[Kunzite]

I don't expect gushing praise from them - I'd just like them to understand what Pentax is doing and see the part of our world where that is useful.

Nobody does. They're just continuing to mock us, because we're few and they can afford to Other us.

[Class A]

Wouldn't matter if he had an f/0.5 lens, photography is the recording of light. Very low light means little to record - unless there is a clean ISO 204,800 option.

Yes, actually having light to record is important.

There is a common misconception that the noise one sees on high ISO images is caused by the sensor. Typically, the noise one sees is owed to the stochastic nature of light. The more one reduces light levels, the more it becomes obvious that every image is formed from discrete photons, as opposed to a weak but otherwise continuous signal. Raising ISO levels is usually done to still cope with very low light levels or to compensate the artificial reduction of light levels through small apertures and/or high shutter speeds. Either way, light levels are low and one sees the noise of light itself, not sensor noise.

[Class A]

The 'accelerator' also expands Dynamic Range {and probably Color Depth} beyond what you would normally be limited to at higher ISO values.

The "accelerator" does not expand the scene-referenced dynamic range and the latter is the only one worth talking about.
Otherwise, denoising an image in post-processing could also be described as expanding dynamic range.

The same holds true for "colour depth". The "accelerator" almost certainly does not have access to data that isn't available to any image processing software that you can use for post-processing.

I keep asking about the preferences of Japanese photographers and keep not getting a clear answer.

Well I, for one, don't know their preferences.

From Pentax's actions, they seem to be oriented towards those who don't want to spend much, or any, time in PP.

That's fine, but that shouldn't impinge one those who want to keep all their options open. If one could opt-out, it would be very much appreciated.

[Kunzite]

The "accelerator" almost certainly does not have access to data that isn't available to any image processing software that you can use for post-processing.

Hmm... we don't know that.

In the Imaging Resource interview, Mr. Takashi Arai talks about a kind of processing which cannot be reproduced when post processing the RAW files:
"Takashi Arai/Ricoh:
The accelerator unit initially processes the output signal from the sensor, meaning that the accelerator comes right after the image sensor. And then it conveys it to the PRIME IV -- PRIME IV is the name of our image processing engine -- and then an accessory unit does a kind of signal processing which cannot be obtained by just software processing mechanism without degrading the resolving performance of the sensor."
Ricoh @ CP+: Rewarding K-1 fans with a major upgrade, plus what comes next after the K-3 II?

Not exactly the accelerator, but where there's one, there can be more.

The accelerator isn't even the only component performing noise processing.

By the way, here's a good, detailed explanation why the accelerator isn't optional:
Challengers | PENTAX K-1 Special site | RICOH IMAGING

[biz-engineer]

Otherwise, denoising an image in post-processing could also be described as expanding dynamic range.

Actually not. That's practical miss-conception that reducing noise increases dynamic range. Reducing noise improves SNR but not DR. When shooting at ISO 800 , DR is reduced by 3ev because the 3 most significant bits of each pixel value aren't used due to sensor being underexposed, and noise reduction doesn't change that. For example, if you have the choice between a crop sensor with f4 lens or a full frame sensor and f5.6 lens, the crop sensor and FF sensors (both using the same pixel tech) will deliver the same IQ in terms of SNR, but the crop sensor will deliver 1ev better DR thank to the crop lens being 1 stop faster!

[clackers]

Actually, it didn't explain why it was compulsory, Kunzite.

In fact it identified different scenarios photographers face.

And we already know it kicks in only after ISO400, and in the K-70 is only half strength compared to the KP and K-1, so it's configurable by Ricoh but not by the camera owner.

[reh321]

The "accelerator" does not expand the scene-referenced dynamic range and the latter is the only one worth talking about.
Otherwise, denoising an image in post-processing could also be described as expanding dynamic range.

The same holds true for "colour depth". The "accelerator" almost certainly does not have access to data that isn't available to any image processing software that you can use for post-processing.

I don't care what you call the 'jog' where the 'accelerator' cuts in at about ISO 600 ..... I call it "improving DR that would otherwise be available at a given ISO" - and it is what matters to me

Attached Images

 

[SirTomster]

Come on. Did anyone expect anything differently?

