[drewprops]

I wanted to prime the pump for a conversation about next year's camera offerings at the Consumer Electronics Show. I'm a consumer-camera user, so my interests will be on what the latest offerings will be in the compact category. I was just looking at the Canon S95 online and wondering if it might be my next camera and wonder what features are likely to be upgraded in its next outing.

Anyway, hope to see some good posts about what rolls out here in a week or so....



...

[Chinney]

Paging Dorian....Paging Dorian.....

[PB PM]

Since the S95 was just released in September, I'm guessing the replacement is a ways off yet.

CES is normally sees updates to basic point and shoot models, and maybe some entry level DSLRs, but all the big updates came out in September before/at Photokina. I think we might see an update to the Olympus m4/3s lineup.

[Matsu]

So far it sucks. Nothing of note. This pobably isn't the show for it, but I'll be interested to see new Full frame DSLR/sensors. At least I thought we might see some new innovative large sensor mirrorless cameras. Blah...

[Dorian Gray]

Ha! I agree with Matsu. The DPReview homepage is a good place to get an overview, and the CES landscape is pretty barren. Generic 16 mp compacts as far as the eye can see... The best of them are probably only marginally better than the worst.

One welcome trend is the move towards integrated GPS, but not everyone will see the point of geotagging, and many will find it difficult even if the manufacturers use SiRFstarIV in these new cameras (and I don't know that they do).

The Oly XZ-1 has impressed Richard Butler, but I see it as only another LX3 copy - minus the LX3's useful 24 mm-e wide-angle.

The Canon S95 is a pocketable camera with good picture quality, and in that strict niche it has no peer. The Panasonic LX5, Canon G12 and Nikon P7000 have more features, better lenses, and better build quality, but they're heavier and bulkier. If you want an S95 I think now's the perfect time to buy: it's been out just long enough for the price to come down from the initial gouge-point.

[Matsu]

I wonder where the DSLR market may be headed. Nikon has filed about 3 dozen patents related to a mirrorless camera system (lenses and body). I'm sure some of this is just smoke, but it makes sense. A new system will have to be virtually complete out of the gate in order gain support in the market. What then are the consequences for the F-mount? Somewhat counter-intuitively, I think it could be worse both for the F mount and future EVF system if a new mount offers unclear market signals.

On 35mm formats, conventional film era wisdom tells us that mirrorless camera lens designs have advantages up to about 75mm focal length, some say 90, and that SLRs are operationally superior and optically equal beyond that point. I'd take an affordable RF

Leaving aside sensors and EVFs, which will eventually be cheap commodities, the contemporary digital market is further complicated by other considerations - video and zoom lenses which favour divergent designs depending on market and application. An EVF design primarily for video would look very different from one designed as a new era digital range finder. m43 somewhat combines those worlds but is hampered by its small sensor when it comes to stills. Variable formats under one mount might make sense: with lenses and bodies designed to automatically crop/correct depending on the combination detected.

If there is a new mount, I could see the DSLR heading up market and the EVF filling the consumer space...

[drewprops]

I've read some detailed reviews of the S95 which laud many of its improvements over the S90, but that camera's video has taken some slagging, particularly over the way it can't hold focus during a shot.

Any of you camera jocks ready to hazard any guesses as to whether Canon will address the outstanding criticisms of the S95 in the same way that they went down the bullet list of complaints to the S90?

I just read about Photokina on Wikipedia... it happens every TWO years?
Perhaps that's how it has lasted since 1950?

"Too much of a good thing is a bad thing"

[PB PM]

Originally Posted by Matsu If there is a new mount, I could see the DSLR heading up market and the EVF filling the consumer space...

I agree, the "entry level" DSLRs of the future will be D7000s and Canon 60Ds. Mirrorless cameras make more sense for most people who are in that price range anyway. I see a growing number of low end DSLR shooters working only with liveview already.

[Xaqtly]

I looked through the camera offerings at CES, I didn't really see anything too revolutionary that was going to be out any time soon, aimed at my level of skill i.e. novice/hobbyist/prosumer. So, I went ahead and bought a Canon Rebel T2i to replace my aging XT. Mostly I'm looking forward to having better AF, higher/less noisy ISO range and HD video.

Also the T2i uses SDHC/SDXC where the XT uses CF. I got a 32GB Class 10 SDHC card to go with it. When it arrives (tomorrow) I'll see about getting some shots with it.

