Saturday, January 14, 2012

Fuji EXR – Dynamic Range Test Mark 1 — Oikos

If you haven’t tasted this spectacular Greek style of yogurt than you owe it to yourself to do so. Wow.

But I digress …

I was asked on the Fuji forums to perform a simple test of dynamic range to prove once and for all whether or not EXR technology has any real value at all. In other words, can other cameras do as well with their software based DR extension?

I am handicapped by the fact that I have only one serious compact camera that is not a Fuji, that being the Panasonic ZS3. It has the iExposure feature that seems to me to be very good at getting the right exposure, so I set that on. It shoots only JPEGs of course.

The F550EXR is the Fuji representative, and it is shot in what I now consider to be the proven best way to shoot any EXR camera, and that includes the X10. M size, DR400, lowest possible ISO.

The D7000 was shot the usual way … control highlights with –EV in RAW and process the result.

What was interesting is that I made one shot with the Nikon at –1EV (black background and sunshine lit white plastic container makes for some –EV as mandatory.) The Fuji was shot at –1EV and –2EV, and the –2EV image was used. The Panasonic was also shot at both, and the –2EV image was used.

The results are fairly predictable … the Nikon image was by far the best.

The Fuji did hold the highlights, but just barely.

In fact, the OOC JPEG is pretty good too …

And the Panasonic was hopeless. This is not a strength of that camera.

These results are pretty conclusive, although it is possible that some newer cameras like the S100, G12, P7100 etc will have better algorithms. And, of course, these others can also benefit from RAW.

The D7000 has a much larger sensor and was shot in RAW. It required very little recovery in ACR and the image is very detailed with excellent handling of the bright areas. The Fuji required significant processing to save the highlights, but the result is not bad at web sizes. The Panasonic blew the bright areas sky high even at –2EV.

Here are some crops for you see see how they would compare in an enlargement. Not much of a contest.

Remember that these images are equivalent to a 49” print viewed from 20 inches. The Fuji did remarkably well when you consider that it started as a 6Mp image. And it held the highlights, which was the most important requirement for this shot. Think of a bride in a white dress standing next to a groom in a black tuxedo and you now know why this matters so much.

For those carrying pocket cams around to weddings etc, EXR technology can make the difference between a bunch of great memories and a lot of disappointment. However, if you are willing to control exposures yourself and shoot and process RAW, then EXR does not have as much bearing on the result. It helps, in my opinion, but there is a small price in reduced resolution and the need to shoot RAW+JPG. And you still need to dial in compensation.

I’ll close with an example from a camera with really poor dynamic range, the Canon G10. I traveled on business with a coworker to Washington D.C. and we hit the mall to see what the fuss was about. If you have never stood on the mall and looked at the Capitol Building at one end in the later afternoon sun then you owe it to yourself to do that one day. Mind blowing sense of power …

Anyway … later afternoon sun blazing off the building in behind my coworker, who was wearing black and has dark black hair. The G10 of course could never have shot this successfully in JPEG, but I shot it in RAW and the result was perfectly acceptable.

Some of the tiniest photosites ever put into a compact camera and the dynamic range shown here has nothing to apologize for …

Note: There might be another test or two from the same shooting session … that’s why I called this Mark 1.