[BigMackCam]

What "considerably better"? K3 is already great. K1 is considerably better than K3, but still 1.5-2EV. K5 a bit better than K3. KP still APSC and likely based on Alpha 6300 or equivalent that is not much better than other APSC. This isn't CMOS vs CCD change man.
...
Yes a tiny bit more but likely less than if you invest in a lens that one stop faster, a great RAW processing software, a bigger sensor or a flash. I don't see it as KP would make a photo that K3 or K5 would break if you have a good workflow. Already people don't all agree on APSC vs FF that FF is really necessary, but there a bigger difference.
...
Good enoug is relative, but I guess 1600 iso on KP go to look better than 12800 on KP and 12800 on KP isn't going to look that great most of the time... So you still want to avoid 12800 isos when you can... Like you'd want to shoot at low iso when you can anway. What I know is that if I have to show high iso often, I'd buy an A7s or a K1 and ensure I have f/2.8 zooms and f/1.4 primes. I'd ensure I have the best raw processing software too and I'll invest into lighting equipment. I am not going to just buy a bit better APSC body.

Nicolas, you and I think a little differently - that's all

Yes, the K-3 is great. It's my main go-to camera, and regardless of these discussions, I'll more-than-likely stick with my K-3 and K-3II for some considerable time. However...

If the KP provides more detail and less noise than the K-3 at the same ISO, that should translate to better IQ through the need for less aggressive post-processing in Lightroom or similar. Let's not forget that with flash photography away from the studio, we may often need to shoot at higher ISOs in order to capture ambient light (your flash shots look fine, but would have benefitted greatly from a better balance with ambient light - for which you'd have to push up the ISO). If the KP produces equivalent detail / noise 1.0 - 1.5 stops higher than the K-3, that increases the number of situations where acceptable results can be obtained, and - more importantly - will mean a tangible improvement to every high ISO photograph. That difference may not be compelling to you... It is for me.

As for the high-res full-frame vs APS-C argument... I have full-frame and APS-C cameras, and with Pentax I choose to shoot APS-C. The relatively compact and lighter-weight form factor of the K-3 is pretty much spot-on for me. I'm also very happy with the K-mount lenses I already own and use, and a number of those are either designed for APS-C format or perform much better on APS-C because of border performance. I don't want to buy a K-1 (not at this point, at least), and despite the fact that I also own some big, heavy, fast zooms and a few fast primes in Sony A-mount, I don't want to buy an A7S either. I'm interested in how much better an APS-C sensor Pentax camera can be at higher ISOs.

Finally, we agree on something - "good enough" is relative. But whatever the level of IQ you consider to be "good enough", it is likely to be 1.0 - 1.5 stops of sensitivity higher for the KP than the K-3. You may not find that appealing, but I do

---

Last edited by BigMackCam; 02-05-2017 at 03:57 AM.

[Simen1]

Finally, we agree on something - "good enough" is relative. But whatever the level of IQ you consider to be "good enough", it is likely to be 1.0 - 1.5 stops of sensitivity higher for the KP than the K-3. You may not find that appealing, but I do

Well, it should be mentioned that its not 1-1,5 stops better in general. Thats only valid in jpeg format and only at ISO 100 000+. A ISO latitude I find rather useless, and in a file format I don't shoot.

At better quality ISOs and in raw I have not seen any good comparisons yet. I have a gut feeling based on physics that the difference will be much lower at lower ISOs.

[mecrox]

Nicolas, you and I think a little differently - that's all

Yes, the K-3 is great. It's my main go-to camera, and regardless of these discussions, I'll more-than-likely stick with my K-3 and K-3II for some considerable time. However...

If the KP provides more detail and less noise than the K-3 at the same ISO, that should translate to better IQ through the need for less aggressive post-processing in Lightroom or similar. Let's not forget that with flash photography away from the studio, we may often need to shoot at higher ISOs in order to capture ambient light (your flash shots look fine, but would have benefitted greatly from a better balance with ambient light - for which you'd have to push up the ISO). If the KP produces equivalent detail / noise 1.0 - 1.5 stops higher than the K-3, that increases the number of situations where acceptable results can be obtained, and - more importantly - will mean a tangible improvement to every high ISO photograph. That difference may not be compelling to you... It is for me.

