[Clavius]

Still hoping for a 645D "Light" to steal some of the FF show

As I said in the other thread... Is the 645D "Light" going to sport the 16mp apsc sensor too?

(This site needs a grinning-devil-smiley.)

[Pål Jensen]

Still hoping for a 645D "Light" to steal some of the FF show

They need to come up with a CMOS sensor with live view and focus peaking. At 70+ mp in order to compete with the best FF cameras (for now the D800; in the future models from Canon as well)

[redrockcoulee]

They need to come up with a CMOS sensor with live view and focus peaking. At 70+ mp in order to compete with the best FF cameras (for now the D800; in the future models from Canon as well)

Are you saying that a larger sensor than full frame needs to be higher MP by 2X in order to be competetive with the D800?


Eddie: In an earlier post you mentioned some one printing 100 inch prints. A few years ago my wife had an exhibition mainly of an installation piece but also 8 60 by 80 inch prints. She shot them on both a D3 and a Fuji GX680 and only the Fuji images were used due the the huge difference in even smaller test prints. The film was scanned in a Nikon CS9000. At the same time Edward Burtinsky's work of the Oil Sands was shown. 4X5 no doubt drum scanned. The difference in resolution between the two bodies of work was amazing, for my wife's work high resolution was not necessary. I guess what I am trying to say is if some one is going to print so large, unless it is to be viewed from a distance such as one does a billboard, then use the appropiate tool. A large format drummed scanned or a medium format digital of any of the current higher ends is still better than a D800 in a large print from what I have read on line, not from first hand experience. Why should our dslr be expected to compete at this?

Does our cameras need to be able to print high quality 100 inch prints? How many can afford to do so, or to either frame them or in the case of my wife have them mounted on metal? A local photographer has a show right now with large prints from a combination of a 35mm film and digital but the largest ones are printed more like a poster and of pictorialist style so high res does not apply.

[gazonk]

As I said in the other thread... Is the 645D "Light" going to sport the 16mp apsc sensor too?

(This site needs a grinning-devil-smiley.)

No, it's going to support 645 FF through a 10 x 10 array of Q 10 sensors, thus delivering a mind-blowing 1.2 gigapixel

[Clavius]

No, it's going to support 645 FF through a 10 x 10 array of Q 10 sensors, thus delivering a mind-blowing 1.2 gigapixel

LOL! That would actually be really very cool though, altough my computer would seriously hate that!

Seriously though, I've always wondered why they never tried using multiple cheap sensors and inbody stitching to form FF images.

[timcatn]

LOL! That would actually be really very cool though, altough my computer would seriously hate that!

Seriously though, I've always wondered why they never tried using multiple cheap sensors and inbody stitching to form FF images.

My guess is that you would need ALOT of processing power to do that. Not to mention that you would have to interpolate the pixels between the sensors (there WILL be a slight gap). The main reason though i think is power consumption. You could theoretically put the equivalent of an Intel I7 inside a camera, but it wouldnt last for very long on normal batteries. The idea is very good as a concept, but not very practical.

[pinholecam]

LOL! That would actually be really very cool though, altough my computer would seriously hate that!

Seriously though, I've always wondered why they never tried using multiple cheap sensors and inbody stitching to form FF images.

About four K01, ganged up on a fixture and fired off via remote would do it.
Only dowside is the number of lenses needed and the rationale of going to this extreme.

[Ishpuini]

Then it's going to be a very boring Photokina for Pentax. I would hope they have kept at least one item to announce during the show or just before.

Well, the challenge of finding out whether the announced cameras are actually worth upgrading to from the actual offering is pretty interesting I think.

I have both the Q and the K-5, so with both the Q10 and the K-5II(s) retaining a similar sensor (in MP count at least) it will be interesting to find out the significance of the other changes.

But of course, there's always potential to make things more interesting!

Wim

[Winder]

Well, the challenge of finding out whether the announced cameras are actually worth upgrading to from the actual offering is pretty interesting I think.

I have both the Q and the K-5, so with both the Q10 and the K-5II(s) retaining a similar sensor (in MP count at least) it will be interesting to find out the significance of the other changes.

But of course, there's always potential to make things more interesting!

Wim

If you own a K-5 then I don't see any reason to upgrade. For people with a K-x, K-20 or older then it could/would be a viable upgrade. The K-3 will be here soon, so unless you K-5 has stopped working I'm not sure the K-5II would be worth it.

[Class A]

with moire correction through software that can be activated through the menus

Moiré is the result of incorrectly capturing information and can never be properly undone by software. It can be made look less conspicuous, but the information lost cannot be restored.

The effect of an AA filter, on the other hand, is much easier to counteract with just a bit higher settings for capture sharpening.

Falk (falconeye) compared the Nikon D800 against the D800E (no AA filter) and observed that

1. the differences are not huge (no one should expect a "significant increase in resolution" from the K-5 IIs), and
2. the D800 image can be made to look like the D800E image by just slightly changing the sharpening settings.

