Photography (again): sensors resolution

One more photographic thread for you

A dear friend of mine, in my opinion excellent photographer, claims that even today he would not buy a FF camera with a higher resolution sensor than 18-20 Mpx, since he believes image quality, in terms of noise caused by the smaller and closer pixel sensors, would be lower.
Of course, he says he personally does not need the other obvious benefits deriving from a high Mpx sensor (large prints, cropping, etc.).

Based on what I read around, his view seems a bit outdated. What is you take, considering the most recent camera bodies and sensors? I realize this may be a fairly complex topic, but that is precisely why I posted here.

If he was talking about compact cameras then I'd agree with him. The manufacturers stuff in extra pixels will seemingly little* regard for increased noise. They want to have a bigger megapixel number.

With DSLRs it is different though. The manufacturers tend to add extra pixels, while at the same time improving other parts of the sensor/processing to compensate for the smaller size sensors. So each step increase in pixel density typically comes with the same** or lower noise that the predecessor.

The sensors in today's 18 megapixel APS-C DSLRs have less noise at 800 ISO than my 9 year old 6 megapixel APS-C DSLR has at 200 ISO.

* that isn't to say that they aren't also making improvements in noise along the way though, the small sensors have over time got less noisy generally as well

** though even with DSLRs there is the occasional exception of sensors in a range getting noiser, but that isn't typical

He was specifically speaking of DSLR.

He owns two Canon 5D MarkIIs, and he was particularly interested in the 6D to replace one of his 5Ds specifically because they kept it at 20Mpx.

When I have the chance, I'll ask him more about this. He usually is a knowledgeable guy, certainly not prone to "fanboysm", and usually speaks with some solid tech/scientific research at support of his arguments, but we did not have the chance to talk more.

I did search the topic, and I found the same you're mentioning, Andy, in "historical" terms.

But, comparing current cameras, more or less belonging to the same generation, I found my friend's point even less convincing.

If we consider the 5D Mark II and the D800, it seems to me that Canon went for the very high ISO with lower noise, while Nikon for the Very High Mpx of similar quality, but at lower ISO. This alone seems to suggest that the whole noise issue cannot be effectively simplified to the "resolution" alone. Which is obvious, maybe, but I guess what I am saying is that other factors such as the NR algorithms and the design of the sensor itself are factors that play a more important role than resolution (if we look at a specific generation of sensors/img processors/bodies).

Am I making sense?

I have Mk II's and a Mk III and the noise performance on the Mk III is much much better than on the Mk II. So I am not so sure I would listen to your friend. Hire one out yourself and make your own mind up

Cheers

Cris

I may actually do that but already now I really don't see either how that could be given the samples available everywhere showing the opposite of what he was saying.

In any case, just to be clear, I am just intellectually curious about this, I am not saying that the hypothetical differences in noise, whether it is there or not, would play any relevant role in one's ability to take a good picture.

For what it's worth, my 2004-era Nikon D70 and my 2008-era Nikon D700 have roughly identical pixel sizes (with the D700 having twice the sensor size). The difference in quality (pixel noise, etc.) is night and day. The D70 at ISO 800 is worse than the D700 at ISO 3200. My six month old Fuji X-Pro 1, with 2.5x the pixels of the D70, crammed into the same space, has comparable quality to the D700, maybe slightly better even.

Moral of the story: pixel pitch, across cameras built around the same time, may well say something useful about image quality. Pixel pitch across multiple years doesn't tell you anything. And, of course, it's an entirely separate issue of keeping your camera steady and using quality glass to make sure that all those pixels have something useful in them.

(DPreview used to nicely tell you the pixel pitch of each camera next to the megapixel number, but this seems to be gone now. Quite a shame.)

And we all agree on this of course. Technology does evolve, constantly.


It may, but does it? That was my original question, actually.
For example, do we know if Nikon D800 and Canon 5D MkIII are very different in terms of noise due to the dramatically higher pixel density of the Nikon, and do we know if Nikon has to apply stronger NR to achieve the same quality as Canon?
By looking at DP review tests, even comparing Nikon image crops so to compare to same global resolution of Canon Images, it would seem that Nikon's mg quality is not at all worse at lower ISOs, and it is at higher ISOs (and one would guess that it would worsen significantly at those ISO values that the camera does not allow, just for that reason).

The only apples-to-apples comparisons you'll ever get are when the two sensors you're curious about came off the same fab line and were engineered at the same time. You can't say anything useful, based purely on sensor resolution, about the latest Canon vs. Nikon.

For what it's worth, the Nikon D800 has a consistent pixel pitch with many of the current-generation cropped-sensor cameras (which clock in anywhere from 12-20 megapixels). As I mentioned above, my Fuji X-Pro 1 modestly outperforms my D700. As such, it's reasonable to expect the D800 to work well, and the DPReview test images support this. Furthermore, unless you're doing wall-sized prints, you're going to be shrinking the D800's output down, which will cover a multitude of sins.

I was not referring specifically to Canon and Nikon (I used them as an example of sensors with very different resolutions), but in any case, it seems here as well nobody seems to have, or know about, any data whatsoever available to support my friend's claim (which in short is just: still today resolution itself is a dominant factor in determining noise and therefore img quality, to the point that one who does not need a very large image, will be best with a lower resolution sensor).

Interesting.

Well stated like that your friends claim is actually true. If one of today's sensors was available at the same size with half the pixels, it would undoubtedly have less noise per pixel. But since no such sensors are available it is not that useful a claim...

And it isn't the same claim to say that sensors back in the day when they had half the pixels they do now were less noisy than today's sensors.

Here's some fun math.

On the D700, the base ISO is 200 and I wouldn't go past 3200. That's four stops. Another way of looking at it is that the D700 gives you 12 good bits per pixel. (The actual sensor gives you 14 bits per pixel, but the bottom two bits are crap.) The D800 gives you roughly the same, 12 good bits per pixel, but you've got 3x the pixels. If you downsample a D800 to the same resolution as the D700, then you're averaging those three pixels together, giving you somewhere between 1-2 bits of additional useful signal per reduced pixel.