As for the high-res full-frame vs APS-C argument... I have full-frame and APS-C cameras, and with Pentax I choose to shoot APS-C. The relatively compact and lighter-weight form factor of the K-3 is pretty much spot-on for me. I'm also very happy with the K-mount lenses I already own and use, and a number of those are either designed for APS-C format or perform much better on APS-C because of border performance. I don't want to buy a K-1 (not at this point, at least), and despite the fact that I also own some big, heavy, fast zooms and a few fast primes in Sony A-mount, I don't want to buy an A7S either. I'm interested in how much better an APS-C sensor Pentax camera can be at higher ISOs.

Finally, we agree on something - "good enough" is relative. But whatever the level of IQ you consider to be "good enough", it is likely to be 1.0 - 1.5 stops of sensitivity higher for the KP than the K-3. You may not find that appealing, but I do

Interesting. I've more or less stopped bothering to correct for noise below at or about ISO 800 with my gear. It can look dodgy on screen if pixel peeping but is simply not noticeable on prints, and noise treated as grain may often actually enhance the atmosphere on BW. I get more from applying extra microcontrast and colour (to compensate for shrinking dynamic range at higher ISOs) than from sharpening per se or noise reduction, and if I do want to reduce noise, a pass though DxO's amazing Prime engine will suffice. I will deal with noise if I am pushing things really hard, like raising shadows 80 per cent or something, but that isn't needed with most images.

[D1N0]

Some High ISO samples from Ricoh's facebook page:

[BigMackCam]

Well, it should be mentioned that its not 1-1,5 stops better in general. Thats only valid in jpeg format and only at ISO 100 000+.

I believe it will be at least a stop better in RAW (and that's the only format I'm interested in, frankly).

Attached below are comparisons of the K-3 and K-70 in DPR's image quality comparison tool. Each one shows the K-3 at ISO 6400, with K-70 at ISO 6400, 12800 and 25600 respectively. The K-70's colour noise manifests itself quite differently - almost as if the colour aspect (but not the luminosity) is being removed. That aside, with both cameras at ISO 6400, the K-70's output wins hands down. With the K-70 at ISO 12800 I'd say it's as good as the K-3's ISO 6400 output (although that may be somewhat subjective - colour noise is better on the K-70 image, luminance noise is a little better on the K-3). By ISO 25600, the K-70's output is considerably worse than the K-3 at ISO 6400.

I realise this doesn't take into account dynamic range and fine detail reproduction, but in terms of noise performance alone, the K-70 appears to offer a one stop advantage over the K-3 in RAW shooting. I'm hoping that the KP will, at the very least, match that advantage - and perhaps do a fraction better...

[ogl]

It was said about RAWs not JPEGs, I'm sure.

new noise reduction in RAW ?

[frogfoot]

new noise reduction in RAW ?

The same as in K-70, or tweaked a little bit i think.

[monochrome]

If this is what we get on the mid-range (is K-70 now the entry level and K-1 the Flagship?) what will we get on an APSc Flagship? Just room on a larger body for the full complement of external controls and connectors, or is there room for more processing HP on a larger motherboard?

[photoptimist]

Overall, I often see this trade-off in taking a photograph in dim lighting:

1. Slow the shutter by 1 stop and double the motion blur, or
2. Open up the aperture by 1 stop and lose 30% of the in-focus scene to DoF blur, or
3. Increase the ISO by one stop and incur more noise and less DR

Option #1 usually sucks unless there's some advantage of letting the moving parts of the scene blur more as a dynamic contrast against the stationary parts. Option #2 is tricky one because some images look better with less in focus, but a lot images don't. Option #3 depends on the current ISO and my perceptions of whether the added noise and lost DR will be "unacceptable". Given the effects of these three controls on the image, a camera that lets me go to a slower shutter speed does always not help, a lens that lets me open to brighter aperture does not always help, but a camera that lets me pick a larger ISO without suffering "unacceptable" loss of IQ is a boon!

Put it another way: if the issue of image noise were eliminated by some magical technology, what shutter speed and aperture would one choose in any given situation? And if Pentax (& their sensor maker) provide another stop or so of ISO for a given "acceptable" noise level, doesn't that increase the percentage of photographs that can be shot at optimum aperture and speed?


(PS: I see Pentax SR as a tool that provides long shutter times without added motion blur for scenes such as architecture and landscape that have only stationary elements. But there's a lot of scenes involving animals, people, machinery, and wind-blown vegetation where SR or a tripod does not help.)

[MrB1]

There is another factor to add into the list, as I mentioned in another post, and that is image stabilisation (SR). This is very effective on the K-70 so it should be expected to improve with the KP's 5-axis SR technology, and could influence what is regarded as the "optimal" shutter speed for many types of scenes. This factor, together with better ISO performance (in JPEG and Raw), should make the KP a very useful tool for hand-held low-light shooting.