But, I am not against what they said if you consider the "normal" upgrades for other camera companies.

But they are viewing the K-1 Mark II as a true successor to the K-1. For any other camera, a release generally is a full upgrade. New sensor, new this, new that, etc.

With the Mark II, Pentax did an update. Not a full upgrade you would expect going from a D810 to D850. Or 5D Mark III to Mark IV. You expect a significant upgrade since the camera has been released for many years.

The K-1 Mark II does not represent a traditional upgrade in a camera. And it seems that has been the norm for years. K-5 vs K-5II(s) K-3 vs K-3II.

But if you try to treat the Mark II as a true upgrade, then yeah, it sucks. Imagine a D860 released in 2020 that had a slight improvement to the auto focus and improved low noise handling. Nikon would be roasted. (Like Canon was for the 7DII)

But if you treat it as a tweak of the K-1, then your viewpoint changes. But every other camera manufacturer out there (that I know of) does not do it like Pentax. And that can (and does) bite Pentax because they are not doing it like everyone else. Reviewers are treating it as full release like they would from any other manufacturer.

If they did, then the next upgrade to the K-1 is still 2 years out (2020ish) The Mark II should not have been released. But instead, Pentax released an update. They did not have to build a camera from scratch around a new sensor. Instead, the tweaked some things.

If anything, I would make fun of them for not realizing what Pentax has done and acknowledge it.

[Kunzite]

Actually, it didn't explain why it was compulsory, Kunzite.

Of course they did, and very thorough - just count how many times they're describing the K-1 II as getting superior images

---------- Post added 31-12-18 at 09:21 PM ----------

And we already know it kicks in only after ISO400

The accelerator itself is permanent, sitting between the sensor and the PRIME processor - everything passes through it.
What we know it kicks in only after ISO400 is a kind of processing detected by Bill Claff's analysis.

[reh321]

The accelerator itself is permanent, sitting between the sensor and the PRIME processor - everything passes through it.
What we know it kicks in only after ISO400 is a kind of processing detected by Bill Claff's analysis.

and "we" don't know - at least I don't know - what the 'pin-out' is for the 'accelerator' - I don't know whether there is a single signal line saying "do/don't do anything" or whether there is a 'command bus' conveying the ISO value which causes it to do something internally.

[Breakfastographer]

Ah, I see. I must come clean and say I've never used PRIME. I've heard about it many times, often held up as the "Holy Grail" of noise reduction, but it seems like it has quite specific applications and isn't necessarily the best choice in all situations... Is that a fair statement?

DxO does promote PRIME as the ultimate in noise reduction. At this point, let me just say that I do think it's worth having in one's arsenal.

[Kunzite]

and "we" don't know - at least I don't know - what the 'pin-out' is for the 'accelerator' - I don't know whether there is a single signal line saying "do/don't do anything" or whether there is a 'command bus' conveying the ISO value which causes it to do something internally.

Sure, there's this issue of being able to actually disable the accelerator - data passing through unchanged.
Not a small feat, in the interview you can read about carefully balancing and fine tuning various aspects of image processing. I'm not convinced it can be done in a firmware update; not unless someone in the know would say so.

But there's the bigger issue of them wanting to: they worked hard, and are convinced the results are superior.

[Breakfastographer]

But there's the bigger issue of them wanting to: they worked hard, and are convinced the results are superior.

I think the astrophotography case is just very clear - that community simply does not accept any noise reduction at all, so they would need this fixed. Or they can keep buying the K-1 at a lower price.

Most would probably suspect Ricoh would rather have them purchase the K-1 II. And Ricoh seem to keep going after the astrophotography market, with the recent lens warmer announcement.

[biz-engineer]

Sure, there's this issue of being able to actually disable the accelerator - data passing through unchanged.

True, some people shoot high ISO because they like images with more noise. Do you realize the K1 II causes extra efforts because you have to add noise with your post processing software, it slows down your workflow.

[clackers]

True, some people shoot high ISO because they like images with more noise. Do you realize the K1 II causes extra efforts because you have to add noise with your post processing software, it slows down your workflow. DPR meant the image is so good you have to cripple it in post, whereas you don't need to do it with an XT-3 because the image is crippled out of camera.

(Laughs)