Now, imagine a hypothetical D700+, fabbed with current generation technology, but still 12 megapixels. The bazillion dollar question is whether the D700+ could have more than 14 useful bits per pixel. If you could give me (dreaming now) 16 beautiful bits per pixel, then that would yield HDR goodness in every shot. Or, assuming a base ISO of 100, you'd be able to shoot without any noise whatsoever at ISO 25600. (Some cameras claim to work at this speed today. They don't do it well.)

Lastly, here's your tradeoff. Are you more likely to want 12 megapixels with 16 beautiful bits per pixel, or would you prefer 36 megapixels with 12 beautiful bits per pixel? The former gives you glorious HDR. The latter gives you outrageous high resolution, if your lens supports it, and can be downsampled to give you 14 beautiful bits per pixel at resolutions you actually care about.

My guess is that somebody inside Nikon thought long and hard about these tradeoffs, and had hard numbers for both options, and decided that more pixels was preferable.

Actually, I need both. I've primarily been shooting live theatre productions. Due to the stage lighting, I get an extremely wide dynamic range. When there's anything shiny (think sweaty actor's skin), or white on stage, the shadows disappear. Sometimes, that's good. Sometimes, it's not. HDR for the win. Except that when we print photos for lobby display, we're pushing the limits on resolution, particularly with low-light shots that are much noisier. And then there's the trick of finding photos that still hold up when printed on a 7' banner. Resolution for the win.

I would (naïvely) assume that your 7' banners can use photos that you've carefully posed with more control over the lighting. Under such circumstances, you can turn the lights up or bring in big strobes. You can also bring in a large format camera, giving you far more resolution than any digital camera could ever hope to provide.

Since the 7' banners are an advertisement for the theatre, we want to use photos from actual shows. We don't have a special photo-shoot just to get a shot for the banner, so what I shoot during the live dress rehearsals over the course of the year(s) is what we have to work with. Obviously, changing or augmenting the stage lighting is a no-go, as it can totally destroy the very essence of what you're trying to capture. Although some stage photographers will do posed (or semi-posed) photos of key moments, I dislike doing that. I find the actors have much better in-the-moment emotional responses during an unhindered performance than when doing a static pose, or even when repeating a portion of a scene for the umpteenth time. Especially when working with non-professional actors. I like the end results better, when shooting this way, and sometimes the most dramatic shots end up being from scenes (or portions of a scene) I would not necessarily have chosen for a posed photo.

It's much like wedding photography during the ceremony and reception -- what you get is what you get, and there are no do-overs. There's no interrupting the priest with an "I'm sorry, can you re-do the 'you may now kiss the bride' part?" I do, at least, get to shoot multiple dress rehearsals for a show.

edit -- tl;dr: my approach to stage photography is a photojournalism style

Yes, sorry if I wasn't clear in my original post. Also, I just re-read my posts in the beginning of the thread and I referred to 5DMKII vs D800, when I intended 5DMKIII vs D800.
In other words, my friend wasn't making historical comparisons (nor he was comparing current sensors to hypothetical versions of the same sensors with lower resolution, but I see what you're saying).
He was comparing current sensors. He was saying that, he would prefer the 6D to the 5D Mark III or, by the same token, the D800, because 6D lower-resolution sensor would produce better images due to less noise.
I was puzzled by this statement having read on-line reviews that seemed to disprove this, at least in such general terms. While this seems to be true at much higher ISO, it's clearly not true at lower ISO, and I am not sure how true it is between 5DIII and 6D (I suspect not true at all). In short, there is much more than just resolution to determine a sensor's noise, so it makes little sense to consider resolution alone.



Very interesting. While instead Canon decided that 16 beautiful bits per pixel, to stick to your example, would be preferable. Hence, Nikon D800 and Canon 5DIII . Do I understand correctly what you're saying?

That sounds more like poor lighting and makeup to me, and is the reason why I would do a setup performance if I were doing something that would end up on a banner that big. The actors can still act the same, but you need full control of the lighting to get the best results, using HDR is just a bit of a bodge really and by setting the lighting for the camera rather than the audience for a one off performance not for public eyes would give you far better results.

I''l say this first before I say what I am about to say as I don't know your photography at all, so I am not talking about you here. I hear it all the time here in the UK "I shoot weddings in a photo-journalist style" when what they really should be saying is "I have no idea how to light stuff so I just snap what is there". Of course there is an art to using available light, but it far more rewarding when you yourself have crafted the whole image.

Whilst it can be true you get better reactions when people are largely unaware they are being photographed I think the benefits of lighting things properly out weigh that.

There is also an added benefit, if things are lit properly it doesn't matter what camera you use, it will always look good

Cheers

Cris

Kinda, sorta, maybe. Since we're talking about the 5DIII, it's actually 22 megapixels. Nikon offers the D600, with a 24 megapixel sensor (allegedly from Sony) and the D800 with a 36 megapixel sensor (allegedly home grown).

If you use the DPReview Compare-o-Tron-2000, go for the raw mode comparison at ISO 12800 and bring up the 5DIII, D600, and D800. This is enough for you to have a good look at the noise. I tend to like looking at the Fujitsu batteries in the lower-left corner.



You can also have a look at their measured numbers.



What you see is that the D800 has slightly worse noise but big buckets of extra resolution. If you scaled it down, you'd probably be in very good shape. The D600 is very close to the 5DIII in resolution, and the D600 slightly beats the 5DIII in noise. I tossed in the much pricier Nikon D4, which is a great example of "16 beautiful bits per pixel" (the sensor is 16 megapixels -- the smallest in the bunch here). Again, if you shrank down the D800, you'd probably come out about the same as the D4; somebody must have tried this experiment, but I'm not really sure. On the flip side, the D800 raw file is 53MB while the D4 raw file is 34MB. If you don't want the super-high-resolution, the D4 would seem to have an advantage.

