Panasonic Lumix DMC-G5 and DMC-FZ200 previewed, DMC-LX7 with samples

That is a very good observation. Even an "old" D40, produces better overall results, with far lower cost, and higher IQ lenses choices(lower cost), and similar IQ sensor; except for much better (IMHO), Nikon Color, output. A D40 is not that big, so why give up so much, especially at a greater cost, for a bridge cam size?

I have long said, those wanting bridge cams, could greatly benefit from micro 4/3rds. Sure it would be slight less telephoto, than tiny, crappy sensors, but so what? Now, think about that constant f/2.8 lens, built for 4/3rds. Now that is a quality, and do all, bridge cam. But oh, they have to save stuff that makes since, for the fool, and who's money is soon parted. The problem is, there is no shortage of fools. Many need that bragging right (excessive Mpx, X factors); on their crapper cameras. Not that pixels, and telephoto are bad, just bad, when imbalanced.

Well, the trend is towards bigger sensorsv(Sony) in these so-called enthusiasts' cameras so IMO this is not a step forward. Samsung managed to keep sensor size at least comparable to earlier enthusiast cameras and still use a fast lens, so I am fully aware of more specs than just sensor size.

Tom Goodman:

And the Sony RX100 raws I've seen (and extracted) aren't really usable beyond ISO 1600, that's a flaw in Sony's pixel cramming.

I am astonished how many people actually shoot at ISO 1600. C'mon, people, get real.

mpgxxvcd: because some times one only has a compact along and not the obviously preferred large format camera. You know, as in "the best camera you have is the one you have with you." I wouldn't think of preferring a compact for making prints but sometimes that's the only option available.

Tom Goodman:

I shoot beyond ISO 1600 much of the time. And for that reason the Sony RX100 is a problematic camera for indoor lowish light shooting.

Yep, I've even shot at ISO 51200, printed too, though that was a Nikon D4.

To be really serious the Sony RX100 should have about 12MP and shoot noise free at ISO6400, though that would cut into A77 and Nex 7 sales.

I have to assume that Panasonic testers like the performance of the camera, and their marketers know that people will react negatively to a sensor size decrease in this segment, but hope that it tests well. Personally I love the emphasis on the lens for this and the FZ200.

HowaboutRaw: I assume you rarely print anything larger than a snapshot, if at all, shooting at those ISO's with an "enthusiast" camera. Your comments aren't really helpful but I know you will keep on making them.

Tom Goodman:

Never said I shoot beyond ISO1600 with my LX5. That's you picking a meaning that suits your end. You didn't limit your comment about how radical it is to shoot above 1600 to small sensored digicams.

So yep, I shoot above ISO 1600 really frequently. I note too that you skipped the point I made about shooting high ISOs with a D4. So that means you knew that my comment was not confined to " 'enthusiast' " cameras. And no I didn't go for tiny prints at ISO 512000, though did use the 600dpi trick and frankly that ISO is stretching the the D4, but ISO 12800 is nearly noise free on the D4, so prints can be done at 300 dpi. Maybe you've learnt something.

Suggestion to DPREVIEW: consider LX7 and FZ200 for next reviews. Ok RX100 is finishing, but the lenses specs of these cameras are incredible, creating a strong expectation, as is possible to see in the comments.

I would have that Nokia PureView would have disabused everyone of the notion that reducing MP reduces noise - at worst it is a wash, and at best increased MP allows for more noise reduction and better images.

@ppastoris

I think you are wrong about the light gathering part. The luminance does not lessen for the same amount of sensor area at the same f-stop ratio even if the physical diaphragm aperture of the lens is smaller.

Please read Notation section:
http://en.wikipedia.org/wiki/F-number

If what you wrote was true, these tiny sensors must be outrageously more efficient than a full frame DSLR sensor!

@ppastoris

Your are wrong!.

In your above example, a 25 times less 'liight' is shining on a 25times less area. The 'brightness'per unit sensor area is the same - and i think this what f-number is proxying..

