Another generation of high-end compacts square off.
The Back Story (Continued)
If you're just tuning in, I covered the G10, P6000, and LX-3 in a previous comparison article. The competition at the high-end of the compact camera market is an ongoing one, and the results change a bit during each generation. Read that article for more background.
We're now almost two decades into the competent digital compact camera era, but many things remain the same. The big issue has been and remains the small size of the image sensor in these cameras. While there has been a strong increase in image quality over the years, the small sensor sized used just doesn't allow the high-end compact camera to produce DSLR-like results in dynamic range or high ISO shooting.
Nevertheless, there's something compelling about having a competent and controllable camera that you can stick in a pocket, even if it is a large pocket. With the typical DSLR and lens combo weighing in at a hefty two pounds or more, there's a big difference between carrying that and a compact camera everywhere you go. Simply put, you won't carry a DSLR all the time, but you might carry a quality compact all the time.
Each of the camera makers here has come at high-end compacts a slightly different way. Canon has had a long and consistent line of G series cameras. From about the G5 onward, the body size and design hasn't changed a lot for Canon. Today's G12 design is clearly recognizable in, say, a G8 or G9. Canon early on got most of the size and design issues squared and has concentrated mostly on iterative aspects of controls and the imaging system. Changes in sensors used has changed Canon's optics a bit from generation to generation, with the last generations offering slightly less focal length range than some previous ones.
Panasonic, too, has taken a consistent path from the original LX-1 in 2005 to the current LX-5: size, shape, and basic control designs have remained relatively intact across generations, again with the main changes being iterative improvements to controls and the imaging system. Like Canon, there have been small tweaks to lens specifications, often linked to slight changes in the sensor used.
Nikon is the odd man out here. Nikon has been completely and utterly inconsistent in high-end compact camera design. We went from the twist-and-shoot style pioneered by the Coolpix 900, to the juxtaposed odd couple of the big-and-square 5400 and bridge camera-like 8700, to the teeny (but no raw) P5000 style, to the gimmick laden P6000, to the current P7000. Is is any wonder that Nikon has fallen out of favor in the high-end compact crowd? All their design decisions seem to be dead ends when looked back at historically, so that doesn't exactly exude confidence in the future of the current offering. As it turns out, this may have produced an operational aspect that comes into play and hurts Nikon, too, but I'll get to that in a bit.
Last time around (G10, P6000, LX-3) we saw a number of clear differences in various areas of these cameras. Each seemed to have a wart or two, though the Canon tended to be the most "on balance" of the bunch. This time around, each company has had another chance to refine their entry, and interestingly, the cameras are coming slightly more towards a "center" in terms of capability.
So let's meet our contenders. For this quick test we've got the Canon PowerShot G12, Nikon Coolpix P7000, and the Panasonic Lumix DMC-LX5 (also sold under the Leica D-Lux 5 name). While testing these, I was comparing them against the previous generation, against mirrorless cameras like the Olympus E-PL1, Samsung NX100, and Sony NEX5, and against a small handful of other high-specification compacts that have appeared recently. But the three cameras in this test all fall into a fairly narrow range that's right in the sweet spot of the "competent pocket camera" market.
So let's look at the key specifications:
||24-90mm f/2-3.3 (3.8x)
||80-3200, 6400, 12800
||2.8", 461k dot, positionable
||3", 921k dot
||3", 460k dot
||15s - 1/4000
||60s - 1/4000
||60s - 1/4000
|Significant Other Features
||16:9, 4:5, 4:3, 3:2, 1:1 formats, optical viewfinder, 720P/24 max video, ~1 fps
||16:9, 4:3, 3:2, 1:1 formats, optical viewfinder, 720P/30 max video, ~1 fps
||16:9, 4:3, 3:2, 1:1 formats, 720P/30 max video, 2.5 fps
Interestingly, on these key elements it's clear that camera companies are watching each other, as the specifications have gone from being fairly divergent (G10, P6000, LX-3) to being remarkably similar. Pretty much every other feature not listed in this chart you can regard as being close enough to being interchangeable. They all have PASM and scene modes. They all have macro capability, image stabilization, built-in flashes with hot shoes, raw file capability, and a host of other similar features. The bottom line is that they all have relatively rich feature sets and lots of user (manual) control ability. More interesting is that they are all using a similar image sensor, chosen mostly for its ability to do a bit better in low light than the previous generations (in the Canon and Nikon).
