Fear not, we are not about to divulge any secrets about where you can be found running around with nothing on except an old mac, unless of course you want to warn us so we can miss the event !
You often hear folk say things like 'expose for the shadows', or 'get the highlights right', but what are they on about, and why ?
It's often a technique to aim to get the exposure right for the shadows in the picture and let the highlights sort themselves out, something possibly more relevant with traditional film photography which is more tolerant of highlights. Sounds easy but with extreme contrast the highlights may be overexposed, or 'blown out' to use the correct term. This subject has been done to death for film and I am certainly not going to dig into this rather deep subject, but what is interesting is how things are today with all this digital photography that needs a new set of rules ?
Without going into boffin mode the digital sensor has far less tollerance of the highlights that film capture offers, things are getting better as the technology improves but they are still unable to handle too much light unlike film that can retain detail well into the brighter lit bits. Problem is that light meters look at light and can only quantify brightness, be it average scene or spot areas but its basically a volume sensor for light.
So, what we are told to do now is expose for the brightest areas, use that retched histogram on the camera to make sure we get as much light captured as possible before 'clipping' kicks in, the point where the sensor is saturated with light and we get pure white. Thankfully RAW came along which if used properly gives us more tolerance by being able to pull back about 2-3 stops of light that may be overexposed, problem is to pull back this details means that something has to give and it's probably the subtle graduations or texture of the brighter parts of the image. "Use the Recovery tool" I hear someone say at the back of the class - if desperate ok but this can be the best way to flatten the highlight detail, and sometimes the whole image - should be nicknamed the 'Flatten' control.
Ok, so if we want the best exposure then the more light we capture in a digital photo the better the range of image detail and in turn the better the exposure we get in the dark areas ? - That's that sorted !
Well, the problem is you see so many images exposed to the max and the result can be total loss of reality, if the brightest part of the scene is a mid light level then that's what it should be captured as, why turn grey to almost white. Far better to look at the scene, how dark are the dark bits and how bright are the bright bits, turn that histogram off and look at the scene..
I could at this point get into the practice of using a 50% Grey card but that's for another day, I would rather share my own thoughts through personal examples.
Example 1 - light flooding into an old Croft.
Exposure was set for the light hitting the floor whilst wanting to retain the detail in the shadows. Had I used an auto mode or reflected light meter reading the light hitting the window surround would be almost correct and everything else black. I personally like the flooded light effect, not for the purists but as an image it works for me.
Example 2 - same location but no direct light in the scene.
This time an auto mode or reflected light meter would expose the brightest area, the floor at full exposure so the whole image would be over bright and to me lack any sense of reality. This is a classic shot for worrying about the shadows and mid tones and not the highlights ...... cos there aren't any !.
Ok, we could set exposure to max and wind down in post production but isn't that just being a little bit lazy...yes, and there is a little more to it ... read on.
Lets look at portraits. 2 examples with the camera set to expose for the actual ambient light and not whats reflected off the subject. In both cases images are shown 'as captured' with no exposure or contrast (gamma) correction.
The first example image if shot in auto mode would give the brightest part of the pic to be close to full exposure which would be the clothing top, which although white was not very reflective so appeared a dull white. However, the near white background would raise the average light level hence risk under-exposing the image so how can the camera get it right ? Here we have a nice even skin tone, detail in the top and also importantly the dark jeans.
The second is a similar light setup but here the black background in auto would give less average light hence risk over exposure, and the lack of highlights would further increase the exposure as the camera will try and make the darkest area as bright as it can, which will be the side of the face and the right arm.
In both images looking at the black and white versions the skin tones appear mid grey which is what I would expect. When did natural skin look white ?
So, whats the point of all this rambling ?
Simple, try looking at the scene and 'seeing' the light and setting an exposure to capture the scene as it really is and not what some research technician has decided as the best vanilla flavored exposure for every picture their camera is aimed at. Many folk starting in photography with digital get very confused about exposure so they rely on the camera's brain which as good as it is can get it very wrong as it can only see the quantity of the light. This is the basis of what is taught on my training courses so that the pictures taken by my students actually look like what their eyes see.
Oh, was that a plug - Yup ! and here comes the obligatory link to details of my training courses !
And finally one more wee rant..... many point and shoot cameras are very clever and have face recognition to help the photographer. Wow, how useful is that ? I think that by the time I forget what a face looks like then I will probably be too old to or blind to want to use a camera.
What next, a camera mode that tells you what to take a picture of ?
Sunday, 6 February 2011
Subscribe to:
Posts (Atom)