And if you want to bring in the image quality stuff in, then it is because of the smaller sensor size and hence smaller pixel area for a given sensor MP count. that smaller pixel area mulitply by the same 'brightness/area give a low intensity,, blah blah...

DPReview is correct to present the values this way. Do it your way is not correct.

For now, I will ignore all the other comments, unless someone has a sincere question (not just trying to demonstrate their ignorance :)).

As to the above comments:
1) ManuH, you are right, I think. The smaller sensors indeed seem to be more efficient per area than the larger ones. You can see that for example from DxOmark numbers. Intuitively it should be the opposite. I don't know why it is so. I could speculate that there's more pressure on manufacturers to improve the small sensors since they have to deal with more noise (photonic noise) to begin with.

2) Jay Kim, image brightness is the same for the same F-number, image size is proportional to the focal length ==> same F-number, shorter lens, smaller image, smaller *total* amount of light recorded ==> higher image noise (roughly speaking).

3) Tomskyair, a) apparently not everybody knows, b) "light-gathering capacity/power" is an actual term in optics (and e.g. one of the main characteristics of a telescope), RTFM :).

May I know what is equivalent f-number??

I wish to know my own ignorance so as not be confused expert.

ppastoris: In a compact camera, I'm more interested in whether I can hand-hold a camera in dim light, while bumping up the ISO as little as possible. My Sekonic hand-held light meter happens to agree with the meter built into my pocket-sized Ricoh GRD-2 camera, which has a maximum aperture of f/2.4, and a 1/1.75" sensor. How does my Sekonic know I'm using a compact camera?

ppastoris: all your smart words can't deny that your "equivalent f-stop" is nothing more than the old DOF comparison between different formats. You better RTFM of a handheld exposure meter if you can still find one.

BTW: I didn't say that there was no such thing as "light gathering capability" in optics. But all your scientific argueing just proves my point about the misleading use of these terms in photography discussions.

Kai Griffin, the reason to use an "equivalent F-number" would be to be able to quickly compare the lens+sensor photographic capabilities, not to operate the camera. Though you could do it that way too, you'd just need to use the "equivalent ISO" = ISO*(crop factor)^2 as well.

Here is an example: you have 5mm f/1.4 lens on a 5x crop sensor. Your meter tells you that you must use ISO 100 at shutter speed 1/25. If you put the "equivalent" 25mm f/7 lens onto a full-frame sensor to get the same DoF and AoV, set the same 1/25 shutter speed, the meter will tell you to use ISO 2500=100*5*5. The resulting images are going to be essentially the same in terms of AoV, DoF, sharpness due to handholding. They will ALSO be roughly the same in terms of noise. There is a good physical reason for that, but simply compare for yourself the noise produced by various crop sensors of the same generation at equivalent ISOs, e.g. Canon G9 at ISO100, Canon 400D at ISO860, Canon 5D at ISO2200(=100*4.7^2).

OngNikon, I could not understand if you are being serious or ironic :). But indeed one could imagine how such a confusion is beneficial to "crop" camera makers, the more beneficial the higher the crop )) : certainly advertising say a 14-35 f/2 zoom for 4/3 system as 28-70 f/2 equivalent is easier than advertising it as 28-70 f/4 equivalent which it really is :).

"...easier than advertising it as 28-70 f/4 equivalent which it really is"
It is from the point of view of DOF. It isn't from the point of view of exposure. What's more important to market, I wonder? I have no statistics, but only a hunch that the majority of people in the market for a fast lens, are in that market because the lens is fast, not because it offers shallower DOF. And I'd wager, were it possible, that 99% of customers for a small-sensor compact camera would be into f/1.4 purely for the exposure advantage. The other 1% thinking they might get some kind of SLR-like shallow DOF effect would simply be mistaken. The 35mm DOF equivalence for a 1/1.7" sensor compact camera is totally irrelevant, other than to instruct those who have the wrong expectations about DOF.

