K20D landscape photos affected by random hotpixels w/ MLU?

So now that the hotpixel issue is sort of fixed (it still randomly hotpixels so they can't map all of them out),

has this bugged any of you that do landscape photos w/ 2sec MLU? Setting MLU to 12s isn't that workable because I

sometimes use HDR and stuff moves enough between frames w/ 2sec MLU as it is.

 

And does the weird banding at high ISO show up in low light images of landscapes?

 

Thinking of upgrading to the K20D from K10D now that it's close to my $900 threshold and wondering if it's worth

it...

Only if your set with your lenses.....

Well, soon after I got my K20D I took an startrail of about 4.4 minutes exposure time @ ISO 1600, at a really

dark site (professional observatory) The noise has a band structure, but I would say that working on the RAW

should be sufficient to get rid of it. A 1600 pixel version of that picture you can see at <a

href="http://www.riviclaudia.de/gallery2/main.php/v/test/imgp0188.jpg.html">my webpage</a> This is the jpg that

came out of the cam software, not a reworked raw! It is not quite as much a quantum leap as I had hoped for in

low-light, but it is certainly better than my old camera, an *istDS. Already at <a

href="http://www.riviclaudia.de/gallery2/main.php/v/holidays/shorttrips/imgp0998.jpg.html">this</a> light level

(1sec at ISO3200) you don't see that strucure anymore.

Thanks, Thomas. That banding looks pretty bad so I guess I'm lucky I'm not into star trails; did the K10D have similar issues for you? The K20D like the other CMOS APS-C cameras seems to only buy you 1 stop in crappy light and 2 stops in good light.

And Javier, I'm all set on lenses except for a nice 70-200/2.8 but it looks like Sigma/Tamron both flubbed it so I have no idea what to do in that range. I thought about a Tamron 28-75/2.8 as well, but it seems like too limited a range. I have a Sigma 17-70 that I use most of the time, Tamron 70-300 that I'm happy enough with for the price/size (the 60-250 and 55-300 would be 2x as much and larger AFAIK), and a 50/1.4 for ultra low light. The K20D would buy an extra stop or two of speed for outdoor sports w/ the 70-300 and I'm hoping the extra pixels help for landscapes where some trees end up being 3 pixels..maybe they'll be 5 pixels w/ the K20D ;-)

Ken, before you arrive at theories about K20D you might try one.

 

IMO it's got around 2 stops advantage over K10D, ie 3200 looks like K10D 800.

 

If "crappy light" means indoors to you, K20D at 3200 looks deep into shadows and doesn't produce a lot of noise there.

 

As you're evidently limited by Sigma/Tamron zooms, you may not see much improvement in detail resolution with K20D.

I've uploaded two more images to my <a href="http://www.riviclaudia.de/gallery2/main.php/v/test/">test

folder</a>.I don't have the K10D, but an *istDS. The picture of the walkway was taken with 1/4sec@ISO3200. Again

this file is the jpg as produced by camera software. The K20D one with the "firewhirl" was cropped, I therefore

uploaded another from that series, a) uncropped, and b) with 1/4sec@ISO3200 just as the *istDS one. I'd say the

banding is indeed less, and at the light-level of the *istDS picture the K20D would not have produced any (or

rather: it would have been below the black-point).

PS: all were taken with the Pentax kit lens, i.e. the DA 18-55mm/f3.5-5.6, all at 18mm. f/3.5 for the *istDS, f/4 for the star-trail, and f/5.6 for the fire (to get fore- and background sharp). Except the startrail, they're all handheld.

John: are you saying you're hitting the theoretical 15mpix lens resolution limit for APS-C w/ the K20D?

 

Thanks for the samples Thomas. The 2nd iso3200 on the K20D is quite a bit better than the grainy *istDS one...

Ken, I don't know anything about theoretical lens limits, just that zooms offer lower detail than common prime lenses from the 70s and seemingly today's Pentax DA primes... higher resolution than K20D with primes won't matter much unless we make prints far larger than 13X18 (my own largest inkjet due to printer limit).

Zoom lenses definitely have less sharpness than primes (unless you're talking about Nikon's latest zooms which seem prime resolution from the reviews I've seen). I guess what I'm wondering is whether the K20D really out-resolves good zooms (not poor zooms). I know it definitely outresolves the old 18-55 kit lens which is why they needed a new one, but I'm not sure if it outresolves the Sigma 17-70 which is much sharper or even the Tamron 70-300 below 220mm (it's not that sharp after 220mm). I think my macro lens (Kiron 105) would benefit from the K20D but I don't use it as much as my other lenses (including the Sigma 10-20 UWA I forgot to mention :-)