The Follow Up to the K7

I know it has been a while since I discussed what Pentax may or may not come out with in the relatively near future. One of the reasons I did not upgrade to the K7, has nothing to do with the merits of it, but more to do with what I do have. As I own 3 x K20D's and a few K10D's, I haven't felt the need to upgrade as yet. I don't feel that the attributes of the K7 are sufficient for me to abandon the K20D, as the output is on par with the K7.

However, if and when Pentax come out with a camera that can give me a few more mega pixels with good noise performance, I will consider upgrading. I have a large arsenal of exceptional Pentax lenses and am very well covered in that area.

So here goes my speculation. The next APS-C format Pentax camera will have around 18.2-19.2 mega-pixels and will have similar output attributes as the K20D-K7. As I generally work within the 100-400 iso range, I suspect that the output will be outstanding and will compete quite handily with the Nikon, Canon and Sony counterparts. It is at that time that I will upgrade.

Well, there you have it in a nutshell. Oh, and I suspect that we might see the new upgrade in the late Spring to early Summer of 2010.

Ben

There must surely come a time, and not too long hence, when even the best lenses will not provide sufficient resolution for more and more densely-packed sensors.

Pentax has already replaced one lens - the original 18-55 - because it couldn't cope with the K20D's sensor. I would guess that at 18MP quite a lot of other lenses would fall by the wayside.

You may be speaking with the benefit of inside knowledge, but I really would prefer Pentax to concentrate more on image quality than pixels.

G

We have the 'Megapixel race'

We have the "High ISO race'

What I would really like the next big technological advancement to be is a 'Dynamic Range race.' If Pentax could concentrate on that one instead of following the herd, it could make things very interesting!

18.2-19.2 mega pixals on an APS-C sensor will be the last nail in the coffin for Pentax. Unless they have an ex Canikon noise reduction engineer lined up for a job they will flunk big time.

I would much rather see an APS-C sensor (or FF) with a pixel density of 1.5 to 2 for either size. I base this on the quality and noise of the 5D Mk1 and my own D700 images.

I agree with the timing. Wonder what the response will be if they do pull a FF out of the bag? Would Ltd lens sales rise??

Regards

One thing that bugs me with my K10D is the noise. I dont want them to go race like crazy. Just get me a good 800 and a quite usable 1600 and I'll be happy I think.
I understand than benji needs more MP, but most users, even those going for a top of the line camera will not really need all those pixels. I would be happy with 10MP-12MP done right. And for the same price point, get me improvements in other areas...
Pentax could do what Canon and Nikon do, which is to launch two versions of the same model, but that would mean sufficient demand at that level to warrant the logistics required. On the other hand, if the 645D does go ahead, that would be Pentax's high MP option.

I don't know what people are complaining about with the noise on the K10D and K20D, in RAW it's comparable with or better than the Nikon D300 and Canon 50D.

You can read this in the review on DPReview on the D300s, and I also have practical real life experience as the 2 photographers I take most of my pictures with have those 2 precise cameras, and their images usually have more noise. (I do expose to the right though).

With JPEG's its a different matter as Canon and Nikon apply more noise reduction in-camera, and so their images will appear to be less noisy (and less sharp).

Shooting up to 1600 in RAW and using something like Nik Dfine works fine for me, and produces prints without any perceptible noise at A3 size.

My PC (Mac actually)has far more horsepower and better software for reducing noise, than relying on the less powerful software and hardware in my camera and the small amount of time is has to do noise reduction.

I just keep taking lots of pictures with whatever I have available, and will worry about the next model is when it comes out.

I'm glad you pointed that out timbo, because the Pentax route of higher resolution is the one I prefer.

I have always said that a sharp grainy image is preferable to a soft mushy one without grain/noise.

Anyone who ever put a 200 ISO colour film in their 35mm film camera and was happy with the results has nothing to complain about noise wise in any event.

I use the 100-400 range just as Benjikan does, usually 200 to get an extra stop of shutter speed.

I couldn't agree more and it will come, it has already been an issue for a very long time in movies/television and the latest generation of HD cameras from Arriflex promise 35mm film-like dynamic range (not quite the resolution, though, of 35mm film) and the promise of HDR. Interestingly the sensor gets so hot that the back of the camera is essentially a big chimney with a cooling fan but the noise (sound type!) is no more than an Arri film camera.

Getting back to the OP, mostly I shoot in the 100-400 range too, but I work a a lot under available light in theatres at 1600 and I dont have a problem with it at all. I've always liked grain and use Tri-X film at 1600ASA.

Grain can give a natural crispness to edges and increase apparent sharpness.

How many of us actually make very large prints on a regular basis where noise really is an issue?

Poor ISO performance is poor ISO performance.
I've used 400 ISO film and like the results I get for SOME pictures.
Film grain is not the same as electronic noise. Although I will concede that cases (rarer than for film) were it can work.
I still prefer having the detail than losing it everytime. The default NR in Canon and Nikon can loose too much detail depending on the model and settings. But Pentax is the same if you apply NR.
What I'm saying is to work to avoid noise altogether. You can put it in later. And btw, they should work on getting filters to reproduce grain not noise.
The fact that RAW is as good or better in Pentax, just shows what a good camera it is. The poor results in JPEG, just show how software is failing in Pentax. Because RAW is good, does not mean than that you can forego good jpeg.

My issues with noise are not from printing, its from looking at the screen (not pixel peeping), its not from expanding either. A tiny image can still look like garbage.

IMO...

I just disagree that the Pentax premium JPEG from the K20D is poor. It isn't.

When you look at A3 competition or exhibition prints from various marques frankly there isn't a problem with any of them, whether shot in JPEG or RAW.

Obviously less noise would always be nice, just like more pixels would be, but we are at a pretty good point right now.

I actually believe that it would be very sensible for Pentax to produce two variants of the same body, one aimed at low noise (K-x sensor?) and the other aimed at very high resolution (Ben's 19MP suggestion) similar to the Nikon D3/D3x concept.

A K-8 and K-8x maybe? I think that there would be sufficient demand for both to make it very viable both economically and from ease of production/tooling etc.

In fact, I could see many users opting to have both bodies to cover different aspects of their photography, especially with identical user controls and accessories.

It makes total sense to me, but maybe this is too logical a step?

...or do something similar to the Phase One backs.
Super high resolution for "normal" ISO range, and use clever "bucketing" of the photosites to increase the ISO & quality while sacrificing pixels.
(Though their method of using diagonally arranged G pixels is obviously patented)

Who knows...
Matt

They can use also other patterns to great effect with the higher pixel count sensors.
RGBW might be an interesting choice to increase sensitivity and dynamic range, at least if they work so good

Mat, I believe the K-M could lower pixel count by coupling them (by software) and lowering the noise.
I don't believe it was really effective though, you can better run a noise reduction at home and then half the pixel amount to get better result.

summer 2010? isn't it too early for the next K7? i'd love that as K7 price would drop within my budget range...

I'm thinking along those lines too.

The K-X is very well priced if you ask me - I have been tempted...but I really want a grip and weather sealing!

If the K-7 drops in price to the level I was able to buy my K20D at, I'd be a happy bunny.