![]() |
Test Photo (click to enlarge) Nikon Z7; 24-70mm f/4 S lens @ 24mm; 1/125th Sec. @ f/8; ISO 64 The area shown in the images below is the small island of dead grass in the right center of the frame. |
Revolutionary in-camera features have already crept in to our current crop of cameras. Features such as Hi-Res imaging where the camera makes an exposure then moves the sensor by a very tiny amount (1 pixel or less), makes another and another, then automatically combines as many as 16 of those image files into a single high resolution final image file (in some cameras you have to later combine the images in editing software), an electronic neutral density filter that can create water blur like would be achieved using a glass neutral density filter, focus bracketing and focus stacking (combining the focused bracketed images in-camera), the ability to buffer numerous images upon half press of the shutter then write them to the memory card once the shutter button is fully pressed so you never miss peak action, eye-tracking autofocus, animal eye-tracking autofocus and on and on There are others but you get my point. Some of these features may have employed machine learning or artificial intelligence (AI) but they have not been labeled nor marketed that way. It seems to me these kinds of features are so amazing that they could be.
Additionally, after the image files have been made, there are now software features given us by image editing software companies that allow us to automatically combine images to make panoramas or high dynamic range composites, combine multiple images for extreme depth of field, move and/or remove objects, etc. Now it seems AI or machine learning is emerging in editing software as well.
Frankly, I am amazed at how much photography has changed in the past several years! However, are all of these features and abilities a good thing or is it too much? I think each of us has to decide how (and I'll use the word which may or may not be the best but I can't come up with a better one right now) "pure" you want your photography and photographic processes to be. Do you keep things simple or incorporate every new technology into your photography or are you somewhere in-between? As for me, it "depends" as is the answer for so many photographic questions.
I've been a fan of Topaz Labs plug-ins for a long time. Topaz Labs has provided some valuable and innovative plug-ins to Lightroom and Photoshop—at least for my editing. Now, they have introduced a new generation of plug-ins with the suffix "AI." Obviously, they refer to the use of "artificial intelligence" in their new products. I can't judge whether or not the company is taking some liberties with their product naming, but I do know that the AI products I've tried so far have been terrific.
The newest one I've tried (and subsequently purchased) is Sharpen AI. That is the one I want to highlight in this post as I think it has a very useful and practical application in potentially making your lenses better and more useful. I won't go into a full description of what capabilities Sharpen AI has, you can read about it here or here (in fact if you are interested in reading about a couple of Topaz Labs other AI plug-ins reviewed by an accomplished professional photographer, look here and scroll down to the posts about the Topaz products). Sharpen AI does have three different settings, however, Sharpen, Stabilize and Focus. In my experience with the plug-in, most of the improvement is done pretty much automatically with no manual adjustments needed by me. Even better!
One aspect I've tried with Sharpen AI and was pleased is that it appears that the plug-in can negate at least most diffraction softening of your images when you need stop down your lens for additional depth of field. If you have been hesitant to stop down your lens past a certain point (say to f/11 for m4/3 or f/16 for APS-C or f/22 for full frame) because of physics of light and lens design which results in your images not being as sharp as at the best apertures, Sharpening AI may help.
I've done some tests on several of my images using a couple different sensor sizes and am providing a couple of examples here. These particular tests (above) were done with files from my Nikon Z7 and the new Nikon 24-70mm f/4 S lens. The ones below with the Nikon 14-30mm f/4 S lens @ 14mm, which I think is its weakest focal length. Both lenses, however, are overall excellent performers.
As with almost every consumer lens, image edges are not quite as sharp as the centers and very small apertures cause softening of the overall image. The problem I found in my initial testing with both of these lenses is that at f/16 the sharpness just starts to fall off and at f/22 it falls off even more. Looking back at my notes for both lenses indicates that I wouldn't use f/22 unless there is a dire need. Generally, too soft. This is due to the extreme bending of the light coming through a very tiny aperture opening resulting in what we call diffraction softening. After my preliminary testing with Topaz Sharpen AI, I now feel comfortable using any aperture in any of my lenses. That may be a premature statement, but I feel pretty good about this software's ability to improve a lens' performance.