[Matsu]

PB PM, I agree. And, as consumer oriented models these will feature:

1.) at most APSC sized sensors
2.) hybrid video functions
3.) large ratio zoom lenses

So, basically, not the AF digital M I'm hoping that someone will eventually make. Given in 35mm full frame format with a selection of primes covering 24-75/90 I'd get one in a heartbeat., but that's besides the point right now.

I think it would be wise to de-couple sensor size from the mount spec on a future EVF camera mount. This way, it could be built for a range of sensor sizes and lenses covering field, studio, video and cinema functions. Think of a concept not unlike medium format bodies that mount 645 to 6x9 film backs. Not all lenses would have to use the whole frame of every camera, not every camera would use the whole image of every lens. They could be matched as need and circumstance dictate. Not every combination would be "ideal" but there would be ideal combinations for every purpose, and flexibility and growth within the system.

The mount would need sufficient throat diameter for larger frames (at least 36x24, but perhaps even more, like 45x30?) and enough control and electrical/optical coupling for all of the following: AF, VR, focal plane shutters, continuous aperture adjustment, powered zoom. Again, not every lens would use/need them, but they would be there to use just in case. With such a mount, because you're not worried about the intrusion of a reflex mirror, you can make lenses for smaller imagers protrude further into the camera body, and those for larger ones extend further outward. They might all have more girth, but you could rather easily have larger 35mm+ format primes and large ratio 4/3rds/APSC lens body combinations all on the same mount. Completely flexible and future proof.

[Matsu]

Of course this is all a bit forward looking and pie-in-the-sky. The show to watch this year will be PMA, and it's not until September... I'm hoping we some indication of new models before then.

[Xaqtly]

Here's one of the first shots I took with my new T2i. First thing I noticed about these shots is how much better low light shots are coming out, compared to my Rebel XT. It's a - ahem - night and day difference. This was I think the 5th shot I took with the camera, after having charged the battery and played with the settings a bit. I was amazed by how good by direct comparison the T2i's pictures were. It's a HUGE difference from the XT.

I wasn't going for photographic awesomeness, I was just screwing around and caught my dog with a funny expression. The camera just did a really good job of capturing it. This is the kit 18-55 IS lens.

[zippy]

Cute dog. I think he's sticking his tongue out at you because you spent all that money on the camera instead of yummy dog treats - and then expect him to pose for nice pictures.

Guard your shoes.....

[Xaqtly]

Originally Posted by zippy Cute dog. I think he's sticking his tongue out at you because you spent all that money on the camera instead of yummy dog treats - and then expect him to pose for nice pictures. Guard your shoes.....

Ha! Luckily for me I also bought him a huge bag of bully sticks, so he's happy.

[AWR]

Originally Posted by drewprops I've read some detailed reviews of the S95 which laud many of its improvements over the S90, but that camera's video has taken some slagging, particularly over the way it can't hold focus during a shot.

I'm no camel jockey but I've been putzing around with a new S95 for 5 weeks or so. Picture quality is good in most instances (no DSLR but excellent for a P&S) and I can't recall focus holding issues during video (I might not be paying attention enough). I have taken plenty of video although I have not spent much time critiquing finer details. Will try to have a closer look tonight or tomorrow.

[PB PM]

I think the issue people have is that the Canon S95, like the G12 cannot auto focus while recording video. That is the one area that the Nikon P7000 has a slight edge over the Canon models (including external mic input for audio recording).

[Matsu]

I saw posted some patent applications from Olympus and Panasonic on lens mount adaptors for 4/3rds. What's interesting about them is that they include a focal length reducer of .5x. This is cool because it restores the correct field of view to legacy 35mm lenses while simultaneously speeding them up to provide equivalent apertures. It's reported that Olympus essentially used such a formula grafted onto the back of their f/2 zooms. That is they used an f/2.8 full frame design with a .5x reducer on the back, and then closed the aperture down a bit to f/2 because they couldn't quite make it f/1.4 without some problems. People who've played around with them say there's a bit more than f/2 aperture available by fooling with the lens release mechanism.

Panasonic's version is interesting because it provides for a .5x focal reduction and a mirror-prism for phase detection, so, in condensing the light by a factor of .5x, you create a 2 stop brighter aperture. Same field of view and same depth of field. A 24-70 f/2.8 on 35mm, through the reducer becomes a 12-35 f/1.4 on m4/3.

