[Saltwater Images]

The high ISO performance is impressive. I invested in the K-5II about 6 months after the K-3 launched opting for larger photo sites on the CMOS for a slight edge in dynamic range and sacrificing some resolution. It wasn't a difficult choice because I was upgrading from a K-100D Super:-)

The KP looks appealing, the control layout superb, I don't care that the top deck is poly-carbonate, one SD Card slot is no big deal either. It looks like this new camera body was designed specifically with my HD DA pancakes in mind! The only things holding me back are; buffer size and shutter life. The current flagship K-3II has a 200,000 actuation shutter life. My aging K-5II has a larger buffer for RAW.

Got to hand it to Ricoh; if you check out some of the other brand forums, the KP is causing a stir in the Sony, Olympus and Fuji camps. Ricoh also positioned the KP where flagship APS-C used to be priced. This means that the next flagship will cost more. Brilliant move to increase the value of products in the line without making those entrenched in the system feel ripped off.

Will I pull the trigger on a KP? Hard to say. Most likely I'll wait over the next several months to see if a new APS-C flagship is released and consider the cost/performance benefit first. How about you, will you be an early adopter of the KP?

---

Last edited by Saltwater Images; 01-28-2017 at 09:29 AM. Reason: added text

[derekkite]

This is mostly dependent of the way you approach photography. I got fantastic wildlife picture in Tanzania with K3 and 55-300 and most of the time I was at iso 400-1600. It got out perfectly, you don't see any issue even on a 4K TV or a print.

The cityscape photo shown at the first post could have looked much better with a tripod even taken with a compact camera or a smartphone. Visibly better. It was taken at f/8 too just to show how the high iso. Nobody prevent you to use f/5.6 or maybe even the f/3.5 of the kit lens in low light if you were shooting handled.

Often people approach the thing the hard way: they go for increadbily bad situation and try to get not too terrible photos out of it and are amazed when the new gear make it look a bit less terrible. That's a way to approach it but another way could be simply to put yourself in good conditions to get the shot so you get a great shot. Not just a not too terrible shot. You could have taken the picture with better lighting a down or dusk just at the right moment and it would have looked much better...

What count is the end result. While I agree fully that the incremental improvement we get are quite nice, you need several generation for it to become a game changer. Even K5 against KP, the difference would be mostly JPEG but K20 against KP would be very visbible.

If we get 4-5 time the low light improvement from K3 to KP, then we get a game changer. We should welcome every improvement, for sure, but that doesn't require to change camera more often than every 5-10 years to get the most out of technology. People were all around K70, now KP. Both seems to have basically the same performance for high iso even if of course K1 is far ahead. Don't hold your breath all future APSC camera are going to get at least that or better and price will drop quite soon. This is technology as usual.

You should come where I live and try this. That lens is good at f8. I don't use my K3 because of the 1600 iso top limit for a shot that can be cropped. I was shooting this week with the K1 at 1/1600 f5.6 and iso floating between 800 and 6400 depending on the light. My preoccupation was biasing the metering for the haze.

If this aps-c can give me k3 1600 quality at 3200 or higher, then it is a great improvement.

[stevebrot]

Japanese manufactures are required to use one of two sensitivity specifications: REI or SOS, both of those specifications are incorporated into iso12232:2006.

True

REI and SOS are both about the brightness of the jepgs produced by the camera

Sort of true. The full answer is more complicated. That is why I provided the link to the Wikipedia article, which in turn provides the appropriate references. Pentax uses SOS which is based on the sensor response relative to the metered EV traceable to an 18% gray standard. The REI alternative allows the manufacturer to arbitrarily designate the ISO based on color-specific response at various sensitivities. The important distinction is that REI is traceable to the manufacturers definition of acceptable response rather than an outside standard. FWIW, REI is analogous to "Exposure Index" (EI) as commonly used by film photographers when wishing to expose at other than box speed when using specialty developers.*

The CIPA DC-004 specification (PDF) may be downloaded from: http://www.cipa.jp/std/documents/e/DC-004_EN.pdf **

The "Overview" section is particularly helpful in that it plainly states that both are indices of "practical sensitivity".

This misconception is cause by Dxo’s use of the term measured iso.

DXO is evil...


Steve

* The ISO film speed on the box is determined by densitometry studies plotting density (response) against exposure to produce the film's "characteristic" curve when processed according to manufacturers recommendation. The relation of ISO film speed to digital ISO sensitivity is fairly arbitrary since sensor response is highly mutable unless one is working from a so-called linear RAW conversion to TIFF where the actual response of an exposure wedge may be directly evaluated. Direct analogy to the film standard is difficult, though the SOS option is close.

** The link in the Wikipedia is out of date. If I get a chance, I will do the appropriate edit. I would have provided the link to the ISO draft spec as well except that the CIPA version is explicit in regards to our Japanese cameras.