That said, I've noticed naughty camera makers implying that their compact cameras with fast-ish lenses will let users create "attractive background blur", which you & I both know is horse hockey.

not just DOF, but also image noise at the same shooting conditions, and also the diffraction limited aperture. Pretty much everything that defines a photograph in technical terms (AoV, DoF, image noise, diffraction effects). And again there is no exposure advantage of e.g. a 5mm f/1.4 on a 5x crop camera vs a 25mm f/7 on a 1x camera. At the same shutter speed you will simply have to use 5*5=25 higher ISO setting on the larger camera to get exactly the same exposure, which will give you exactly the same image noise in theory (because it simply is the result of the particle nature of light -- the number of photons hitting each "pixel" of a sensor is instrinsically random, see http://en.wikipedia.org/wiki/Shot_noise ). Sensors only increase this noise, and in practice the sensors from the same generation increase this noise by approximately the same factor (+-1EV or less).

So to reiterate, practically you will not get any advantage (+-1EV at most) using a bright 5mm f/1.4 on a 5x crop at say 1/25 sec, ISO 100 vs 25mm f/7 on a 1x crop at the same 1/25 sec but higher ISO 2500. Using however the larger sensor camera you *will* get a large improvement in image quality of you trade some DoF and use 25mm f/1.4 lens at ISO 100 on the large camera.

If something is still not clear let me know, please. I'll explain.

ppastoris, whether your math is right or not, it just seems to me that you're stating the bleeding obvious with regards to noise: who really cares how high a full-frame camera's ISO needs to be in order to match the noise of a compact at ISO 200? What relevance is there in saying that there is no exposure advantage vis-a-vis noise between a compact at f/1.4, ISO 200 and a full frame dSLR at f/7, ISO 5000? Clearly, I can't cram the dSLR into my jeans pocket as I'm heading out for an evening stroll with my wife along the riverside of an exotic city. Naturally, I could take the massive dSLR instead, but my wife would probably find that about as romantic as me lugging along a tome called "The Pedant's Guide to the Mathematics of Photography".

No, I'd normally have my Ricoh GRD2, which maxes out at f/2.4. It shares the LX's sensor size, and there is *definitely* going to be an exposure advantage between the two when wide open, and I won't even mention the Ricoh's lack of IS.

Tomskyair,
short answer : pretty much ANY full-frame camera will indeed have a comparable (+-1EV) quality at ISO5000 as a 5x crop camera at ISO 200 :).

Longer answer. Need data? See DxOMark measurements for e.g. 1) Lumix LX3, 2) Nikon V1, 3) Nikon D7000, 4) Nikon D700, 5) Nikon D800 (in the order of sensor size). These cameras achieve the same noise level (SNR=30dB) at ISOs 1) 94, 2) 345, 3) 1167, 4) 2303, 5) 2801 respectively (the metric is called "Low-light ISO" on DxO website). Their respective crop sizes are 1) ~4.3x, 2) 2.7x, 3) 1.54x, 4-5) 1x. Hence the equivalent ISOs ( ISO*crop^2 ) are 1) 1739, 2) 2560, 3) 2801, 4) 2303, 5) 2801. As you can see they all are pretty close, except the LX3 underperforms even taking into account the small sensor. LX3 is a well known poor performer in noise department, which I can confirm since I owned both LX3, Canon 350D, and Canon 5D MarkII at the same time (ISO 400 on LX3 is a bit worse than ISO6400 on 5DMkII).

Any further questions?

Kai, an f/7 FF lens would be a lot more comparable in size to the f/1.4 lens for a 5x crop sensor provided the same IQ. Ok, it may be a bit of a stretch for a 5x crop, but if you compare lenses for the systems with less crop difference you can see that the "equivalent" lenses are similar in dimensions, weight, and price. One example is Olympus 4/3 14-35 f/2 (28-70 f/4 equiv.) vs Canon 17-55 f/2.8 (27-88 f/4.5 equiv) vs Canon 24-105 f/4 (fullframe). They for example weigh 900g, 645g, 670g respectively.