Lastly, for what it's worth, DxOMark seems to think that the D800 is the best sensor they've ever measured.

Sometimes it absolutely is. I'm shooting for community theatre, not professional theatre (though, around here, there's sometimes seems to be very little difference). Sometimes, however, it's not. For example, your hero may be sitting in a bright pool of light, while secondary characters are skulking around in the dim background. You want those characters in your shot, because it would change the tenor of the scene if they weren't there, yet if you expose for the hero (1/125s, f2.8, ISO 1600), they're virtually black, and if you expose to get reasonable detail of the characters in the shadows (1/60s, f2.8, ISO 1600), your hero ends blown out. If I had a bit more dynamic range, that's not as significant of an issue, just like it's not an issue for our eyes.

In an ideal world, yes, you'd know in advance what might be ending up on a banner, and do some hero shots. I don't, however, have that luxury. I'm basically given two or three nights to shoot a show during tech rehearsals (the week before it opens), during which I'm required to be as unobtrusive as possible (I may, or may not be allowed on stage). I hand over a CD/DVD of images, and some time in the future, someone designs promotional brochures, including a banner, and says "I love this shot, and want to use it on the banner -- how big will it print?" At that point, I can't re-stage that shot. It may have been taken last week, it may have been taken two years ago. The actors may not be available, sets are likely destroyed, costumes and props have been returned, etc.

Except that you've changed the scene -- that's a big no-no in theatre photography. That's one of the challenges of doing theatre photography that I really love -- I don't have any control over the lighting. I don't get to walk in and say "I'm going to put a strobe light here, and a back-fill light here, and..." I'd have a very irate lighting designer on my hands, because my lighting for a photograph changes the atmosphere of the scene, and the image I capture is no longer true to the performance, even though it might be "better," and would, consequently, be deemed unusable.

Absolutely agreed on the first part (when my wife and I were getting married, and looking for a photographer, we noticed that a lot of "photo-journalistic" style photographers were crap), but I disagree on the second. That style (of shooting with available light) was critical for us, because our wedding is about emotions. People's reactions were our top priority -- it's more important for us to remember the joy in grandma's face and laugh, than how beautiful an empty dress looks when hung in front of a window (to pick out one of your recent blog images that I really like). I don't give two hoots of a rat's patootie if the photographer is more rewarded by being able to craft the image entirely themselves, that's not what they're there for. That's not to say that we would accept crap when it came to images where people's reactions weren't in consideration -- the "glamour" or still-life type shots, if you will -- but those types of images were of secondary importance for us. In the end, we hired the (former?) head of the photography of a large-ish local newspaper, and were quite happy with the results. We got the emotive shots we wanted, and we got a few good glamour shots (though nothing, I think that quite matches your style, but I haven't looked at them for a while).

Similarly, when I'm shooting a show for the theatre, I'm not there to craft an image with total creative control, myself. The director has complete creative control over staging (nothing worse than having an incredibly tense and dramatic moment with the significant players at opposite sides of the stage), the lighting designer has complete creative control over lighting, and both have been locked before I ever enter the scene (haha). My creative control as photographer extends to framing and the controls I have on my camera (aside from flash, which is verboten). I'm even limited in where I can stand (I spent much of last night balanced on top of the armrests of the chairs in the 2nd row). My job is to figure out how to get the best picture I can, and do it before the opportunity disappears. My primary function is to get photos that can be used in lobby displays, and to document the show. Getting photos usable for advertising purposes is secondary, though, for most purposes (banner being the exception), ends up coming from the the set of photos used in the lobby display.

True, but a) if you were too concerned about the lighting, you may have missed the moment altogether, and b) just because the photo looks good, doesn't make it suitable for a wind range of uses.

The key for me, when shooting for the theatre, is flexibility, since I can't plan for everything in advance.

You could easily be talking about me.

At one time in my life I fancied myself to be a semi-professional photographer. I had a full color darkroom back in the days when that was virtually unheard of in a home setup - there were no digital cameras back then! I was the official track photographer at the local race track. I even did a couple of photo shoots for a (no, the) local modeling agency. People familiar with my work at the race track had me photograph their weddings. I figured I was pretty hot stuff.

Then... I attended a wedding as a guest, not a photographer, and I watched a real pro at work. This guy was magic. He was always just where he needed to be to get the shot he wanted, he seemed to be everywhere at once and yet never intrusive or in the way. He saw (and made!) photo opportunities that I never would have thought of. He had an assistant whose job was to always know which camera the photographer needed at that moment (it's a lot quicker to grab a different camera than to swap out lenses) and to keep fresh batteries in all the cameras and strobes. Over a three hour period that covered the wedding and the reception, he took more than 2,000 pictures. He was working hard, the sweat just rolling off of him, and he never relaxed for a moment.

Oh, he did some set pieces with proper lighting between the wedding and the reception, but for the most part it was "journalist style" and he certainly captured the spirit of the wedding.

I never had the audacity to photograph another wedding after that, and still am embarrassed about the absolutely pitiful job (reckoned with 20-20 hindsight) that I did on the ones I shot. That night was a real education for me.

tanstaafl.

As an aside, the fastest film I ever shot was T-Max 3200 (discontinued!), which I'd then push a stop further using proper Kodak T-Max developer. For my purposes at the time (high school yearbook photographer), we never, ever printed larger than 8x10 (typically 5x7), so it just didn't matter at all. I actually made some 11x14's for myself, and so long as you're not sticking your nose into the picture, they look fine.

Today, any of the cameras we're talking about give you radically better speed and noise and resolution than any film emulsion could ever do, particularly at high speeds like we're talking about.

Back to the topic at hand, since you're stuck with the lighting you've got, you need to look at all the other techniques you can bring to bear on the problem. You can rent/buy brighter lenses ($$$ + heavy weight + tight depth of field). You can use a tripod or monopod to allow for a longer exposure (but longer exposures only work when the actors aren't moving much). You can also get fancy in post-processing; Lightroom is pretty good, but supposedly you can do better with things like Noise Ninja (which has now evolved into a full-blown raw conversion program: Photo Ninja, which seems to get strong reviews). Also, I'll note that when you convert to black & white, you can get away with more because it just looks cool. A little bit here and a little bit there can add up to a big difference in the quality of your result.