Please note, this is by no means a more compact set-up. It's even bigger (lens plus reducer) but it gives you the same exposure through a smaller sensor, at least in theory.

Why would anyone want this? Cinema applications with expensive cinema lenses comes to mind. As a gateway camera from DSLR to EVF cameras, but perhaps something a bit expensive for that? For a future with a sort of sensor size independence?

Same could be done, for example, using APSC sensors and a .67x reducer. A 24-70 f/2.8 full frame becomes a 16-46.6 f/2 APSC, again with the same depth of field and field of view.

I don't see it happening, but it is interesting. A looooong time ago, Nikon tried this built into a camera, but it was a huge frankencamera, and probably the focal reduction was too great. Curiously, the optical reduction system sucked up a lot of light (the effective T-stop of the lens-camera combination being a lot less) which is not my understanding of how a reducer should work - ie. concentrating the same light into a smaller area; thus larger relative aperture...

Anyway, interesting nonetheless. Maybe I'll update this post with some links later.

[PB PM]

I think the biggest letdown of CES is the E-PL2. Olympus is still using the same old 12MP sensor that is three years old now with almost no improvement in image quality at high ISO. Does Olympus really think that adding new "art filters" is going to sell the camera?

[Dorian Gray]

Originally Posted by Matsu What's interesting about them is that they include a focal length reducer of .5x. This is cool because it restores the correct field of view to legacy 35mm lenses while simultaneously speeding them up to provide equivalent apertures.

Another thing it would do, unless badly designed, is increase by nearly a factor of two the spatial resolution for equal MTF. Yay! Which is not to say you'll get better sharpness in the final photo of course (twice the magnification, etc.), just that you won't get much worse sharpness (as is the case when using legacy 35 mm lenses on small Micro Four Thirds sensors now).

Originally Posted by Matsu Please note, this is by no means a more compact set-up. It's even bigger (lens plus reducer) but it gives you the same exposure through a smaller sensor, at least in theory.

Though we should be clear that it doesn't solve the problem of noise/dynamic range caused by small sensors at base ISO. But in light-limited and depth-of-field-limited scenarios it would equal a full-frame setup, in terms of noise - assuming equal sensor efficiency (still not achieved by Micro Four Thirds).

Originally Posted by PB PM I think the biggest letdown of CES is the E-PL2. Olympus is still using the same old 12MP sensor that is three years old now with almost no improvement in image quality at high ISO. Does Olympus really think that adding new "art filters" is going to sell the camera?

It looks like Panasonic isn't selling their new 16-megapixel sensor to Olympus. Since Four Thirds and Micro Four Thirds need all the momentum they can get before the big boys release competing products, I can't imagine Panasonic is doing this willingly - maybe they have yield issues with their new sensor?

[PB PM]

Either that or Panasonic is keeping it to themselves to prevent Olympus from grabbing more market share.

[Matsu]

Originally Posted by Dorian Gray Another thing it would do, unless badly designed, is increase by nearly a factor of two the spatial resolution for equal MTF. Yay! Which is not to say you'll get better sharpness in the final photo of course (twice the magnification, etc.), just that you won't get much worse sharpness (as is the case when using legacy 35 mm lenses on small Micro Four Thirds sensors now). Though we should be clear that it doesn't solve the problem of noise/dynamic range caused by small sensors at base ISO. But in light-limited and depth-of-field-limited scenarios it would equal a full-frame setup, in terms of noise - assuming equal sensor efficiency (still not achieved by Micro Four Thirds).

This is so interesting. I guess it depends on how close the sensor can get. Sony's new APSC sensor is very quiet and has great DR at base ISO, so running that an ISO or two lower than a FF would be a viable option. f/2.8 ISO 400 on APSC+0.67x reducer (net f/2 speed) vs f/2.8 ISO 800 on Full frame. Nothing available in 4/3rds is really as good, yet, but on a 4/3rds+.5x combination the comparison would be f/2.8 ISO 200 +0.5x reducer (net f/1.4 speed) vs f/2.8 ISO 800.

What does the dynamic range of a good 4/3rds sensor look like at ISO 200 vs a full frame sensor at 800? Of course you reserve the option to shoot the FF at lower ISO as well.

