[mrNewt]

K-5 wins that battle... still such a great camera even when compared with the new top of the line K-3.

---

Last edited by mrNewt; 01-10-2014 at 01:50 PM.

The four images again... this time uncropped with both sharpening and NR turned on but no brush effects, only global effects..

To my eye, even cropped to the same size, at 1600 the k-3 is noisier, and since 1600 on the k-5 was my upper limit, and the K-3 goes beyond that, for me personally, 800 ISO will be the top for using the K-3, although I have been sneaking up to 1000 lately in critical situations.

There is alot of color noise in the K-3 picture. That I do not understand. I did a test of my own when I got the K-3, can´t do it again because I sold the K-5 but still have the images. When looking at them, taken on a tripod inside in medium illumination, I do not find that there is more noise in the K-3 images. The noise is more fine graded than the K-5 and above 3200 the K-3 has a smoother noise. Your pictures show the opposite. I wonder how that can be.


K-5 and then K-3
ISO12800 and ISO6400

[normhead]

K-5 wins that battle... still such a great camera even when compared with the new top of the line K-7.

I assume you meant K-3, ya, I guess we couldn't have very thing improved in every release forever. Still good for what I bought it for, but maybe not a K-5 replacement for everything.

[RonHendriks1966]

Does that make a difference in RAW? I usually never pay attention to the jpeg settings.

Some people expect it, some people argue that it's not true...

Oh yes, K-5 also applies noise reduction to RAW images.

[normhead]

There is alot of color noise in the K-3 picture. That I do not understand. I did a test of my own when I got the K-3, can´t do it again because I sold the K-5 but still have the images. When looking at them, taken on a tripod inside in medium illumination, I do not find that there is more noise in the K-3 images. The noise is more fine graded than the K-5 and above 3200 the K-3 has a smoother noise. Your pictures show the opposite. I wonder how that can be.


K-5 and then K-3
ISO12800 and ISO6400

I have seen other cameras show results similar to yours, so I'm a little mystified myself...I'm hoping to be able to figure out if this is just my copy, or copies of both cameras, or if this is a universal kind of thing. Unfortunately, like you, I can only test what I have.

[mrNewt]

I assume you meant K-3, ya, I guess we couldn't have very thing improved in every release forever. Still good for what I bought it for, but maybe not a K-5 replacement for everything.

Oups... yes, I ment K-3... no idea why I typed K-7
Most probably a future replacement will address that issue.

Great camera nonetheless! Those ISO selections is something that I never work with anyway.
I'm a low ISO guy... the lower the better . K-5 for as long as I had it was stuck on 80 ISO.

[normhead]

Oh yes, K-5 also applies noise reduction to RAW images.

Both cameras are set to Auto high ISO noise reduction... maybe i should I've it a try with NR turned off.

Oups... yes, I ment K-3... no idea why I typed K-7
Most probably a future replacement will address that issue.

Great camera nonetheless! Those ISO selections is something that I never work with anyway.
I'm a low ISO guy... the lower the better . K-5 for as long as I had it was stuck on 80 ISO.

Ya, I'm almost theologically opposed to high ISO images, but sometimes I need high ISOs for BiFs.

I may have another go at this tomorrow, and do some lower ISOs as well,

[philbaum]

Thanks Normhead for the thread!

Lightroom V4 and V5 have this neat process for sharpening with the "masking" slider. One presses the alt key, then at the same time slides the masking key till the sharpening is eliminated from non-critical areas. I think that would help clean up the negative space areas in the early k3 shots.

[philbaum]

Both cameras are set to Auto high ISO noise reduction... maybe i should I've it a try with NR turned off.

Where have you been :-). That guy "Ed" who did the videos on the early blur concern on the K3, showed pretty conclusively that one should stay away from the Auto setting for K3 high iso NR. The K3 AUTO setting jumps up the NR aggressively for even relatively low iso's and results in the more blurred images that folks were getting. Turn it off!!!!

I even have it turned off on my K5 as well. LR does a much better job anyway.

phil

[Rondec]

I don't understand it either. This is an iso 3200 photo with no noise reduction in Lightroom 5 (no sharpening either). With K3.

[RonHendriks1966]

I don't understand it either. This is an iso 3200 photo with no noise reduction in Lightroom 5 (no sharpening either). With K3.

Basicly you can make excellent hi-iso images as long as you have some nice light on the spot you want to be the centre off your image.