[stevebrot]

I do wonder how much of a difference between the KP and the K3 II will really be visible when you use RAW images. Half a stop, maybe?

My question as well.


Steve

[richandfleur]

dang! . if they could wait a little and put this sensor tech in FF for K1, then likely we have a very usable iso 204800 in K1, an instant A7s killer.

Remember the A7s can shoot video well...

[stevebrot]

Remember the A7s can shoot video well...

Dang! You would have to bring up that camera's special purpose!!!


Steve

[interested_observer]

Very good point. I use very high iso to calculate exposure time when exposure time is beyond 30 sec. with automatic exposure (out of range). Using very high iso in LV mode to find correct exposure, then drop the ISO value and increase exposure of the same amount of stops.

In shooting landscape astro, there is never sufficient light for a good exposure - you just take what you get and be happy. At wide focal lengths, with earth rotation you are at 10 to 13 seconds of exposure (rule of 200) and then you start trailing stars. With the GPS/AstroTrack - it's advertised at up to 5 minutes, however in reality you are at 60 seconds max**[1]. And at 60 seconds with the astrotrack, you are blurring the landscape, so you are going to be compositing (star and landscape elements from multiple images).

With some higher ISO's you should be able to collect light a bit better, with out GPS/AstroTracking with a single frame (especially with a cropped sensor), and that will (at least for me) make astro landscape panos much easier and nicer - with better light (both from the stars, and collected off the landscape elements).

Also there is a thread over in the AstroPhotography area - that discusses that ISO 1600 is somewhat the knee - up to 1600 and you are collecting photons, above 1600 and you are just amplifying what you collected (with the K5 class of Sony sensor). So, there is a fine line. There will be another threshold with the K70 and KP sensors that will need to be determined. I'm guessing that it would probably be higher - 3200, 3200++? I don't know.

• How to avoid wasting photons with too-high ISOs (?) - Astrophotography - PentaxForums.com

**[1] The GPS/AstroTracker as good as it is, still (in my opinion) has a problem with wide angle lenses. When you take a lens and put it on a physical tracker, you get perfect star points (when calibrated and aligned), because the entire camera, lens and sensor are all moved in unison. With the GPS/AstroTracker, the stars are moving, but the camera and lens are not moving, but the sensor is. So, the light from the star is taking a different path through the lens, which causes some odd trailing around the extreme edges and corners of the frame. This coupled with less than perfect information from the GPS/calibration (compass/pointing error, elevation error, etc.), all combine to provide this trailing around the extreme edges. Pentax has been silent on this, and I doubt that we will ever hear anything about it.

At higher ISO, I'd say at least a stop - the K-3's sensor was a step back from the K-5's, but the KP (apparently) is a step forward. The K-70 has about a stop over the K-3II.

As good as my K5IIs is, it does look like there is an emerging advantage with the new bodies. I am really wanting to see what the real K3II replacement brings us. Just a better packaged KP sensor, or the next new sensor beyond the KP with better supporting in camera image processing. Unfortunately, we really will not know until about 6 months after the K3II replacement hits the streets and folks start to get some hard data from all of these new bodies for comparisons.

[Ian Stuart Forsyth]

Sort of true. The full answer is more complicated. That is why I provided the link to the Wikipedia article, which in turn provides the appropriate references. Pentax uses SOS which is based on the sensor response relative to the metered EV traceable to an 18% gray standard

Pentax uses SOS which is based on the brightness of the jpg nothing getting around this If as you say that pentax uses SOS that is based on the sensors response relative to the metered EV traceable to 18% gray, the raw file would not pass the SOS standard. This can be seen very easily as any raw viewer and will show that 18% is under exposed (by 1- 1.2 stops far greater than is allow under SOS standaed) and then brightened to meet the SOS standards for sRGB in the jpg image. The SOS standard only applies to the outgoing jpg and not to outgoing raw files.

"The Standard Output Sensitivity (SOS) technique, also new in the 2006 version of the standard, effectively specifies that the average level in the sRGB image must be 18% gray plus or minus 1/3 stop when the exposure is controlled by an automatic exposure control system calibrated per ISO 2721 and set to the EI with no exposure compensation. Because the output level is measured in the sRGB output from the camera, it is only applicable to sRGB images—typically JPEG—and not to output files in raw image format. It is not applicable when multi-zone metering is used."

---------- Post added 01-28-2017 at 07:38 PM ----------

DXO is evil...

DXO is not evil they need to have a way of measuring a sensors performance that ignores the cameras manufactures predetermined highlight headroom and to do this they had to created a term "measured iso" that has nothing to do with iso but rather a saturation exposure expressed as an iso.

