[MJKoski]

All 36MP Sony chip cameras get more or less messy after ISO3200. Chroma noise appears all over. D810A is an exception to this. And now with certain limitations K-1 mk2.

[i83N]

All 36MP Sony chip cameras get more or less messy after ISO3200. Chroma noise appears all over. D810A is an exception to this. And now with certain limitations K-1 mk2.

fair enough.

[interested_observer]

Thanks Interested_Observer. I'm definitely curious to see, as I like night/astro photos as well. Though I'm maybe more curious about the results at 20-30 secs and ISO 3200, 6400 and 12800.

I have 6 files for you. The large file transfer service that I usually use (file dropper) keep throwing errors when I upload the zip file. So, does anyone have a service to suggest, that's available, so that I can upload the straight out of the camera .DNG files for whomever.

I'm going back to bed, got home at 2am. I need some more beauty sleep.....

[SteveinSLC]

I have 6 files for you. The large file transfer service that I usually use (file dropper) keep throwing errors when I upload the zip file. So, does anyone have a service to suggest, that's available, so that I can upload the straight out of the camera .DNG files for whomever.

I'm going back to bed, got home at 2am. I need some more beauty sleep.....

Thanks. I think I can give you access to my Google Drive if I get your email. I will send you a PM.

[interested_observer]

Please post your results, or at least your opinions? @MJKoski has what I believe are valid concerns around the noise reduction on slightly-out-of-focus detail. I don't consider the issue to be as bad as he does, but it's a subjective assessment. I think the effect is there, even if (in my view) it's quite minor. It would be interesting to see if it affects your star photos...

Since you asked, here are some of the preliminary results. I shot 150 frames over 3.5 hours. I was going mainly for a 30mm stitch of either 2 or 3 rows. I optimized what I was doing and was able to get everything in to 2 rows of 7. Essentially one row for the sky and one for the landscape. Plus some extra of the ground elements in the sky. At the very end, I decided to do an experiment, so I went to 15mm and shot several vertical sets of 3 for some test stitching. The image below is the result of this little test - essentially a quick look of what I was trying to achieve. It only took a year.

Overall, this is by far the best I have done with the Milky Way, color and light - especially since I have been weather out for 4 months now. White Dots - and I have read everything that @MJKoski posted before buying the K1. This is by far the worst I have encountered. I believe that the white dots are as bad as they are due to the 50 seconds and not taking the time to shoot 120 second ground segments - so, I'm essentially making the problem worse. I guess I should shoot the ground elements with NR - dark frame subtraction enabled. I hate to take the time, but I really am asking myself if I want to start sending the camera to Pentax and trying to get it fixed and having a good fix.

Anyway, this is a stitch of 3 images, shot at 15mm, f2.8, 50 second exposure astro tracked at ISO 800. Post processed for the sky prior to stitching and then the ground processed in the stitched result. This was a test to see if I was able to capture the essence of what I was trying to capture. Shot at midnight. No real special post processing - just standard stuff for the stars - levels and curves. Light dome to the left is the little town of Superior, AZ - 5 miles away, with the light dome to the right, the Ray open pit copper mine 40 miles away. The Phoenix metro area is 60+ miles away directly behind me.

An original Arizona cattle watering hole at Picket Post Mountain.

[madbrain]

Yes. mk2 is ISO100-400 camera when exposing an image now.

There is an ISO 500 step also on the K-1 II which has no NR.

---------- Post added 05-19-18 at 02:52 AM ----------

And mk1 is iso100-1600?

Mk1 has no detectable NR at any ISO, per Bill Claff's tests. Of course, there is very noticeable noise at higher ISOs, such that you almost certainly would be applying NR in post.
I think if there is any case where you wouldn't want to apply any NR in post, it would be ISO 640-3200 range or maybe even ISO 640-1600 range. But it really depends what your target is. For web sizes or home-printable sizes, it's pretty much irrelevant unless it's a small crop.
It may become an issue for very large prints or crops of very small areas of the photos. For the later, it probably means you were not shooting with the most appropriate lens for your subject, unless we are talking about astrophotography and shooting planets, in which case you might use the longest focal length you have and still end up having to crop. Personally, I have been using RAW converters software which has fairly poor NR, such that I almost never enable it because it blurs the pictures so much. So, I welcome the lower NR of the K-1 II. However, I can totally understand why others might want to turn it off in some cases.

