[Class A]

If (and I know we don't really know what exactly it does) the accelerator unit just boosts the signal to the processor rather than does any image processing per se,...

The "accelerator" processor is not an analogue unit. There is no "boosting" (as in amplification).

Fourier transform (taking the image from the spatial domain into the frequency domain) analyses show that there is high-frequency attenuation starting with ISO 640. This does not happen with regular noise reduction (such as dark frame substraction). This only happens when denoising techniques are applied that use smoothing which itself is based on determining the value of a pixel depending on its local context (cf. "nearest neighbour" approaches).

One can debate the visual impact all day long and even come to the conclusion that one prefers the treated output, but objective measurements don't lie.

[Kunzite]

Objective measurements can and often do lie.
I have an issue with the way those measurements were done - on patches of uniform colors, then the conclusion being that there "must" be some detail loss, in completely different circumstances. So if they're not confirmed to have a "visual impact"...

[MJKoski]

Lenstip measured lpmm values. Those are about the only number you are going to get if you want to watch numbers instead of photos.

I already did image difference comparison in Photoshop using layer blending modes just to note what goes missing after accelerator does its job. Someone could do the same with Matlab easily and export a matrix containing differences. There are some numbers then. :O

[Kunzite]

Differences is not the same with detail loss. Differences will appear just because of the random nature of noise, imprecise alignment with the test chart, etc.

[MJKoski]

No. You can see calculated image difference as faint edges in near-sharp areas which vanish in ISO800 image versus ISO100. Not just random noise.

[Kunzite]

And that doesn't happen with the K-1? In completely identical conditions?
Higher ISO will destroy some of the detail. Other factors will do that, too.

[sbh]

Perhaps, but maybe not.

I was impressed by the detail you recovered from one K-1 II image, but I'm sceptical that the same technique will be successful with different scenes where the "accelerator" may have smoothed irregular details (such as the dust particles on Adam's electronic component images).

I also noticed blotches in your processed K-1 II image that weren't present in the K-1 version; I guess that's what DPReview refers to as "artefacts".

I find some noise to be quite natural but smoothed surfaces and artefacts to be rather unnatural.

I did some tests with Adams raw files (the ISO 6400 ones). Yes, some details are reduced but I also found that the K-1 II actually shows more details in other areas, especially when colour is involved. Or the other way around: The noise of the K-1 obscures some detail which is captured by the K-1 II. However, the raw files of the K-1 II need a little more sharpening to make this visible.

Both have their dis-/advantages and I think it depends on what the individual user prefers. I slightly tend towards the II just because I find the colours more precise and like the overall result.

[reh321]

Well, even if converter X mangles the "RAW" in unwanted way during the post-processing stage, user has already began his/her work to get it ready for print / publish. However when the data is already mangled...

Consider this: Ideal RAW gives 100% quality. Data Mangler Product X reduces it by, let's say, factor 0.75. 1 times 0.75 = 0.75. When one puts Data Mangelerator Chip in the chain which messes the quality by, let's say, 0.75 factor the outcome after post-processing stage is 0.75 x 0.75 which aint 0.75 anymore as in the first case. Very bad.

Stacking plastic images results in ultra plastic images.

The evidence is that "manipulating" the 'raw' data before you get it is more common than uncommon.

https://www.pentaxforums.com/forums/190-pentax-k-1/367132-dprs-k1ii-re-shoot...ml#post4301643

[Rico]

Here is a quick real world image from my upgraded K-1. It was a bright sunny day so the light source was very strong. It was very windy outside. The image was taken at ISO 8000 using ƒ11 at 1/5000 sec to maximize DoF and freezing movement from the wind. Daylight white balance. The lens was an A series 50mm ƒ2.8 Macro. The JPG was created from a DNG file.

The subtle changes in the gradation for the plant leaves from shadow to light etc are still impressive even at ISO 8000. The continuous tone without splotchy patches of noise handle post processing with far greater ease than the files from the K-1 without the Accelerator Unit. In no way is this file inferior to what the K-1 could produce. The Accelerator Unit is definitely giving a more natural presence to the files. I have been dealing with K-1 files since July 2016. This type of output I could only hope for from the K-1.

[zapp]

If a $500 update promises three specific enhancements and we have to discuss one of them to death (see above), two are left. Handheld pixel shift and faster AF tracking. For AF tracking we have some subjective comments - fells better. Handheld pixel shift is OK, but nobody is raving about it. The accelerator is not a killer feature - I even assume Pentax thought they could do more with it two year after K1 release. That means, K1 is already quite pefect and the accelerator does more or less OK. Add some new features marketing wise, any camera should get a CLA with the upgrade (costs extra, but helps), and Pentax can collect some funds for real improvements.

