[reh321]

drache let me get this straight. People are using "review" site images to make conclusions on image "quality" from places like DPR and IR that have taken their test images in unscientific manner but my real world example isn't good enough. My image is reproducible. I can take dozens and dozens of similar images.

So, you handle each image exactly the same? What is the advantage then of PPing a 'raw' file instead of using JPEG created by the camera?

[Rico]

Your example image is excellent, but what it doesn't give us is the same RAW shot taken with both cameras in using the same lens and shooting parameters, both without and with processing applied to the best of the user's ability - and copies of the unprocessed RAW files for us to play with. It's a big ask, I know, but that's what's required. Otherwise, what your image does is tell us that the K-1II is capable of taking really nice images - and well know that anyway.

I asked previously if you could process the Imaging Resource K-1 and K-1II files I referenced and process the K-1II file to the point where it provided equivalent detail in the bright red material swatch. In earlier posts, I showed how, even with colour noise reduction applied, my attempts at processing demonstrated slightly more detail in the K-1 file. I'm genuinely interested to see if the same level of detail can be brought out in the K-1II file in that specific area of the image, since it's the reds where it seems to struggle based on my own efforts with those files. If it can, fantastic. If it can't it doesn't mean the K-1II is a lesser camera, but only that we accept there is some cost as well as benefit to the accelerator's noise reduction.

BigMackCam this is very tedious timing wasting endeavor and getting boring. Here is what you are asking. These images are JPG's from DNG files created using Digital Camera Utility 5. I did not alter them in any way sans sizing etc.





I am not seeing any loss of detail in the red swatch. There is a difference in hue for that particular swatch between cameras but there isn't any "less" or "more" detail one to the other. The question I ask is why aren't you seeing "loss" of detail all over the K-1MKII image in the same way you are seeing it in the red swatch. What is happening in the red swatch is happening equally across the entire file. This goes for all such claims of "loss" of detail from the K-1MKII.

Here is something nobody has mentioned while 200% and 300% pixel peeping the difference in the color to the circular proportional pinwheel scale lower right:





One camera has to be way off or IR substituted in a white one for the beige one.

[BigMackCam]

BigMackCam this is very tedious timing wasting endeavor and getting boring. Here is what you are asking. These images are JPG's from DNG files created using Digital Camera Utility 5. I did not alter them in any way sans sizing etc.

I'm sorry you feel that way

I'm not trying to challenge anyone, trip anyone up, or prove that I'm right and they're wrong. I'm genuinely interested in this because I think it's important for future development in Pentax cameras. If there's truly no cost to the accelerator's noise reduction, no-one would be more delighted than me. As it is, working with the available files myself in Darktable, I keep swinging back and forth between a preference for the K-1 and K-1II files. My current feeling, based on nothing more than my own post-processing attempts, is that there is minor detail loss in very isolated areas of the K-1II's images, but considerably better noise handling. I would love to find that the minor detail loss I've noticed is a limitation of my post-processing techniques rather than the accelerator's NR.

Thank you very much for providing the photos (these are the ISO 1600 shots discussed, yes?). May I ask for one last thing? 100% crops of the bright red swatch from both the K-1 and K-1II images, as the reproduction size in your post doesn't allow for comparison of fine detail. It's at 100% reproduction that I can clearly see the difference in my own processed images.

Thanks in advance

[Rondec]

I'm sorry you feel that way

I'm not trying to challenge anyone, trip anyone up, or prove that I'm right and they're wrong. I'm genuinely interested in this because I think it's important for future development in Pentax cameras. If there's truly no cost to the accelerator's noise reduction, no-one would be more delighted than me. As it is, working with the available files myself in Darktable, I keep swinging back and forth between a preference for the K-1 and K-1II files. My current feeling, based on nothing more than my own post-processing attempts, is that there is minor detail loss in very isolated areas of the K-1II's images, but considerably better noise handling. I would love to find that the minor detail loss I've noticed is a limitation of my post-processing techniques rather than the accelerator's NR.

