[FozzFoster]

Hi all!

We made a video comparing the K-3 III and the K-70.

Results were pretty much expected - the K-3 III holds it's noise incredibly well.
And so does the K-70 really.

But the K-3 III is clearly better controlled especially at high ISOs. However, anything over 102400 (the top ISO for the K-70) the ISO on the K-3 III isn't really workable.

Still incredible results and we wanted to share them with you here

Curious thing, it looks like my K-70 is overexposing by about 1/3 stop. The exposures are pretty darn close so I hope the images are still good for testing and you find the comparison useful

Let me know what ya think!

Youtu.be

Youtu.be

[Martin Stu]

good test, interesting results

[FozzFoster]

good test, interesting results

thanks for the kind comments

[reh321]

Curious thing, it looks like my K-70 is overexposing by about 1/3 stop. The exposures are pretty darn close so I hope the images are still good for testing and you find the comparison useful

With my KP, the results seem better when I intentionally 'overexpose' at higher ISO values.
Perhaps the K-70 takes that into account.

[stevebrot]

But the K-3 III is clearly better controlled especially at high ISOs. However, anything over 102400 (the top ISO for the K-70) the ISO on the K-3 III isn't really workable.

This conclusion does not surprise me. Good work!

Curious thing, it looks like my K-70 is overexposing by about 1/3 stop.

That is why comparisons should be done in M mode with identical settings using the same lens. If there is still over/under exposure, something else is amiss, perhaps in PP.


Steve

[FozzFoster]

That is why comparisons should be done in M mode with identical settings using the same lens. If there is still over/under exposure, something else is amiss,

I'm wondering if my K-70 sensor is calibrated correctly?
The issues seems consistent and I am definitely using the same lens and settings
The only test I have left would be some depth of field work to see if the aperture is closing down the same on both cameras with the same lens.
After that, I'm going to guess a sensor calibration issue.
Thanks for the kind comments!

[stevebrot]

I'm wondering if my K-70 sensor is calibrated correctly?

If shooting RAW, the cause may be traceable to the camera profiles used by your RAW processor and/or "custom image" settings.

OK...I reviewed the video and it looks like the difference in brightness is obvious on the rear LCD, but that may just be due to the rear display settings. Moving on to Lightroom, the Library view LR may be using either the preview JPEG from the DNG or a LR-generated preview depending on configuration and whether edits have been done. If LR-generated, the camera profile embedded in the DNG will be used for the K-3iii while the KP will use Adobe Standard by default.

Lens corrections are turned off, right?

A blank white wall histogram evaluation might be your next step, though I don't think I would be too concerned about a 1/3 stop difference in sensor response between two very different image processors. (Remembering that live view metering is based on image processor data.)


Steve

---

Last edited by stevebrot; 05-14-2021 at 01:31 PM.

[FozzFoster]

I reviewed the video and it looks like the difference in brightness is obvious on the rear LCD, but that may just be due to the rear display settings.

In the footage, on the back LCD, I'm in manual mode and you can see the light-metered exposure stops at the bottom of the LCD screen. It's not about the brightness of the screen itself, but seeing how the light meter is reading the scene. You can see the K-70, at the same settings as the K-3 III, is 1/3 stop over exposing.

In Lightroom, there is a JPEG preview, but once zoomed 100% the .dng loads and you can still see the K-70 is a touch brighter.

I can triple check my settings... but I am sure my results are consistent.

---------- Post added 05-14-21 at 04:30 PM ----------

Also totally agree that a third stop isn’t a big deal and can be expected between cameras with different image processors

[stevebrot]

You can see the K-70, at the same settings as the K-3 III, is 1/3 stop over exposing.

I was doing some more thinking about this and was reminded that ISO on Pentax cameras is referenced to sensor response with sOOC JPEG.* As such, I suspect that M-mode comparisons using identical settings between different model Pentax camera will likely show some variance based on ISO definition for each. I suspect that the ISO calibration of the K-70 relative to the K-3iii is about 1/3 stop low.


Steve

* Standard output sensitivity (SOS) is one of three methods allowed by CIPA standards. Their methodology uses a standardized setup to evaluate OOC JPEG to match an 18% gray standard.

[FozzFoster]

and was reminded that ISO on Pentax cameras is referenced to sensor response with sOOC JPEG.

* Standard output sensitivity (SOS) is one of three methods allowed by CIPA standards. Their methodology uses a standardized setup to evaluate OOC JPEG to match an 18% gray standard.

makes sense! good explanation for variance among sensors. Seems a bit silly to call it a 'standard' when you allow three different methods to calibrate haha

[luftfluss]

Good work!

Having both male and female narration is a very good idea - adds a bit of oomph to a rather dry subject.

[stevebrot]

makes sense! good explanation for variance among sensors. Seems a bit silly to call it a 'standard' when you allow three different methods to calibrate haha

I was wrong...CIPA only allow two of the elements of the ISO12232 standard:

Recommended Exposure Index (REI): Manufacturers allowed to define ISO as what produces best exposure at the given setting.

Standard Output Setting (SOS): Using a standardized light source, the numeric rendering must equate to 18% gray (RGB 118, 118, 118 on average across the frame).

It should be apparent at some point that ISO on digital cameras is less about sensitivity and mostly about metering to respect traditional notions regarding EV appropriate for a given subject lighting (e.g. Sunny 16 rule) as well as existing photographic lighting conventions and meters. Metering is the tail that wags the dog.


Steve