I think this would be a discussion better held over a pint of beer, I think you are missing my point a little bit.

Lots of my photography involves flash in some way, and not all the posed stuff. Lighting the reactions during the speeches for example, really important and I am always amazed by photographers who don't have a clue and just put up with gloomy lighting. If you know what you are doing you can add lighting to a situation AND get the natural reactions.

You don't see too much of that stuff on my site, as quite frankly it doesn't sell. People buy the setup couple shots that look so "natural".

I get what you are saying, but it sound like you are working with people who don't appreciate or understand good photography. If they did they would let you have some input at some point of the production.

Cheers

Cris

For sure.

Quite possibly.
I agree with that -- shooting with a photo-journalistic style doesn't preclude use of flash, in any way. For non-theatre photography, although my taste tends to run closer to "available light", I do use flash when I think it's necessary. Sometimes I take one shot with, and one without, just so I can appreciate the difference, and adjust my notion of when is "necessary". It does help, even when I'm just using the dinky little built-in. For theatre photography, though, it's not an option for me -- it's specifically disallowed, as I'm shooting during the production (even if it is "just a rehearsal"). Flash would be disruptive, and potentially dangerous (don't need an actor falling off the stage, getting poked in the eye with a sword, or spraining in ankle due to a missed landing on a leap, because they're momentarily blinded by flash). If I were shooting Broadway or West End shows for glossy commemorative brochures, etc. then I'd expect to have an entire day set aside for photo-shoot day, and, when the stage lighting isn't adequate for a shot, have a discussion with the producer and lighting designer about a) whether it's a shot they really require, and b) how it might be improved. But the theatre I'm working with has neither the time, nor the budget for that (a show only runs three weekends), and from what I've seen on the internets, that sounds pretty typical, even for theatre's larger than ours.
I think they're just striving for a balance, where their primary focus isn't the advertising. The majority of the time, everything honestly works out just fine. It's just that every now and then the stars align in such a way that I go "hrm... I wish I had a higher res version of that shot" or "I wish I'd had a bit more dynamic range, so that I could boost the exposure of the shadows in this shot by 1/3 of a stop." I don't lose any sleep over it.

Well, that is precisely what the D800 review says. Scaled down to 5DIII resolution, the D800 performs like or slightly better than the 5DIII.
On the other hand, why look at ISO 12800 only? 5DIII reaches higher ISO values, which seems to me is the feature the 5DIII is betting on.
In other words, and as you were suggesting in your previous post, Canon and Nikon either went for higher ISO or higher resolution. Graphs also seems to confirm this.

Also, I assume that D800 and D600 do not allow as high ISO as the 5DIII because noise at those ISO value would be unacceptable. Is this a wrong assumption?

For a while, the lobby displays were getting printed at 11x17, which was pushing it. We're now down to 8x10, which is much more forgiving.
Did that. I'm now shooting with an f/2.8 lens. It helped immensely.
Too unwieldly. I'm moving constantly. My monopod and tripod stay at home.I didn't know that Noise Ninja has evolved. I'll look into Photo Ninja.

I think my next photography-related purchase will be a D800 body. I'm currently using a D80. For the most part, it's fine, but I want an FX sensor, and I want a sensor with better noise characteristics. That it has video capabilities will make it an easier sell with my wife, who wants a new video camera. With Child #2 on the way, it's a very good possibility.

The next step, assuming you can get closer to your subjects, is something like a 50mm f/1.4. Two extra stops are a non-trivial improvement, and the lens is cheap.



Needless to say, it's a world of difference, or at least that's what I thought when I upgraded from a D70 to a D700. I'll caution you that the upgrade to FX implies an upgrade in carry weight. And I don't have to caution you that kids have a habit of forcing you to haul around things besides cameras. This sort of logic led me to my Fuji X-Pro 1 and might similarly lead you to one of the various "mirrorless" cameras out there.


While all these cameras hypothetically support shooting at these immensely high ISO values, in practice, you'd never want to go so high unless you were only trying to get something barely usable when you shrink the picture down mercilessly.

Also, here's a trick to keep in mind. When you shoot raw, there's no such thing as the camera restricting the ISO. Instead, the sensor gets what it gets. If you shoot at the base ISO, then the high bit may well be one or zero. If you shoot one stop faster, then (assuming you're not blowing highlights) the high bit will always be zero. Basically, you're left-shifting the results by one bit. You want to shoot four stops faster? Left shift by four. The more you shift, the more you're digging into the low-order bits, which give you progressively less useful information.

Lastly, since I know somebody will bring it up, left-shifting the bits is much dumber than any actual raw conversion program. Instead, you want to see if there's actual signal up there and play HDR games of some sort to preserve that highlight detail that you'd otherwise be pushing to pure white. That's more or less what all the major commercial RAW converters do.


Nutshell summary: the current crop of fancy pants D-SLRs are all very good. This year, the Nikons are marginally better than the Canons, but not enough that you shouldn't get a Canon if you otherwise prefer them, perhaps because you prefer their lenses or something. In four more years, Canon may well have leapfrogged Nikon, but good glass will always be good glass.

With specific regard to the exceptionally high resolution of the D800, it's effectively giving you a choice. You want the bit-depth quality of a lower-resolution sensor? Then just downsample your image. Done. You prefer incredible high resolution, albeit with more noise in the dark bits, then you've got that choice as well -- a choice you don't get with any other camera. The only real price you pay for this flexibility is that the raw files are big and presumably are that much slower to process. Also, you may feel an unbearable need to purchase the most expenses lenses that money can buy, never mind a fancy tripod, expensive strobes, and so forth, so you can actually get the most out of that lovely sensor.

