[audiobomber]

EDIT: Just searched, according to this link, the K20d does no NR on raw files, only jpgs. https://www.pentaxforums.com/forums/pentax-dslr-discussion/32501-high-iso-noi...thing-raw.html
I've read the same for the K-5, so I would assume the K-7 doesn't either. So ignore where I talked about NR settings, because I guess they're not doing anything.

Cliffnotes: There is no noise reduction in these images.

Depending on the raw convertor and how it is set, the NR settings may or may not be translated during conversion. Adobe programs can be set up to import the camera settings or not. You can easily tell by shooting a shot with and without. The same goes for the Image Tone settings, sharpness, contrast, saturation etc. The raw convertor may either import the camera settings or not. If the settings are not imported, unprocessed jpegs will be very flat looking.

I use Pentax software for conversion. It's true that the NR setting is not applied to the raw file, but the camera setting will be imported to the jpeg unless I change it in post processing. IOW, I shoot raw, but I generally leave the camera set for Weak NR and it is applied to the final image.

I found the default NR settings on the K-x produced a plasticky look that I detest. I have it set for Low starting at ISO 1600. If I need more, I use Noiseware.

---

Last edited by audiobomber; 05-01-2012 at 08:40 PM.

[Laurentiu Cristofor]

Just some observations:

The 100% crops show that the shots are not focused identically. That will impact comparisons at 100%.

In practice, what will matter more for low light shots is focusing and metering accuracy and not sensor noise. SR effectiveness can matter too.

[alstauffer]

Depending on the raw convertor and how it is set, the NR settings may or may not be translated during conversion. Adobe programs can be set up to import the camera settings or not. You can easily tell by shooting a shot with and without. The same goes for the Image Tone settings, sharpness, contrast, saturation etc. The raw convertor may either import the camera settings or not. If the settings are not imported, unprocessed jpegs will be very flat looking.

I use Pentax software for conversion. It's true that the NR setting is not applied to the raw file, but the camera setting will be imported to the jpeg unless I change it in post processing. IOW, I shoot raw, but I generally leave the camera set for Weak NR and it is applied to the final image.

I found the default NR settings on the K-x produced a plasticky look that I detest. I have it set for Low starting at ISO 1600. If I need more, I use Noiseware.

I will definitely have to look into that, I wasn't aware they did that, and it might explain some differences between what I see in Lightroom while I'm editing, and the final jpg I get at the end.. I actually just set both cameras to that setting - Every photo I take goes through Lightroom (and Topaz if needed), so it doesn't make sense to do it in camera I guess.

Just some observations:

The 100% crops show that the shots are not focused identically. That will impact comparisons at 100%.

In practice, what will matter more for low light shots is focusing and metering accuracy and not sensor noise. SR effectiveness can matter too.

I know they're not focused identically, sorry. You can still get a general idea looking at the OOF background I think..

I don't get your last sentence, though. If that were true then K-7 would outperform K-x in low light/high ISO, and it doesn't? Are you just saying that a noisy picture is better than a misfocused/poorly exposed/blurred picture?

[Marc Sabatella]

As far as in-camera noise reduction, I had it set to medium on both, starting at ISO800. I had assumed (probably incorrectly) that they would be the same, since the bodies came out close to the same time

They are totally different cameras, with totally different sensors and hene totally different JPEG engines. While they might strive to make "medium" have more or less the same overall effect on both, that.s going to be impossible. Much like trying to get "medium" level of speciness mean the same thing for both an Indian vegetable curry dish and and a Vietnamese beef noodle soup.

I didn't want to do anything at all in post-processing because IMO that's no longer comparing cameras.

As long as you do the *same* thing in PP, there is no issue at all. But even if you set NR to the same valuesfor both image in your PP software, it's still the case that you are dealing with different sensors with diffeent characteristics, so the same settings might not actually be expected to provide exactly the same results.

But the advantage of fine tuning the NR to get optimal results for both - not maximum NR, but optimum in terms of your own personal preference for the inventiable tradeoff between noise and detail - is that then you see a teat that resembls what you will actually see in real life. If one camera does better than another with NR turned completely off and no NR applied during PP, that,s all well and good, but is that how you normally shoot? And if so, why, if noise is so important to you that you are going to the trouble of doing the comparison? Why not compare *the actual images you would likely to be using* of the scene, rather than some images processed with settings you'd never use in real life?

