Great K3 review
This was an interesting comment elsewhere on the site:
Comment link
Why is it that the 'independent' reviews of the K-3 almost all state its excellence at high ISO's (as this review also does), but a large proportion of opinion on here seems to think that it is 'crap' at high ISO's?
At reasonable ISO levels I don't notice mine being worse than my K-5iis. Up to 3200 the K-3 is fine and I certainly wouldn't want to use my K-5iis at any higher than this level. I did a shoot recently at a live music venue in near darkness and the K-3 produced excellent results, also as a bonus it is brilliant at focussing in low light, even better than a K-5iis in this regard.
Sorry to go off on a tangent!
Regards
David
Regards
David
Further to post above and to richandfleur's link, I'm not trying to claim the K-3 is anywhere near as good at high ISO as Sony A7, just observing that it's no worse than my K-5iis in my opinion.
Regards
David
You have experience with both, so you carry a lot of weight in these discussions.
I think the vibe was that the K-3 was no better that the K-5iis with regards to low noise/high ISO abilities, and that simply urked some who thought it would/should have been. The fact remains that the K3 has a newer processor and much bigger burst memory buffer etc. It's a different camera, with advances in some fields, but not in others.
Given the plateau of APS-C sensor advancement, I see either lower res offerings or a larger FF sensor type as the way forward in ISO ability terms. A certain A7s for example is in a league of it's own for the price. One can only hope that Pentax opts to emulate it's strengths in this ISO area in a subsequent product development/release. There are options to stay in the traditional DSLR formfactor, or to drop the mirror box, but regardless it has been shown that low light performance CAN be delivered.
To address your comments directly though, I think they're essentially the same. The noise appears to be different in nature, but fundamentally they're in the same ball park. A new approach will be required to deliver something with a significant advancement in this ISO specific area.
https://www.youtube.com/watch?v=1flm65f2Gy8
http://www.ephotozine.com/article/samsung-nx1-expert-review-26178
There maybe something else going on here but I can't for the life of me see it being over a stop better in high ISO than the K3, as stated in the CSTV vid, the sensor on the K3 is very efficient, some 50-60%, I have read recently, not even if Samsungs BSI sensor made that 100% would it only equate to about a stop, the review I posted in the K-S1 thread, the user reckons he preefers the output of that camera to the K3, things move on in a year, it's hard to keep up, if Samsung make a move sony will follow, whatever comes after the K3 will be better no doubt, whatever they decide to bung under it's bonnet, not too soon though Ricoh eh, I'm always skint! I'd be very interested in what a K60 might be able to do though, we're due a more video oriented Mid range DSLR sooner than a new K3 I think, and the 12 bit sensor in the K-S1 could be interesting from that point of view. The review sites want better video and Ricoh will have to provide it if only to stop them having yet another reason to diss the brand at every opportunity
P.S. Never shot a video, not even once, never entered my head to do so but apparantly I am in the minority...
My viewfinder is 576,000,000 pixels.
My other viewfinder is 5.76,000,000.
www.exaggeratedperspectives.com
Video wise people want real stabilisation back, and higher bit rates. Oh and the ability to visualise audio levels, if you actually use use on board audio.
The rest is actually there in my opinion.
To balance Jules and David's opinions. The K3 is worse at high ISO than the K5. And the K3 isn't capable of lifting shadows, like the K5 can.
I use both, and usually have them both in my bag side by side. And match the body to the conditions
Anything over 100 is high to me being almost purely a Landscape shooter these days!
Agree with the lifting of shadows though, you can get acceptable results from a K5 lifting almost five stops, I know because I've done it, takes lots of PP and some decent software (Used Topaz Denoise), problem is with the K5* that you sometimes need to do that as the meter can be so flakey! The K3's meter is far far more accurate and reliable, it's usually my fault if I get results that are sub par not the camera, it was not that way with the K5*
My viewfinder is 576,000,000 pixels.
My other viewfinder is 5.76,000,000.
www.exaggeratedperspectives.com
Not keen on the meter in my K3. I routinely use +0.7 EV
ISO's that I'm talking about are 1600 +. I can often see a difference at ISO 800.
This isn't meant as a downer on the K3. Its general the camera I reach for. Just not the one I reach for when the light goes, and I need to keep the shutter speed up. Nor the one I'd choose for astro photography
An open admission from me is that even when I used to shoot motorsport professionaly in all that time ISO 1600 was the highest I ever went, and then only once! I'd rather just not get the shot than have something a picture editor might like but I'd rather not have viewed by all and sundry, I recognise that I am in a very small minority there, possibly of one?
I'd be the same if I still had a K5*, I'd want to keep it, none of my K3 comments are meant to be disparaging about the K5* I just find the K3 better but would love a chance to own both, never tried Astro, something I'd like to get into someday, you've got good dark skies over there for that...
My viewfinder is 576,000,000 pixels.
My other viewfinder is 5.76,000,000.
www.exaggeratedperspectives.com
To balance Jules and David's opinions. The K3 is worse at high ISO than the K5. And the K3 isn't capable of lifting shadows, like the K5 can.
Both great cameras.
Best regards
Add Comment
To leave a comment - Log in to Pentax User or create a new account.


2342 posts
16 years
Storbritannia
http://www.itsjustlight.com/reviews/pentax-k3-dslr-camera-review/
My viewfinder is 576,000,000 pixels.
My other viewfinder is 5.76,000,000.
www.exaggeratedperspectives.com