Thus, there really are only two things we have to talk about:
- What's the difference in image quality? Are any good enough for a serious shooter to carry around all the time?
- What's the difference in handling? Do any of them get in the way of the serious shooter who wants to control the camera?
I'll give you the short answer to each:
- Differences in image quality do exist between the cameras, though some of those are being driven more by the lens than the sensor. Overall, things have improved in modestly high ISO values with these cameras, so they work well in modest to bright light, less well in low light.
- All three cameras cater to the serious photographer in handling, with Nikon and Canon taking a more retro approach, Panasonic staying with a more modern approach.
Choosing between these cameras, therefore, is not as easy as it was a generation ago: they've all come to a very competent center.
Four things stand out as key differentiations in the feature sets: lens, color LCD, optical viewfinder, and size/controls.
The lens differences are the most obvious to a serious shooter. Nikon has taken the "superzoom" approach (28-200mm), Panasonic has taken the "wide kit lens" approach (24-90mm), while Canon sits somewhere in between (28-140mm). But there are two winners here, depending upon how you assess lenses: the Nikon definitely has more reach than the other cameras and excels for the telephoto-obsessed user; the Panasonic is catering to the indoor, low-light user. This one difference is almost enough for you to pick your camera from. The Panasonic starts with a full stop advantage indoors (f/2 versus f/2.8), and pretty much retains that as you zoom in. Couple that with the wider 24mm and indoor shooters are going to be happier with the Panasonic than with the Nikon or Canon. Outdoors, it depends upon whether you're a landscape shooter (the Panasonic's 24mm) or a more general shooter (the Canon's and Nikon's larger focal length range). I'll examine the optical quality of the lenses later, but in terms of feature sets, you want to pay close attention to what you need from the lens, as it is a key differentiation of these cameras.
Likewise, there is some difference in the color LCD that comes into play. Only the Canon's LCD is positionable. That means that it can be flipped to provide protection while carrying the camera (yes!), or flipped to be viewed in front of, behind, above, or beneath the camera. If you're into weird, extended, or contorted camera positions, the Canon is the only one that's going to help you see your framing. Again, a clear differentiation, and one that may trigger your purchase choice. The Nikon has the best resolution on the color LCD, but in practice I find that it only makes a modest difference in shooting and reviewing images.
If you don't want to look at an LCD to frame shots, the Canon and Nikon both have terrible optical viewfinders. Yes, they deserve the adjective (terrible). They zoom with the lens, but they provide ~80-85% views that aren't parallax corrected. They can't be used for precision framing. Panasonic takes a different approach, allowing use of the VLF-1 electronic viewfinder that originally appeared with the GF-1 m4/3 camera. This is a low resolution (202k dot), 100% view, tiltable accessory that, despite that low resolution, still provides a far better compositional aid than the optical viewfinders on the Canon and Nikon. If you want to look through something to compose, the Panasonic is the only choice, in my opinion, but it comes with the penalty of extra cost and adds bulk to the smallest camera in this test, and it's a pretty grainy EVF.
Which brings me to size. The Canon G-series have always been at the big end of the compact camera size. The G12 won't fit into most shirt pockets, but it will fit into most jacket or vest pockets. It's a semi-pocketable camera. Nikon for some reason went from the diminutive P6000 to a G12-sized P7000. The Nikon P7000, too, won't fit into a shirt pocket, though it is a slightly lighter camera than the Canon. Unfortunately, the Panasonic LX-5 isn't shirt-pocketable, either, because its lens sticks out from the basic camera body enough to not slip into most dress shirt type pockets. But in terms of size, the Panasonic still has a clear size and weight advantage over the others. It slips into a slightly smaller area in my camera bag, it fits into slightly smaller pockets than the others, and it definitely is lighter than the Canon G12.