[biz-engineer]

Pentax uses SOS which is based on the brightness of the jpg nothing getting around this If as you say that pentax uses SOS that is based on the sensors response relative to the metered EV traceable to 18% gray, the raw file would not pass the SOS standard. This can be seen very easily as any raw viewer and will show that 18% is under exposed (by 1- 1.2 stops far greater than is allow under SOS standaed) and then brightened to meet the SOS standards for sRGB in the jpg image. The SOS standard only applies to the outgoing jpg and not to outgoing raw files. "The Standard Output Sensitivity (SOS) technique, also new in the 2006 version of the standard, effectively specifies that the average level in the sRGB image must be 18% gray plus or minus 1/3 stop when the exposure is controlled by an automatic exposure control system calibrated per ISO 2721 and set to the EI with no exposure compensation. Because the output level is measured in the sRGB output from the camera, it is only applicable to sRGB images—typically JPEG—and not to output files in raw image format. It is not applicable when multi-zone metering is used." ---------- Post added 01-28-2017 at 07:38 PM ---------- QuoteOriginally posted by stevebrot Quote DXO is evil... DXO is not evil they need to have a way of measuring a sensors performance that ignores the cameras manufactures predetermined highlight headroom and to do this they had to created a term "measured iso" that has nothing to do with iso but rather a saturation exposure expressed as an iso.

No sure I've followed it all. What's the reference level for measuring ISO? If the output should be 18% gray, what should be the luminescence level of the light source for ISO100 ? Why use saturation level instead of 18% gray?

[LightBug]

Just to make sure, I ran the available High-res samples through NeatImage8, set to highest quality and I find ISO12800 delivers top results, ISO25600 is still very usable (at least on on one of the examples with a night scene) with ISO102400 no longer usable in any meaningful way. That is absolutely very impressive! That tells me the structure of the noise lends itself very well to analysis and removal up to a certain point. I'm not in the market right now for a new body but if I were, I would give the K-P some serious consideration.

Literally the only downside to my kind of shooting is the limited battery-life but that can be resolved with a couple of spare batteries in the pocket. Still puzzled about the initial price level the K-P was inserted - it seems to sit in the slot previous top APS-C bodies occupied. What that means for the K-3 II replacement (I saw someone mention 6 months away) I can only guess.

I'll post some crops of those files below:

ISO12800 (left before NI8, right post nr)


ISO25600


EDIT: ISO51200 added

Thanks for posting the samples! The noise does seem to clean up very well!

[biz-engineer]

I got fantastic wildlife picture in Tanzania with K3 and 55-300 and most of the time I was at iso 400-1600.

That's one of the situations of the sunny f16 rule. In short, the f16 rule tells you that for the most common still shooting situations, you don't need more than ISO1600, even with a Pentax long lens (f5.6 / f8). All current apsc cameras can cover this range. You need at least one full stop ISO improvement when shooting moving subjects with >200mm long lenses, that where full frame helps a lot.

[f22]

I think it's important to distinguish between daytime shots and night shots, like in the examples which were taken at night. Night shots, or poorly lit situations, is where this bit of technology is really going to improve results.

I think the KP is going to equal or perform better than some of the medium format camera's when it comes to resolving detail and contrast on areas which are dark to begin with.
The sample shots are very impressive, and like most cameras, the highest 2 to 3 ISO setting usually render unusable levels of noise that even PP has a hard time correcting.
I think this high ISO in an APC-S format is truly a practical advancement that will result in more keeper shots taken in less than ideal lighting conditions, i.e. indoor concerts, museums, indoor venues, etc.). It also allows you to use faster shutter speeds, and less chance that blurring will happen if you aren't using a tripod or support.

And I agree with pixelhdr, if this boost Dynamic Range at the same time, that is a great benefit too.

The depth of field and shutter speed bracketing sounds extremely interesting too, for portrait, macro, etc. where you want to quickly take a few versions at different f stops without e-dialing it.

[MrB1]

I think this high ISO in an APC-S format is truly a practical advancement that will result in more keeper shots taken in less than ideal lighting conditions, i.e. indoor concerts, museums, indoor venues, etc.). It also allows you to use faster shutter speeds, and less chance that blurring will happen if you aren't using a tripod or support.

I agree. The K-70 is already very good at higher ISOs and the KP looks as though it should be even better. The SR on the K-70 is also very good and looks to have been improved on the KP. Taking both effects together, it's looking great for hand-held low-light shooting.

Philip

[f22]

I agree. The K-70 is already very good at higher ISOs and the KP looks as though it should be even better. The SR on the K-70 is also very good and looks to have been improved on the KP. Taking both effects together, it's looking great for hand-held low-light shooting.

Philip

I agree with you and I think the edge would go to the KP however. If you can increase the ISO on the KP, you could get away with using a faster shutter speed on some shots, and still get a less noisy image. That DoF and shutter speed bracketing are interesting as well. I like that there is an onboard flash, and not thrilled they opted for a battery with less shot capacity, just to make the form factor smaller. The AF is improved too (more focal points, etc.) and the AF is now -3 to 18 EV, and the electronic shutter that is 1/24,000, that is another creative extension.

[Eric Auer]

I wonder how well it will AF in low light?

Eric