Philip

[Nicolas06]

(PS: I see Pentax SR as a tool that provides long shutter times without added motion blur for scenes such as architecture and landscape that have only stationary elements. But there's a lot of scenes involving animals, people, machinery, and wind-blown vegetation where SR or a tripod does not help.)

SR isn't meant to fix motion blur but hand shake. If you shoot with a 300mm, you may only require 1/100s instead of 1/500 and if 1/200 is sufficiant for your subject movement, that still 1.5 stop more to play with for iso or apperture. That very usefull tool for widlife photography !

[Nicolas06]

I believe it will be at least a stop better in RAW (and that's the only format I'm interested in, frankly).

Attached below are comparisons of the K-3 and K-70 in DPR's image quality comparison tool. Each one shows the K-3 at ISO 6400, with K-70 at ISO 6400, 12800 and 25600 respectively. The K-70's colour noise manifests itself quite differently - almost as if the colour aspect (but not the luminosity) is being removed. That aside, with both cameras at ISO 6400, the K-70's output wins hands down. With the K-70 at ISO 12800 I'd say it's as good as the K-3's ISO 6400 output (although that may be somewhat subjective - colour noise is better on the K-70 image, luminance noise is a little better on the K-3). By ISO 25600, the K-70's output is considerably worse than the K-3 at ISO 6400.

I realise this doesn't take into account dynamic range and fine detail reproduction, but in terms of noise performance alone, the K-70 appears to offer a one stop advantage over the K-3 in RAW shooting. I'm hoping that the KP will, at the very least, match that advantage - and perhaps do a fraction better...

Essentially the luminance noise is the same at the same isos, but the chrominance noise is more visible on K3 on that processing software. On DxO, both would have not an hint of chrominance noise, even at 25K. But both have the similar loss of sharpness due to visible grain at the same iso setting.

And insterrestingly, the same comparison tool still show K1 as much better than K70 at same iso, not only less noisy but with more details thanks to its 36MP...

K70 is better than K3 for noise handling, but not the silver bullet some here think it is.

[BigMackCam]

Essentially the luminance noise is the same at the same isos, but the chrominance noise is more visible on K3 on that processing software.

Those images aren't using any post processing software, Nicolas. They are the RAW samples. I'd encourage you to download them and play with them in Lightroom or DXO (I've done so in Lightroom, so I'm pretty informed on the levels of luminance and colour noise in both the K-3 and K-70 files). Lightroom also gets rid of the colour noise very easily.

And insterrestingly, the same comparison tool still show K1 as much better than K70 at same iso, not only less noisy but with more details thanks to its 36MP...

Yes, if you downsample the K-1 image to match the dimensions of the K-70 image. With both at full size, they're remarkably similar in the noise department. But that's not really relevant, since we're talking about APS-C here (specifically the KP). it's good to see just how well the K-70 fares at high ISO against pretty much any other APS-C camera. It's actually remarkably close to the D500 at ISO 12800, 25600 etc. And the K-70 already beats the latest Sony APS-C offering (the 6500).

I remain hopeful that the KP will be at least as good, if not slightly better than the K-70, for high ISO imaging.

Oh, and there is never a silver bullet. Some bullets, though, are more effective than others

[mecrox]

If this is what we get on the mid-range (is K-70 now the entry level and K-1 the Flagship?) what will we get on an APSc Flagship? Just room on a larger body for the full complement of external controls and connectors, or is there room for more processing HP on a larger motherboard?

A lot of the points emerging on this thread do beg the question of what exactly an APS-C flagship would have over the KP which would make it worth buying - or manufacturing. Not much all that much, it might appear. $1500 for a bigger battery and a slightly better buffer? That's quite expensive I'm sure another Pentax camera will appear before long but it may not be the camera people are expecting just as the KP isn't really what folks were expecting.

[BigMackCam]

A lot of the points emerging on this thread do beg the question of what exactly an APS-C flagship would have over the KP which would make it worth buying - or manufacturing. Not much all that much, it might appear. $1500 for a bigger battery and a slightly better buffer?

As a minimum - bigger battery, bigger buffer, faster continuous shooting, dual SD card slots, IR remote ports, top panel LCD and HDMI output (that would bring it back in line with the K-3 / K-3II). Perhaps, additionally, the K-1 style articulating screen, SAFOX 12 AF (found in the K-1) and some other improvements and/or goodies...