For me to switch to a 50mm prime, I'd have to use two cameras, or I'd be spending all my time switching lenses. I haven't pulled the statistics from my exif data, yet, but I'm generally closer to the 70mm end of my lens. It's an FX lens, so there's the 1.5x crop factor to account for, too. (I know, I know... I shouldn't really be using this lens with a DX sensor, but we bought it knowing we'd eventually move to a body containing an FX sensor.)

Needless to say, it's a world of difference, or at least that's what I thought when I upgraded from a D70 to a D700. I'll caution you that the upgrade to FX implies an upgrade in carry weight.[/quote]Yes, but I'm already carrying with a fairly large lens on it, so I don't think the extra ~1/2lb is going to be that significant. I guess we'll have to see.

Yeah. I've been looking at the rugged cameras. Waterproof/crushproof/diaper-bag proof. I have an Olympus model that went tits up when the seals went out (splashdown on a 2-storey waterslide), but I was never really fully satisfied with it. It was the only serious waterproof digital P&S available at the time I got it, though, so I'm happy to see Nikon and Canon enter the market.

Dollar-no-object, then, how about an 85mm f/1.4? Perhaps rent one for a week to give it a shot and see whether it works for you. Two extra stops may make a world of difference for you if you can deal with the much more narrow depth of field.

Not necessarily.

I know that this is not a big-bucks interchangeable-lens DSLR, so most of you will dismiss it out of hand. But if you read the review and the specs, you will see that it has some pretty impressive capabilities, some of which might suit your needs very well.

1) HDR capability. It takes the "best" parts of three near-simultaneous exposures to produce a high dynamic range final image.

2) Extended zoom. How much of your noise is because you are losing 50% or more of your original image to cropping due to the fixed focal length lens you are using now? How many shots do you not even take because you were in the wrong place at the right time?

3) Speed. This Leica-built lens holds f2.8 across the entire 25-600 mm [equivalent] range. That is an incredible specification.

4) Control. You have full manual control over everything: P-S-A-M, manual focus, forced flash (not an issue in your case ), auto exposure bracketing in 1/3 EV steps, white balance pre-sets and/or bracketing, burst mode at 15 fps, shutter speeds to 1/4,000 second... read the specs and you'll see what I mean.

5) Cost. Less than 1/3 the cost of the lens that Dan linked to.

6) Versatility. Incomparable. Even RAW output.

But, it's not a DSLR. People won't be impressed. If you can live with that, it might pay you to take a look at it.

tanstaafl.

Urmmmm the whole point of an SLR is that you can swap to the appropriate lens as needed. I just don't get why you wouldn't swap out lenses. I swap 100s of times in a day, nothing has lens caps on or stupid UV filters, quick easy and if you buy good lenses in the first place they will stand up to the knock about no problems at all.

I do put an lens cap on my fisheye actually, but only so I can stack another lens on top of it in the bag.

If you are shooting at f1.4-f2.8 dust on the sensor just isn't a problem either.

Edit - I can highly recommend this bag, I can fit everything I need for a wedding in this... http://www.thinktankphoto.com/products/retrospective-lens-changer-3-black-shoulder-bag.aspx

Cheers

Cris

Looks like a fantastic camera!

Near-simultaneous isn't sufficient when you have moving subjects in low light.


None -- I'm not using a fixed focal length lens, and I don't crop images. Or rather, when I crop images, I do so because I want to change the aspect ratio, not to crop-and-zoom. I don't do digital zoom.
Many. But that usually has little to do with the focal length, and more to do with needing to be 6 feet higher in the air, or 8 feet to the left, or transported to the middle of the auditorium from where I am at the front of the stage.

But the rest looks very nice.

My answer to that is "it depends". Some scenes, I'd be quite confident that I have all the time in the world to swap lenses, and not miss anything important. Heck, I could saunter from the stage front to the sound booth for a shot to capture an overview of the staging/blocking, and not miss anything terribly exciting. Other times, thinking back to some of the scenes I've shot, I'd be attempting to change lenses in the dark, while on the move, navigating around people, stairs, and auditorium seating. Not doing this on a daily basis, I'm not so confident that I'd get that swap down in time, before I have to be where I want to be for the next shot. This is where having that staged shoot would come in handy, as I'd just switch lenses and have them re-run that part of the scene.

I've never been one to worry about dust. I've swapped lenses in a windstorm, while at the Oregon Sand Dunes. My wife wasn't so thrilled with that.

That bag looks awesome... thanks for that pointer. That's a case (NPI) where having the right gear is helpful. My camera bag is a backpack style bag, with a side-pocket access. It's great for walking around on a family outing, but not so great when you're in a single location for a few hours, needing rapid, frequent access while on the move. But then, like all of our gear, we really bought it for the former, with the theatre work being a secondary use.

It's all been a huge learning experience -- I've only been doing theatre photography for ~4 years, and it's been 100% learn-as-you-go, so I guarantee you there's lots I don't know about the subject.

The great thing about that bag is it's silent! No zippers or velcro (well there is but it can be silenced), for me it's ideal.

Changing lenses becomes as easy as changing gear, I don't even notice it now. You can also anticipate the lens you'll need most of the time. I shoot weddings, so the risk of missing moments isn't any greater. But having the right bag is key, without the LC3 my photography life would be much harder. I have tried a shootsac before but it grips the lenses.

I put small cotton flannels in the bottom of the 3 LC3 compartments, I can then just drop stuff into them without thinking. The rear compartment is big enough for batteries, a flash gun, flash trigger of some sort etc... Also has a slot of business cards which is really really useful to be honest.

I have a larger shoulder bag that I transport things in, it stays in the car now with the spare bodies etc... and the LC3 sits on top of everything.

Cheers

Cris

Sounds like a good rig for your needs!