Had most consumers understood that a bright f/1.4 lens for e.g. m4/3 is as capable photographically as just as large (and cheaper) f/2.8 lens for FF perhaps the market would not be as driven towards the small sensor cameras (like m4/3) as it is now. On a small FF body you'd at least have an option of using a f/1.2-1.4 lens when you needed shallow DoF and/or low light ability, while it's essentially impossible to make an equivalent lens for a smaller sensor.

@ppastoris

Noise level differs between cameras. Even when they have identical sized/resolution sensors. So with your argument, you should also need camera manufacturers to specify "equivalent ISO" values and not just equivalent f/stop. But then it gets REAL messy as older DSLR's are no where near as noise-free as the current crop of DSLR's. Heck, my old Canon 20D camera is more noisy than most current compacts with 3-5x smaller sensor size.

FoV equivalent makes sense as people want to know how much that lens can see. But equivalent F/stop just don't make any sense unless you really care and measure how much or little DoF you are going to get. Most people don't notice the difference of DoF up to around 2 stops anyways even with a DSLR while they notice big time about the extra 4mm focal length difference at wide angle for example regardless of the sensor size. That is the reason why focal length is stated in equivalent values. I'm sure you know this also.

@ppastoris

Also, small sensor cameras are not as useless as you say as most consumers don't care about the differences that you are talking about. Many just prefer it being smaller, more versatile and cheaper.

Last time I checked, no FF DSLR fits in your pocket. Having a FF DSLR with a 24-70mm 2.8 lense and a 70-200mm 2.8 lens might be awesome but a compact camera with a 24-600mm F2.8 equiv. zoom lens would be far BETTER in many situations for its size, weight, versatilityand the noise/image quality difference would be practically nil if used under bright sunlight situations. But yes, for dimly lit cafe's, the FF DSLR with 24-70mm would win every time for IQ/noise.

But really, we are talking as enthusiasts/pro's/hobbyists. Most general consumers are actually very happy with their iPhones and Galaxy S's and don't even care not having a dedicated camera at all. =p

@ppastoris: You said "an f/7 FF lens would be a lot more comparable in size to the f/1.4 lens for a 5x crop sensor provided the same IQ", but again this is completely hypothetical and irrelevant. There are no f/7 FF lenses on the market, though you might find one in a former Eastern Block flea market, perhaps made of bakelite. Even without the lens, the FF body is gargantuan - the smallest FF body I know of is the Leica M9, and neither pants nor most people's budget is that big. Compact 1/1.7" sensor size cameras do one thing that no full frame camera can ever do: they can make you forget you're even carrying them. It's the #1 reason people gravitate towards them. It is beyond ridiculous to be getting mathematical about the technical equivalence of such cameras to cameras approaching the size of a house brick. That is why quoting "equivalent F stop" for a little camera like this is a such a silly idea; it says nothing about what people are actually interested in.

Kai, for a small 35mm camera see Minox 35. With a 35 mm f/2.8 collapsible lens that camera was just a bit thicker than Lumix LX5, and (more relevantly) about exactly the same size as a m4/3 Lumix GF5 body w/o a lens. So yes it IS possible to make a full-frame compact pocketable.

Jay, just saying you will be carrying a 24-600 f/15.6 eq. not f/2.8 eq. Very different from if it actually were f/2.8 :). Someone with say a seemingly "darker" Nikon 18-300 f/3.5-5.6 would actually be carrying 27-450 f/5.3-8.4 equiv., and should they crop a bit -- 27-600 f/5.3-11.4 equiv. That's, a lot brighter on the wide end ans still almost a stop brighter on the long than 24-600 f/15.6 (naturally 24-600 f/15.6 eq. is smaller and cheaper).

Which is exactly my point, would the effective F-numbers been published along the effective focal lengths the consumers (enthusiasts(?)) would be able to make much more informed decisions IQ/weight/size/price tradeoffs when selecting their hardware.