Panasonic Lumix DMC-FZ18 Review
In addition to the auto white balance mode the FZ18 offers five presets (daylight, cloudy, shade, halogen and flash). There is no preset for fluorescent lights, but there is a manual white balance mode that allows you to point the camera at a white or gray card and create up to two custom settings.
In use - especially outdoors - the FZ18 delivers consistently accurate color. Under artificial lighting the results are not as convincing. While the performance of Auto White Balance under incandescent light list just about average you'll need to use manual (custom) white balance under fluorescent lighting.
Fluorescent - Auto WB
|Incandescent - Auto WB
Red 4.0%, Blue -8.2%
|Incandescent - preset WB
Red 8.4%, Blue -16.8%
No real complaints here. The range is good (around 1 to 20 feet with auto ISO), and color and exposure very reliable. Flash recycling, when the battery is full and your subject isn't too far away, is pretty good too; you can get shot-to-shot times of just over two seconds.
Good color, +1 flash compensation required (due to the white wall).
Good color, good exposure
The FZ18's Macro mode allows you to use the full focusing range of the lens (from 1cm at the wide end of the zoom and from 200cm at the long end), though leaving it turned can slow down non-macro focusing a little. At the wide end of the zoom you're able to capture an area just over 25mm across; at the long end it's 94mm - this is surprisingly good for a 504mm (equiv.) focal length. You'll see some fairly strong corner softness and some fringing in the wideangle test chart, but this is less of an issue in most 'real life' macro work.
Barrel and Pincushion Distortion
Given the huge (28-504mm equiv.) range, the lens shows remarkably little distortion. The 0.7% barrel distortion is just about visible in wideangle shots, but not a major problem in real world shots unless you shoot a lot of architecture at the wide end - and is a lot lower than many cameras with much less ambitious zoom ranges. There is virtually no measurable distortion once you start to move into the middle and long end of the zoom. If you open the full size images you'll see that there is a tiny bit of corner softness at the wide end of the zoom, but again this should not translate into any real life problems. Again, there's some fringing in both these shots - more than we saw in any 'real world' samples.
|Barrel distortion - 0.7% at Wide angle
Equiv. focal length: 28 mm
|Pincushion distortion - 0.0% at Telephoto
Equiv. focal length: 504 mm
Specific Image Quality Issues
Like the FZ5,FZ7 and FZ8 before it, on the 'camera' side the FZ18 is a remarkably reliable picture-taking tool, with the 'hit rate' (of properly exposed and sharply focused images) only starting to drop in low light, when shots at the long end of the zoom become a little more hit and miss focus and camera-shake wise, and noise becomes more of an issue.
Of course nothing is ever perfect, and the FZ18 - like so many similar models - produces results that, whilst fine for prints up to about 5x7 inches, don't bear close scrutiny that well. At ISO 100 (and in good light) the only real complaints are a slight softness combined with rather excessive sharpening and the usual tendency to blow out highlights in very contrasty scenes. The huge 18x zoom range is obviously pushing things a little too far, and even the Leica badge won't hide the fact that contrast and sharpness drops visibly as you reach the longest focal lengths. And despite the fact that Panasonic has turned down the noise reduction at lower ISO settings we were still seeing NR smearing in shots, and in some cases (when the light was a bit dim) noticeable shadow noise too.
But, pixel peeping aside, it's unlikely anyone will find much cause for complaint when producing standard sized prints from low ISO shots taken in good light; color is excellent (bright without being unnatural) and the default contrast, though a little high for our tastes, is perfect for the 'take it, print it' user. The extensive in-camera image parameters and the ability to shoot raw mean the more discerning user has more than enough control at his or her fingertips if the default settings don't produce the desired result.
In low light the FZ18's biggest asset - that huge zoom - becomes considerably less useful. As the maximum aperture starts to drop so the need to turn up the ISO becomes more pressing, and - like all cameras using such a tiny sensor - this is something you really want to avoid, leaving you with the stark choice; camera shake or excessive noise and desaturation. The focus also slows down in low light and to be honest I'd simply not recommend the FZ18 for the kind of social snaps taken in low light that make up a large proportion of the pictures taken by the average user. Take it walking, take it on safari, just don't bother taking it to a dinner party.
We found a little purple fringing and a little chromatic aberration when looking very closely at high contrast edges at the wide end of the zoom (and in macro mode), but it's not a major problem.
