[davidreilly3207]

Today was very overcast and went to test my new K-3 at our local wildlife refuge. The attached photo was a RAW DNG taken at 1/1000, f8, iso1600 by a Tamron Adaptall 500mm f/8 SP Mirror (55B). No post processing except cropping including setting sharpening and noise reduction to zero in Lightroom 5. There appears to be excessive noise. I was expecting better from all the samples I've viewed in the reviews.

1. Does my photo exhibit higher than normal noise or are the samples in the reviews post-processed?
2. Can an individual camera be defective in this regard?

What do you think?

Attached Images

View Picture EXIF

File Date

 

[LaurenOE]

Today was very overcast and went to test my new K-3 at our local wildlife refuge. The attached photo was a RAW DNG taken at 1/1000, f8, iso1600 by a Tamron Adaptall 500mm f/8 SP Mirror (55B). No post processing except cropping including setting sharpening and noise reduction to zero in Lightroom 5. There appears to be excessive noise. I was expecting better from all the samples I've viewed in the reviews.

1. Does my photo exhibit higher than normal noise or are the samples in the reviews post-processed?
2. Can an individual camera be defective in this regard?

What do you think?

Any mirror/cat lens, and especially one that uses an older technology like a Tamron Adaptall system will not be as sharp.
With the resolution of the K3, you are going to see everything that is wrong with mirror and cat lenses.
ISO 1600 is also near the end of where the K3 produces clean images, but in your case, I attribute most of what you see to the lens.

With the K3, good glass has finally become critical.

[LaurenOE]

Also, with all the good lenses in your signature, why would you be testing out noise with a 20 year old cat/mirror lens? That doesn't make too much sense.

[Painter]

I guess I'm looking at it at 100% on my iPad it looks like you've turned down the chroma noise which I think you mentioned. That cleans right up easily it doesn't seem out of line though but I think the optics aren't helping with the evaluation.

[ChrisJ]

Your not giving the K3 a chance to show what it can do with this lens, they're just simple (and cheap) Schmidt Cassegrain mirror lenses which aren't even colour corrected which is why they are never truly sharp, your sensor may well be out resolving it and, because of the small aperture you are probably getting diffraction too. All these will exaggerate any noise in the image.

Any cheap 500mm refractor would be better.

Chris

[Miguel]

What do you think?

I think I'm far more interested in the same shot with your post-processing skills applied. I've got about 140,000 images in Lightroom, say 80% RAW. I don't think I like any of them much with all settings set to 0.

M

[davidreilly3207]

I wasn't testing for noise but the distance to the birds required the 500mm which I have had better results with lower isos on my K20D. I also shot at iso 100 with my DA*50-135mm (Tamron 70-200 is away at Tamron for service) and the were good. Hopefully on better lit days shots with the 500mm will improve.
I'll be shooting a play tomorrow with the DA*50-135mm where iso will probably have to be in the 400-800 range for adequate shutter speeds. I will update this post afterwards.

---

Last edited by davidreilly3207; 11-22-2013 at 07:00 PM.

[stevebrot]

I see the noise you are referring to and it does seem excessive based on the 1600 ISO in the EXIF as compared with published photos from the K3 at the same setting. Is there a reason why you defeated the default noise reduction on import of the DNG?

The lens (sharp or not) is not pertinent to the question..."noise in the channel", so to speak.


Steve

[Ole]

Viewed at 100% crop (pixel peeping) the K-3 images are noiser at high ISO than the K-5 as explained in PentaxForums' in-depth review. Scaled down to the resolution of the K-5 the difference diminishes/vanishes, however. Which means that the K-3 is every bit as good as the K-5 for equally sized images.

[stevebrot]

Any cheap 500mm refractor would be better.

Are you familiar with the Tamron SP 500? While not as good as the second version (55BB), the OP's lens was no slouch when it was made (originally released ~1980) and is a much different beast than a Russian MTO or the usual no-name cat. As for "color corrected", I think you may be referring to poor correction of chromatic aberration. Some is visible in the posted sample, but considering that it is a full resolution crop from a 24 megapixel sensor, the CA is not excessive even by current standards. Diffraction issues at f/8? Well, I guess we should all pack up our long lenses and go home.