Thank you very much for providing the photos (these are the ISO 1600 shots discussed, yes?). May I ask for one last thing? 100% crops of the bright red swatch from both the K-1 and K-1II images, as the reproduction size in your post doesn't allow for comparison of fine detail. It's at 100% reproduction that I can clearly see the difference in my own processed images.

Thanks in advance

But if you have to squint at an image at 100 percent in order to see the difference in detail, what does that tell you?

[normhead]

But if you have to squint at an image at 100 percent in order to see the difference in detail, what does that tell you?

It tells you it's not relevant to a print file, or a reduced size image. People have argued pixel peeping is relevant if you print large. Personally I doubt it.

I had an image taken with the second softest lens I own, there was sign that was a little fuzzy and would have ben hard to read. If it was unreadable however, it would have made absolutely no difference to the overall image. The sign was stapled to the door i a boathouse. Like most detail of the really fine type, it was irrelevant to the image.

[BigMackCam]

But if you have to squint at an image at 100 percent in order to see the difference in detail, what does that tell you?

The things is, I don't need to squint to see the difference. In my processing attempts, at 100% reproduction on my 17" laptop screen, there's a clear difference in that red swatch. All I'd like to know is, is it my processing or the K-1II accelerator's that is causing the difference?

I've maintained all along in my numerous posts on the subject - in this thread and others - that at 50% reproduction, it's not an issue - so in most circumstances, the better noise profile of the K-1II is actually preferable to me. I've defended the K-1II's accelerator module, because I like the reduced need for colour noise reduction in post-processing, and the additional headroom it offers for further colour noise reduction without impacting image colours adversely.

But if there's a cost in detail, however minor, it's better to know that... and better for Ricoh Imaging to know that, so that next time they'll hopefully make the feature switchable or adjustable.

Either way, we know the K-1II is an awesome camera with great IQ...

[SteveinSLC]

At least in what Rico just posted, I can't see any red difference in the photos in the thread itself. No detail differences and if there are hue differences, I can't tell.
I'm viewing on a newish LG 27" Ultra HD monitor, just calibrated last week.

And they clearly did something relative to that pinwheel. Not only is the color different, but the blue brush on the right is much closer in the MkII shot.

---

Last edited by SteveinSLC; 06-29-2018 at 11:42 AM. Reason: Added content

[BigMackCam]

At least in what Rico just posted, I can't see any red difference in the photos in the thread itself. No detail differences and if there are hue differences, I can't tell.
I'm viewing on a newish LG 27" Ultra HD monitor, just calibrated last week.

Nor can I, at this size of reproduction. Hopefully, the same will be true for the 100% crops

[biz-engineer]

Will Pentax K1 Mrk II blur the sharpness of the new DFA*50 outstanding lens ?

[Rondec]

The things is, I don't need to squint to see the difference. In my processing attempts, at 100% reproduction on my 17" laptop screen, there's a clear difference in that red swatch. All I'd like to know is, is it my processing or the K-1II accelerator's that is causing the difference?

I've maintained all along in my numerous posts on the subject - in this thread and others - that at 50% reproduction, it's not an issue - so in most circumstances, the better noise profile of the K-1II is actually preferable to me. I've defended the K-1II's accelerator module, because I like the reduced need for colour noise reduction in post-processing, and the additional headroom it offers for further colour noise reduction without impacting image colours adversely.

But if there's a cost in detail, however minor, it's better to know that... and better for Ricoh Imaging to know that, so that next time they'll hopefully make the feature switchable or adjustable.

Either way, we know the K-1II is an awesome camera with great IQ...

Sure. And my personal opinion is that there probably is a very small amount of detail lost in some situations. This is the cost of having a significant improvement in noise at higher isos.

I just would hesitate to say that the K-1 II is "a step back," as has been reiterated multiple times in this thread and elsewhere. I would describe it as a small step forward and would further say that if you never use (or are currently satisfied with ) the K-1's high iso and auto focus performance then you don't need a K-1 II. But that's no reason to disparage what is an excellent camera that performs quite well in the real world.