Posted here are examples. I took the one of the sharpest apertures, f/8 and compared the sharpness to an identical image made at f/22. It is apparent that the f/8 image is much sharper. I then edited the f/22 image in Topaz Sharpen AI and compared the before sharpening image to the image after it was sharpened. The improvement of the sharpness is remarkable; not identical, but pretty darn close; close enough that you would have to pixel peep and look very closely to see a difference. I did this at the center and the edge of the frame.
The below images were made using the Nikon 14-30mm f/4 S lens.
![]() |
Edge of frame. Left is f/22 just edited in Lightroom and the right identically edited but also run through Topaz Sharpen AI. (click to enlarge) |
![]() |
Same as above but at the center of frame; f/22 before and after editing in Sharpen AI. (click to enlarge) |
![]() |
Comparison at edge of frame; f/8 (left) and f/22 (right). Obviously, the f/8 image is natively much sharper. (click to enlarge) |
![]() |
Same as above but now the right hand image has been processed in Topaz Sharpen AI. A remarkable difference, in my opinion. (click to enlarge) |
Will I be using Sharpen AI in the future? You bet! For all of my images? No, just for the ones where I think it will improve the image to meet my vision for the image.
Another company that is now implementing artificial intelligence is Skylum Software. They are best known for their software Luminar and Aurora HDR. Skylum provides digital image editing software that incorporates AI abilities into their "filters," as they call them. Filters is another name for the sets of adjustment sliders that impact the image in specific ways.
In the current version of Luminar they incorporate artificial intelligence to help you improve the overall image (Accent AI) by "intelligently" adjusting contrast, color, saturation, clarity and luminance all with one slider. Also, the AI Sky Enhancer filter acts just on a sky to improve its color, contrast, saturation, etc. I am a Luminar user and I think what these two filters provide is quite remarkable. But that isn't the end of the story.
Today, Skylum is introducing a new version of Luminar, called Luminar 4. In this new version they are taking the idea of using artificial intelligence to the next level by adding a number of new machine learning capabilities. I have not had a chance to try them out but here is what is coming. After I get familiar with these new AI filters, I'll let you know what I think.
Along with Accent AI (being renamed to AI Accent?) and AI Sky Enhancer, they are adding AI Structure, AI Sky Replacement (yes, from what I've seen on YouTube it will be very easy to replace an entire sky and re-light your scene with just about one click, not that I will want to do that), AI Skin Enhancer and AI Portrait Enhancer. All of these new capabilities are said to use machine learning technology. Also, although Skylum has not attached the AI label to them, two other adjustments are quite remarkable. They are "Sunrays" and "Smart Contrast."
I think you will see more of these types of advances from the other image editing software companies in the future. I believe Skylum has set the bar higher for image editing software and the other major companies will have to match them and even add more similar capabilities. Whether or not these types of features do too much and change the digital image more than you feel is "true" to the original subject is up to you to decide.
As we move forward with our digitally dominated craft, I think that there will be some very interesting things coming for us to consider, both in-camera and on the software side of things. Personally, I can't wait to find out what the future holds for us digital photographers.
DISCLAIMER: I am not sponsored nor supported by Topaz Labs, Skylum or any other company nor do I get any products for free or at a discount. I do wait for companies to offer their products at a discounted price to the general public before I buy, but I pay what you pay. Also, I have no affiliate links and get no commissions from any company.
Join me over at Instagram @dennisamook or my website, www.dennismook.com.
Thanks for looking. Enjoy!
Dennis A. Mook
All content on this blog is © 2013-2019 Dennis A. Mook. All Rights Reserved. Feel free to point to this blog from your website with full attribution. Permission may be granted for commercial use. Please contact Mr. Mook to discuss permission to reproduce the blog posts and/or images.
No comments:
Post a Comment