Can you elaborate a bit more on this? This is very interesting.
I understand what you're saying, but isn't that true only limited to one specific sensor? In other words, if you take SensorA and shoot at ISO 800 instead of ISO 100 specifically to reduce the time the shutter stays open, then you are obviously exposing SensorA to less photons, therefore losing information (high bit will be zero) compared to a longer exposure at ISO100.
But, aren't different sensors responding differently? Isn't the point of sensors/cameras reaching ISO 25.000, for example, precisely the ability of those sensors of retaining more information out of that shorter exposure to photons?

In other words, would I be wrong in assuming that a Canon 5DIII "SensorB" will have LESS zeros at high bit at ISO 25.000 at 1/320 shutter speed, for example, than the Nikon D800 "SensorA" at those same settings (if it could reach that high ISO)? If so, then I'd rather have that possibility, which I assumed was the whole point of such sensors compared to competition. And I *think* I saw some nice pictures at high ISO from the 5DIII, but may be wrong there.

What am I missing?

Let's simplify the world down to a single pixel sensor. It has a very simple API (this being modestly simplified, but not by much):

interface Pixel {
void SetZero();
uint16 GetValue();
}

If you want to take a picture, you tell the pixel to zero itself, then you open the shutter, wait a while, close the shutter, and finally read out the value of the pixel. You'll note that the pixel has no clue about ISO speeds and whatnot. It's just a little capacitor that builds up charge from photons that arrive.

When you read it out, you get an integer. If too many photons arrived, you'll get MAXINT. If none arrived, you get zero. To keep things simple, let's assume that the sensor is perfectly linear in its response to photons. Double the number of photons, double the integer you get back. (In real life, there's an analog-to-digital converter in there that tries to convert the charge in the capacitor into something resembling linear response. Let's ignore that for now.)

Instead, let's talk about mapping this linear uint16 to color values in a JPEG (which go from 0-255). Every sensor has a "base ISO" of some sort. This indicates that an exposure for that base ISO should get close to MAXINT and a standard 50% grey card should come out to MAXINT>>1 (i.e., half of maximum). The simple answer is to just take the high eight bits and call it a day.

When you're going to a higher ISO value, all you're doing is shifting the 8-bit window of that uint16. The high bits are zero anyway, so ignore them. Every bit you shift is exactly equivalent to doubling the ISO value.

So.... if your pixel gives 16 perfectly absolutely awesome noise-free bits, then you can shift 8 times. If your base ISO was 100, then your final ISO is 100 * 2^8 = ISO 25600. In reality, the low-order bits are noisy, and the lower you go, the less useful information you get. Furthermore, say you only had 14 real bits of information (i.e., the bottom two bits of that uint16 were always zero). You could still shift by 8, and you're still technically shooting at ISO 25600, but you've only got 6 bits of non-zero signal, and the low-order bits of that are going to be noisy crap.

To complicate things modestly, when you have a larger pixel, your base ISO might be higher. The Nikon D700's base ISO is 200, versus 100 on the D800. That means the D700 kinda gets the first shifted bit for free. (Downside: if you're shooting something that's stupidly really bright, you might find yourself wanting a neutral density filter on your lens. For most people, this is a non-issue.)

Whew!

What this boils down to is that many cameras will claim to support ISO 25600 or higher, but they're not exactly doing it very well. That's why DPReview (and others) conduct various tests to measure noise, resolution, and so forth. Many camera vendors will play post-processing games to try to clean up the mess they get from their sensor. You can run all manner of noise reduction algorithms, but to really clean up a mess, you're also going to get rid of very fine details. Or, you can just downsample the image altogether, pooling several adjacent source pixels into one final pixel. You give up image resolution, but you gain back the ability to have a non-crappy high ISO exposure.

If you're hoping to print something 6 feet tall, sharp as a tack, you'll be unsatisfied with this tradeoff. If, however, you're shooting for a newspaper or web site, and you don't need anything more than 1000 pixels across for the final image, you're going to be perfectly happy shrinking things down from the source resolution. (Which is exactly what newspapers and so forth do.)

After all that, I can finally address your question. If you're comparing a hypothetical 22 megapixel vs. 36 megapixel camera, otherwise made with the same generation of sensor technology, the 22 megapixel camera may have less noise in the low bits. The 36 megapixel camera has more pixels. When you reduce the image from both cameras to your target output resolution, they'll probably perform very similarly.

When you want to talk about the specifics of the Canon this vs. the Nikon that, you instead need to look at the test measurements. Those data seem to suggest that the Nikon D800 kicks everybody's ass. Today. We'll see what comes out tomorrow.

Ok, more interesting info, so thank you!
And of course more interesting questions to come (at least for me)
I am on my mobile phone so i'll just ask one which i can't keep for later: so if i understand how the machine works, ISO setting only actually matter for jpeg output? To put it differently, if i could take two identical shots, at the same time, all settings being the same except ISO, will the generated raw files be identical?

The ISO setting affects the voltage that is fed to the amplification circuitry on the sensors. Higher voltage means more sensitivity, but also more noise. Lower voltage means a cleaner image, but longer exposures required.

So, no, not just for JPG images. It also matters for RAW.

-ml

I'm not entirely sure that's correct about the voltage variation. If anything, it will vary from one camera to another.

Certainly, when I've radically underexposed a shot at a low ISO then just cranked it up in Adobe Lightroom, the effects are indistinguishable from just shooting at the proper high ISO to begin with.

I've read up extensively on this stuff in the past, but don't have any current links to provide you. Do a little research, though, and discussion of voltage/amplification/noise should pop up.

I thought I'd read that the amplification was only turned up when you enable those extended ISO modes that most DSLRs have.

But a quick bit of Googling proves I remembered wrong

Changing the ISO settings is basically changing the gain on the amplifier between the sensor and the A-to-D converter.

http://photo.stackexchange.com/questions/2946/how-is-iso-implemented-in-digital-cameras

Yes, I thought I read the opposite some time ago, but I was not particularly interested in the topic at the time and may be remembering wrong. In any case, I thought that the bit-shift Dan mentioned was in fact the "extended" ISO modes, while the ISO setting per se consisted in - that is the way I remembered - changing some "hardware" operating parameter on the sensor itself. Mark's amp voltage change explanation seems to fit well what I remember. Of course, I really don't know. I am mostly learning how this works as we speak.