Once you get beyond these four main differentiation, the features sets look remarkably similar for all three cameras. None has a megapixel or any real video advantage. None has a real ISO advantage (the Panasonic allows Auto ISO to go into levels the other cameras can't hit, but that's mostly unusable). Heck, all three cameras have at least two user settings positions on the Mode dial. It's clear that the three companies are looking closely at the other's offerings and trying to match key feature items.
One small nod for some Nikon users: the P7000 uses the new EN-EL14 battery, currently shared by the D3100. So if you're shooting with a D3100 as your main camera, the EN-EL14 simplifies your battery and charger choices. I hope that Nikon comes up with some other DSLRs using the EN-EL14 battery (the D7000 uses the EN-EL15 battery, unfortunately.
Therefore, if none of the four things I outlined above immediately tell you which of the three cameras you should buy, it'll come down to handling and image quality.
As I've noted, the feature sets of these three competitors have mostly converged, so handling becomes more important in the evaluation of which might be preferable.
The Canon approach has long been "retro" with the G series, and the G12 is no different. Dials for exposure compensation, ISO, and mode; a Front Command dial and back dial around the Direction pad; not a lot of buttons. It's easy enough to get in command of the Canon, and the fact that they've been refining this retro approach for quite awhile (mostly) shows. The indicators on the dials are lit when the camera is on, which makes it easy to see their position in low light. That little touch shows a level of refinement that's nuanced and makes the camera handle infinitesimally better. One glaring dissonance: zooming isn't allowed while taking videos, but I'm not going to really get into the video capabilities of these cameras, as it would take a whole article in itself to get to all the small differences and what they really mean.
Nikon seems to have abandoned the Nikon UI and just copied Canon. Why we have an exposure compensation dial when every Nikon shooter expects a button+dial approach, I don't know. Moreover, labeling a button (right where we'd all expect an exposure compensation button) with the Canon Av/Tv nomenclature just makes Nikon's G-envy all the more obvious. I can just see the engineering meeting now: "If it worked for Canon, let's just do that." This is not my father's Nikon. It appears to be my Uncle's Canon. Nikon does have one interesting twist to the dial theme: they have a dial with a button in it. Twist the dial to a setting (ISO, WB, BKT, QUAL, etc.), press the button and use the Rear Command dial to make a direct setting. Now that's a bit more Nikon-like. But somewhere in the specs they copied from Canon it said "three dials" so Nikon created three dials. Three different styles of dials. So much for the retro "direct set" thing that Canon gets down to a T.
Panasonic has backed a bit off of their previous UI (remember the thumb joystick?). We've now got a Rear Command dial and the more familiar Panasonic button array (Direction pad buttons + dedicated buttons). We still have the Quick Menu idea, it just gets pulled up with a dedicated button now rather than pressing the thumb joystick.
Serious shooters want to set a few things a lot, and do so quickly: exposure mode, aperture, shutter speed, ISO, and white balance. They want to set a few other things easily, too: quality, focus modes, and aspect ratios. They want to completely set the camera to one master set of settings with a single control.
On this last item, the cameras are identical: all three companies now have user setting spots on their Mode dial (Canon and Panasonic: C1, C2; Nikon: U1, U2, and U3). This basically accumulates all your sub settings into one master setting you can flip to quickly. So if you shoot your JPEGs one way and your raws with the camera set differently, you can simply dedicate those user positions to your two master settings (you can still override individual settings as needed).
But what about those individual settings? Are there problematic handling issues with any of them?
Not really. Of the things I've listed above (other than exposure mode, aperture, and shutter speed), all three cameras can be fine-tuned quickly. Panasonic mostly uses their Quick Menu to get you access to those critical items. The Nikon mostly uses the dial with the button in the middle to get you to them. The Canon requires you to either dedicate buttons to oft-used things (a couple of the buttons are configurable), press the Func Set button and navigate to the major items (much like the Panasonic's Quick Menu), or use the menu system. Of the three, the Canon can be a little slower for someone changing a lot of things constantly. Not really enough to worry about, but it is a consideration.