[normhead]

If a $500 update promises three specific enhancements and we have to discuss one of them to death (see above), two are left. Handheld pixel shift and faster AF tracking. For AF tracking we have some subjective comments - fells better. Handheld pixel shift is OK, but nobody is raving about it. The accelerator is not a killer feature - I even assume Pentax thought they could do more with it two year after K1 release. That means, K1 is already quite pefect and the accelerator does more or less OK. Add some new features marketing wise, any camera should get a CLA with the upgrade (costs extra, but helps), and Pentax can collect some funds for real improvements.

Given the level of Digital Photography these days, who's done better with a release in the last 5 years. I'm still debating whether my K-1 is better than a k-5 for many images. And we have some image shot side by side that would suggest there are times when it isn't. The low hanging fruit is gone. We have Nikon Relaseing the D850 that isn't really better than a K-1, We have Canon releasing the Latest 5D incarnation that still isn't better for IQ than a K-1. Where is this camera company that has made great advances in their latest model release pf a 36 MP + camera?

No one expected increases in IQ when they upgraded their SLR. Same lens, same film, same image. It's all about the refinement of the process now. In a way, it's not even really designed for K-1 users, it's design for the guy looking for a current FF body. So what's on your shopping list?

Richer colour at high ISO? Slightly better AF? You can happily buy a K-1ii knowing the engineers have been busy since the K-1. But until whoever perfects the organic sensor, this looks to be pretty much it. It's unfortunate that people have become so use to huge changes things model to model that if something doesn't slap them in the face they feel cheated, but the simple fact is digital is maturing, and the low hanging fruit is gone.

It's simply unfortunate that some people are too incompetent at what they do to properly evaluate the changes. For me it's a lot simpler, Do K-1 users who have a K-1 like them? And the simple answer is a rough guess is about 10-1 do, and the one who doesn't is a kook. One of those self proclaimed scientists who thinks he's so good he doesn'tt need no stinking peer review, or relevant data.

Its tough on the test sites. They have to hype cameras that are essentially more of the same. SO, in the case of DPR, they just make stuff up.

Pentax mae a unique offer to give folks a chance to upgrade to the latest tech for $500. That in itself is pretty special. Who else has done that. But that's the thing Mr. Zapp, it doesn't matter how special, how unique what Pentax does is, it's not good enough for you. What you are asking Pentax to do here, is do something no one else can do. At this point in time Pentax is ahead of the game, where it counts, in the IQ of the final image, not frills like AF or FPS who's effects are not directly apparent in the final image.

I'm getting really tired because people try and present themselves as qualified to comment on the direction Pentax engineers have taken. They aren't. And their shoddy attempts to discredit Pentax are at best, inconclusive, at worst, grandstanding and playing the crowd like snake oil salesmen.

So enjoy your coffeee, go take some pictures.

[gaweidert]

If a $500 update promises three specific enhancements and we have to discuss one of them to death (see above), two are left. Handheld pixel shift and faster AF tracking. For AF tracking we have some subjective comments - fells better. Handheld pixel shift is OK, but nobody is raving about it. The accelerator is not a killer feature - I even assume Pentax thought they could do more with it two year after K1 release. That means, K1 is already quite pefect and the accelerator does more or less OK. Add some new features marketing wise, any camera should get a CLA with the upgrade (costs extra, but helps), and Pentax can collect some funds for real improvements.

Zapp, I cannot find it at the moment buy there was a video by CameraVille that compared the K1 and K1-II focus speeds posted in another thread on this forum. The K1-II was a fraction of a second faster, but for a moving object, that may be all that is needed. It is not advertised as a whole new autofocus technology just an incremental improvement.


I have used handheld pixel shift and found that it works as advertised. It is a nice feature and I have no problems waiting 20-30 seconds to see the results. It's a heck of a lot faster than waiting a week when I had to send the film out for processing. I know that people complain about the delay, but I cannot envision a scenario where you are going to be using it where rapid fire action shots are needed.


In the Pentax scheme of things, the K1-II is evolutionary and not revolutionary. They added a few features for those who want them and allowed a lower cost option for those who want them and have a K1 already. To me that is revolutionary. I have one of each and like them both. No pans to upgrade the K1. Both are great tools for capturing images.

[texandrews]

I asked and they emailed back. It's probably taking a little longer than usual due to the multiple things DPR got wrong (including post processing...), and IR's interest in checking all that. They also have a resident Pentax person. Think I'll be spending more time there...

[KX5]

No news about firmware upgrade for the noise issue in K1ii?

[Rondec]

No news about firmware upgrade for the noise issue in K1ii?

I don't think there will be one. The whole point of the K-1 II is the accelerator chip and Pentax isn't going to release a firmware to allow that processing to be turned off. By all accounts they are focused on further camera and lens development and don't see it as an issue.

If someone doesn't like the accelerator, they are best off buying a K-1 original which is still available and is a little cheaper than the K-1 II.