[monochrome]

Honest question:

• How many people in how many circumstances will actually realize an actual loss of VALUE in terms of money or contest submissions or something tangible from these slightly less detailed RAW’s?
• How many people will have better looking OOC jpeg’s to display or upload from a scene that previously would have been too noisy to be worth bothering over?

I just don’t believe this argument has any real, practical conclusive value. Is anyone really going to lose a display-sized print sale or really going to lose a wedding customer because ISO 12,800 noise reduction doesn’t justify a hypothetical bit of loss of edge detail isible at 400% at ISO 800, or because Bill Claff uploaded a graph with some triangles slightly below a line? I see data points, but so what? I gather the sensor would allow me to underexpose and push, but so what? What does any of that prove?

What is the ratio of now-usable-prints to now-lost-sales or now-not-usable-prints?

Or are you subtly inferring that I should use my KP for my amateur stuff and leave this argument to the professionals who need to use a K-1 / K-1 II, because clearly no one is arguing about the AU effect on RAW’s from the KP?

Put another way - should i, personally, upgrade my K-1 or not, given that I have the AU in my KP? For which actual buyers does this really matter and how, actually, in the real world, and what will be the effect in lost camera sales to Pentax of this seemingly near meaningless explosion of FUD?

---

Last edited by monochrome; 05-19-2018 at 07:07 AM.

[Kunzite]

I call it the "every new Pentax camera is worse than its predecessor" syndrome.

[Rondec]

Honest question:

• How many people in how many circumstances will actually realize an actual loss of VALUE in terms of money or contest submissions or something tangible from these slightly less detailed RAW’s?
• How many people will have better looking OOC jpeg’s to display or upload from a scene that previously would have been too noisy to be worth bothering over?

I just don’t believe this argument has any real, practical conclusive value. Is anyone really going to lose a display-sized print sale or really going to lose a wedding customer because ISO 12,800 noise reduction doesn’t justify a hypothetical bit of loss of edge detail at visible at 400% at ISO 800, or because Bill Cliff uploaded a graph with some triangles slightly below a line? I see data points, but so what? I gather the sensor would allow me to underexpose and push, but so what? What does any of that prove?

What is the ratio of now-usable-prints to now-lost-sales or now-not-usable-prints?

Or are you subtly inferring that I should use my KP for my amateur stuff and leave this argument to the professionals who need to use a K-1 / K-1 II, because clearly no one is arguing about the AU effect on RAW’s from the KP?

Put another way - should i, personally, upgrade my K-1 or not, given that I have the AU in my KP? For which actual buyers does this really matter and how, actually, in the real world, and what will be the effect in lost camera sales to Pentax of this seemingly near meaningless explosion of FUD?

I think people for some reason have focused a lot on pixel level comparisons which to me are fairly meaningless. The question really isn't can you find miniscule differences between a K-1 and K-1 II detail content at iso 1600. It really isn't even whether if an "expert" looks at a K-1 II image at pixel level if they can see evidence of noise reduction. It isn't even whether an experienced photographer could do a little better with high end noise reduction software and skills. Because those things miss the point of photography, which as you say, is the final photograph.

The basic point at hand is if the K-1 II files look better printed with less time and effort, then the accelerator has a benefit.

Let's say I come back from a trip to a museum or aquarium with my family. I have shot the majority of images at iso 3200 and 6400 because the light wasn't great. I have 500 images, most of which are snap shots but have meaning to me. Do I want to process every photo to get 3 or 4 percent more detail out of them? How much time would that actually take me and how much benefit would I truly see using standard presets in Lightroom or Nik Dfine? Clearly (in my mind) for hobby photographers like myself there is benefit. But the case is broader yet because my wife shoots weddings and when she comes away with a couple of thousand photos, she really doesn't want to spend a lot of time with each image processing it. Having relatively clean images in the 3200 -12K iso ranges would save considerable time and it is guaranteed that the clients would not see a difference -- formal shots are done at low iso anyway and it is reception shots and ceremony shots, which aren't blown up to large sizes.

Maybe this is a negative for wildlife and astro photography, but I am a bit unclear on that and would like to see more real world examples before I would comment on those.

[Tan68]

... BUT... I'd like to mention something else which I feel is a little bit of a 900lb silent gorilla in the room:

When I first got my K-1, coming from a K-5 classic, I was surprised a little by the amount of noise at base iso and 100% viewing compared to the K-5, and I find myself applying NR in post to RAW in ACR to files as early as iso600 or so. This takes time, and effort... if the built in NR just mildly puts back what I was used to with the K-5 as to noise, and doesn't smear it any worse than I would do with ACR in PP, then is there really a problem??? ...