I'll use Manual exposure mode to illustrate how apertures and shutter speeds get set. On the Panasonic, the Rear Command dial controls both. You select which by pressing the dial in (it's a button as well as a dial), which toggles you between the two things the dial can set at any given time. In Manual exposure mode, that's the aperture and shutter speed, and a manual exposure bar appears just above the numbers at the bottom of the display. Simple enough.
On the Nikon it works like a Canon. Yep. Really. The dial surrounding the Direction pad controls the aperture, the Rear Command dial controls the shutter speed. A very small exposure bar appears along the left edge of the display. The Canon, well, it too works like a Canon ;~). The dial surrounding the Direction pad controls the aperture, the Front Command dial (normally it would be a top dial on the Canon DSLRs) controls the shutter speed. A small but readable exposure bar appears in the lower right corner of the display.
So at the basic "I want to be in control" aspect of handling, the cameras all pass my tests for directness and ease of use. The Panasonic takes a slightly more modern approach, but I can live with that. The Canon is a Canon in approach, and the Nikon as I noted, seems more Canon-like than Nikon.
So let's talk about some of the other aspects of the cameras. Flash exposure compensation is buried on page three of the shooting menu on the Panasonic. It's on the quick Func Set menus (item four) on the Canon. It's on the first page of the Shooting menu for the Nikon. While that looks like a win for the Canon and Nikon, I can find other things that are wins for various other combinations (or single) cameras. For instance, the Panasonic has a direct switch for both MF/AF/Macro and Aspect Ratios, while the Canon and Nikon bury the later in the menus and do the former from a menu triggered by a button on the Direction pad. So if there's a non-basic function you use a lot, it probably does pay to take a close look at how you access that on each of the cameras--it could make your choice dramatically easier.
Panasonic uses a switch for the On/Off function of the camera. I still prefer this to the recessed buttons Canon and Nikon use because it's harder to accidentally change it during shooting. Still, be aware that it sticks up enough that it is easy to move the switch when sticking the camera into a tight camera bag. Panasonic's "record movie" button is right next to the shutter release, too, so it's easy to accidentally start recording a movie, too.
Both the Nikon and Panasonic have a "flash up" button or switch, and the flash rises above the top plate of the camera. This gives a slight bit of separation from the lens that you don't get on the Canon, where the flash is fixed into the front of the camera body.
But let's talk practical for a moment. Let's get out the three cameras and use them in the field for awhile and see what happens. I'll preface this by saying I tend to be an Aperture-priority user who dips into manual focus a lot and occasionally is changing exposure compensation. Every now and then I like to look at the 16:9 composition instead of 4:3. Most of the other settings aren't changing often, if at all.
The Panasonic is remarkably like my other Panasonics, including my GF-1. Nothing really distracted me in actual use, and the dedicated focus and aspect ratio switches took care of my most common secondary settings. I find that I "just shoot" with this camera. It has a dedicated lens cap, which I anchor to the left camera strap post so I don't lose it. The camera nags me when I forget to remove it before shooting. As I mentioned, the On/Off switch can be dislodged easily when putting the camera away into a tight camera bag, and I hit the record movie button too often accidentally, but I find that I don't think much about the camera when using the LX-5.
One thing Panasonic needs to do a better job of, however, is figuring out what to display on the color LCD and why. With the histogram on the display, I've got 13 icons also on the display (or 11 and the date). The display is a mess. So you end up hitting the display button and getting a clean display for composition. This is about the only shooting handling problem that bothers me. It would be nice to have more control over what does and doesn't appear. Panasonic's clutter-the-display approach is too nanny-like and too amateurish.
The Canon feels comfortable to me in shooting because I've been using G series cameras for a long time now (back to the G6). I like the positionable LCD and use that more than you'd think, even when shooting "normally." But like the Panasonic, the display is too busy. I do NOT need a giant Av on the display while shooting (or the ISO value) when I can see those setting on the dial at the top of the camera. The virtual bubble level is quite handy, but the histogram could use some work.
