Reevaluation, The X100S, Etc.

It's hard to believe the winter is here. Seems like just yesterday I took that road trip to DC armed with only my X100S, my D600, and a 50mm prime or two, and maybe another couple of lenses. Actually too many lenses for the purposes of the trip - re-evaluating my bias towards the 50mm focal length vs. the 35mm welded to my Fuji X100S. I did end up staying true to my purpose and that is re-evaluation of a bunch of things that tend to build up biases over time.

I did all sorts of things I never do. Like crop at ankles and knees, use Aperture priority auto and intelligent evaluative metering where it makes no sense to me. Most of all I used the little Fuji in a lot of ways I would not use a 35. It worked out for the most part as I've mentioned. I have a new and healthier point of view on the 35mm field of view. In fact I used the X100S almost exclusively for my second road-trip to DC.

Those kinds of re-evaluations are not a bad thing to do every once in a while to put your head in a different place than where it goes out of habit. It definitely gives you some fresh perspectives. Even if it fails miserably the exercise can generate some new ideas might not fail. They maybe fantastic. Having said that and having found that my 50mm bias was a bit overblown there are a couple of notes I wanted to share where my other biases were confirmed - as in yep same ol' same ol' story and no need to change my current working method for the most part.

There are a dozen things to evaluate that I tried. I tend to go back after some time has passed and look at those things again. Your thought process changes so it's definitely something worthwhile - heck even the selects you made tend to change quite a bit. Today I'm going to share only one tiny little thing and my take on it - auto exposure and intelligent metering.

When I shot indoors for my little personal project I shot manual. When I moved outside to more thoroughly explore the a bunch of other things I never do a big one was setting both cameras to aperture priority auto-exposure and just let it rip. I wasn't too too worried as both cameras can easily deal with a stop of correction either way during post in most shooting conditions. What did I find? Exactly the same thing I've found every single other time on every camera ever - auto exposure gives you random shot to shot results most of which are wrong and attempting to fight with it via exposure comp is a fool's game. Far more distracting and tedious than just shooting in manual and paying attention to the light.

Yes there are circumstances where auto exposure might get you a better result but those are kind of rare - at least for me. Like just picking your camera up and firing it with no evaluation of the scene. Maybe if the light is changing wildly shot to shot - rare for sure. Otherwise let's take a look at what happens for the vast majority of shooting conditions.

Here's what I found going through those images again. Generally both cameras under-expose when not sure what to do. Shot to shot variation happens with even a minor framing variation. Light and dark colors or high contrast absolutely cause these supposedly "spot on" evaluative metering systems to be absolutely wrong in 80%+ of scenes. The only way to know is to chimp every shot or pay very close attention to your viewfinder readouts and use exposure compensation to make shutter/aperture the exact same thing every shot that it was when you chimped and adjusted it. Why would you want to play that game? More specifically why would I.

Okay so here's my normal working method. Manual exposure set to somewhere around what the camera meter tells me with either more or less based on experience (ie light colored background etc). Take test shot. Make minor adjustment if required. Then shoot shoot shoot paying attention only to the subject and the light. If I change the scene or the light changes then time to re-evaluate.

Let's compare that to a working method in auto exposure that would give me the same results. Okay… set camera to auto of some sort - for me that would be aperture priority. Take test shot, evaluate and adjust exposure comp accordingly. Now here's where you have two choices on the next part. Ether take note of the shutter/aperture combo and pay very close attention to make sure that it stays the same as long as the light is the same and if not mess with the exposure comp to cause it to be the same - hmmmm. this is just like manual well that's friggin' stupid. Option 2 would be even worse - take test shot and chimp every frame - if it's off then re take the same exact frame with some compensation adjustment and check it again???? really? WTF? The only reason to do that is if you had no idea what the shutter/aperture was that was what you wanted.

I didn't do either of those - I did what is supposed to help you focus on the subject. I just set it for auto aperture and shot. The results? All over the map. Take the shot at the top. That was a whole stop over exposed - not too bad because the light was lower contrast. That might be deadly in light with more contrast. How did that happen? Well three shots before that I happened to chimp for the first time which was about a dozen shots in - I just had to. Ooooops light background it's at least a stop under so I dialed in +1 compensation and went back to shooting. The shot at the top was a minor recomposition and boom now it gets it right. Too bad I now had +1 dialed in. In a lot of ways evaluative metering can be less predictable than center weighted.

Okay so my reconfirmation of my working method for most conditions produced what I already knew. I wanted to see if my new smarter cameras produced wildly better results than any other camera I have ever had with auto exposure. Answer = no unless it's conditions where the only way to keep up is with auto and produce better results than you would otherwise. I'm not a complete curmudgeon though there are situations where it will be a help instead of a hinder but those are pretty clear. Those knobs and buttons on cameras are there for a reason - choose the appropriate setup for what you are actually doing not some one size fits all thing.

The same thought process is appropriate for TTL auto flash as well. I won't go into that again but here's a good place to start… If the flash is actually on the camera and you are moving in relation to your subject auto TTL is great. It's faster and better than you will be. If your flash is not on the camera and your subject is not moving a whole lot but you are then manual flash control is far better and more convenient with no surprises based on where you are, how you frame the subject, etc. It's really not rocket science it's much more common sense.


Ps. Last note that Fuji X100S 23mm lens is really quite delicious. Check out the way it renders at f/5.6 both subject and background. Here's a screenshot at 100% of a different scene at f/4 - mind blowing. It's not just sharp it looks good.

blog comments powered by Disqus