One capability not mentioned is to restore blown/clipped highlight detail. Use the quick selection tool to select the blown area, expand the selection slightly, then hit generative fill without a prompt to recreate the detail in the selected blown highlight area.
This is great. I am in school for graphics design, and we just doing an AI image project in my 101 Photoshop class, where you have to create a background realistic, and fill it with 10 objects generated with AI, I am enjoying this a little too much, and this video is very helpful for my photography editing. Thanks
It's so synthetic, you can't even tell anymore what's real and what isn't. I miss the days when someone showed me a photo and I was impressed by the technique and creativity. Now I'll never know who's photo is honest and which one isn't.
Thats true, but one thing: If the image is high resolution with deliberate detail, it's most likely human made as theyd never work in low resolution on purpose. But Ai will only render in windows of 1024x1024 so typically it's lower resolution or lower detail. Not a great way but the best I know of!
Might seem like a small thing but the borders hack is very neat indeed, thanks! I find gen fill otherwise so handy for adding little props and details that would simply add to the shoot time. So adding a bracelet or knick-knack, changing a cup of coffee to a coke because it fits the scene better, generating spooky stuff for a dream sequence shot, whatever … truly game-changing.
No problem, I deselected the nose so it wouldn't change and it can help keep the generation more confined to the eyes only. Hold down Alt (PC) or Option (Mac) whilst using the select tool to deselect.
@@thewebstylist Note taken! Great idea. If it helps, I've worked on a video for text effects which I'll release soon enough. It's not as punch or valuepacked as this one but could be useful for many :)
Thanks to your video I was able to re-edit 2 pictures for a client by removing something & adding back in part of a huge structure & it worked beautifully for me, however I do have a question. When I need to select my area to delete & let the prog find me replacements, It doesn't give me the Lasso tool to select my object, I use my brush tool and it erases my pixels, but it does work. I was curious why it wouldn't let me use the Lasso tool ? Any advice ? Thank you for your video, it helped me do what my customers wanted. 😉😀
Thank you, glad I could help! I'm not sure whats happening there sorry. Thats confusing try rasterizing or selecting different layers maybe? Sorry Im just not sure.
It did take a few attempts, however, Id.experiment was the the selection placement (closer or overlapping bottom of text). Sometimes you can leave the genfill field blank or type in water reflection
It's probably a stupid question... but can you use generative fill to obey colors, i.e., take a black and white line art and turn it black and gold/ white and gold, etc.?
Not really, although you can try using a semi transparent selection and try prompting that way? Here's what I mean : ruclips.net/video/ErgQD3ieUDs/видео.htmlsi=N9NVK3INyKp8Ms95
@WadeMcMaster thanks for the fast reply. I just didn't want to recreate my same "line art" in different color schemes... so I was hoping i could select and recolor it into a black/ gold color scheme like that of inverting colors etc. One last question... is this feature included with Photoshop, or is it an exlcusive offer. Dont want to buy if this isn't incl.
How do u duplicate a generative fill? For example I'm trying to change some rims on a car but when I click on more than one it doesn't do a good job but if I do one at a time gives me amazing result but it won't create the same one.
Yeah that is tough, at this stage the only way I can see you getting an identical rim, would be copying the layer itself and masking it manually..I've found getting the same result difficult also. Hopefully they improve this. One idea though - roughly cut out the first rim, place it over the new second wheel. then create a semi transparent sslection and use the same prompt - it tends to refer to the underlying image more. This video may help if you want to try that: ruclips.net/video/ErgQD3ieUDs/видео.html
Have you faced the wrong sizes of the generated background? Im doing product photos for website. I have an object that is around half meter diameter, its a roll of a garden hose. I remove background and ask photoshop to put it on standing on grass. And photoshop creates close-up background and eventually it looks like a macro shot. And my half meter diameter roll of a garden hose looks like 10 centimeters. Any advice to specify size of generated background?
Oh sorry mate I havent had that issue, but maybe you could get a photo from pexels or unsplash and blend the bottom in instead so it stands on that grass? or shot of 'lawn' instead of grass - that term may be translated as further away? i think it would be a play on words or framing photography terms if you know any that are suitable
No worries. Ctrl click the layer (selects the shape) Go to select - expand selection by 15px or so Ctrl + alt click the layer (deselects the shape) Expand selection again by 5 px or so. If your on mac, Ctrl = command Alt = option
*You showed how to make an image into a sketch or a paint, but how about backwards, making the sketch or a paint into a realistic image? Like the painnt of Tanos, I thought you were making that painto into a realistic image representation.* Maybe next time?
