without ND. The white clip IRE of underlying S-Log3 on the waveform or zebras will go down with lower ISO. Many people who use CineEI shifts work with the Rec709(800%) LUT. I often shoot CineEIed down a stop or a stop and a half. I generally use a version of the LC709A LUT scaled to extended range. I've gone into the reasons at tedious length elsewhere. ;-)

You can actually make the camera work in the manner of a C300, baking in exposure shifts, by setting 'SDI 1 & Internal' to 'On' in the 'Video' menu and selecting 'S-Log3' as the LUT option. This permenantly bakes in the LUT and hence the EI shift, but again, will reduce the dynamic range captured if off-native ISO. You (and probably your client) will also be looking at uncorrected, flat log. Make sure that 'SDI 1 & Internal' is 'Off' if you are shooting 'proper' CineEI log.

On the other side of the coin are filter options.

I would steer clear of a variable ND, but a decent Pola is something I find myself using all the time in the sun. I have a Schneider Tru-Pol and a Tiffen Ultra Pol and both are very good. You lose around a stop and a half as a base [EDIT UTTER TWADDLE - WASN'T THINKING STRAIGHT, IT'S MORE LIKE JUST UNDER .6] (somewhere between a .6 and .9 ND) [END OF EDIT], but you can also control reflected backlight on skin as well as glass and sky to further reduce contrast. A variable ND is basically just a pair of circular polarisers bolted together and as others have mentioned is not a particularly good idea unless you really need it.

NDs are a good idea, but if you get very heavy with them then you can find blacks developing a red/magenta tinge from infrared leakage. The camera has good IR control on the internal NDs and as I say, it's an issue for very heavy external NDs and the fix is IRNDs, which are NDs which also block IR. They are considerably more expensive than conventional NDs, particularly for decent ones.

Four stops would be what I would class an extreme ND, but if others are finding that they have conventional NDs which work well, I'd take a look at their suggestions,

Ben
 
Last edited:
All depends where you are shooting I guess. I know people in California that laugh at the puny amount of ND I use in dreary London most of the year round.

Current 'best of the crop' NDs are the Formatt Firecrests which have little to no colour cast.

I still use Tiffen IRNDs but they do have a slight green tinge to them.
 
Ben ,
I've never had any problem with 3 stop ND's from B+W or Heliopan.

BTW- the Pol knocks out 2 1/2 stops not 1 . I've heard complaints that it can hurt skin tones, but personally i've never seen it, but one should watch carefully on that. It can certainly help backlight or darken the sky on a backlit interview .

Variable ND is actually a circular Pol in back with a Linear Pol in front. If you use one spend a lot of money on it as the cheap ones are utter garbage. We had to use them with DSLR's but they can do odd things so one needs to be careful.
 
The issue with pola on skin tones is that removing the reflective nature skin can have in the same way it does with reflections on water/glass can lead to skin appearing flat/plasticky.
 
Thanks Ben, that's kind of what I figured. though if you're aware of that you should be able to control it . Key would be only to use the Pola when it can do you some good. Better to use ND for general iris control.
 
Some great suggestions here! Thanks for the advice. Just out of interest (not because I'm looking at cutting corners) if an additional ND did created a colour cast, would it be possible to create a lut in resolve that would both expand the contrast back to 709 levels and deal with the ND's colour cast at the same time? Or would it be like trying to correct the green cast of fluorescent lights, which is, of course, an absolute nightmare.
 
In short - No. If you have color cast from IR (actually far red) in the blacks it can be very tricky to deal with. Actually dealing with color cast from fluorescents is quite straightforward in my experience. How have you found it to be a nightmare?

However I don't think far Red contamination is an issue on the FS7, though it has been for RED . EX-1 and a lot of other cameras. I could be wrong about that though.
 
Lenilenapi said:
How have you found it to be a nightmare?
Click to expand...

I'm not talking about professional fluros, kinos etc, but standard fluros in uncontrolled environments. I find getting skin tones just where I want them is a PITA. Luckly I don't run into the situation that often, but invariably it's on an institutional (school, hospital...) doco or promo.
 
Hmm... I 'm pretty picky about skin tone and usually find that white balance does the trick though I always carry a 2" square gel swatch book to dial it in - typically might balacne through a 1/8 CTB + 1/8 Green. That would be Custom of course. If you're talking CINE EI of course its all post. Though with LUTCalc now you can create a version of 709TypeA LUT corrected for Flos or anything else.
 
cameramanben said:
Soldier,

I think that there are two sides to this. First off is the way that CineEI is designed to be used. Whilst 'gain' is not available in CineEI, this is because the camera is capturing the entire dynamic range that the camera is capable of. Baking in gain - for example as the C300 does to get around an 8-bit codec - means that the full dynamic range would only be available at native ISO.

Instead of gain which is baked in, the idea is that you change the ISO rating, the camera records it as a piece of metadata and the post software later does the exposure compensation required. In the real world software currently often does not read the ISO metadata, so in that instance exposure compensation LUTs will do it properly (its not just conventional gain). I wrote a web app to make them called 'LUTCalc'. The big advantage of CineEI is that you always have 14 stops, it's just that at low CineEI ISO you have extra shadow detail with reduced headroom, at high ISO extra highlight headroom with reduced detail in deep shadows.