The most frustrating camera in shooting was the Nikon, but not always because of the design (okay, Nikon's virtual horizon display is more annoying than Canon's). There are enough firmware bugs that cause the camera to do something other than you think it will (or should) that I can't even begin to list them. Then there are the odd interactions that weren't well thought out. For example, I couldn't figure out why my histogram kept appearing after I unassigned it from the Av/Tv button. Well, that's because it was also set in Monitor Settings (which get ignored when you toggle the histogram off via the Av/Tv button). I understand the interaction now, but it was one of many "do I have to read the manual?" things that came up during one shooting session. A really good design doesn't have many of those, but I keep discovering new ones on the P7000 ;~).
But one thing that really bothered me (other than the repeated Lens Initialization messages) during a cross country trip by plane was that the color LCD was very slow to respond to light changes. Very slow. For instance, my keyboard is in darkness right now and obviously my monitor is well lit. The Canon and the Panasonic recognize the difference in brightness when I do a quick pan from one to the other and adjust immediately. The P7000? Almost 3 seconds. That's enough of a delay to miss a critical shot, and I don't like that kind of flaw on a high-end camera.
Overall, the Canon feels polished. It should be, as it's the twelfth iteration of a design. The Panasonic feels polished except for the font and icons it uses on the display (the font is slightly crude, and the icons too GameBoy-ish). The Nikon feels a bit unpolished, as if a lot of ideas were thrown into the stew, but the cook didn't do enough taste testing and the final results have some conflicting bits.
That said, they all offer a serious photographer pretty much everything they want from a control stand point.
Conclusion: all are photographer's cameras. But if you opt for the Nikon, make sure you're that photographer first.
I'll cut to the chase: despite dropping back to 10mp, none of these cameras are going to win an award for low noise. If you're thinking about using them above ISO 800 (some might say 400), think again. DSLRs continue to spoil us in this respect, and no compact I've tested to date makes it into the range of even low-end DSLR capabilities. That said, the latest generation of compacts, like the latest generation of DSLRs, is getting better. Part of that is the fact that Canon and Nikon have backed off the megapixel wars (probably temporarily) and have gone with a relatively large "small sensor." Part of it is that all the companies are getting better at handling noise in JPEGs.
I've got a lot of things to talk about regarding performance. A few things I won't talk about are battery life (not spectacular on any of these cameras, but I don't have enough field experience with any of them yet to dial that into a reliable number other than to say it'll be in the low hundreds for heavy users of the LX-5, more for the Canon and Nikon), and I won't talk about the movie performance, how effectively the scene modes work, and a host of other things. In this section, I'm going to talk about four basic things: autofocus performance, lens quality, low ISO image quality performance (i.e. "best possible picture"), and high ISO degradation.
Before we get to image evaluation I should point out that I didn't spend a lot of time evaluating JPEGs. Since I'm looking for the compact that can get me the best possible image quality, I did most of my evaluation using the latest version of Adobe ACR, and then trying to tweak out the last bit of detail I could from each raw file. I'll make a few comments about JPEGs, but ultimately my assessments rely upon what I could get out of the raw files.
The Panasonic is respectably fast and very consistent for a contrast-based system. It's not DSLR quality--no compact is going to tweak contrast focusing to the levels you get with DSLRs--but what I find about the LX-5 is that it is consistent. It drives to the focus point and stops, and it does this relatively fast no matter whether we're talking about bright light or low light. On rare occasions the LX-5 will pause for a moment before driving to focus or will drive beyond focus, which obviously slows it down a bit. That seemed happen more with low contrast dark subjects than anything else. Low light doesn't seem to make much difference, at least if the subject is within range of the assist light. Overall the autofocus performance is something I can live with. Manual focus shines. Flip the switch to MF, use the buttons to move the focus. Alternatively, use the Rear Command dial while judging the automatically supplied zoomed area in the middle of the frame. Pretty simple, pretty direct, and if you can see the screen clearly, very effective.
Surprisingly, the Canon G12 was less consistent than the Panasonic. If there's good contrast, it focuses a bit faster than the LX-5. But at certain focal lengths in tough contrasts, the Canon will slow down and get a little less consistent. I learned quickly that if the assist lamp had to come on, the camera was sometimes, but not always, going to be more leisurely. But on well lit scenes with contrast, the Canon was ready as fast as any compact I've tried with autofocus. Manual focus is good, but not perfect. Press the MF button, use the control wheel while judging the automatically supplied zoomed area in the middle of the frame. Just as simple and direct as the Panasonic, though I had a much more difficult time judging focus accuracy on the Canon's LCD than the Panasonic's.
