Sunday, October 24, 2010

Nikon D3100 versus Fuji HS10 and F200EXR / S200EXR -- How many stops of detail retention from an entry dSLR to a bridge cam?

I have had some discussions recently regarding entry level dSLRs and bridge cams. The consensus is that a bridge cam can keep up under some kinds of light … and I tend to agree. By keep up, we always have to add caveats … for images where subject isolation is not all that important. You can do a portrait with a bridge camera, but you’d better know how you will handle your backgrounds. And get the lit properly. That kind of thing …

But one area that always stands out as a weakness is the low light image. And we’re not just talking about high ISO images … but rather any image. If you take a bridge cam into low light, the noise on the sensor – even at 100iso – is going force some very heavy noise reduction, which will lead to smearing of low contrast details.

This is especially bad with hair, and I always like to prove my point with the rather nice images from imaging-resource.com … they shoot a mannequin in low light for every camera they test. Now, I thought we’d compare Nikon’s new and very inexpensive D3100 against the two Fuji bridge cameras … the HS10, which sports a Sony 1/2.3” BiCMOS sensor, reputed to be quite good in low light because it’s backlit nature allows much more light to hit the photodiodes.

The S200EXR is a hybrid bridge cam with the S100fs body and lens (mostly) and the F200EXR sensor. A 1/1.6” EXR sensor. Quite large when compared with the HS10. The problem we have here is that imaging-resource never shot any samples with the S200EXR … possible because they already shot the F200EXR, which has the identical sensor. So I will use that camera’s image for this test. And, I will use the SN mode image on the site so that I get the binning advantage as well. The only image they have in SN mode is at 200 ISO, so we will use that. The HS10 will be at 100iso, its best. And the D3100 will show 100iso and 1600iso to see if the bridges can beat it wtih 3 and 4 stops advantage.

Se here are the crops:

hs10_vs_d3100_vs_f200exr[1] My analysis:

  • This is not a contest at all. Even at 1600 ISO the D3100 crushes the two bridge cams for hair detail. A slight amount of grain has crept in above the eye, but that is easily vanquished with NR.
  • The F200EXR beats the HS10 handily. There is some hair detail, although the clumping is rampant.
  • The HS10 is embarrassing here. In low light, the contrast is low enough that the cam’s jpeg engine simply goes nuts on the hair and makes it look blurred.

So … if you shoot low light, it is pretty obvious what you should be shooting …

2 comments:

Yeroc said...

I'm curious how the Sony NEX series would hold up. Have you done any testing with one of them?

Kim Letkeman said...

Yeroc: No, I have not tested a NEX camera. but I did include some crops from imaging-resource to contrast to other similar cameras ...

http://kimletkeman.blogspot.com/2010/10/is-d3100-improvement-over-d3000-at-high.html