2. User interface and usability

Good overall responsiviness makes D70 joy to use. One can shoot RAW images at a fast pace without any noticeable delays. There are, however, user interface issues I'm not totally satisfied with. I don't have any previous experience with Nikon DSLRs, but I was surprised to find out that image previewing on camera LCD is not thought out better. Especially since D70 is far from being the first DSLR from Nikon. First thing I'm dissatisfied with is the overlay texts on those pages which show textual information together with preview image. The font of the text is somehow soft and difficult to read compared to much smaller LCD display of Canon Digital Ixus i for example. Also some of the text could be easily replaced with symbolic information (metering, white balance etc.) and thus all the data now presented in two separate pages could be put to one page. There is also some totally useless information here, like the camera make and model. Why would I want to know that the image I'm previewing was taken with Nikon D70?

Histogram had quite good resolution, but in a camera of this level it would not be too much to ask to have different colour channels shown separately. The biggest problem in image preview is zooming. One button, marked with magnifying glass, is needed to active zooming and zoom image to approximately 2x the size of display. But when I want to zoom in from there, I need to press another button and keep it pressed while rotating main command dial. Similarly, I can also zoom out, but not enough to see the whole image. Instead I need to press the magnifying glass button again to cancel zoom mode. After getting used to this two button and command dial system surely works without thinking too much, but it still needs lots of work to get to maximum zoom when I want to check image sharpness. Phil Askey's Nikon D70 preview says pressing magnifying glass button goes directly to last used magnification, but this is not true for my D70 at least.

Another quirk also relates to viewing zoomed preview image. For some reason I cannot understand the whole display area is not used. And if I select to see portrait images rotated this feature causes wasting of half of the display!

3. Exposure metering

I've always been totally satisfied with the performance of matrix metering in my Nikon F80, and also in older F601 for that matter, but now that I can immediately see the histogram of each image I take using D70 I'm having hard time getting used to fact that even 1005 pixel colour matrix meter is not real artificial intelligence after all. First two days I had feeling that I need to dial in positive exposure compensation, either +0.3EV or +0.7EV, almost all the time.

But yesterday I was shooting a dance performance where there are brightly lit people in white clothes against black background and then I for the first time saw also need for negative exposure compensation when most of the image was dark background. When shooting film in the same conditions before I've always had positive exposure compensation of +0.5EV all the time and forgiveness of Fuji Press 800 has usually been enough to deal with it. Of course I was aware that with film replaced by CCD this all could change dramatically, and it sure felt like it did. Sometimes very small changes in the composition of a shot resulted in need to alter exposure compensation 2/3 stops to either direction. The good news was that most of the time no compensation was needed.

Photo 1. In this kind of situation one cannot blindly trust matrix metering (Tokina 20-35mm f/3.5-4.5 @ 24mm, 1/20s f/5, 800 ISO, -0.3EV).

4. Autofocus

Based on somewhat inaccurate timings AF motor speed seems to be similar to that of Nikon F80. But the feeling I got while shooting both in the same conditions yesterday (using only central AF sensor on both) D70 feels like having a slight edge when it comes to snapping into focus. Also the sound of the AF motor is more civilised and silent.

5. Hot pixels

My D70 does not currently have any hot pixels in CCD or LCD. Dark frame shots exposed for 30 seconds show noise, which is depending on the ISO setting. In and near the upper left corner there are two areas where noise is more pronounced, some other website said that this is because of warmth caused by some amplifier near the CCD. 

Photo 2. Equalized image of dark frame noise 30 second exposure at 800 ISO. Noise reduction was switched off. Please notice that JPEG compression has removed much of the noise in this small image. 

6. Colour Moiré and NikonView

There has been lot of talk in some discussion forums about so called colour moiré problems in D70. I haven't had time to investigate it any further yet, but I did took some RAW photos to see if it would show up easily. It did, but not so easily I had thought. Taking a closer look at that image I found out that moiré looked completely different when viewed with 300% zoom NikonView or when imported to Photoshop for editing.

Photo 3. Colour moiré seen in zoomed view of NikonView and Photoshop.

It's obvious that NikonView applies heavy sharpening when NEF images are shown at the magnification of 100% or more. But still I cannot understand why the image looks so much different when it is actually imported to Photoshop for editing (I did that from File-menu of NikonView if it matters). It looks like completely different CFA interpolation algorithm is used. I guess NikonView must use some quick-and-dirty interpolation to faciliate faster image viewing.

7. Colour Moiré and NikonCapture 4.1

I've been using 30 day trial version of NikonCapture. Unlike NikonView it offers special filter for colour moiré reduction. After shooting almost 1000 images I have not encountered real problems with colour moiré too often. But in the image below is a real world example of colour moiré in hair. In this image I noticed the problem right away when I was looking at the image at 100% magnification.

