Sunday, December 22, 2013

Fuji XQ1 Review – Part 5 – More on Dynamic Range and photographing back lit objects like Amaryllis (Hippeastrum)

To recap the impact of the XQ1’s X-Trans sensor on dynamic range – it does not have hardware dynamic range extension as does the EXR line of sensors. Instead, it uses Fuji’s software implementation, which demands that ISO is raised stop for stop of dynamic range extension. So adding a stop of dynamic range requires that you shoot at minimum 200 ISO, and 2 stops requires 400 ISO. That’s how it has been since they introduced the feature in the S100fs and that does not seem to be changing.

Note: The software DR extension in other brands generally does not have ISO restrictions, being implemented purely in software. I have no idea why Fuji chose this method, but it does lessen the effectiveness of the technique somewhat.

So I have this Amaryllis bulb that I planted perhaps 6 or 8 weeks ago. Being a fairly large bulb, it has done amazingly well. I have it in front of a kitchen window so that it can get access to as much light as possible and it has responded well.

Here is the first image I got from the XQ1 of the plant in full bloom. It is backlit (obviously) and I shot it at 100 ISO, which forces DR100. Thus, it has the best noise profile and the worst dynamic range. This is the JPEG and you will note the blown out background and fairly dark foreground. Not my favourite result.

Fuji xq1  10.9mm  100iso  f/3.8  1/15

Now here is the same image from the raw file:

A much better result, but the background is harsh with many blown out areas. Bummer. So even in raw the DR100 files have poor latitude. No surprise, just be aware of it.

So what happens of we shoot at 400 ISO with DR400 in effect?

fuji xq1  6.4mm  iso400  f/4  1/40

Now that’s more like it …

The same angle at 100 ISO and DR100 to refresh your memory …

fuji xq1  6.4mm  100iso  f/2  1/40

These were shot minutes apart, so there is no hocus pocus going on here. And both are raw images processed pretty much the same. I think the DR400 is a tad warmer, but that’s not the relevant difference.

The plant is quite full here as it has two stems in full flower. Eight large blooms together is a beautiful show. This backlit scene never had a chance to retain detail in the background at DR100, but in this case the result is ok …

Clearly, DR extension works. And works well. The question is whether it creates too much noise and edge destruction for your taste. Here, you cannot tell the difference in the RAW images and I can say that the JPEGs show no difference either. I will show more detailed comparisons later.

That evening, I shot a few images at high ISO with only the under-counter fluorescent lights across the kitchen for lighting. I corrected white balance on the flower itself, using the white parts of the petals as neutrals.

At 3200 ISO, I think the XQ1 is serviceable …

fuji xq1  25.6mm  3200iso  f/4.9  1/13

That’s one seriously full Amaryllis.

A few days after that series was shot, the first stem had lost all but one of its blooms. And a third had popped up. A very impressive performance by this bulb I must say.

I shot this series from across the room at full zoom on the XQ1, which means that the slow lens came into play. I also stopped it down a bit to f/5.6 to maximize sharpness. Here is the JPEG and the background is fully realized at DR400.

fuji xq1  25.6mm  640iso  f/5.6  1/30

Now here is the raw image, processed to my taste.

That’s more like it … just a bit of opening of the shadows and a slightly warmer presentation. Note, however, that I still have the auto white balance tuned to add warmth as documented in another part.

So how about this for a treat? The Fuji F200EXR takes on the XQ1. Let’s see how the DR and image quality compare …

fuji f200exr  28mm  3200iso  f/4.9  1/180

Huh? Why would the auto in P mode choose 3200 ISO when it easily could have dropped safely to 800 ISO? Well, who knows. Fuji have had problems with such things in the past and when you compare cameras you sometimes see bizarre issues.

Of course, the background has been held. Sort of. The camera just seems really insensitive next to the XQ1.

Anyway, it’s acceptable, but you have to take control of the F200EXR obviously, since it makes some pretty bad decisions with backlit subjects. And there is that cold Fuji white balance that has plagued the line forever. In this case, were I to reshoot with the F200 I would choose cloudy white balance.

Finally, just to show you a really unusual twist with this second stalk of the Amaryllis, here are the five blooms. Wicked …

Note: Macro mode used for both cameras, 3200 ISO chosen by both cameras (correctly, this time.)

fuji xq1  6.4mm  3200iso  f/5.6  1/30

fuji f200exr  6.4mm  3200iso  f/3.3  1/45

Not much of a contest here. The jpeg engine in the XQ1 is vastly superior. The sensors are pretty close in size, but there is really no contest after downsizing. I did get a bit closer with the F200 and that shows in how nicely blurred the background is. But I think I could have replicated that with the XQ1 … I will test them head to head more seriously in a future part.

So … conclusions:

  • Check your Amaryllis bulbs before purchase. Bigger is better, always.
  • You are safe enough with the XQ1 in DR400 despite the forced ISO 400 … it is clean enough to make it worth while with much better backgrounds and real improvements in dynamic range, even in raw.
  • Tune your JPEG engine if you want to get rid of the cold presentation. As in:

    This might be a tad strong for some … just tweak it until you like it.