EDIT: Just searched, according to this link, the K20d does no NR on raw files, only jpgs.

to be more accurate, the specific form of NR than can be switched on or off via the menu is only for JPEG, and that's true of everymcamera I have ever heard of. However, most cameras also have some amount of NR that is automatically applied even for RAW after a certain ISO that cannot be switched off. Sinceit cannot be switched off, there is also no point in worry about it, but I point this put just in case you later run into what appears to be information stating that your camera *does* apply NR to RAW and you are wondering which statement is in error. the answer is, neither statement is in error; theynare just talking about different types of NR. The type you can switch on or off is JPEG only; the type you cannot turn on or off is also applied to RAW.

So ignore where I talked about NR settings, because I guess they're not doing anything.

except that as some have noted, many raw converters will look at the exif or other metadata, see that you had nr turned pn in cmwa, and default to applying what they judge to be equivalent amounts of NR. You wold have to check the NR settings in your RAW processong and explicitly return them to zero to be sure. But again, I don't find that to be a useful comparison. Why compare images that are not shot and processed the way I wold actually shoot and process them? The trick is is in figuring out how you would actually use each camera - which camera settings and PP settings suit *you* best for each, so you know in your heart of hearts that you are comparing images as you wold actually see them in real life. Of course, that's going to be diffierent for each person, so unfortunately, you can't make such a taste truly objective. You kind of have to choose between *useful* and *objective* in this respect, I'm afraid.

[alstauffer]

You kind of have to choose between *useful* and *objective* in this respect, I'm afraid.

That's what I'm realizing.. and it sounds like mine is somewhere in the middle, so that it's neither objective nor useful

Exam in 30 minutes, I will read the rest more closely and respond afterwards.

[Laurentiu Cristofor]

Are you just saying that a noisy picture is better than a misfocused/poorly exposed/blurred picture?

That's part of it. The other part is that the K-7 can keep up with the K-x up to ISO 800 and even at 1600 the differences are small. So, combined with the better ability to focus, this made me prefer the K-7 - I rarely go beyond 800. In the very rare scenarios where I knew I would shoot constantly at 1600 or higher, I would pick the K-x, but those are very, very rare situations - mirror lens use mainly.

[georgecape]

Here's my K-7 shot at ISO 6400

[Laurentiu Cristofor]

Here's my K-7 shot at ISO 6400

Why did that shot require ISO 6400?

[georgecape]

No good reason - more coincidental.
I was practicing on birds on our patio with the Sigma 170-500, so at a high speed and in TAv I aimed at the pots (all in the shade) and only on developing the bird shots this one popped up at ISO 3200 and +1 EV which I thought was quite amazing?.
Are you saying it does'nt count? I was very chuffed seeing the K-7 gets such a bad rep etc.

[Jasvox]

Try that above shot in low light and see how much noise you have. Heck, bump it down to 3200 or even 1600.

Jason

[georgecape]

Oh, I do regularly and as long as I keep it overexposed and not too much cropping and some noise reduction in LR it turns out just fine.
But the moment I misread and underexpose I get into trouble.
Its still the best camera I ever had.

[luftfluss]

No good reason - more coincidental.
I was practicing on birds on our patio with the Sigma 170-500, so at a high speed and in TAv I aimed at the pots (all in the shade) and only on developing the bird shots this one popped up at ISO 3200 and +1 EV which I thought was quite amazing?.
Are you saying it does'nt count? I was very chuffed seeing the K-7 gets such a bad rep etc.

Image noise is worse in low-light situations than in brightly-lit situations.

The K-7's "bad rep" WRT high-ISO noise is relative to the competition and it's predecessor, which has output that's a little better, but often not noticeably better.

[alikaytan]

I have K7 for about 3 months and if i want a real claer photo i can't go beyond iso400, so i usually go after the light and it doesn't hurt me. I like to chase the light.

[Laurentiu Cristofor]

I have K7 for about 3 months and if i want a real claer photo i can't go beyond iso400, so i usually go after the light and it doesn't hurt me. I like to chase the light.

You will find out in time that you can use higher ISO settings than 400, but you need to get the exposure just right.

Here are a few ISO 1600 shots:

[alikaytan]

You will find out in time that you can use higher ISO settings than 400, but you need to get the exposure just right.

Here are a few ISO 1600 shots:

Thanks for the real inspring shots, they are great!! I'm really amazed that these are ISO 1600 and that much clear and you thought me that i have to study more about the exposure.