I have been using your latest version of Topaz AI V2. I do like most of the improvements.
However I don’t like the Adjust Lighting or balance. When I used the lighting it turns my photos Green, this isn’t adjusting the lighting its adjusting the colour on the whole photo. Lighting should just adjust the brightness either darker or brighter is it not? as for the colour balance the AI doesn’t really adjust anything, from what I have noticed I have to do this myself. Maybe some of the other users have better results. If I do get better reults on other photos I will be happy to let you know. However I will say I do like Topaz overall and It has even fixed quite a few of my really bad blurry images.
One other thing I think Topaz could do is when we crop if our horizon needs to be straighted, could this be included in the the crop feature?
I’m not sure where you want comments about the Noise Remove Standard V2 and the Sharpening Standard V2, but…
So far, especially on images where I’m having to upscale 2X, I think they BOTH work better than the old standard versions. They seem a bit less twitchy than the old Standard versions, and so far I’m not getting halos when I sharpen.
AND, the ability to brighten an underexposed image while processing in Photo AI (as long as I remember to turn it off when I process the image and send it back to Lightroom), is nice. It makes it a little easier to see how my adjustments are affecting the image.
I’ve noticed a bug making a comeback. If you crop a photo directly in photo AI and then export the new, cropped image, it will retain the same thumbnail as when it was uncropped. But it’s beautifully cropped once opened in any image viewer, albeit with a slight bug. On opening, the viewer will show the uncropped image for 1/2 second and then correctly display the cropped image.
So basically the answer is no. I can customize how they work, but if I wish to disable it altogether, I still cannot. In your screen capture, I do not see the lens blur model. I really do need to disable that model altogether.
New adjustments of lighting and color are very slow. They are not fit to work with real files in my opinion. Today, large photoshop files is not an exception. So, waiting for a minute or two after an adjustment until the file updates is counterproductive IMHO.
I don’t know if you have Adobe Creative Cloud, but the Photoshop Beta can be a good step at restoring older photos. The neural filters do an excellent job restoring photos. The Color Halftone filter reduces moiré patterns, and the Scratch Reduction does a good job removing scratches and dust particles. There are some other restoration tools as well, but I find those are the ones that I use most for my particular photos. Pair it with Photo AI and your results will turn out very nice.
Thank you, I do, though I usually mostly use Colorize. I’ll have to test some other filters going forward.
I’m also enjoying Generative Fill using text-to-image from scratch or for expanding Stable Diffusion renders (which are enlarged in Photo AI either before or after!).
Lost the ability to revert to auto-pilot settings.
Sharpen Standard v2 severely over-sharpens images. I find I need to dial it back by at least half.
While I appreciate that the denoise model will fix some minor blur and the sharpen model will reduce some minor noise, I find that if both models are used each are active anyway. This makes no sense and there is no logic as to how to correct it.
Lighting and color adjustments are of zero value to anyone working with RAW images. They are extremely gross adjustments and does not obviate the need to edit in Lightroom or other editing programs. Combine that with the inability to export the image as a DNG and those features are less than useless.
Here’s a sample that I did in ten minutes this morning, combing Photoshop Beta and Topaz Photo AI. I’m still trying to get the order of each program right, but the results are pretty good, considering most of the time was taken up getting the Generative AI to look plausible.
That’s great! Reminds me of some manual retouches I did about 25 years ago in PS for a local wedding photographer. She was shooting medium format and had the little paper strip that holds the roll together stuck in the film plane, leaving a diagonal rectangular blank spot on each image…
I was able to reconstruct the bride’s dress, flowers and veil. AI would have been perfect for this and so much easier!
Guess what, I just got a notification for a very highly produced YouTube ad re: capturing images & using Photo AI for processing. In it, the text testimonial quoted, “Over 1 million ‘Costumers’ are using Photo AI.” I didn’t realize the people who design costumes would be such a huge market for them… that too is impressive.
The reference to over 1 million “Costumers” using Topaz Photo AI follows spoken testimonials by a number of guys. I guess mostly “costumers” (ie, costume designers) & guys use the software.