The Nikon was certainly the fastest Coolpix I've seen in a long while when it came to autofocusing. Unfortunately, it often lies to you visually. The camera will beep and indicate focus has been achieved but the display is frozen briefly and clearly seems to indicate otherwise. If you hold the shutter release partway you'll see the display refresh and indeed focus has been achieved (usually). It appears that Nikon may be pausing the stream to the LCD before it actually shows the focus that's been achieved. Worse still, there's even less consistency in the P7000's focus speed than the Panasonic and Canon. Just running the three cameras side by side here in my low-lit office generates situations where the Panasonic and Canon quickly lock on focus, the Nikon doesn't and sometimes never achieves focus. Yet in that latter case, releasing the shutter button and re-engaging sometimes works to get focus. I also noticed that the P7000 sometimes has problems shooting through glass, such as the window of an airplane. Even with a clean window and the camera nearly against it, the P7000 would generate all manner of fuss where the Canon and Panasonic simply performed normally. On one cross-country flight I generated more than a dozen Lens Initialization messages trying to focus the P7000 while the Canon G12 simply never complained. If considered by itself, the P7000's autofocus is usable and okay, but sometimes adds uncertainty to the equation. I could live with that, I suppose, if I hadn't seen that Canon and Panasonic in the exact same circumstances didn't have the same issues. Moreover, the P7000 seems a step backwards from the P6000 in autofocus performance. Manual focus isn't as direct as the others. Press the Macro Focus button on the Direction pad until you've selected MF, then use the Direction pad (not either of the dials!) while judging the automatically supplied zoom area in the middle of the frame. Like the Canon, the Nikon's zoomed area is more difficult to judge than the Panasonic's. But things don't end there. If you now want to get out of MF mode, you have to non-intuitively press the OK button, then the Macro focus button again, navigate to AF, and click OK again. Worse still, twice I've encountered situations where getting out of MF mode does not take, which is almost certainly a bug in the firmware.
Conclusion: While I can live with any of the three in autofocus for my kind of work, the Canon tends to frustrate me the least outdoors, the Panasonic the least indoors. As with the previous generation, I'd judge the Nikon as being the laggard of the three for autofocus, but now with the added burden of some clear feedback and firmware problems. You won't be shooting action with any of these cameras without doing some focus anticipation, but the Canon and Panasonic outperform the Nikon in this area. Manual focus users will likely prefer the Panasonic slightly over the Canon due to the clarity of focus in the enlarged section. Nikon's manual focus implementation leaves much to be desired, both from a handling and a performance standpoint.
The images you'll see in this section are all raw conversions, though I've tested both JPEG and raw on all the cameras.
The Panasonic, in particular, has substandard in-camera JPEG processing compared to the Canon and Nikon, though both the Canon and Nikon leave some image quality on the table when they convert to JPEG, too. As always, go by what I write, not by what you see. By the time these images get converted for the Web I can't guarantee that you see what I see. Indeed, I don't spend any real time trying to optimize such images for the Web, as I've found that just the JPEG conversion alone tends to add a level of disguise to what the cameras are really capable of.
Here's a tough shot for any camera, here taken by the P7000 at ISO 100:
Pressing ACR about as far as I dare go in sharpness and noise reduction, here's one of the more challenging areas of the image at 100% (note that this is taken a steep angle through aircraft glass):
Pushed to the extreme like this you can see some of that "painterly" aspect you get with heavy doses of noise reduction and sharpening together. Here's the same area with sharpening but no noise reduction:
Considering the speed of the plane, the aircraft glass, that wide open lens, and the small sensor, there's quite a bit of usable detail here. The trick is going to be getting the detail out while lowering the noise. Still, if you're noise averse, you're just not going to like these small sensor cameras.
