[JohnBee]

John, is the banding on that green felt?

The last image had different types of light shining on the exhibit, it is impossible to get WB right there. Also, the magenta cast should not change noise pattern.

But, it seems like the banding has worsened from some of these first shots.....these are just sample shots at iso3200, nothing to showcase my photographic skills

Yes on the right of the wheel in the felt... there is noise banding present in that area.
It's nothing to worry about, though it looks like you could most likely wash it out with some noise carefully applied correction.

I totally understand your plight on WB in sub par lighting.
Which is why I always keep a set of WB cards around my neck when I shoot and some WB caps(use them mostly). They really do help.

The tinge is easily corrected BTW

[pcarfan]

John, that's not the intent of my post. I can adjust the WB easily, I was just explaining why the camera got it wrong. I didn't want to do any Post processing including WB. This is just cropped and posted, and please try to ignore everything else other than banding and anything that would influence banding.

It seems to me, these earlier images I took with the original formware have less banding than the ones I posted previouslyn (which were all taken recently with the updated firmware).

P.S: The k-7 IMO is good up to about iso800 for jpeg shooting.

[JohnBee]

John, that's not the intent of my post. I can adjust the WB easily, I was just explaining why the camera got it wrong. I didn't want to do any Post processing including WB. This is just cropped and posted, and please try to ignore everything else other than banding and anything that would influence banding.

It seems to me, these earlier images I took with the original formware have less banding than the ones I posted previouslyn (which were all taken recently with the updated firmware).

P.S: The k-7 IMO is good up to about iso800 for jpeg shooting.

I understand, though, as I've found a direct relationship between that of WB and MT, I figured I'd mention(WB MT) as part of the discussion.

I was talking with my wife(just now) and she too feels that the banding on one of our K20's. started after a firmware upgrade. Though I can't substantiate it, I will look through the archives to see if I can't identify some before and after behavior.
Also interesting, is that our "better" K20D(in terms of banding) is the one with the original 1.0.1 firmware in it. So... there definitely is something something to look into on that.

FTR. Here is some K20/ ISO4000 output.

I wouldn't flaunt this as an "as is" image, I fought artifacts all the way home to get this image out of it. But... we did get the shot, and so that 1/2 stop or whatever the K20 had to offer really came through for us that time.


K20D/Series-1/2.8/Monopod

@javier,...Did you put those lenses for sale here?, I think the Sigma 100-300 and the bigma should sell quickly, but not too many can afford the Sigma 500/4.5, but that is one stellar lens though....

@Gus, are you being sarcastic there

?

Not being sarcastic at all... do I have a rep for it?

I believe if I was shooting raw with the K7 and getting your ISO 3200 shots after a pass with some NR program. I would not spend a nickel on the K-x. Yours are very usable and the K7 is more than 2x the camera the K-x is in feature set alone.
My goodness the Kx is a single wheel menu driven entry level...

So if I had the k7 and was getting your results I would spend that 500 bucks on glass... (dont know what in the hell Pentax has for $500 that you and I don't already have) Now that was sarcasm!!!

I just have a dead K20 sittin' on a bench in Arizona right now... and I ended up with this Kx and I am singing the Hi ISO Hallelujah chorus. It is shocking to me how good the jpegs are.
But it is an entry level, crawl thru the menu cam.

I do think the K20 will lose some face time ( if it ever gets back) But if I can figure out the balance between the two, it will be a great combo.

Peace Love and Understanding...

FTR. Here is some K20/ ISO4000 output.

I wouldn't flaunt this as an "as is" image, I fought artifacts all the way home to get this image out of it. But... we did get the shot, and so that 1/2 stop or whatever the K20 had to offer really came through for us that time.


