Heather Iles Loc: UK, Somerset

Thank you for your response. I will give it a try.

JimH123 Loc: Morgan Hill, CA

I should mention that the image you used for the CIZ experiment is the type that CIZ works great. When it upscales the image to generate the extra pixels, everything in the image is quite predictable. Where CIZ gets in trouble is with very busy images with lots of fine detail. In this case, the algorithms that it uses have trouble constructing a good upscaled image. But most of the time, you don't have an image that busy.

In comparisons I have done in doing the upsizing in SW on the computer, the computer wins. Photoshop has the means to do it. And so does ON1 Resize. Of course, the complexity of the image again plays a part. If the image is real predictable, then they both come out fairly close. But if the image is full of fine, unpredictable detail, the post processing effort will produce better results.

One thing that CIZ is really good at is movies. I have used it with 2.0X CIZ of birds in the distance such as cormorants, and the results are great. It is hard for the eye to see any fine detail problems with a moving image.

gessman Loc: Colorado

Yep, Jim, I get that and I appreciate the consistency in your comments from the first time I heard them and I'm sure anyone wanting to know what you're saying will be miles ahead by listening to you. But, let me give you a hypothetical scenario and you give me your response assuming that none of the aspects of the scenario are subject to change. You have an early learning "straight out of the camera" (SOOC) shooter who has a shot at a Grizzly at 400mm on a snowfield, where there's no foliage or other small detail to get confused, and the shooter, who is not perfectionistic and who has no intention to post process images only has a 210mm lens mounted and nothing bigger in the bag. Shooting Raw and bringing the shots back home to post process after buying and learning a couple of pieces of software is not going to happen. Would you discourage the shooter from using CIZ?

imagemeister Loc: mid east Florida

So very WELL stated ! ......
.

JimH123 Loc: Morgan Hill, CA

No, I wouldn't discourage the shooter to just go for it with CIZ. Plus, post processing is not for everyone.

imagemeister Loc: mid east Florida

I am sure most people who love to PP would love to justify their time and SW expenditure by "judgeing" that computer SW "wins" and gets "better" results than in camera CIZ !

I did testing - not against stand alone SW - because I knew I would never be using that - but against 1.4 and 2X extenders on a Red Belly Woodpecker - not on type written signs - with plenty of random detail. My conclusion is that in a blind test you could maybe tell there was a SLIGHT difference in the appearance of the images - but I could not say one was "better" than the other !
.

gessman Loc: Colorado

Thank you sir. I appreciate your candor as I think some lessor experienced folks inclined to use CIZ may
but are reluctant to do so with the current levels of mystery and criticism. It is true, there is a time and a place for everything, even CIZ apparently. At least, thanks to you and a6k, anyone reading this thread who comes up with some "frankenstein" aspects to their images after using CIZ will understand why and will know how to avoid it in the future. I appreciate your contribution.

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

Which it does very nicely!

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

Me too!

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

I did zero processing on th CIZ image, but I did crop, upsample and applied an automatic tone adjustment to the arw file. I did not adjust the color. I think the whole point of my "test" was to see if CIZ, SOOC, could be equaled or made better using a raw image and upsampling. The CIZ image was done with camera settings for contrast, saturation, sharpening and noise reduction - pretty much as I think Sony intended, since many of the other adjustments and settings are disabled when using CIZ. I think the test was a fair comparison if the best image quality was the goal.

BTW, the goldfish and bike were not mine - they were in the links.

JimH123 Loc: Morgan Hill, CA

This is not to downplay CIZ, but I am going to provide a post processed resized image, before & after. I used an Olympus EM5ii with a 40-150 lens at 155mm (with 2X crop factor, was like using a 250mm lens). Not the $1200 pro version of the lens. Just the slower, consumer version often found for $99 on sale.

The camera is 16M pixels and the original was 4608 x 3456 pixels. I then cropped it down to 2036 x 1311. This is aprox. a 2.5x crop. Then I resized it by 250% using Photoshop's Preserve Details 2.0 which now gives it 5090 x 3278 pixels. This is a bit more zoom than CIZ offers at 2.0X but can be useful to see it doing its thing.

The process I followed was:

1. Load original RAW into Lightroom (Gorilla was at the LA Zoo - and was giving me the evil eye)
2. With no adjustments other than Lightroom's defaults, Crop as shown
3. Send cropped image to Photoshop
4. Resize the image using Preserve Details 2 (using 250% and default settings)
5. Return to Lightroom
6. Run Topaz AI Clear on the image using defaults.