I have also seen the gorilla in our midst

I used the K-5 at ISO 800 and would later lift exposure 1-1.5 stops and this gave good brightness for most low light scenarios I found myself in. I didn't really like using ISO 1600 in camera and this was just the way I liked it.

The K-1 has seemed to exhibit more noise than I would expect to see compared to my K-5. But I haven't done side-by-side comparisons and this is just a feeling I have.. I am very familiar with K-5 and some familiar with K-1. It is entirely possible I am pushing the K-1 a little harder because I know it should perform some better, not at a pixel level, but at the same viewing sizes I typically use. So far, I haven't been disappointed with K-1 /output/.

I used to really be into technical stuff in what I hoped was a practical way.. sometimes it was just to satisfy my curiosity. Before ISO-lessness versus ..ISO-ful was a thing, I had determined the best low light IQ was available from a K-20 when first setting desired aperture and slowest possible shutter and then exposing right using the ISO setting but only up to 1600. This was empirical study :^) I was learning about digital stuff...

Anyway, rambling hard now. I feel like i have noticed what you have between K-5 and K-1 but don't see me testing things out to a firm conclusion; I feel less of this desire to know with a dead blow hammer as I get older and just want stuff to work.. [ pointing to Windows 10 :^) ] and I think the K-1 II might be an okay bet for me.

* Knowing how to squeeze performance from a K-20 was helpful sometimes because operating the older cameras with less better performance was like driving a stick shift car with no choke while newer cameras with performance to spare are like driving cars with comfort-glide suspensions...

[Not a Number]

I think people for some reason have focused a lot on pixel level comparisons which to me are fairly meaningless. The question really isn't can you find miniscule differences between a K-1 and K-1 II detail content at iso 1600. It really isn't even whether if an "expert" looks at a K-1 II image at pixel level if they can see evidence of noise reduction. It isn't even whether an experienced photographer could do a little better with high end noise reduction software and skills. Because those things miss the point of photography, which as you say, is the final photograph.

You've got the right idea. To quote RIcoh's K-1 Series Special site:

PENTAX engineers are always working toward the realization of images that are rich in detail and gradation; achieving this goal often means valuing sensory evaluations over numerical evaluations.

Realizing the ideal imaging concept included extremely detailed activities, because PENTAX places as much importance on the human perception of beauty as the company does in numerical evaluations.

[Erictator]

I think you are spot on. The shortest ownership I ever had of a digital Pentax slr was the K-7. I got some good shots out of it at base ISO, but I never warmed up to it because I couldn't push it hard at all. I missed my Ds sorely, and dumped the K-7 the minute the K-5 came out.

It all boils down to practical application. If it works for me in my typical real life scenario's, then I'm a happy camper. The K-1 MkII may do a little more NR, or, maybe the files are just plain cleaner. Noise can mimic detail and give the false impression of improved sharpness or detail.

Don't all sensor read-outs do at least some cleanup? We just get used to the output as "raw" and assume it's untouched, but I'd be very surprised if they are not at least minimally cleaned up. Pixel mapping for instance, must happen at the pre-raw output level. We prolly never notice much because they usually get better with each release of a given sensor.

Speaking of gorilla's, I think another one I haven't seen mentioned is that the new sensors like the 36mp and newer, have so much resolution that the test environment has become uber critical. Focal plane, lens calibration, lens sample variations, etc etc etc all become a huge factor in testing at the pixel level. I have a feeling that is why the new DFA 50mm has been delayed... Pentax's internal testing is probably uber critical (unlike DPR), and found issues in sample variation that had to be fixed in the production line level.

Eric

I have also seen the gorilla in our midst

I used the K-5 at ISO 800 and would later lift exposure 1-1.5 stops and this gave good brightness for most low light scenarios I found myself in. I didn't really like using ISO 1600 in camera and this was just the way I liked it.

The K-1 has seemed to exhibit more noise than I would expect to see compared to my K-5. But I haven't done side-by-side comparisons and this is just a feeling I have.. I am very familiar with K-5 and some familiar with K-1. It is entirely possible I am pushing the K-1 a little harder because I know it should perform some better, not at a pixel level, but at the same viewing sizes I typically use. So far, I haven't been disappointed with K-1 /output/.

I used to really be into technical stuff in what I hoped was a practical way.. sometimes it was just to satisfy my curiosity. Before ISO-lessness versus ..ISO-ful was a thing, I had determined the best low light IQ was available from a K-20 when first setting desired aperture and slowest possible shutter and then exposing right using the ISO setting but only up to 1600. This was empirical study :^) I was learning about digital stuff...