The FZ18 has the usual problem associated with limited dynamic range and a steep tone curve (high default contrast), which means that in very bright, contrasty situations you will see some highlight clipping. To be fair the problem is a lot less prevalent than we've seen with other models, thanks mainly to an exposure system that tends to slight underexposure in such situations. You can also (as discussed later) get some of the highlight detail back by shooting raw, though don't expect miracles. The only point at which highlight clipping becomes a serious issue is on the thankfully rare occasions where the exposure system gets it wrong and over exposes.
|100% crop||69mm (equiv), F3.2|
Noise and NR at low ISO settings
Shooting in anything less than perfect light brings out the worst of the FZ18's sensor, with visible shadow noise and obvious noise reduction artefacts even at ISO 100. It's actually slightly better than the FZ8 (and for that matter better than the Sony H7/H9), but that doesn't make it any less unpleasant. Once again we see the limitations of a 1/2.5 inch sensor when it comes to capturing the texture and fine detail in anything but glorious sunshine.
|100% crop||504mm (equiv), F4.2|
In the course of our studio testing we came across one rather unusual side effect that perhaps reflects the sheer amount of processing the FZ18's Venus III is doing to JPEGs. In what appears to be a mix up between the demosaicing and noise reduction systems some mild moiré (the FZ18 appears to have a fairly light anti-alias filter) has resulted in unsightly blotches of gray squares (desaturated).
The artefacts are most visible at ISO 200. We have not been able to reproduce these artefacts in any of our real life shots, it is highly unlikely they will create a problem in everyday shooting.
|100% crop||114mm (equiv), F3.6|
The MEGA O.I.S image Stabilization system used on the FZ18 (and most other recent Panasonic models) works, and it works well - perhaps better than anything else on the market. There are two modes: Mode 1 (IS on all the time) and Mode 2 (IS is activated at the moment the exposure is made). Mode 1 makes framing easier - the IS system steadies the preview image. Mode 2, which minimizes the amount of movement needed by waiting until the actual moment you press the shutter, is marginally more effective, but there are occasions when mode 1 works better - perhaps because it seems easier to hold the camera steady when the display isn't jiggering around.
That all said, there doesn't seem to be a hard-and-fast rule; sometimes both modes are equally effective, sometimes one works better than the other (though as the charts below show, overall mode 2 is the best). In all cases I'd recommend shooting several frames if your shutter speed is really low - that way you've got a much better chance of success.
But one thing is for certain; image stabilization works when it comes to avoiding camera shake; we certainly found the 3 or 4 shutter speed steps gain (over the recommended minimum unstabilized shutter speed) claimed by Panasonic to be justified - and were able to shoot at 504mm equiv. at speeds as low as 1/25th successfully. Impressive stuff.
The stabilization test
In this simplified version of our SLR IS test, eight hand-held shots were taken of a static scene with the stabilization off, on (mode 1) and on (mode 2). The shutter speed was decreased by a stop and repeated (from 1/1000 sec to 1/15 sec). The zoom was set to its maximum position (504mm equiv.), the test chart was 4.5 m away from the camera. This procedure was repeated ten times.
The resulting images were then inspected and given a blur score - 'Sharp' (no visible blurring at 100%), 'Mild Blur' (the kind of camera shake that is tolerable at small print sizes) 'Heavy Blur' (unusable due to camera shake) and 'Very Heavy Blur' (little discernible detail).
As the charts below show we were able to get a measurable two to three stop advantage. More importantly we were able to get 'usable' results in two out of five shots at shutter speeds as low as 1/25th second (Mode 2).
Hand-held, no stabilization (504mm equiv.)
As you can see from the chart below once we dropped to below 1/80th second we had little or no chance of getting a usable shot, and only at 1/1000th second were we guaranteed of a sharp result every time.
Hand-held, stabilization mode 1 (504mm equiv.)
With mode 1 we were able to get usable shots (with only mild blur or no blur at all) down to 1/25th sec. If you take several shots at the borderline shutter speeds (for me, 1/80th and 1/40th at full zoom) you should get a few sharp ones. Once you drop to 1/25th sec or below the chances of producing usable material deteriorate quickly.
Hand-held, stabilization mode 2 (504mm equiv.)
Mode 2 is obviously more effective, though the difference is fairly small, and at shutter speeds down to 1/80th sec you have a better than odds chance of getting a completely sharp shot. Even down as low as 1/15th second you are getting usable, if not completely sharp, results.