Perhaps everyone is really looking at the wrong end of the spectrum. Using a reducer could make for an interesting path to a unique and more affordable "medium format" option. If it turns out a bit too pricey, then that's the perfect niche within which to work out the kinks - turning 35mm full frame cameras into ersatz medium format ones...

[PB PM]

I think I'd go for the newish Pentax medium format camera (hey its only $3000 more than D3x), rather than using a reducer.

[Matsu]

But would you take it over a 5Ds2 ? Or, better question, would you want to own a medium format back at all? A lot of the time these are just rented, while most people own their 35mm bodies.

Here's where I see the niche. A guy has a medium format film rig left over from back when before he moved most of his work over to 35mm DSLRs. He might rent a medium format digital back from time to time when he needs those big files, but operationally, it's just not even close to the conveniences of a modern DSLR. So, now he buys this reducer and uses his old lenses on the new body when he's after a certain look but doesn't feel like renting out a back. Properly design, he would keep all the conveniences of his DSLR (minus AF)

I think that an even more fun thing to do would be to use an astigmatic reducer to create a modern digital XPan out of a DSLR. Something that squeezes the frame horizontally, but otherwise leaves it in tact vertically. You then re-stretch it in software, so you don't lose any of your pixels to the crop. Though your final image would have greater vertical than horizontal DPI.

[PB PM]

Originally Posted by Matsu But would you take it over a 5Ds2 ? Or, better question, would you want to own a medium format back at all? A lot of the time these are just rented, while most people own their 35mm bodies.

The Pentax 645D is an SLR style medium format body, with a 40MP sensor, not a digital back for a film body, and if I was a dedicated landscape photographer who made big bucks, I'd pick the device that got the job done best.

I wouldn't want the 5D MkII, I'd pick the D3x and 14-24mm F2.8G.

[Matsu]

This is true, the smart money may be on used Pentax gear if/when newer models come out, though I think that the update cycles for this level of equipment are a lot longer, so we don't get to answer that question for a few years. Of course, I look at the concept of reducers, for now, as a novelty, but an interesting one. We know they've been used built-in for at least some lens designs.

It would be really interesting to see some f/2.8 standard zoom designs reformulated into f/2 APSC designs. With an extra stop of brightness, and 50% greater spatial resolution, they would bring essential parity to APSC platforms relative to 35mm, when using the standard zoom lenses. Would look like this:

APSC kit:

16-46 f/2
50-135 f/2

35mm kit

24-70 f/2.8
70-200 f/2.8

Roughly equivalent FOV and DOF, probably just about the same size/weight lenses and maybe even bodies, and I wouldn't expect the lenses to be any cheaper either. That said, if we saw such lenses, I think that wold be a strong indication that 35mm full frame remains expensive for some time to come. Not that it's too expensive for what it is, but perhaps that it's too expensive for the update cycle of contemporary cameras - and they become more like medium format in the marketing as well. It's nice that some well heeled amateurs, moonlighters and hobbyists bought 5d-d700 level cameras, but having those around gives pros a cheaper alternative, and that may not really be a good idea. Those guys can and will pay, so they'll just make them.

Most of the current crop of full frame cameras utilize one of only three sensors. Nikon's 12MP, Sony's 24MP or Canon's 21MP, and all have been updated far less frequently than their APSC-m4/3 brethren where there are 12-14-16-18 MP sensors in 2X, 1.6X, 1.5X and 1.3X

Rumors now suggest a forthcoming 25MP APSC Sony This has nothing to do with any possible high speed APSC zooms, just shows that the update cycle for APSC is faster at the moment. We might see 30-40MP full frame sensors soon as well, then all this means nothing.

Right now a pre-level APSC body costs from 1200-1800. A pro level 35mm goes from 2500-7000, but at the bottom are some compromises. 5ds2 has semi pro level body on a pro level sensor, D700 has it the other way around. A fully pro-spec 35mm camera body starts at 5000.

After a fashion it makes sense to sell high-level, expensive APSC lenses and relatively cheaper APSC bodies. Lenses I will keep for a long time. Bodies I will recycle every 2-4 years... I don't believe that the price differential between APSC and 35mm should be so great, but maybe it is, and maybe that's where the manufacturers want to keep it?