Steve

(...simply goes ballistic when people dis these quality older lenses...)

BTW...specs and Modern Photog test results (done with film) available at Tamron SP Adaptall-2 500mm F/8 Model 55B


...

[normhead]

Looking at my DA* 60-250 images, I have no idea what's going on there.

Working today with the DA*60-250… some serious cropping going on here.
1600 ISO, grainy but bearable, F4, Shutters from 1/40 - 1/60, too slow for these birds. But I got a few.
There is no noise reduction on these first four images.









And the last one with selective noise reduction applied.


From this thread
https://www.pentaxforums.com/forums/172-pentax-k-3/241198-k3-sample-shots-pos...ml#post2587056

[dosdan]

Viewed at 100% crop (pixel peeping) the K-3 images are noisier at high ISO than the K-5 as explained in PentaxForums' in-depth review. Scaled down to the resolution of the K-5 the difference diminishes/vanishes, however. Which means that the K-3 is every bit as good as the K-5 for equally sized images.

That's right. When showing the amount of visible noise in an image you need to also mention what the reduction in original image area (if any), through scaling and cropping, is. (That's besides any in-camera or in-PP NR applied.) For example, a 4500x3000 original image will look less noisy if you downsize it for web display to say 1500x1000. But if you crop it so that 1500x1000 of the original pixels are shown in a 1500x1000 image (i.e. 100% crop), then such an image from a small-sensel APS-C 24MP sensor is noisier than the one from a larger-sensel APS-C 16MP sensor.

See https://www.pentaxforums.com/forums/16-pentax-news-rumors/238970-pentax-k-3-h...ml#post2534630

Dan.

[normhead]

That's right. Any image demonstrating noise performance in an image need to mention what the reduction in area (if any), through scaling and cropping, is. For example, a 4500x3000 original image will look less noisey if you downsize it for web display to say 1500x1000. But if you crop it so that 1500x1000 of the original pixels are shown in a 1500x1000 image (i.e. 100% crop) then a small-sensel APS-C 24MP sensor is noisier than a larger-sensel APS-C 16MP sensor.

See https://www.pentaxforums.com/forums/16-pentax-news-rumors/238970-pentax-k-3-h...ml#post2534630

Dan.

I'd also point out that the noise is much more noticeable in out of focus areas. If that is all you have in your image then... but, I've also accepted that 1600 ISO on my K-3 is about the same as 3200 ISO on my K-5. It doesn't mean you won't get a good image, it means you'll get fewer good images. I personally like 800 ISO images best, on both cameras, and try and shoot at 100 ISO whenever possible. K-3 images don't get really ugly until you get to 12800 ISO. Up until then, you might tweak out a keeper or two.

[rawr]

Norm, a couple questions: what's the crop level to those images? How do you do your NR? And do you do apply any masking to your sharpening? If not, it's something to perhaps look into.

BTW, DPReview has the K-3 in it's image test scene now. Here's what the K-3 RAW looks like against the D7100/D610/K500 at ISO 6400. Noise seems well controlled and detail remains high.

So I'm still puzzled by all the noise talk. Maybe we need to wait for all the RAW processors to properly manage K-3 RAW's before we close the book on K-3 noise levels in JPG.

[normhead]

Hey rawr, no noise reduction applied... pretty much straight off the camera. The image went through my normal sharpening (and contrast) routine, which makes the noise worse.The bottom image, which is my normal noise treatment is masked in a sense, but technically masking is not possible in Aperture so it's a work around. I'm not making any serious decisions until the Aperture profiles arrived, but that might take a while. I received the latest update from Apple a couple days ago, and the K-3 wasn't on it.

Oh and the images were cropped from between 1/3 to 2/3 of their original size, before downsizing to 1024, By the way, I'm quite happy with these images, for 1600 ISO. That being said, I never felt I could reliably shoot 1600 ISO on my K-5. I don't like 1600 ISO, I don't like shooting over 800 ISO on either camera, and if I could have reduced the images to 400 ISO I would have been down in my comfort zone. I just didn't have enough light. To me, no 1600 ISO image is as good as the same image shot at 100 or 200 ISO everything else being equal. So with any 1600 ISO image shot on any camera, you could have done better.