Anyway, rambling hard now. I feel like i have noticed what you have between K-5 and K-1 but don't see me testing things out to a firm conclusion; I feel less of this desire to know with a dead blow hammer as I get older and just want stuff to work.. [ pointing to Windows 10 :^) ] and I think the K-1 II might be an okay bet for me.

* Knowing how to squeeze performance from a K-20 was helpful sometimes because operating the older cameras with less better performance was like driving a stick shift car with no choke while newer cameras with performance to spare are like driving cars with comfort-glide suspensions...

[Tan68]

Honest question: ...

Ask an honest question, get an essay :^|

... <ul>How many people in how many circumstances will actually realize an actual loss of VALUE in terms of money or contest submissions or something tangible from these slightly less detailed RAW’s?...

I expect not many

... Or are you subtly inferring that I should use my KP for my amateur stuff and leave this argument to the professionals who need to use a K-1 / K-1 II, because clearly no one is arguing about the AU effect on RAW’s from the KP?

Put another way - should i, personally, upgrade my K-1 or not, given that I have the AU in my KP? For which actual buyers does this really matter and how, actually, in the real world, and what will be the effect in lost camera sales to Pentax of this seemingly near meaningless explosion of FUD?

-- there may be an answer in this somewhere --


Of all the reasons available to want a FF camera, I want the flexibility to crop extensively (I don't use zooms..) and reduce the number of lenses I carry and lens changes I make. I count grams when I travel and a K-1 kit with a lens omitted weighs the same as my old K-5 kit. I am plenty happy with my decision to crop 1.5x to achieve angle of view I rarely used anyway. ta

Many times in low light scenarios I was DOF limited even with the K-5. Versus K-5, the FF camera should have around a stop better performance at comparable viewing sizes but I would give that stop up to stopping down to maintain preferred depth of field with the FF camera. Make sense..? I haven't really gained much in low light performance in many of the places I use my cameras. >> a side note for Clackers & crew : I will not respond to any esoteric arguments about FF DOF fru-fru because, as I have stated in other threads, /subject framing/ is important to me and I require no education how FF, in a pomme to pomme de terre comparison, can actually provide greater DOF than APS-C. Over and out. :^)

If the KP and K-1 were released together, I would have bought KP.

With the KP, I would have enjoyed a stop better low light performance over K-5 and had the option to use that stop for greater DOF in my favored low light scenarios... as noted above, DOF with FF and same subject framing are about a wash moving from K-5 to K-1. The KP kit would have given me more flexibility where I would like to have it for the same weight.

When the K-1 II upgrade was announced, I was become excited. If the K-1 II had a stop better performance as does the KP then I would have greater flexibility with DOF in my favored low light scenarios! I could have my KP and eat it, too :^)

Looks to me like the K-1 II does have that stop better performance.

I haven't used external NR (NIX or Photo Ninja or etc) on a Pentax camera since a few fluffed K-20 shots... I typically dial back Capture 1 default NR. I do have to futz with NR for a fairly current 4/3 camera (what is a blue sky Panasonic??)

Point being, I am not an NR wizard. Of course, I am happy enough when a wizard passes by but do wish him spread his magic elsewhere. I am satisfied with what I get from my software as is. I have been writing 'my favored low light scenario' because this varies from person to person but it is around f/2, 1/60 at ISO 800 lifted 1-1.5 stops later.. with occasional high DR scenes such as a single image : back lighted stained glass with reasonably well exposed interior.

Looking at the KP samples, I expected I would be happy with the results at a stop less exposure compared to K-5 and without the need to bone up on NR technique. I expected to be just as happy with K-1 II.

I was plenty happy with KP samples as are many people. I do think the same pixel popping stress would surround KP if we had on hand both a KP & KP II, one with and the other without the accelerator. I floated this idea in another thread but it has been judged unpopular :^|

Now FUD

If the K-1 II NR started /after/ ISO 800, all would be a non-issue for me because I could shoot as I had in the past and use ISO 1600 in-camera if I wanted to 'turn on' NR. But the threshold aint 800 and I got a decision to make. I see some point in ISO-less but have so far found matching ISO closely enough in-camera such that I only have to later lift a couple stops works better for me than a lower ISO and later lifting 3-4 stops...

I do see K-1 to K-1 II comparisons in which first K-1 looks better and then K-1 II looks better in either this or that portion of one or the other test image... I hate to give up any hard won detail at all! but... to me, these comparisons mean things are a wash.