[Matsu]

And just one more possibility for the focal reducer slash phase detection module. Either of these could be built into lenses independent of each other. One way to put really fast AF into mirrorless cameras might be to build a phase AF sensor into the lens itself. You could use a small transluscent mirror embedded in the optical path of the lens, and so long as it's capable of communicating the correct information with the camera, bingo, phase detection lenses! These could augment the camera's AF in certain scenarios or take over entirely depending on use.

[Dorian Gray]

Originally Posted by Matsu f/2.8 ISO 400 on APSC+0.67x reducer (net f/2 speed)

A quibble only, but you mean a 0.71x reducer, i.e. 1/(square root of 2). Edit: or more likely, you do mean 0.67x, but the f/2 speed is rounded. Fine.

Originally Posted by Matsu What does the dynamic range of a good 4/3rds sensor look like at ISO 200 vs a full frame sensor at 800?

They'd be equal if the sensors were equally sophisticated, but at the moment most full-frame sensors, although not cutting-edge, are still far better than the 12-megapixel Four Thirds sensor.

Originally Posted by Matsu Of course you reserve the option to shoot the FF at lower ISO as well.

Yeah: this is where the argument runs aground, I fear. And physics say we'll never overcome this either. Mid-tone grey at ISO 100 on an imaginary Four Thirds sensor with 100% quantum efficiency would still be noisier than mid-tone grey recorded by today's relatively inefficient full-frame sensors at ISO 100. No amount of new technology can change this: incoming light is just too noisy! The only way for Four Thirds to catch full-frame is by allowing much longer exposures, i.e. offering ISO 25, which can only be achieved by deeper electron wells; and since capacitance is closely related to area this is a hard ask (the old digital cameras with very low ISOs didn't have deeper wells, but rather less efficient micro-lenses — or none at all). Worse, any improvements to small sensors might be applied equally to larger ones.

Originally Posted by Matsu It would be really interesting to see some f/2.8 standard zoom designs reformulated into f/2 APSC designs. With an extra stop of brightness, and 50% greater spatial resolution

41%, since gaining a stop would imply a 0.71x reducer. (And of course it would be a bit lower than 41% because of imperfect optics.) Edit: I should leave rounded figures alone!

Originally Posted by Matsu After a fashion it makes sense to sell high-level, expensive APSC lenses and relatively cheaper APSC bodies. Lenses I will keep for a long time. Bodies I will recycle every 2-4 years... I don't believe that the price differential between APSC and 35mm should be so great, but maybe it is, and maybe that's where the manufacturers want to keep it?

The idea of reducers, etc., is interesting, but it sounds a lot like Four Thirds, which failed because the cost of APS-C sensors plummeted, and even full-frame sensors were put into $3k cameras. Full-frame sensors admittedly have things against them from a manufacturing point of view, which makes them disproportionally expensive and forever doomed to be one step behind the cutting-edge of sensor developments. But this is more than made up for by their real performance advantage at base ISO, which cannot be regained by clever optics on an APS-C sensor, and by the strong emotional attachment to "full frame" — note my use of this loaded term throughout this post, for example!

[Matsu]

Yeah, the market's the real determiner of what gets built. Personally, I'd take a full frame 35mm DSLR, but I'm not paying full freight for a new one at this point in the cycle, not even for a used one. Not until we see it's replacement. Buying any D700 now is too much of a guess from the sale-resale end of things. So we wait.

The design build cycle for sensors looks to me that over the span of one generation of full frame sensor we see multiple APSC revisions, and then, just at the point where the overall performance of APSC seems to catch-up, we get a new generation of full frame. Well, APSC sits at 16-18MP right now, and with dynamic range and ISO performance better than old generation 35mm, and pretty close to the current generation.

Sony is suggesting a 24MP APSC soon... If it's time for a next generation of full frame sensors, my guess is we'll see them before or at PMA in Sept. My hope is that somehow this gets the major players to intro a $2K body (street), even that body rocks revisions to the older 21 and 24MP Canon/Sony sensors.

[drewprops]

My camera has a zoom!!

(every camera thread here quickly escapes into a blinding flurry of technobabble)


...

[Matsu]

In the absence of cameras we want or could/would actually buy, we babble. Let me add that Sony were the only company saying anything interesting at CES: new NEX cameras and lenses, and new Alpha's with more carryover video technologies.