The TRUTH about Adobe Generative AI | Cost and Performance
Vložit
- čas přidán 18. 09. 2023
- Is Adobe charging for using it's AI technology? What are Generative Credits? How well does it fit into a photographer's workflow? What's the deal? Watch and find out!
Adobe FAQ on generative AI credits
helpx.adobe.com/firefly/using...
Don't forget to visit my website at www.jamesparkerphoto.com to sign up for my newsletter and check out my upcoming workshops and 1 on 1 tutoring. I teach classes on using your camera, landscape photography, lighting, post processing, and more.
Links:
Website - www.jamesparkerphoto.com
Facebook - / jamesparkerphoto
Vero - vero.co/jamesparkerphoto
#adobeAI, #adobefirefly, #generativecredits, #workshops, #lightroom, #adobe, #photoshop, #LandscapePhotography, #PacificNorthwest, #educational, #howto, #composition, #postprocessing, #photography
Finally someone spitting facts about generative fill instead of making viral tiktoks or reels where you can’t see that it actually sucks
Very informative, I agree, it is still not ready for prime time, certain it will improve and the resolution hopefully will increase (imagine it will cost). I use it in many abstract designs but when it comes to photos or standard images, it often just looks terrible. It is a pity that Adobe are spending so much effort on this to the detriment of adding or improving many of the features it has.
For a year now I have been paying $19.99mo for 1TB because I somehow never saw the $9.99mo plan. So is there anything being stored in the cloud without me physically doing it myself that having 1TB would possibly be needed for? I store my LRC catalog and all images on an External SSD. From what I see, both plans are identical besides the cloud storage accounts.
TY
James - you will be bumped down to 20 gig of online storage and lose the rest. If all of your files are local you won’t lose a thing that matters.
Agreed that it's not quite ready for prime time yet. You may have gotten SLIGHTLY better whale results if you'd tried "humpback" instead of "hump back", but I'm just guessing. They say they're working on better resolution for generative actions, but it'll likely come at the cost of additional credits.
I did try several iterations but it never worked to an acceptable level. I'm aware you can create small 1024x1024 boxes and have it generate just that little bit and then repeat it over and over and over and over to eventually maybe get a usable image, but that is not something I'd do or expect customers to do. And I am sure they will improve it with time - BUT as the resolution increases so will the computing time on Adobe's end and I can promise there will be a point where they charge. It's not realistic to expect them to host servers churning out 30-60 megapixel images based on millions of prompts without an additional expense.
Thank you James . Clear explained and demonstrated
Let’s pretend that it is past Nov 1 . How many generative credits do you reckon you used in the demos ?
On Adobe site they state
Feature
Generative credit use
Generative Fill, Generative Expand, Text to Image, Generative Recolor
1 credit*
The * has the further explanation of
* For standard images of up to 2000 x 2000 pixels
What do you reckon they mean by “for standard images up to 2000x 2000 pixels
The resolution remains at 1024x 1024 so it is not that
I’ve tried to get the answer from Adobe but no luck so far
Thanks
Graham UK
Graham - Thanks for the comment. The way I understand it is every time you click the GENERATE button that's a credit. If you regenerate the same prompt because you didn't like the results, that's a credit. If you use the Generative expand that's a credit. In the video above I'd guess I used 5-6 credits. As for the 2000x2000 pixels, I believe it refers to the size limitations where it would work best - i.e. for use with images 2000 x 2000 pixels. For example, generating a whale inside an image sized at 2000x2000 pixels is the recommended size constraints. Anything larger than that (as with any modern camera or phone) and it the results will be... less than ideal since the generated content will be limited to 1024x1024.
I could probably spend hours using this tool just for the amusement value.
There’s that. Try using people photos and adding hats, glasses, costumes, riding a tiger, etc.
Wait, so you're saying we should believe what Adobe tells us directly rather than what some random dude named Fotoguy6969123 said on social media? Sheesh ;)
It's horrible - the whole concept is. If you want artificial photographs then give yourself up to AI. If you want to be a photographer then don't. Especially if this worked perfectly and we could agree upon the technological achievement of extending photographs with false additions, who would not question his own ability to create a real photography based on personal skills? Only the superficial content creators would not. The world is already overloaded with mountains of worthless content! and fewer and fewer serious photographers and their photography. As we already have witnessed - this will end up with a content creator giving a few instructions to AI - and the creepy thing will create one or more scenarios fitting the input. Great. Hallelujah. How long will everybody be satisfied? You can't be proud any more.
You still have to be an artist who knows what to do, when you successfully implement an AI workflow. The generation itself is like paint. Just because everybody can buy some, doesn't mean everyone knows how to use it properly. And I think you don't quite understand what it is. And what it can be.
Editing photos for example takes time and effort. And this time and effort is made easier with tools. And the tools available aside from the "generative AI" are also based on AI. And generate. But that's not a problem for you (simple patch tool for example)? Or even using a PC in combination with editing software is also cool with you? It's "video killed the radio star" all over again. But that's normal when new stuff comes into the overall meta.
While in beta I used this to remove objects, backgrounds, etc.. and fill in minor details (grass, branches, etc..) without any effort. It did basically the same thing that I'd usually do, but MUCH faster. It was amazing for that!
Don't focus on the uncanny generation this thing can do. I am sure you enjoy photos where AI has been used and don't even know it ;)!
@@schpeidermann I was not referring to born artists like you who without doubt creates great art without telling anybody about the AI part. No, I fear the wannabe photographers who are now left with yet another viral tool in their hands and about the collective product - a new tsunami of crabby images.
@@Eigil_Skovgaard Your statement was:" If you want artificial photographs then give yourself up to AI. If you want to be a photographer then don't".
Many "gave themselves up" to it to create beautiful images but still can handle their camera, the "third" eye and nature's light, environment, etc.. That's how they do it. Gotta know your tools. Otherwise you stay a wannabe.
And why fear them? Wannabes stay wannabes. No matter the tools.
I also was worried about TikTok, but in the end I'm not affected. I don't watch it. Simple.
@@schpeidermann I think your adaptive nature, which is very common for a certain progressive movement - is blurring the distinction between shit and Chanella - and that represents a very deliberate political attempt to destroy normality. You can go on with your woke idiocy and push for changing black and white into brown - but please stop wasting my time with your educational babble.
With all due respect, you're talking about the same thing for the first 4 minutes of the video.
I was ranting a bit. 🙂