Alan Milnes
Well-Known Member
Anyone using this? Interested to hear what people think about it and how they are using it.
I will say, there is not a single photo I take that is so important that I would need to spend money on Cloud Credits. That's really insane. It will be interesting to see if this model survives. I can't imagine it does.I have it and have been a Topaz client for well over a dozen years now. My primary use for it at this point is to access their sharpening modules since it is an agglomeration of what used to be DeNoiseAI and SharpenAI. It also includes what used to be their Clarity and Adjust applications. My own take on PhotoAI is that it an attempt to make their older applications into a one size fits all application with a lot of automated processing.
I am not much of a fan of where they are taking this to be honest. I have a large collection of Topaz applications that have been dropped by the roadside over the years while they integrate the technology into their next do-it-all app that never quite measures up to what they left behind.
Disclaimer for what I am going to say next - I was a beta tester for DxO PureRAW 4 which is DxO's denoise application that is a direct competitor to Topaz DeNoise. Yes I have bias here. DxO's approach to cleaning up noise from high ISO images and applying corrections for lens distortion could not be more different from how Topaz approaches the same task. DxO tests every camera and lens the major camera manufacturers release and has built one of the most extensive data bases of how to correct for them in the photo business. Their PureRAW application applies those corrections to your RAW files taking into account what lens and camera you used for the shot. Topaz relies almost entirely on an AI based approach with little regard to what camera or lens you used so it's a best guess as to how they apply corrections. On a one to one comparison I don't see a use case for using PhotoAI DeNoise solutions for processing RAW images compared to what I can do with DxO PureRAW.
Do note the emphasis on RAW in those opinions - DxO only works with RAW images on files where they know about your camera. New camera models can often be off the table for a month or so right after release. PhotoAI has no such limitation and can do some very workable processing with JPEGS and TIFF files so I use it to play around with older digital media that didn't support RAW outputs. Sometimes (but not always) it does some amazing recovery work with older files.
In the most recent updates Topaz is venturing out into AI that requires considerable computing resources so their answer is centered around a cloud based compute model that uses their servers to handle the tasks. The downside is that they are charging for their server time and now want you to buy cloud compute credits in order to use their shiny new models. You do have the option to still process locally but the compute requirements are a low bar. I have a Mac Studio with an M1 Max processor and 64 GB of RAM. I tried locally processing one of their sharpening techniques on an image (not a stitched pano) that ended up taking over 20 minutes to complete. I am not sure I am going to renew my update subscription if this is where they want to take this now....
So far PhotoAI only has a minimum footprint for the cloud processing which shows up un their sharpening enhancements as Super focus BETA. Most of the rest of its functionality can all be run locally. GigapixelAI went deep into the AI weeds with the newest version and wants you to use the AI cloud server for almost any of their newer enhancements. I get a lot of Topaz emails prodding me to buy cloud credits since they seem to be pushing things in that direction for PhotoAI as well.I will say, there is not a single photo I take that is so important that I would need to spend money on Cloud Credits. That's really insane. It will be interesting to see if this model survives. I can't imagine it does.
Interesting. I still haven't seen any real reason to use PureRaw over ACR though. From my testing with PureRaw3, the results were similar except the AI in PureRaw kept inventing shapes in the shadow area's that couldn't be there in real life, like when PureRaw inserted rice terrace type of shapes in the shadows of the mountains in my image.So far PhotoAI only has a minimum footprint for the cloud processing which shows up un their sharpening enhancements as Super focus BETA. Most of the rest of its functionality can all be run locally. GigapixelAI went deep into the AI weeds with the newest version and wants you to use the AI cloud server for almost any of their newer enhancements. I get a lot of Topaz emails prodding me to buy cloud credits since they seem to be pushing things in that direction for PhotoAI as well.
You can combine DxO and PhotoAI as long as you run DxO on the image first and output as a DNG file. DxO PureRAW will let you recover at least 2-3 stops of shadows on a RAW file which would work wonders for the eagle you posted today. You can run PhotoAI on that DNG file to use the sharpening enhancements that haven't been subsumed to the cloud servers which does include fixing motion blur. I typically run the file into DxO PureRAW and then call up PhotoAI from inside of Photoshop as a filter plugin so I can do the sharpening in a layer in case I need to mask out some of the oddities that can arise with sharpening recovery like that.