Test driving a camera

I need to have use of a second camera body this Saturday for a wedding, and a friend in the camera club was kind enough to lend me his Nikon D700 last week for a couple of weeks.  That camera is probably topping my photography equipment wish list, and also on paper ideally suited to wedding photography as it is supposed to outshine its baby brother, the Nikon D300, which I shoot with currently, in the area of high ISO performance – something that is important for the low light situations that crop up at weddings (and especially winter weddings).  A little test, including another photo, below.

The camera I was loaned comes equipped with a Sigma 24-70mm full frame lens (it being a full frame camera), and I was glad to see my f/1.8 50mm works nicely on it too and makes it nice and compact – this could well be my chosen setup for many of the detail shots in the bride’s house on Saturday morning.

I took it home with me on the Luas the night that it was given to me and so while enjoying the 20 minute spin out to Sandyford thought I might as well start my test drive.

With my D300 I’m very comfortable shooting at ISO 800 – if I need to I can push it to perhaps ISO 1200 like I did during Aileen & Neil’s ceremony a few weeks ago.  I start sweating if I’m forced to ISO 1600 as I know there’ll be lots and lots (and lots) of noise – colour noise at that – and the image can be compromised.  But not going there can often compromise the image anyway with a less than ideal shutter speed.  I’ve accepted a long time ago that f/2.8 is a fact of life for wedding photography.

What I was hoping I’d discover with the D700 is that these limits are pushed higher, and that there’s a little more leeway if it’s required.  So my first test was to take a shot in ambient light at ISO 1600 – where I’m not happy with the D300 – and see how the D700 performs.  The resultant image you can see above, and I think it looks great (in terms of ISO performance I mean).  A 100% crop showing the level of noise is at the right, and you can see that it’s very usable.  That’s without any noise reduction and just some basic RAW processing.

Straight away that’s a result – there’s a lot of black in this shot due to the fact that it is a night shot, and even in those dark shadows the noise is, to my eye, very acceptable.  At ISO 1600 on the Nikon D300, if I was going to use the shot, it’d probably be as a monochrome in the hope the colour noise might look more like grain. And it rarely does.

So let’s see how the ISO performes when pushed to 3200.  For this second part of the test I made a late-night mistake, and although I increased the ISO by just one stopped, I closed the aperture down by 2 stops (so it’s f/5.6 @ 1/60s, ISO 3200).  The resultant file was therefore 1 stop underexposed and needed to be pulled back a stop in post-processing.

In a way this is a good real-world test in that it reflects what might happen on a wedding shoot where RAW might help you correct a poor exposure – for instance one that might happen if a guests flash fired as you were metering – in this way.

First let’s take a look at the 100% crop of much the same area as above, and you can see that it’s certainly noisier, but it’s still actually usable.  This one, especially on skin tones, would be more likely to be considered for a monochrome conversion, though again remember this crop is without any noise reduction done, and with the RAW file exposure adjusted by +1 stops, which I would suspect doesn’t do a noisy image any favours in the shadow regions.

For web use, I would have no problem with images at ISO 3200 from the D700.  In fact, later on that night I took a blurry shot from the passenger seat of the car of a cool moon that was out that night at ISO 12800 and even that shot wouldn’t have been going straight to the bin (if it had been sharp).

A blog-sized version of the ISO 3200 Luas shot is below so you can judge for yourself how you think it performs.  I would safely say the D700 is a stop better than the D300, and probably closer to two stops.  I look forward to a little more testing with it over the next week, with its big day out coming on Saturday.  I’ll keep you posted as always about the results.