A regular Bayer-array sensor requires an AA filter to properly function (unless it outresolves any lens it is used with, which is not the case yet). I believe companies like Nikon and Pentax are offering AA-filter less camera versions because some customers want them (against technical reasoning), not because they think that they are a valid technical alternative or even technically superior.

BTW, Rob Galbraith has also done a D800 vs D800E comparison and I cannot understand why people are willing to accept all the ugly colour moiré for an advantage that can almost equivalently be achieved by slightly more aggressive capture sharpening (that can be blindly applied to all of the image, instead of requiring attention to local areas).

[ogl]

[*]the differences are not huge (no one should expect a "significant increase in resolution" from the K-5 IIs), and[*]the D800 image can be made to look like the D800E image by just slightly changing the sharpening settings.

The difference also: D800E has better, righter and clearer colours, better contract and better halftones. Especially, it's easy to see at landscapes and nature photos.
Skintones are more natural, not plastic like with cameras with thick AA filter.
My K200D is better than K-5 in terms of colours, sharpness and contrast.

Aggressive sharpening kills the picture. It's not cure for camera with thick AA filter.

Thick AA filter is answer of manufacturers to the wish of users to shoot in JPEG with acceptable quality and without thinking about post-processing.
There were more cameras with weak AA filter before. Pentax had K10D, K200D, K-m, K-x, K-r. Maybe K100D. Kodak FF, Nikon D70 and e.t.c...
It's seems to me K-30 has weaker AA filter than K-5
AA filter has also bad influence on the corner sharpness (of lens) at the final photos.

[Pål Jensen]

The effect of an AA filter, on the other hand, is much easier to counteract with just a bit higher settings for capture sharpening.
.

Maybe. But I have never seen a digital image that looked tack sharp right out of the camera. I always thougth that was due to the AA filter. If these images had been shot on film and seen through a loupe, in the pre digital days it would have been regarded as unsuitable for large printing. Surely there must be some loss of finer detal by using an AA filter? I wonder if people are comparing edge definition, which shouldn't be significantly affected, as oposed to texture and structure of the image?

[ogl]

By the way, D800E has better low-light ISO and a bit better colour deptht han D800 in test of sensors .

DxOMark - Camera Sensor Ratings

[Winder]

LR4 has a morie tool that works fairly well on RAW files. Once a file has been turned into a JPEG there is not much you can do with morie. There are some Leica M9 files you can download on Adobe's site to play with the morie removal tool. I'm sure Adobe will continue to improve this tool.

There are an awful lot of lenses out there that are soft enough at wide apertures that morie will never be a problem with them. You only see it with certain lenses, at certain apertures, and with certain subjects. It has not been a complaint with Leica.

I really was not expecting an AA-less body until we hit 24MP on APC-C. I hope we see a K-3 with 24MP sensor an no AA filter.

[Class A]

The difference also: D800E has better, righter and clearer colours, better contract and better halftones.

I am convinced that the absence of an AA filter cannot introduce the differences you describe. Where have you seen the respective images? Were they the result of a controlled experiment?

Aggressive sharpening kills the picture. It's not cure for camera with thick AA filter.

1. Neither K-5 nor D800 have "thick" AA filters. Both cameras still produce a bit of moiré in extreme cases.
2. There is no need for agressive sharpening. Just a bit of capture sharpening is required to increase the micro contrast a little so that the slight softening of the AA filter is counteracted.

Thick AA filter is answer of manufacturers to the wish of users to shoot in JPEG with acceptable quality and without thinking about post-processing.

This is not true. AA filters are required for RAW files just as well. You can think of an AA filter as a poor man's three beam colour splitter. Each photosite in a Bayer matrix sensor only captures one colour and the filter's effect can be regarded as splitting the photons that would have hit one sensel in a Foveon sensor to many Bayer matrix sensor sensels, so that all colours are captured, not just one.

Surely there must be some loss of finer detal by using an AA filter?

A well dimensioned AA filter only minimally affects true spatial resolution but mainly avoids colour moiré.

A monochrome only sensor could get away without an AA filter due to the high fill factor of today's sensors, but omitting an AA filter for a Bayer matrix sensor is just inviting trouble for very little gain.

By the way, D800E has better low-light ISO and a bit better colour deptht han D800 in test of sensors .

The scores are very, very close to each other and some differences may be just due to measurement tolerances. For instance, the D800 has higher DR than the D800E but I'm sure that the difference is not real. Even if the other differences are real and due to the AA filter, they are definitely inconsequential in terms for making a real world difference for image quality.

LR4 has a morie tool that works fairly well on RAW files.

I haven't used it yet but judging from my other Lightroom experiences, it won't be perfect. The main thing, though, is that you always have to apply the moiré tool manually in a localised fashion depending on an individual image, whereas you can set the global sharpening to an appropriate capture sharpening level once and then forget it.