I was recently asked to clarify the exposure levels I use when shooting S-Log, especially why mid grey is set at 38IRE. Here’s my reply.
Exposure is not something that is set in stone. The idea behind setting reflected mid grey at 38 IRE is that you place the middle of your tonal range slightly below the mid point of your full latitude range. Reflected white should then come out around 68-70IRE. This then gives you further headroom for direct light sources such as lamps, the sky or specular reflections. In addition, video cameras tend to perform better in shadow and underexposure compared to highlights and over exposure, so it’s generally considered better to be slightly under rather than slightly over.
But, all this theory is based on shooting some imaginary generic scene and as a result is a generalisation and may not be optimum in many lighting conditions. You also need to consider how S-Log works. Each stop is getting allocated roughly the same amount of data as the next. But when you consider that each brighter stop is has twice the brightness range as the previous stop then you need to consider that to some degree, each brighter stop is being recorded more highly compressed than the previous. This means that you can stretch and pull the darker parts of the exposure range more effectively than the brighter part, so you don’t want your images to be too bright as this may not grade as well as the same scene exposed a little darker. Now on top of all this you need to consider noise. Underexpose too much and you will have issues with noise when you start lifting the shadow areas and darker parts of the image. So what I’m trying to say is that you don’t want to overexpose and you don’t want to underexpose, obvious perhaps, but where exactly you set your exposure will depend on the tonal range of the scene you are shooting. A dimly lit interior may have a much reduced range compared to a sunny exterior and this is where the skill of the operator comes in, knowing when it is desirable to push your exposure up or down.
Normally, mid grey at 38IRE works well and will put you in the right ball park. But if you don’t have any direct light sources then you can probably lift mid grey a little, but in my opinion you must always, always protect your highlights. Small specular highlights or a small bright cloud may not show up on a small monitor or VF and might not register on the histogram or zebras so often it’s good to have a bit of headroom in reserve. A small area of overexposure can easily spoil an otherwise correctly exposed scene and it’s impossible to correct a blown out highlight in post production, once it’s gone, it’s gone.
Sadly I can’t say you should always expose “x” at “y” as it’s just not that simple. You have to judge for yourself depending on the scene you are shooting and the look you want to achieve. With S-Log, I would not get too hung up on white not reaching 100%, reflected white should only be at around 70%, it’s direct light sources you need to watch as these are what will give you your biggest exposure headaches.
Having recently acquired S-Log I believe you should clarify your recommendations. I have found that depending on whether LUT’s are turned on or off there is a different reading for IRE % and at this suggests that if exposure is set with LUT’s off, then 38 IRE for neutral grey is 45 for light tone skin and 70 for reflected white. With LUT applied, I find that the settings are similar to non-S-LOG cinnegamma with Grey around 50, skin at 65-70 and white at 100. When I bring the footage to post with the LUT settings, the scopes show of course the non-LUT values, but after de-log the settings return close to those of the LUT, which is I guess the idea of using a LUT. My question is then why talk about setting neutral grey with the LUT off, when it is so helpful to have the LUT on?