K20D/Series-1/2.8/Monopod

I would flaunt it! That is a beautiful shot at ISO 4000! I respect how well you fought this one! It was worth whatever tweeking you did.
I would love to sit next to you as you work your magic cleaning up an image like this. I am one of those guys who learn better by watchin' than readin'

[jgredline]

The issue that Javier has is that he is unwilling to shoot RAW. It is not unreasonable to expect to shoot JPEGs, but the reality is that when shooting high iso on Pentax (other than the kx), RAW and noise reduction software make a big difference.

At the same time, I would have to disagree with Javier. I could not get better results with my K20 than this, certainly not anything that would be visibly different at printing size. In fact, most of my K20 iso 3200 photos look lousy -- in decent or not in decent light. I pretty much keep iso 1600 as the upper limit, unless I desperately need a photo and can't use flash.

I will admit that the last firware upgrade, did bring the High ISO issue closer the K20D, but all n all, Both K-7's I have had, simply don't compare to the K20D. This of course is my opinion and I must admit that I am impressed with what pcarfan has been able to do with the K-7...But he also on another level than me, and hence his wonderful results...

[Laurentiu Cristofor]

Yes on the right of the wheel in the felt... there is noise banding present in that area.

What I see in that area looks more like JPEG compression artifacts.

Frankly, if one needs confirmation from someone else on whether there is banding or not in a shot, then I'd imagine banding isn't a problem for that shot.

If pcarfan's files are scaled down and compressed, what we might be seeing here are just side-effects of the algorithms used for scaling down and compressing the images. I've looked at his 3200 shot in original size and banding isn't half as obvious as it is in his OP.

[JohnBee]

I would flaunt it! That is a beautiful shot at ISO 4000! I respect how well you fought this one! It was worth whatever tweeking you did.
I would love to sit next to you as you work your magic cleaning up an image like this. I am one of those guys who learn better by watchin' than readin'

Thanks for the kind words.
I would be more than happy to share the workflow I did on this image with you.
Though I don't know if I'd have the time to make a video or anything, but I should be able to write something up with screen grabs(at least).

What I see in that area looks more like JPEG compression artifacts.

Frankly, if one needs confirmation from someone else on whether there is banding or not in a shot, then I'd imagine banding isn't a problem for that shot.

If pcarfan's files are scaled down and compressed, what we might be seeing here are just side-effects of the algorithms used for scaling down and compressing the images. I've looked at his 3200 shot in original size and banding isn't half as obvious as it is in his OP.

As with all things relative, noise, and other IQ aspect will ultimately come down to personal choice.
As for the banding, I really have to say, I've been dealing with it long enough to be affected(spotting it) by it, rather quickly.
Though whether or not it is acceptable to others, is of course entirely subjective.

PS. on the other hand, I was not able to the original images, do you have a link? Perhaps it's time to get my glasses changes

[Rondec]

I remember a monkey photo that Pcarfan posted that was pushed a couple of stops to, I think, roughly 12,800 iso equivalent that was very impressive and i don't remember any banding on that one. Of course, for real image quality and dynamic range, you need to keep your iso low...

[JohnBee]

I remember a monkey photo that Pcarfan posted that was pushed a couple of stops to, I think, roughly 12,800 iso equivalent that was very impressive and i don't remember any banding on that one. Of course, for real image quality and dynamic range, you need to keep your iso low...

Yea, I've seen many high ISO shots without it too.
I've been trying without success to identify what conditions invoke this behavior in the sensor, and everytime I thought I was on to something, I'd be proven wrong.
It seems to appear in both dark and light light areas, just as in short and long shutters.
It is really an odd phenomenon to say the least.

[Laurentiu Cristofor]

PS. on the other hand, I was not able to the original images, do you have a link? Perhaps it's time to get my glasses changes

Check pcarfan's answer to this thread - it's right on the first page.

Also, have a look at the samples of banding posted by others in immdiate proximity to his post. I can clearly see banding in several shots, the K20D dpreview ones being some of the most obvious. But it seems to me there's a world of difference between those K20D shots from dpreview and pcarfan's shots.

Let me know what you think.