This is where monitor LUTs (MLUTs) become pretty vital. Set the camera to show MLUTs and the waveform and zebras will shift with the ISO set. Provided you don't clip white, you can drop the ISO to open the lens without ND. The white clip IRE of underlying S-Log3 on the waveform or zebras will go down with lower ISO. Many people who use CineEI shifts work with the Rec709(800%) LUT. I often shoot CineEIed down a stop or a stop and a half. I generally use a version of the LC709A LUT scaled to extended range. I've gone into the reasons at tedious length elsewhere. ;-)

You can actually make the camera work in the manner of a C300, baking in exposure shifts, by setting 'SDI 1 & Internal' to 'On' in the 'Video' menu and selecting 'S-Log3' as the LUT option. This permenantly bakes in the LUT and hence the EI shift, but again, will reduce the dynamic range captured if off-native ISO. You (and probably your client) will also be looking at uncorrected, flat log. Make sure that 'SDI 1 & Internal' is 'Off' if you are shooting 'proper' CineEI log.

On the other side of the coin are filter options.

I would steer clear of a variable ND, but a decent Pola is something I find myself using all the time in the sun. I have a Schneider Tru-Pol and a Tiffen Ultra Pol and both are very good. You lose around a stop and a half as a base (somewhere between a .6 and .9 ND), but you can also control reflected backlight on skin as well as glass and sky to further reduce contrast. A variable ND is basically just a pair of circular polarisers bolted together and as others have mentioned is not a particularly good idea unless you really need it.

NDs are a good idea, but if you get very heavy with them then you can find blacks developing a red/magenta tinge from infrared leakage. The camera has good IR control on the internal NDs and as I say, it's an issue for very heavy external NDs and the fix is IRNDs, which are NDs which also block IR. They are considerably more expensive than conventional NDs, particularly for decent ones.

Four stops would be what I would class an extreme ND, but if others are finding that they have conventional NDs which work well, I'd take a look at their suggestions,

Ben
Click to expand...

There are so many things wrong with this post Ben I don't know where to start. Have you actually used the FS7??
 
To answer the OP; I just use additional ND. I often shoot with fast primes outside in the sun and it's not unusual for me to have a 1.2 or even 1.8 ND on the lens in addition to the internal ND.
 
Hi Liam,

I have to admit that my own camera is an F55, but I have used an FS7 several times (primarily baking in, so I have used the technique of a varient of the LC709A LUT scaled to extended range, then changed CineEI with the H/M/L switch). If I have described something factually wrong, I apologies and would appreciate it if you gave me any corrections.

I have a feeling I've expressed myself poorly, rather than incorrectly. First and foremost, I do not wish to tell people what they should or shouldn't do. That being said, I thought I'd throw out a few options for experiment for with these extremely flexible cameras.

What I stated is factually accurate with regard to log recording, but the baking-in in particular is distinctly not the way Sony intended log to be used. I personally wouldn't recommend the baked-in log approach (what I described as being like the C300) - what's the point of shooting log then? That doesn't mean it wouldn't work for some people.

I mentioned exposure compensation in post, then my web app for generating LUTs, thinking it best that if someone wanted to play, they could make their own. That could be seen as a hassle, so in some downtime today I've knocked up a few LUTs to play with, all for with the camera shooting CineEI, S-Log3/S-Gamut3.cine:
S-Log3-MLUTs.zip - Basic extended range scalings of LC709 and LC709A. About 4 stops headroom in legal range at native ISO.
S-Log3-1D_Exposure_Correction.zip - 1D 4096-point S-Log3 Data Range to S-Log3 Data Range exposure correction cubes (show up under 3D LUTs in Resolve). 2 stops under to 2-stops over in 1/3 stop steps.
S-Log3-LC709AExtended-DtoL-Resolve.zip - 33x33x33 3D LC709A Extended Range with 2 stops under to 2 stops over corrections. Data levels to Legal levels as with the basic Sony looks, and the Sony-recommended workflow in Resolve.
S-Log3-LC709AExtended-DtoD-Lumetri.zip - as above but data levels to data levels. An experienced DIT tells me that this is what the Lumetri plugin in Premiere CC seems to expect.
S-Log3-FalseColour-MLUTs-DO-NOT-BAKE-IN.zip - False colour MLUTs for in camera with a bunch of different output looks to the non-false coloured bits of the frame. Off native ISO clip values (white clip in particular aren't reliable done in a LUT like this, but 18% gray (green) one stop over 18% gray (pink) and 90% white (orange) will be on the money. At native ISO blue and purple will give a good guide to black clip and near black clip and red and yellow do the same for white clip and near white clip.

All I'd say to people is have a play, this stuff is so flexible, and Liam, honestly, if you could have a go at correcting my errors I would be genuinely grateful.

Thanks,

Ben
 
Lenilenapi,

absolutely - my bad on that, slipped half way through typing 1 1/3 stops for my Pola - that's what Schneider say for it, I'm a waveform and zebras rather than lightmeter kind of bloke, so to be honest have never checked it outright.

Also, very fair point by someone about 'plasticy' effect of the Pola on skin. I very often have it on in bright sunlight, but not necessarily dialled in to reduce skin reflections. It's just there when I need it and giving me a bit of ND.

As for the IR side of things, I regularly add a couple of stops ND no probs, it's just I thought it worth mentioning as a possible issue if someone is going to 4+ stops of conventional ND.

Sorry if a came over sounding like an ignorant t*t!

Ben
 
You must log in or register to reply here.
Back
Top