Let's try the same thing with the Canon. Here's an overall shot at ISO 80:
Here's the noise reduced and sharpened conversion:
And here's removing the noise reduction with the same raw file:
As you might be starting to figure out, the lenses are sharp, there's a lot of detail available despite the 10mp size, but noise definitely is an issue you'll be dealing with, even at base ISO.
More images and individual descriptions to come.
Low ISO Image Quality
My basic philosophy with compacts is to shoot them at the base ISO whenever possible, shoot raw, and learn how to tweak every last ounce of image quality out of every pixel. In this section, I'm talking about what happened when I did just that.
Panasonic: what amazed me most was how much better every raw file I worked with looked than the camera could do with its Venus in-camera imaging ASIC. The difference is so apparent that I can't really consider the LX-5 a camera that I'd drop down to JPEG for in low light. But other than a bit more noise than the other two cameras, I find nothing that disturbs me much in the raw pixels. Indeed, I thought maybe the lens on this camera was only decent until I started running the raw files. Suddenly, acuity popped way up. The LX-5 does move the bar a bit from the LX-3, but that's most evident the higher you push the ISO.
Canon: Surprisingly, Canon took a step backwards from the G10 with image quality at the base ISO. I strongly prefer the results out the G10 to the G12 (or G11, for that matter) at base ISO. I was also surprised by the amount of noise in the G12 at base ISO; I had expected less, but it's still clearly in compact camera realm (good but not great). Like the Panasonic, the bar on image quality doesn't get moved upward until you boost the ISO.
Nikon: Here's the rub: Nikon has gotten slightly better performance out of the same sensor than Canon. This is visible even at base ISO, and was one of the most surprising things I discovered in testing. The EXPEED processor is achieving this with JPEGs, too. I wanted to like this camera more because of that.
Conclusion: the Nikon eeks out a small win here.
Do I even need to go there? Really? You're telling me that you're actually going to set one of these cameras above ISO 400? Are you nuts? Okay, the camera makers say they've "fixed" this in the latest generation of compacts, so let's put it to the test.
What you'll see in this section come from sample images shot as the highest quality raw and JPEG at ISO 3200 and Auto white balance, with image stabilization, all on a partially lit scene (I turned on the left light on my product table, but not the right, and I tried to hide one of the buildings in the shadow). Basically, I wanted to show what would happen if you shot in low light with the cameras about as aggressively as you'd ever consider.
Panasonic base ISO (50% size): Auto WB was a bit cool and on the green side, but colors rendered well. Overall scene contrast was lower than the Canon and Nikon (look at the whites). Sharpens reasonably well, but has too much JPEG artifacting in it compared to the Canon and Nikon, making it difficult to keep noise off edges during sharpening. Luminance noise was overall lower than the Canon and Nikon in broad tonal areas, but I'd swear they were higher at detailed edges, which would imply that Panasonic may be touching the raw data. This makes it a little tough to sharpen well, much like the problem with sharpening JPEG mosquitoes causes.
Panasonic ISO 3200 (50% size): fairly muddy result due to noise pollution in all tonal values.
Panasonic ISO 3200 (50% size) with noise reduction: surprisingly takes very little color noise reduction, but a larger amount of luminance noise reduction. Unfortunately, whatever the camera is doing to raw files to get better broad tonal areas compromises the edge detail when sharpened. Quite usable, but there's a roughness to the edges that will hurt high detail areas.
Canon base ISO (50% size): Auto WB was accurate though slightly magenta, colors rendered well, though a bit more saturated than the other two cameras, sharpens well but has clear luminance noise in it. While difficult to tell from these samples on the Web, the luminance noise is slightly more prevalent in the Canon than the Nikon, but the Canon's edge acuity is also slightly higher. Some people will prefer that, but it makes getting all the noise out without killing the edges slightly difficult.
Canon ISO 3200 (50%): curiously the Canon does not "block up" the same way the Nikon or Panasonic do, but it shows almost exactly the same noise levels.
Canon ISO 3200 (50%) with noise reduction in ACR: an interesting difference showed up: the raw image didn't need any tonal adjustments to come close to the base ISO image. However, note that for JPEGs edges totally lose their integrity due to JPEG compression interacting with noise reduction (200% view):
You can get away with this with some subjects, but fine detail is compromised by this JPEG mosquitoes interacting with sharpening and noise reduction.
