[rawr]

I'm supposing the metering is the same, or similar, between the two as I can't find any mention on Ricoh.com of any "fantastic new metering system" for the K3 over the K5II/IIs.

The K-3 metering chip is a 86,000 pixel RGB sensor; the K-5/K-5II metering chip meters a scene at a maximum of just 77 segments ... So the metering hardware is physically very different between the cameras, as is the metering software (and features).

If you ever want to see the difference, have a look at a K-5 and a K-3 image in EXIFTOOL, and you'll see data for thousands of metering points in the K-3 file , compared to a relatively tiny number of metering points for the K-5.

---------- Post added 2016-03-24 at 09:18 AM ----------

Both cameras were set to centre spot metering.

Unless it was just for testing, one probably should not routinely use that mode. It only meters about 5% of a scene, from the dead centre, and is best used for special purposes. It also basically switches off the 'intelligence' of the K-5 or K-3 metering system, and gives the camera metering system very little information to work with in terms of scene lighting or colours.

[stevebrot]

I can't find any mention on Ricoh.com of any "fantastic new metering system" for the K3 over the K5II/IIs.

It is a little hard to find on the current Ricoh site, but the RGB meter sensor was a stand-out feature when the K-3 debuted in 2013.

Feature 5? PENTAX K-3 II | RICOH IMAGING (about half-way down the page)

As mentioned above, comparison between cameras is a little difficult unless the target subject is highly uniform (e.g. gray card). I would suggest filling the frame with a uniformly-lit white wall in indirect light using center-weighted mode at sufficient distance to not cast a shadow on the surface. For such a setup, both cameras should return a fairly gray rendering with a strongly centered histogram.

As for attempts to dissect the P mode behavior, here are a couple of things to chew on:

• The exposure system uses the meter information in a dumb way despite the meter system being able to use very sophisticated means to derive an EV
• The program line is two-dimensional only (shutter speed/aperture). Auto-ISO (if enabled) is not directly driven by the program line. The opposite is closer to the truth.
• The program lines are the same regardless of the current ISO setting such that the actual EV(100) applied is the same regardless of ISO
• The Auto-ISO record mode menu settings (Slow, Normal, Fast for the K-3) should not be construed to be equivalent of the various program line options. To paraphrase the user guide, "Slow" means that the camera will take a conservative approach to bumping the ISO. "Standard" (default) means moderate sensitivity to increase the ISO. "Fast" means that the camera will aggressively bump the ISO in response to low measurements from the meter. To be explicit, having the setting on "Fast" may or may not result in a higher shutter speed for a given light level. It depends on the current program line, metered light level, and initial ISO setting.
• The Auto-ISO settings (Slow, Standard, Fast) do NOT apply to TAv mode.
• Attempts to predict or tailor programmed exposure behavior using Auto-ISO will likely lead to frustration and IMHO are not worth the effort*

Steve

* I commonly use the phrase "Auto-ISO is EVIL" to describe my feelings regarding this feature. It is useful and I frequently use it myself, but NEVER when I desire fine control over the exposure triangle or sensor noise potential. Noobs are frequently confused by the feature and advanced users are often frustrated when they attempt to leverage it to a particular purpose.