[JohnBee]

Check pcarfan's answer to this thread - it's right on the first page.

Also, have a look at the samples of banding posted by others in immdiate proximity to his post. I can clearly see banding in several shots, the K20D dpreview ones being some of the most obvious. But it seems to me there's a world of difference between those K20D shots from dpreview and pcarfan's shots.

Let me know what you think.

Okay thanks for that.
I've looked them over and it's all good(no glasses needed )
I've found banding in all three full res shots as well.

I agree on the banding being affected in resize as well.
However, I'm not at all convinced were looking at image distortion(not so fancy).
Given that the patterns are consistent(with whats in the scene), I'd say were most likely looking at the equivalent of stepping back and getting a better picture. Having said that... the banding is well and accounted for in all three images at 100% as well as the resized ones. However, it is rather muted in the ISO1600 shot and since it's a widespread pattern, it makes it harder to identify at 100%. However, we rarely(if ever) view images at 1:1 so... I'm not sure how helpful it is to assess banding at this level tbh.

And I won't argue on the banding phenomenon in other models too ie.K20(and not just Pentax), though it certainly does seem to creep in earlier on the K7 models than that of the K20's. Especially when working with shadow recovery etc.

Another interesting phenomenon in the samples you linked here, is that I'm seeing rather strong wavelet(NR) patterns at 100% in both the ISO3200 and 6400 images.
Which might be a sign that conditions could improve if these were shot in the RAW.

For me(can't speak for others), banding is the worst high ISO bi-product there is due to the difficulties it poses in removing it(see all channels).

I would flaunt it! That is a beautiful shot at ISO 4000! I respect how well you fought this one! It was worth whatever tweeking you did.
I would love to sit next to you as you work your magic cleaning up an image like this. I am one of those guys who learn better by watchin' than readin'

I would be more than happy to share this information with you and anyone else who would cares within the community.

[Mr.Turnip]

... and I ended up with this Kx and I am singing the Hi ISO Hallelujah chorus. It is shocking to me how good the jpegs are.

But it is an entry level, crawl thru the menu cam.

Are you aware that you can access almost all of the functions on the Kx (in shooting mode) by pressing the [INFO]-button?
In that menu you can use the scroll wheel for selecting features...that's really only 'one step'...

Are you aware that you can access almost all of the functions on the Kx (in shooting mode) by pressing the [INFO]-button?
In that menu you can use the scroll wheel for selecting features...that's really only 'one step'...

Dang I really got to read this manual. You mean I can change metering between spot center matrix with the info / thumbwheel? I have been bangin back and forth thru the darn menu. A real pain...

YEP THERE IT IS

Thank you Claus >>> I really have to read the manual!!!

Hahaha what an idiot I am

Next time I am giving any advice or giving anyone any grief... Please link to this post!

[cbaytan]

just FYI..

A guy from PentaxTurkey.net forum asked to Pentax why K-x high ISO performance is lot better than K-7 and if there will be a firmware update for K-7 for high ISO noise problem, here is the official answer (I've bolded the juice part), so K-7 owners can get the hint:


__________________
Dear Cagatay,

Thank you for contacting Pentax.

The K-7 uses a higher resolution sensor than the K-x, but the sensors are the same size. That means the K-7's pixel density is higher than the K-x, naturally leading to more noise.

Additionally, the K-7 owners are assumed to understand the benefit and necessity of shooting at lower ISO. Therefore, the Pentax engineers did everything possible to maximize the performance of the camera at lower ISO settings. This includes giving the camera a wider dynamic range at these settings.

The K-x is an entry level dslr and its owner is more likely to shoot at a wider range of ISO settings, including high ISO settings. Therefore, the engineers tuned the processor specifically to respond better at higher ISO settings at the sacrifice of dynamic range.


If you are in need of further assistance, please respond to this email or call our technical support center at 800-877-0155.

Sincerely,
Randall Schumacher
Pentax Imaging Technical Support