I believe some people are on the fence between K-1 & K-1 II because the accelerator can not be disabled and they must make a choice good forever! People are adrift on a floe of uncertainty headed south with FUD masters predicting warm weather. This makes objective decision making hard and I will admit my initial glee was tempered by predictions of a warm spring. :^(

In spite of all, I have decided the K-1 II will be good for me.

My remaining reservations are over the upgrade process itself. Why must joints be soldered!? What mayhem will be wrought by an upgrade technician with either fat fingers or low blood sugar. Both things truly worthy of worry. I begin my Descent into the Maelström and seal this MS. in a Bottle... Wish me luck... :^|

[monochrome]

@Tan68 In my case the K-1 is for legacy FF manual lenses that I can OVF focus (I did purchase D FA24-70 and 70-200) - pretty narrow use case, I suppose. The KP is for casual use and snapshots at frame rates the K-1 cannot do. I generally don’t take advantage of the crop factor. The KP AU allows higher ISO for faster shutter speed, mostly for indoor shooting. K-1 is noisy at ISO 3200. As things are, since K-1 is intended for more serious hobby photography, if I need to I can learn PP NR for far less money than buying a K-1 II or even upgrading.

My choice comes down to selling KP, all the DA lenses, and upgrading the K-1 forever, waiting, selling K-1 and buying a K-1 II forever - or using what I have. ATM I think I’ll stay with what I have, not because of the RAW smoothing issue, but because my actual use case is satisfied.

The good that has come of this, for me specifically, is the gear lust was tempered by the discussion so that I am forced to make a rational decision.

[Tan68]

You've got the right idea. To quote RIcoh's K-1 Series Special site:

I don't know if this will show up in the nested quotes, so here : Realizing the ideal imaging concept included extremely detailed activities, because PENTAX places as much importance on the human perception of beauty as the company does in numerical evaluations.

This is the same position many people take with the limited lenses

I am perfectly fine with my small lenses
I don't want the increasing popular perfectly corrected soup cans
I realize I am leaving some elements of lens performance on the table
And I am okay with that...

So... why is it so hard to leave some detail on the table with K-1 II, this detail being distinguishable only after exhaustive study...

I think it is because I /could/ buy a soup can lens and rake in the extra lens performance winnings.. while anything left behind with a camera is regained only after buying a new body.

However, I don't see me buying soup can lenses ** and while I want the ultimate better body, I think I will be happy with Pentax K-1 II just as I am with Pentax other-than-perfect-not-soup-can lenses. bless their tiny hearts being driven to extinction

** I recently purchased an IRIX 15
WTF ?!! LOL I like it but I Will Not travel with it back-pack style :^p

---------- Post added 05-19-18 at 10:48 AM ----------

@Tan68 In my case the K-1 is for legacy FF manual lenses that I can OVF focus - pretty narrow use case, I suppose. The KP is for casual use and snapshots at frame rates the K-1 cannot do. The AU allows higher ISO mostly for indoor shooting. K-1 is noisy at ISO 3200. As things are, if I need to I can learn PP NR for far less money than buying a K-1 II or even upgrading.

My choice comes down to selling KP, all the DA lenses, and upgrading the K-1 forever, waiting, selling K-1 and buying a K-1 II forever - or using what I have. ATM I think I’ll stay with what I have, not because of the RAW smoothing issue, but because my actual use case is satisfied.

The good that has come of this, for me specifically, is the gear lust was tempered by the discussion so that I am forced to make a rational decision.

Well, if I had the KP, I don't think I would upgrade to either K-1 or K-1 II. Upgrading a K-1 only gives me, in some cases, performance I would see with the KP.

However, I don't got a KP so my decision tree is little different :^)

I could either become an NR wizard or sell the K-1 and buy a KP. I want only one body to use and will upgrade the K-1 to encompass benefits I could enjoy with the KP while still having benefits the FF body provides versus APS-C such as your use with manual focus lenses or my desire to crop extensively.

The K-1 body is bulkier than KP but not enough to matter (it still just fits in the bag I used for K-5) and the weight between systems for my kit is the same.. so, I get picked from the floe by SS K-1 II while you may be rescued by SS KP... see you in port :^)


+ I noted earlier that if KP & K-1 were released together, I would buy KP. I could still achieve KP-ness by selling the K-1 but selling gear isn't easy and after buying L-bracket and batteries I don't think I would be ahead much. So I think I will stay with what I have since I started down this route.