[Curious]

A friend of mine is looking into getting her first DSLR, and asked me to look at a few cameras for her to help her pick one. Of course, along with the cameras she was considering, I wanted to recommend she look at Pentax (she was only considering Canon and Nikon). However, as I was looking at test images, I noticed that while other brands tend to avoid them, Pentax cameras tend to produce jaggies. I figured this was because of the absence of the AA filter on the K-3, but the K-5 also produces jaggies where others don't. I'm left without an explanation, and it makes it hard for me to give a strong recommendation for Pentax (in terms of image quality). One of her top choices is the Nikon D5300, and it really produces amazing photos (see attached - D5300 left and K-3 right), and even while it doesn't have an AA filter, it seems to be completely free of jaggies. Whats the deal? Why does Pentax seem to have this unique problem?


Thanks for any insight on this.

[Steve.Ledger]

Whats the deal? Why does Pentax seem to have this unique problem?

Never seen this.
Are you viewing all photos at 100%?
The video system of the computer display will create jaggies when not viewed at 100% (for example when fit to screen) or the software has no image smoothing.
(Try viewing the same images using FASTSTONE Image Viewer with smoothing enabled)

BTW: Nothing is attached to your post.

[Adam]

A friend of mine is looking into getting her first DSLR, and asked me to look at a few cameras for her to help her pick one. Of course, along with the cameras she was considering, I wanted to recommend she look at Pentax (she was only considering Canon and Nikon). However, as I was looking at test images, I noticed that while other brands tend to avoid them, Pentax cameras tend to produce jaggies. I figured this was because of the absence of the AA filter on the K-3, but the K-5 also produces jaggies where others don't. I'm left without an explanation, and it makes it hard for me to give a strong recommendation for Pentax (in terms of image quality). One of her top choices is the Nikon D5300, and it really produces amazing photos (see attached - D5300 left and K-3 right), and even while it doesn't have an AA filter, it seems to be completely free of jaggies. Whats the deal? Why does Pentax seem to have this unique problem?


Thanks for any insight on this.

To me it looks like JPEG compression (or scaling in post) is to blame in the photo on the right. In any case, both cameras are more than capable of taking great photos. I think lens selection, features, and handling should be the primary things to look at when making your camera choice, since the sensors used by post Pentax and Nikon these days are about the best you can get for APS-C.

[Steve.Ledger]

To me it looks like JPEG compression (or scaling in post) is to blame in the photo on the right. .

Adam, nothing is displayed for me. If I try to copy the image URL I get;
"Invalid Attachment specified. If you have just posted a new thread in the Marketplace and are a new user, it will be made visible shortly. If you followed a valid link, please notify the administrator."

[Curious]

Both of the images are 100% crops taken from imaging resource's comparometer. You may be right about JPEG compression, but the guys at imaging resource are pretty meticulous at treating all tests the same, so if it is a result of JPEG compression, wouldn't it be present in both images (assuming the photos were treated in the exact same way)?

Imaging Resource "Comparometer"

[Curious]

Is this visible for you Steve?

---------- Post added 10-18-14 at 07:59 PM ----------

I think lens selection, features, and handling should be the primary things to look at when making your camera choice, since the sensors used by post Pentax and Nikon these days are about the best you can get for APS-C.

I agree with what you on that Adam. And in that regard, the K-3 wins hands down. It is just that I thought that the K-3's image quality would at least match that of the Nikon, but if this is indeed a problem, I would consider it a fairly big negative. I am mostly just trying to understand why this would happen in one and not the other (given the close similarity of the sensors). Is it possibly that Nikon has some in camera processing that does a good job of clearing this up?

[Adam]

. I am mostly just trying to understand why this would happen in one and not the other (given the close similarity of the sensors). Is it possibly that Nikon has some in camera processing that does a good job of clearing this up?

I think that perhaps the test photos above are flawed, especially if they were not captured at the same time or under identical conditions. If you'd like to take a look at some more full-size studio samples from the K-3, we have some below:
Index of /645z/files/studio/k3

We also have plenty of real-world RAWs and JPEGs on this page:
Review: Pentax K-3 - Sample Photos | PentaxForums.com Reviews

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

[stevebrot]

I am mostly just trying to understand why this would happen in one and not the other (given the close similarity of the sensors).

I went back to the original images on the Imaging Resources site and I believe the "jaggies" you are referring to are moire'. This is expected for a camera without a low-pass filter. Compare the D5300 with the D7100 (same sensor as K-3). The D7100 has moire' similar to the K-3. Why the D5300 has less with this subject is anybody's guess. BTW, I don't believe that the K-3 and D5300 photos are actually in focus for the items the OP is concerned about. This normally would not make much difference since the still life set is designed to evaluate color fidelity and noise characteristics.