I also returned to step 2 and used Topaz AI Clear on that one too. Then exported step 2 and step 6 as JPEGs. Topaz AI Clear, by the way, is an amazing product in that it reduces noise a bit, and is fabulous on producing a sharper image.

The result is just to do a show and tell of the post processing resize route. No complaint about CIZ. Just showing the alternative.

I am added a second set of images of Canada Geese that were taken in JPEG using a Sony A57 with a Tamron 150-600 with this image taken at 360mm. I do notice some crazy artifacts above the geese in the grass that I don't get when I shoot RAW.

Original image

(Download)

Cropped and then resized image

(Download)

Original Canada Geese

(Download)

Cropped and Resized Canada Geese

(Download)

JimH123 Loc: Morgan Hill, CA

Adding one more example. This was taken using a Sony A57 and a Minolta 500mm Reflex lens and was taken in RAW. Image 1 is the uncropped original. Image 2 is a crop of the center Canada Goose and then resized by 400%. Image 3 is the one above the center with it's wings in the down position. Again, 400% crop.

Adding images 4 & 5. Same camera and lens. 200% Resize using the upper White Pelican.

Image 6 is a turkey vulture resized 200%. Started with a JPEG. Used AI Clear 1st to totally remove the noise, and then resized. This time, it was able to work.

My procedure was modified slightly. To resize by 400%, I had to really knock the noise down, so I used AI Clear on the original and then resized in Photoshop the AI Clear image.

Note: I did attempt to do this on several more JPEG images, but the results just were not as good. The JPEGs tended to have tiny artifacts that the resizing then made worse. The RAW images did not have any artifacts.

What was interesting with these Canada Geese is that they flew in from somewhere circling the small lake that I was standing by several times allowing me opportunity to photograph them multiple times. And finally they landed perhaps only 300 feet from where I was standing. These were hand held using the 500mm reflex lens (750mm equivalent on the 1.5x crop ped Sony A57). I was positioned in absolutely the right spot for the show they put on.

(Download)


(Download)


(Download)


(Download)


(Download)


(Download)

JimH123 Loc: Morgan Hill, CA

Just so you are clear. CIZ will work on ANY lens you can attach to the front of the camera. It is better than just plain cropping as it creates new pixels to back fill so that the image ends up with the same amount of pixels. There is a lot of difference between CIZ 1.4X and CIZ 2.0X in the number of new pixels it has to invent and this can make a difference. For instance, CIZ 2X is creating 3 new pixels for every 1 existing pixel. For 1.4X, it is closer to 1 new pixel for every 3 existing pixels.

The argument I have been making is that if one desires to post process, and preferably shoots in RAW, that you can get even better results.

But if I want to shoot movies of wildlife, I don't hesitate to use CIZ since I can't detect any compromises on moving images.

JimH123 Loc: Morgan Hill, CA

1.4X and 2.0X teleconverters are a mixed bag. There are lenses that are designed to use a specific TC and those are fantastic. And there are generic equivalents that can be tried on most lenses. And for the most part, they don't work so good. I don't own any good lenses with their matching TC as I don't have the budget for them. I do have some generic TCs and I find that I don't get very good results.

CIZ with a decent lens should beat the generic TC's. Plus you don't have to deal with loss of light.

I do have a couple TC's that screw onto the front of the lens. I do find that they work better than the TC's that go between the camera and lens. But they are really heavy. If the camera has image stabilization, the camera does not know about the extra focal length and can miscalculate the amount of sensor movement to use. Plus, all the extra mass on the end of the lens may cause the focus motor to struggle. One of these kind that I find does work well is the Olympus T-CON17 which provides 1.7X. It works best if used with a steady tripod with perhaps manual focusing. These types of TC's were once popular but have fallen out of favor. They were once very expensive and now they are a fraction of the original price. I only use them for experimenting.

CIZ is as good as it gets without doing post processing. Post processing will beat it. Especially if RAW is used. But if you're not into post processing, then it doesn't matter. Just enjoy your CIZ.

imagemeister Loc: mid east Florida

Or paying the expense of specialty software. I would add that the opinion that it is somehow "better" is a very subjective matter - and if indeed it is better in some way , the magnitude of the difference must be quite small - IMO.

IMO, The most pertinent and compelling comparison for MOST people is or should be made against tele-extenders - not bigger/more expensive lenses or the expense (time and money) of specialty PP softwares !

Using CIZ in all respects is FREE.
.