In either case, I wonder why is Nikon D800 ISO sensitivity is reported 100-6400, while Canon 5DIII is 100-25600 (again, I am just curious to understand how it works, no intention to prove any specific product "better").

If ISO setting is as Mark indicates, one may argue that Sensor+amp circuitry on some cameras be good enough to allow workable pictures at higher ISO, while some other sensors+amp are not.

If ISO setting is only software/firmware based, as Dan suggests, then I can't find a reason why Nikon shouldn't push the settings all the way to 25600, a good marketing tool regardless, right?

Maybe... but not necessarily.

People in the market for a Nikon D800 are likely to be somewhat knowledgeable about digital photography. You [Taym] and I are not world-class experts on the subject, but we both know that anybody claiming to create usable photos at ISO 25600 is exercising extreme marketing hype, and not only that claim but perhaps others they might make become suspect.

Such a claim would instantly bias me against the Canon, and their "marketing tool" (in my case at least) would work against them.

tanstaafl.

Yes, of course you're right, and I was exaggerating in "2.", above.

But actually, precisely because these cameras target somewhat knowledgeable market, I seems unlikely to me that Canon is producing a sensor which is just identical in performances to Nikon's, but at lower resolution and with a mere software tweak used to throw at buyers a "fake" ISO 25600 (and dpreview.com does not put it that way, in fact, for what it is worth). While we know that pictures at 25600 are in many cases useless, I would also reasonably expect that the whole ISO range from 6400 to 25600 does bring benefits, just as higher resolution does. One may argue which is a more relevant benefit, but still.
And, this seems to me a realistic possibility since the ISO setting is actually modifying how the sensor responds (by changing the voltage fed to it) before the ADC kicks in, so before entering the digital domain.


That actually is one of the most advertised features of the 5D Mark III. Or, at least, it is in the form of "best camera ever to take pictures in the dark", or something along those lines. Still a bold statement, I'd say, that I would expect carries at least some truth in it.

This is a 5DIII ISO 12800 shot: http://www.flickr.com/photos/davebass5/8040289680/

It looks great to me. Now, unless I am missing something obvious, this could be taken at a twice as fast shutter speed than one could ever do with a D800 (at 6400). That's a feature, and a competitive advantage, that one may find desirable. Certainly the D800 will allow for other things that the 5DIII won't.

Now I, as a novice, amateur, and one who likes to carry his camera around when travelling and "size the moment", I think the ability to use a faster shutter speed is more valuable than pushing the resolution to 36Mpx. Indeed, others will prefer otherwise.

Apparently, I was wrong about how sensors do high ISO. Interesting.

That said, the DPReview tests and DxO people certainly know what they're talking about, and that's that the D800 has a kick-ass sensor that performs very well at high ISO, particularly if you're willing to shrink things back down again later.

On the other hand, the raw files are almost twice the size of other cameras.

At the end of the day, you buy what fits your budget and solves the problem. If you've got a specific thing that you're shooting a whole lot, then you might make specialized decisions about gear that are better optimized for that use case (e.g., purchasing a really bright prime lens of the size that you might otherwise spend most of your time zoomed). If you lack that sort of specific use case, then this whole discussion sounds a bit like premature optimization. It's not necessary.

Lastly, if you just need something to carry around for spontaneous moments, then you should be paying attention to size and weight as much as the quality of the sensor. You may prefer to skip the D-SLR universe entirely and go for one of these world-beating compact things, like the Sony RX100.

Definitely, but I never intended to find the best sensor out there; I was more interested in understanding how the technology works. Looking at what top producers are doing can give you some interesting hints about that. And, on those lines, I was (am) wondering why they do what they do.

For example, while Nikon seems to have pulled out a higher resolution sensor than Canon that also performs slightly better when cropped down to the same res than Canon's, it seems to me there's also more than that in the whole story, which is interesting. In fact, Nikon sensor does not reach the ISO values that Canon sensor does. In light of what the ISO setting does physically, such extra feature is an interesting fact that seems to suggest that Nikon and Canon are going two different paths, to some extent.

So, I am just observing that

A fun thing I've always enjoyed with my cameras, is their ability to take great photos sans flash in situations where everyone around me is scratching their heads in puzzlement.

A lot of that goes down to the optical image stabilization in the lenses I choose, but a lot more goes to shooting at up to ISO 1600.

A newer body would give me 2-4 stops (ISO) more of exposure latitude there, which given the results I'm already getting, would be astounding! Granted that I'm already happy with what I get at 1600, the difference between 3X versus 4X that are pretty much moot.

So perhaps I ought to upgrade. Just need to find the right camera body that's compatible with my (Canon EOS) lenses. Come-on Canon, gimme a great body with onboard flash (or "speedlight" in Canon terminology)!

I found that image stabilization makes a huge difference for me. In some specific situations, I was blown away.
I took some pictures at a concert in particular, with a 70-300 IS mostly set at 300mm (which we've been talking about here on the BBS a while ago). Instead of the expected blurry mess :), I got some amazing pictures. That is entirely due to the good equipment and IS in particular, because I really did not do anything special, I had no tripod or even monopod, and I was essentially lacking any experience in shooting in those lighting conditions. I just gave it a try. Those pics were so successful among friends I ended up putting them on Flickr for them to download!

And so I did. I now have a Canon 7D body, supplanting the 40D.

It's a nice camera, very quick, lots of nice features including real weather seals, etc.. But despite having native ISO 6400, plus an "H" mode of 12800, it isn't really that much better for noise than the 40D was (native 1600, "H" of 3200).

The 7D here seems to be somewhere between 1 - 1.5 stops better than the 40D, though the built-in noise reduction firmware makes it hard to tell.