K-5 II on iso 25.600


I had this image printed on A3 format hanging on my exhibition about womens soccer last year. https://www.facebook.com/fotoexpositieBeNeLeague

[normhead]

Where have you been :-). That guy "Ed" who did the videos on the early blur concern on the K3, showed pretty conclusively that one should stay away from the Auto setting for K3 high iso NR. The K3 AUTO setting jumps up the NR aggressively for even relatively low iso's and results in the more blurred images that folks were getting. Turn it off!!!!

I even have it turned off on my K5 as well. LR does a much better job anyway.

phil

Did you just admit you watch Ed?

Basicly you can make excellent hi-iso images as long as you have some nice light on the spot you want to be the centre off your image.

K-5 II on iso 25.600


I had this image printed on A3 format hanging on my exhibition about womens soccer last year. https://www.facebook.com/fotoexpositieBeNeLeague

I have always found huh ISO works better when you have good light and you are increasing your shutter speed, than when you have low light and really nned it just to get an image.

[derekkite]

Why different noise characteristics? At a given iso, the noise is not a factor of the iso number but of the noise part of the image that is amplified.

At any given level of light, the aperture opening and shutter speed will determine how much light hits the sensor. The iso number is like gain on an amplfier; it takes the signal with noise from the sensor, and amplifies it. If the sensor has seen little light, the signal to noise ratio will be low, lots of noise, and a high iso will be very noisy. If there is little light but the shutter speed is long, exposing the sensor to the light for a long period of time, then amplified by the iso, the noise will be considerably less.

So a 1/8000 f4 iso 800 shot is noisy and awful, where an f4 1/320 iso 800 shot is very nice. The first being very bright, the second being low light.

A stop is not a stop is not a stop. I'm not certain, but I don't think that a stop of aperture equals a stop in shutter speed when it comes to image quality. I'm certain that a stop of ISO does not equal a stop of the other two when it comes to image quality.

The K5 and K3 are similar but different. The curves of signal generation vs noise are different. I have gotten very nice shots in low light with both, and the key to quality even at high iso is a big aperture and a long exposure, especially the long exposure. The K5 would get noisy even at moderately long exposures, but the K3 seems to love longer exposures. A short exposure with a small aperture will give you noise on the K3 even at low iso's, and the fall off is quick, the K5 gives a bit more.

If the K3 is able to give a reasonable shot, the K5 is no better. But there are times when the K5 gives you a reasonable shot but the K3 won't. Not many.

[Class A]

Ok, now there's a difference, using the K-3 at 6400 there is enough noise to produce sharpening artifacts.

Not only are you comparing with different noise reduction applied (in-camera), you also show the K-3 image at higher magnification.

You appear to have cropped out the same sized pixel area for both cameras, which shows pixels at the same size, whereas the K-3 pixels should be smaller by a factor of 2/3 to provide a fair comparison.

There is no question that a smaller pixel (from the K-3) will look worse than a larger pixel (from the K-5) when both are shown at the same size. The question is whether the K-3's larger amount of pixels can compensate the higher per-pixel-noise, accordingly.

Both cameras are set to Auto high ISO noise reduction... maybe i should I've it a try with NR turned off.

Yes, please turn of all in-camera noise reduction.

Please also leave the EXIF data intact. In one comparison you show two crops, one from the K-5 and one from the K-3, but there is no difference in magnification this time. How did you scale down the K-3 image? If you leave the EXIF data intact, we can make sure that no errors happen (things like that happen, as your first post demonstrates).

Making fair comparisons isn't as easy as it sounds. The results from Tjompen1968 look like what I'd expect and are in line with what DxOMark results.

[Class A]

I'm not certain, but I don't think that a stop of aperture equals a stop in shutter speed when it comes to image quality. I'm certain that a stop of ISO does not equal a stop of the other two when it comes to image quality.

Your doubt is justified, because indeed "a stop is a stop" no matter whether it refers to aperture, shutter speed, or ISO.
One stop more always refers to "double amount of light" when it comes to aperture or shutter speed changes.

With respect to ISO, one stop more refers to "double the amplification".

You are correct in assuming that ISO should have no influence on noise, but

1. that would be true only if sensors had no read-out (and thermal) noise whatsever. Since even the phenomanlly good SONY sensors contribute a little bit of noise, raising the ISO is like using a magnifiying glass on that tiny amount of residual noise.
2. raising ISO often goes along with mantaining the same exposure, i.e., you are closing down the aperture and/or increasing the shutter speed accordingly. The latter mean that there will be less photons to collect, so automatically shot noise will increase (with respect to the signal level).

The second effect is the same for all sensors, so if there is a noise difference in high ISO shots then one sees the magnified effect of very small differences in read-out noise (and thermal noise if the exposures are long).