D5300 (ISO 100)

K-3 (ISO 100)

D7100 (ISO 100)


Steve

[rawr]

It is just that I thought that the K-3's image quality would at least match that of the Nikon, but if this is indeed a problem, I would consider it a fairly big negative.

Remember the IR images are default JPEGs. They set everything at the camera defaults as a matter of course. Which means the results are all adjustable by the end user - JPEG compression levels, picture styles, etc. The Pentax defaults will probably be different from the Nikon defaults.

And could you pls draw a circle on those images to indicate the location of some K-3 jaggies to me? I see soft JPEG compression blur, for example on the brush tips, but not jaggies. Perhaps if I looked at 300% I might see jaggies. Jaggies are where all smoothness breaks down and you see little chunky 'steps' in the image.

[stevebrot]

Remember the IR images are default JPEGs.

Yes, this is true. I also found that it makes a difference what monitor is used to view the images.

I believe that the "jaggies" the OP is concerned about are in the middle of some of the bristles of the brown brush and in the small graduations of the circular slide rule. They are not true jaggies, but appear to be an interference pattern (moire').


Steve

[Curious]

Remember the IR images are default JPEGs. They set everything at the camera defaults as a matter of course. Which means the results are all adjustable by the end user - JPEG compression levels, picture styles, etc. The Pentax defaults will probably be different from the Nikon defaults.

So, what you are saying is that with the default settings, the Nikon just has more anti-aliasing and possibly lower JPEG compression levels? And by adjusting the settings on a K-3, similar results could be achieved? I shoot in RAW anyway, so if this is truly just a result of JPEG compression, then it doesn't bother me. My friend on the other hand...

I'll have to look back at some of my K-5 photos though, because I'm pretty sure that even in RAW, leaves in the background have quite a bit of aliasing.

could you pls draw a circle on those images to indicate the location of some K-3 jaggies to me? I see soft JPEG compression blur, for example on the brush tips, but not jaggies. Perhaps if I looked at 300% I might see jaggies. Jaggies are where all smoothness breaks down and you see little chunky 'steps' in the image.

---------- Post added 10-18-14 at 10:09 PM ----------

Admittedly, these are not severe 'jaggies', but in comparison to the clean lines that the Nikon puts out, it is definitely less appealing.

[stevebrot]

You might want to also consider that the D5300 photo was taken with a different lens (Nikkor 60mm macro) than the K-3 and D7100 photos (Sigma 70mm macro).


Steve

[LensBeginner]

The image on the right also looks a bit soft to me.
Misfocused? different aperture? worse lens?

[bobpur]

The Comp. is a good reference point but don't take it too seriously, you are always better to judge by numerous user reviews than
any photo site.
I bought the Olympus EM-5 ( since sold ) based on some and found out otherwise, very bad raw chroma noise even at base Iso. It
was said to be great at High Iso !

[interested_observer]

Just to add a thought here on the general topic of "Jaggies", several months ago I was over at Costco picking up a couple of test 20x30 prints (can't beat the price). I really wanted to see them, so at the counter I pulled them out of the plastic sleeves and unrolled them. About 10 seconds into my gaze - I was thinking that damm they look good, I heard a voice off my shoulder say that "normally I don't like blue cars, but that is a tremendous image". I started talking with the gentleman, who turned out was a retired professional product photographer. He was in printing some images of his neighbor's cat who recently passed away. He was giving her some pictures of her pet, out stalking prey in his backyard.

During the conversation, he added - "but the Jaggies". Yup - they were there, just in the same places as the ones in this post. I stood there dumbfound. I was engaged in the conversation with him, and not really looking at the images, while he was looking at the prints. These were the 2nd and 3rd large prints that I had ever done - just too see what they looked like. They were not in the RAW images, I was sure, I know I checked. After post processing, I had found some purple fringing (that the 12-24 is famous for). So, I went back into LightRoom 4 and fixed the CA - and for good measure just applied the lens correction. He looked at me and said - that's where he would start looking for the culprit.

So, back home I went to look at each step of the process. The RAWs were clean. The tone mapping output was clean (shot in a very dark, heavily back-lit old car museum/factory, so I bracketed). I walked through the LightRoom processing and there going into the lens correction - all was fine, however its result just spread them all over the place - small, but they were there. They were most noticeable in the areas of light to dark transitions. I really did not need the lens correction - it didn't add anything that was really desirable.

To me that was a very valuable lesson - all for $8.99. I was going to toss the print, but my son said that he wanted it for his office wall at work. From a distance it looked great.