Photo 4. Colour moiré in hair. Images are shown at magnification of 300% and no sharpening was applied. Setting colour moire reduction filter to high level seems to work while lower levels do not. Overall colour difference in the images is probably do some mistake I made, not caused by colour moire reduction filter.

8. Colour shift with high shutter speeds

Another problem, or should I say feature, many people have been worried about is colour shift in image when highest available shutter speeds are used. I've seen very bad example images, but then again it seems that my D70 is quite good in this respect as seen in photo 5..

Photo 5. Image of pure white snow taken using 1/8000s shutter speed and  200 ISO sensitivity. Whole original image shrinked on the left, and the same image after setting saturation to +95 in Photoshop on the right. You can see some green and magenta colour cast in left and right ends of the right image.

9. First trials in IR photography

In his Nikon D70 Review Bjørn Rørslett reported good infra red performance. At first I couldn't locate my plastic IR filter so I tried another approach to IR photoghoraphy. In his  another page Bjørn reported getting "false-colour" infrared images using Nikon D1 together with stacked polarizers. The idea is to use two polarizers so that maximum attenuation of visible light occurs. IR (and possibly UV) is not affected as much as visible light and interesting false colour images can be captured.

I wanted to try this with D70 as I happened to have two polarizers and one of them was linear, which is a requirement here (or maybe it could work with two circular polarizers if first of them is reversed, I'm not sure). My linear polarizer was Hoya 67mm  and circular B+W 62mm. I just taped them together and used step up rings to fit them to 52mm filter thread of my AF Nikkor 50mm f/1.8.

We still have snow covering the ground in Finland this time of the year and there are no leaves in any leaf trees, but I was happy to see this is really working as you can see in the photo below.

Photo 6. Same scene photographed using stacked polarizers to maximise visible light (left) and to minimize visible light (right). Right photo clearly shows IR effect in spruce needles (autolevels applied in Photoshop). Exposure for right image was 1.6s f/11 @ 200 ISO.

I also shot photos showing more sky although it was mostly cloudy. The red channel of them gave look very familiar from film based monochrome IR images with almost black sky and very light trees (spruces and pines that is).

Later I found my plastic IR-filter I could try real IR photography with D70. The filter is just cheap piece of acrylic plastic I obtained from company called Teknofokus. Passband of that filter begins around 730nm and it costs only couple of euros, so it's ideal for novices like me to get some feeling of IR photography. Results are looking pretty good considering lack of both leafs in trees and experience of photographer in IR photography.

Photo 7. The same spruce as in previous photo photographed with plastic IR filter. Left photo is directly from camera and right photo was converted to monochrome in Photoshop. Lens was Tokina AF 20-35mm
f/3.5-4.5 and exposure 4s f/11 @ 200 ISO.

 

10. CF write speed

In marketing material Nikon is suggesting that using a fast CF card will help shooting long series of images continuously. Rob Galbraith has already extensively tested CF write speed on Nikon D70. I still wanted to do my own tests to see if I can get similar results. Results are seen in table 1, and photo 8 shows the cards I tested.

Table 1. Comparison of write speeds with different CF cards.

One should not need any other proof to believe that using a fast CF card makes huge difference in this camera.  Write speed to SanDisk Ultra II CF is 7x that of slowest CF I happened to have available for the test.
 

Photo 8. CompactFlash cards used for writing performance tests.

11. CF read speed

Nikon D70 features USB 2.0 connectivity to computer. This together with the fact that D70 writes fast CF cards quickly can make you believe transferring images from camera to computer should be pretty fast just by using the USB cable supplied with the body. Still many serious and professional photographers prefer using separate card readers.

In search of card reader to I tested three different card readers to see if they can keep up with a fast CF card like Sandisk Ultra II. The readers tested were:

• Sandisk ImageMate CF/SD/MMC USB 1.1 model SDDR-73

• Kingston 2.0 6/1 Hi-Speed USB 2.0 Memory Reader FCR-HS26/1

• Hotreader 7-in-1 Card Reader HR1-U27 USB 2.0
  

Photo 9. Hotreader 7-in-1 Card Reader (AF Nikkor 50mm f/1.8, 1/160s f/4, 800 ISO).

Measured transfer speeds from CF card to computer are listed in the table below. Included are also Nikon D70 connected to computer using USB cable and Rob Galbraith's result obtained with Microtech FireWire CameraMate CF reader for SanDisk Ultra II CF.

Read speed in Megabytes per second	MMC	Nikon 16 MB CF (by SanDisk)	SanDisk 1 GB Ultra II CF
SanDisk Image Mate	0.9	0.9	0.9
Kingston Memory Reader	1.6	1.8	2.2
Hotreader 7-in1 USB 2.0	1.6		8.4
D70 connected via USB 2.0			0.8
Rob Galbraith's result			9.3

Table 2. Comparison of read speeds with different card readers and Nikon D70 (connected via USB 2.0).

It's obvious that USB 2.0 connectivity of CF reader or camera does not guarantee fast transfer speeds.