However.. since it has about double the number of pixels (18mp versus 10mp for the 40D), I suspect I can just save my images as half-sized "mRAW (10mp)" and see an extra stop or so of usable ISO from pixel binning. Not tested yet, but hopefully so. I've always been happy with 10mp, so no hardship here in continuing at that file resolution. And doing so keeps the storage/computing requirements more or less unchanged.

I'm enjoying the new camera, but also the second toy:

A new zoom lens in the bag: the Sigma 18-250mm f/3.5-6.3 DC Macro OS HSM super-zoom. This lightweight swiss army knife now replaces the Canon EF-S 17-85mm-IS as my primary travel lens.

Image quality looks to be improved (versus the Canon 17-85mm), weight/bulk are down, the image stabilization works very well, and the zoom range is phenomenal. Focus is fast, but cannot be overridden manually (as with the Canon ring type USM lenses) unless the AF/MF slider switch is adjusted first.

So now I can still cart the heavy 7D body along, but with only two lenses in the bag, both rather light in weight: the Sigma, plus the Canon EF-S 10-22mm for indoor shooting. All without cutting off circulation as I shoulder the kit bag.

Happy days!

Merry Christmas to you!

I loved my 7D, great choice Mark.

And I spoke in the past tens because I have my little story to tell you guys too: in few days I sold my 7D+10.22mm and got a 6D, with just 200 Euro extra.
That's the surprising (to me) power of "annunci.ebay.it" which, I now realize, half of the country is watching day and night.
annunci.ebay.it is a Craiglist type of service eBay.it has here, which I had no idea could be so popular. As I've been thinking of a full frame camera for a while, I decided one night (around midnight!) to post my 7D and the only EF-S lens I have (10/22mm) ads, without thinking too much - after all I was very happy with my 7D.
The morning after, I had tens (!!) of emails from all over Italy for both pieces. Wow!

So, in the end I thought: why not? It's Christmas after all and that will be my present to myself, I guess. In two days the entire operation was done and I am now the very happy owner of a Canon EOS 6D!


Now I am in the market, in the long term, for either a 17-40mm or a 16-35mm to replace my old 10-22mm.

You'd not have have any luck trying to pry my beloved 16-35L from my cold dead hands.. Even on cropped format it's a great focal length and a fantastic lens! Decent size too (not massive like many L lenses).



Happy Holidays!

Time to amend that thought: Having now seen some of the images (at reduced size) on my 1920x1200 IPS notebook screen, I'd say the 7D is giving me at least 2 stops extra for usable ISO, over the 40D. Which is pretty good considering the 40D was widely thought to be less noisy than the 50D, and the 7D is reported to be only slightly better for noise than the 50D.

Maybe it's the built-in noise reduction, maybe just the binning of pixels when images are reduced for display.. dunno, but the ISO 3200 / 6400 shots look fantastic!

Cheers

So, one more very positive opinion on the 16-35 I've been looking for some used ones. Do you own version I or II, Mark?


We're now going to my parent's house in the mountains. I should have sorted out all issues with internet access from there (there's an ISP that offers radio-based connectivity, which I think I mentioned here last August), but you never can tell until you're on site. I will finally test my 6D on the field, which I can't wait.

If I can't read you all, Happy 2013!

I believe mine is the original 16-35L, from about 10 years ago. No "II" anywhere on it. A very, VERY sweet lens!

Some used version IIs are sold at 850-950 Euros here. Iam guessing a cheaper version I could be a deal.
V. II is slightly bigger in size, and has various IQ improvements, based on some reviews I read. A brief review with some good pictures to compare the two versions, here.

Mmm.. The II version would have to be a lot better than the original for me to want that extra bulk in the bag. For me, the original 16-35L is stunning for the quality and the small size of the lens.

I'm sure either will be excellent for you though!

Speaking of which.. I think I'm going to sell my 70-200L-IS f2.8 lens -- it simply doesn't get used enough, despite the incredible image quality from it.

It is likely heading to eBay unless somebody I know wants it.

And ditto for the 70-300-DO-IS compact zoom. My new Sigma ultra-zoom has displaced it as well as the 17-85mm-IS Canon lens.

Cheers

Rightfully so, I think. The IQ improvements are not at all huge, based on what I read, and in any case v I is still rated excellent by virtually everybody I've heard from and by any test I've read.

Interestingly, however, v II weights just as v I, which is good.

Lousy bokeh on that lens, but everything else about it is fantastic. Rugged (dropped it twice already, oops!), very, very sharp, incredible zoom range, authentic macro capability, amazingly affordable price tag, etc. A superb travel and walkabout lens; one that will be on the camera far more often than not.

So I've now gone whole hog and ordered a polarizing filter and an ExpoDisc to fit it.

Until now, I've made do with a single 72mm ExpoDisc for all of my lenses, but this Sigma lens is going to get used a lot, so having the proper size (62mm) ExpoDisc for it will be a boon.

Cheers

The Canon EOS 7D body has a built-in Canon "Speedlite Master" controller, which is accomplished through use of the built-in flash to send commands to slave Speedlites.

Since Canon has been offering point-of-sale rebates on Speedlites here through December 31, I decided to take advantage and try out the system, purchasing a pair of 320EX Speedlites from a local dealer at the sale price.

Using the 7D as the controller, with its onboard flash as a fill lamp (or even not used for exposure), the system works very well with the two slave 320EX strobes. Well enough that I'm inclined now to carry one in the bag on future travels for use as an outboard fill lamp and macro strobe. Which is saying a lot, since I've only rarely lugged along an external flash before now.

Canon's Speedlite exposure system has puzzled many (most?) users, myself included. But it begins to make sense once outboard slaves are employed.

Cheers

Ok, one very first impression after finally taking the 6D out with me for some more extensive time, is that it gave to my 24-70mm lens (yes, Mark, the one I got from you, still a favorite of mine) a whole new meaning. 24mm is a real "wide angle" focal length, while 70mm is still a nice zoom. Quite obvious for many here, but I just thought I'd mention it.

I am probably going to miss the zooming capability of the 70-300mm, but I have not really used it extensively since I have the 6D to tell.