This purchase was really hard to justify. So in the end I didn't bother justifying it and simply handed over my credit card. Actually I knew that I was about to get a good bonus for working over Y2K but, even so, I didn't need an F5. Need is irrelevant. Resistance is futile.

Certainly, there is quite a technological leap between the F4 and F5. The body feels quite strong and robust. The handling is good - but not quite as good as the F4s, in my opinion. The F5, though, doesn't seem to be quite as well put together as the F4. Tolerances seem to be lower: the main switch/firing button assembly wobbles a bit and the rubber doesn't fit the body exactly, resulting in bits coming unstuck (I've noticed this on several F5 bodies, especially around the front aperture selector dial). There is none of this on the F4; everything seems to fit everything else exactly and there is no button/switch assembly wobble at all.

The above may sound as though I don't like the F5. This isn't the case at all, since it certainly has its strengths and I've grown to like it more the more I've used it. Its strengths are concerned mainly with speed and accuracy: compared with the F4 there is little AF hunting and the F5 locks on in much lower contrast situations. The five focus areas are useful, especially with moving subjects, although my main use for them has been with off-centre subjects (focusing on somebody's eye, for example, without having to re-frame too much). The matrix meter is reliable, if a little overly cautious in high-contrast scenes. For example, highlights anywhere in the frame will usually be perfectly exposed, sometimes at the expense of shadows which are too dark. I have subsequently found the F6 to behave differently, in that it will sometimes blow highlights in favour of the main subject; something the F5 will rarely do. With neutral subjects, matrix readings between the two are virtually identical, so it's just the way the F5's matrix algorithm is set to deal with such subjects. I generally prefer the way the matrix meter of the F6 works, although, of course, neither is infallible.

The F5 has superior flash capabilities when used with the SB-25 upwards. A multi-segmented meter pattern is used for flash and the camera is able to take into account the distance information provided by D lenses, which is in contrast to the F4's centre-weighted flash meter and its inability to use D information. I don't use flash an awful lot - I have used it both as the main light source and for fill-in and it works very well. I am certain that the F5 will be more accurate than the F4 for the above reasons but, considering the limited amount of flash photography that I have done, I can't say that there's much of a difference between results from either.

My F5 has a bit of a problem which my F100 (which has very similar functionality) does not exhibit. On its own it's not a major problem and I'm sure it's within tolerances, however, the result of the problem has been accentuated in the past by a couple of AF Nikkors I used to own...

When selecting the lens aperture via the F5 dial, rather than on the lens itself, the result is slight underexposure. This is only observable on slide film as negative film has greater exposure latitude. I have observed between approximately 1/6 and 2/3 stop underexposure depending on lens used and aperture set. After several films' worth of testing, using the F5, F100, F4 and several lenses, I have come to the conclusion that my F5 is out by an amount between 1/6 and 1/3 stop, depending on the value set. The rest (making up a total of about 2/3 stop maximum) has been down to one or two lenses. This has been confirmed by testing the F100 with all lenses. It's surprising how noticeable 1/6 - 1/3 stop underexposure can be (I can only guess the value as the image appears to be exposed somewhere between "correct" exposure and 1/3 stop under). Couple that with, for example, my (now ex-) 35-70mm f/2.8, which tended to underexpose very slightly when the aperture is set via the camera and not the lens ring (confirmed on F100), and the cumulative result is very noticeable underexposure with slide film. Unless the lens in use is a "G", I can get around this by setting the aperture on the lens ring rather than the camera dial, as is required anyway with the F4, F3, etc. I don't doubt that my F5 is within Nikon's error tolerances and the error is not too surprising considering the mechanical couplings and the very small degree of lever movement which equates to quite large aperture changes. It just so happened that slight errors in two of my Nikkor AF lenses resulted in adding to the underexposure. My Tamron AF, on the other hand, has internal error of the opposite sense which results in very slight over-exposure when used with my F100 in aperture-set-via-dial mode compared with using the aperture ring. This effectively cancels out the error in the F5, resulting in images which are virtually indistinguishable when both methods of aperture setting are used. Using my 24mm f/2.8 prime on the F5, results show only a very slight underexposure in aperture-via-dial mode cf. via the lens ring which is due to the camera only (no internal error in the 24mm). The results are acceptable and I probably wouldn't have even noticed that there may have been a problem if all of my lenses were without their own error like the 24mm. Incidentally, my "problem" lenses were two zooms: 35-70mm f/2.8 D and 80-200mm f/2.8 D, neither of which showed any real-world flaws on any other body. The only reason I noticed is that I happened to have a slight adjustment error in both body and lens which reinforced each-other.

I won't put details of my testing here - suffice to say that I am confident that my F100 is correct and my F5 is out by less than 1/3 stop (probably 1/6 - 1/4 EV or so). I don't believe that the problem is uncommon and it seems reasonable to assume that some may slightly over- rather than under-expose. I was in dialogue with somebody some time ago whose F5 had the same problem as mine. His had been to Nikon Service several times and each time came back with the statement that the error was within tolerances. Like me, he set the aperture via the lens to get around the issue (this, of course, won't be possible in S or P modes - only in A mode - and not at all possible with "G" lenses). Other people, if they notice and care, may set an exposure compensation. Many people prefer to slightly underexpose slide film anyway as it tends to saturate colours but I generally don't, especially as it makes scanning them more difficult.

I had to have my F5 repaired a few months after buying it because the LCD display in the viewfinder stopped working properly: three of the segments making up the frame counter were permanently off. The whole display module was replaced under warranty and now all is fine again.

To conclude: the F5 handles well, is fast, robust, responsive and has advanced functionality which I've barely touched on here. I put loads of films through mine and played with it a lot, yet, after well over a year, it was still running on its first set of batteries. Yes, they were lithiums but they do seem to last ages (in mine at least; I've read of battery problems but these were probably in earlier models)* . Better than my F4s? In many ways, yes but in some others, no. The F4 seems to have been made with closer attention to detail whereas the F5 has a somewhat more "mass produced" feel about it. In purely functional terms it is way ahead of the F4 and is a lovely machine to use but, for me, the F5 can't quite match the F4s in terms of handling and ultimate quality feel.

* I had to change the first set of batteries, a few months after writing the above. I tested them with a multimeter and I was happy to discover that the F5 did truly drain those lithiums completely. Many earlier reports that I'd seen of the F5 and its apparent battery problems described how the it would lock up, saying the batteries were dead before they actually were (that is, when voltage had dropped only slightly). So I'm happy that mine doesn't do that and that batteries seem to last a long time. Incidentally, I get by for quite some time using alkalines, too.