Nikon base ISO (50% size): Auto WB was very accurate, colors rendered well, sharpens well with some clear luminance noise.
Nikon ISO 3200 (50% size), no noise reduction in ACR: noise dominates, with color noise being clearly present. Colors lose accuracy and everything goes drab due to the noise.
Nikon ISO 3200 (50% size), careful noise reduction and some exposure and tonal tweaking: the image is surprisingly usable, though fine detail and texture will certainly be lost, and edge acuity suffers.
Conclusion: backing off to 10mp was probably a good idea overall, as all three cameras can produce a "usable" ISO 3200 image now. That image won't be without flaws, and it takes some gentle care to extract everything out of the resulting file that is possible. At base ISO, there's little to distinguish the cameras. If I had to choose, I'd probably pick the Nikon over the Canon over the Panasonic, but it's so close that we're deep into nitpicking here. There are good qualities to all three cameras at base ISO I like, just slightly different qualities. So far, I've felt I had to do a bit more work with the Panasonic raw files than the Canon or Nikon ones, but not enough to worry me.
At ISO 3200 all three cameras turn in a result that four years ago would have been considered incredible. It's here that I start to see some clear differences that turn into preferences for me. In terms of easily getting tight detail, the Nikon wins, but that comes at the expense of having to spend time making other adjustments to recover some semblance of tonal balance. Some people are going to find that a lot of work. Canon gets closer to a good tonal balance out of the camera (both in JPEGs and in raws run through ACR's defaults), but it has that attribute of JPEG mosquitoes that gets in the way for JPEG files. Something about the noise in the Canon raw file makes it a little tougher to get edges quite the way I like them, but I can get there. So essentially with the Canon you're tweaking different things to get to the same place as the Nikon. When all is said and done, I like what I get from the Canon raws at ISO 3200 better than I like what I get from the Nikon, but it's a close call. The Panasonic is strange in that it required very little color noise reduction in ACR, but a lot of luminance noise reduction. This, too, has its good and bad points. Overall, I favored the Canon over the Nikon over the Panasonic, but after doing raw tweaking, again we're back to nitpicking.
I doubt that most users will ever notice enough difference between these cameras in terms of image quality after tweaking to concern themselves. Before tweaking, yes, there are differences, and JPEG shooters I think are probably better off with the Nikon (low ISO) or Canon (high ISO) than the Panasonic.
Which camera you decide upon is going to depend upon how you value various different bits and pieces. There's no clear overall winner, nor are these cameras trying to appeal to different types of users as did the last generation. I've already pointed out many of the small differences that I feel are differentiations to a buying decision. For me, it really boils down to a two-camera race, though: the Canon G12 or the Panasonic LX-5. If Nikon addresses the many small firmware and handling issues I found, it would be an absolute photo finish between the three.
Personally, I'm not swayed a lot by the "retro" notion. Yes, dials are nice (as long as they don't dislodge easily, as the Nikon's exposure compensation dial does), but the interesting thing is that the camera companies didn't seem to follow that notion all the way through. We don't need the mode or the ISO repeated on the color LCD and distracting you from your framing if the dials are providing you that information already (at a minimum, you should be able to remove duplicate information via custom setting). And Panasonic's modern UI doesn't take me any longer to set the camera with (and matches their other cameras well).
For me the decision really boils down to lens. If I need wide, the Panasonic is going to win (especially since Nikon seems to have delayed their wide angle converter for the P7000). If I need a bulletproof travel camera, the G12 tends to win because of that flippable color LCD. But overall I'd be happy with the image quality from any of these cameras, and other than dealing with the P7000's firmware issues, they all allow me the kind of direct control I seek.
The cameras all started out at the same price point: US$499 here in the States. The Panasonic seems to be slipping faster in price than the Nikon, while the Canon still seems to be holding at the original price in most places. Thus, another simple method of choosing may be on price. As I write the original review there's a US$70 difference in street price between the LX-5 and the G12.