[normhead]

Will Pentax K1 Mrk II blur the sharpness of the new DFA*50 outstanding lens ?

I the test images I've seen some shaper mkII images and some not as good. The 400 ISO image was sharper, better micro contrast and richer colours. I'm not convinced yet that the different details are not a function of other factors than the accelerator chip. The thing is lights get changed, the sets and lighting have been changed from time to time. We don't know what happened between them shooting the K-1 and the K-1 mk 2 images. Hopefully when we get the forum review, we get comparable images taken at the same time in the same environment within a minute or two of each other.

[Rico]

The red swatch has a particular weave the K-1MKII is picking up much better especially the spaces between the pattern. There are multiple reasons for this the angle of capture are off creating different angles of incidence for highlights and shadows. The two cameras have used different settings both color temperature (white balance) and Image Tone Natural over Bright which influences the color hue. The lighting and or exposure may be different also as you can see the shadow cast by the color thread in the previous images is much darker for the K-1. This also influences perception.

You can see how the background grey is slightly a value darker between the images. This is happening all over the image. The red swatch color for the K-1 image between the pattern is several values darker than the pattern itself where it almost looks like two different hues giving the illusion of "more" detail. The K-1MKII image the red swatch the pattern and space between have a more consistent hue which is showing more of the actual weave to the fabric.

[BigMackCam]

Sure. And my personal opinion is that there probably is a very small amount of detail lost in some situations. This is the cost of having a significant improvement in noise at higher isos.

I just would hesitate to say that the K-1 II is "a step back," as has been reiterated multiple times in this thread and elsewhere. I would describe it as a small step forward and would further say that if you never use (or are currently satisfied with ) the K-1's high iso and auto focus performance then you don't need a K-1 II. But that's no reason to disparage what is an excellent camera that performs quite well in the real world.

Agreed. And if I've ever come across as disparaging about the K-1II, that's not been my intent. Let me be absolutely clear to anyone reading this - I would be delighted to own one. I consider the NR to be a step forward in some situations, or - at the very worst - a step sideways in others...

[BigMackCam]

The red swatch has a particular weave the K-1MKII is picking up much better especially the spaces between the pattern. There are multiple reasons for this the angle of capture are off creating different angles of incidence for highlights and shadows. The two cameras have used different settings both color temperature (white balance) and Image Tone Natural over Bright which influences the color hue. The lighting and or exposure may be different also as you can see the shadow cast by the color thread in the previous images is much darker for the K-1. This also influences perception.

You can see how the background grey is slightly a value darker between the images. This is happening all over the image. The red swatch color for the K-1 image between the pattern is several values darker than the pattern itself where it almost looks like two different hues giving the illusion of "more" detail. The K-1MKII image the red swatch the pattern and space between have a more consistent hue which is showing more of the actual weave to the fabric.

Thank you! That's really helpful, and valuable

I do notice tiny details missing in each file versus the other, but I'd say it's a wash overall in that sense - and I agree, the pattern differentiation looks better on the K-1II file. In your DCU-processed example, I prefer the K-1II.

One conclusion we can draw from this - and it wouldn't be for the first time when looking at RAW performance in a camera - is that the processing software (and the pilot ) makes a huge difference. Using Darktable, I saw (and demonstrated) more significant differences between the K-1 and K-1II files.

It would be worth alerting @bdery to your findings, as he's writing an article which will compare the IQ between both cameras. Your results should be noted and taken into consideration.

Thanks again

EDIT: Just to confirm, Rico, these are the ISO 1600 NR0 files, yes?

[SteveinSLC]

I'd agree. There are some small differences. The shadow cast by the red on pink fabric is darker in the MkII. But one has to look pretty closely, even at this magnification. And I'd be really hard pressed to say that one or the other is actually "better" in any significant way.

And I continue to be on the fence about paying the $550 for the upgrade.