Oh, nice! This would be very beneficial to add to post-production on Midjourney's final work. I wish Midjourney had a dedicated standalone software with a similar generative fill feature like Photoshop. As of right now, Photoshop's AI model only uses limited stock photos.
Every time I try to change hair color , eye color or clothes I always get “ violates photoshop community guidelines “ and also fill eats up so much memory I have to go purge in the settings.
I've been playing around with Generative Fill for a while, but your techniques are eye-opening! Thanks for sharing. One thing I hope they improve: when you told Generative Fill to change eye color or lip color, it changed the actual eyes and lips to someone else's. That's not what was asked for!
That is true, I didn't consider that until after I did the video too. However, you could try selecting only the pupil of the eye / lips precisely and get a better result? I think they'll get that done in time, being a photo editing app theyd be mad not to work on a solution like that.
This seems like it's finally going to remove my need for a thumbnail editor. I can just come up with thumbnail ideas and generate them myself now. Thanks for tge tips! I assume this is all paid though right and not free?
It does require a paid creative cloud subscription to download Photoshop beta. You can sign up for their online image generator and fill for free now though. More info on getting Photoshop beta is in this video if it helps: ruclips.net/video/beIs3AwNjxk/видео.html
I didn't realise Generative Fill was only for the subscription version of Photoshop :( I've just purchased the 2023 version for £130 and now it wants me to pay £20 month for the beta subscription version so I can at least have access to Generative AI. I was so excited to use it too :(
It's Adobe Photoshop beta, you will need to pay for Adobe creative cloud and then download Photoshop beta. More info in this video: ruclips.net/video/beIs3AwNjxk/видео.html
TY for another great video. Exactly what I need to learn ATM. Q: I'm working on a few projects that involve a lot of image processing. My PS skills are pretty average. I feel like I'm drowning. What do think about using online services like Fiverr? Also, a video on file management for creating videos from stills would be really handy. It's not uncommon to end up with thousands of images, audio files, movie clips, animation clips, PS files etc - all in various formats and stages of processing. What's the best way to keep track of it all?
Hey mate, no problem I'm glad you like it! Fiverr can be good for basics, but I struggled to find high quality Photoshop users on there. They may be there and require some extra searching or maybe what they do might suit, depending on what you need. Either way i search deep. In regards to video management, I go with sub folders upon sub folders. Usually: - folder (date or month) - project folder (name of project) - then more sub folders like: -- editor files -- images and clips (break it up further if needed) -- video output for exporting the whole or portions of the video I hope that helps man :)
Also if you need more, break up each folder even more. OR have a naming system on the file like _INTRO- me talking.mp4 _INTRO- Screen cap footage for broll Etc etc
@@WadeMcMaster Thanks for that. I'm currently doing a similar thing with filing. I find Bridge quite handy too. I thought there might be some you-beaut method I'm missing out on.
"Discover unlimited possibilities for creativity with Photoshop’s This Fill AI-discover 33 innovative ways to enhance, edit, and alter your images effortlessly."
There is a lot more of it, I do think it'll make human made imagery more sought after though. There's definitely way more demand for art and design than there is artists so it's an interesting time.
This is a good video. But what you have shown us here is the result of using Photoshop tools, not how to use them. We, the beginners who constitute the vast majority of your channel’s followers, hope that you will provide us with a video that explains in detail how to use the various Photoshop tools related to artificial intelligence and Generative Fill. Thank you!
Appreciate it, thank you. I have a bunch of those videos on my channel.if you want to check them out! Check this Playlist for ai: ruclips.net/p/PLsaZXe5Zf2eGljO-3t0zX7jiVqhKBVSKh&si=aEupB7lcqp-L-fgT This playlist for general Photoshop tutorials: ruclips.net/p/PLsaZXe5Zf2eFcA9htb5xFkhB70-gICXEZ&si=BMjLwTztpIIfRMgv
it's a big advantage here if you can speak english, not like mine does.. it's really hard for me to use this gigantic tool perfectly with the german language here. grits often come out here.
They don't know how to do their job properly then. Bloody disappointing. Just pop a cup in an envelope, ya know? Haha, thank you very much for the Super thanks mate :) Great appreciated!
@@WadeMcMaster You're top drawer, pal. Great knowledge, good pace of delivery and patient enough for old farts like me. Cheers. The Ai stuff is what draws me in
haha, thanks mate. I've been slowly trying to inject some dry humor in as I get more comfortable being my stupid self in these videos. I feel like dry suits my straight forward delivery too.
There's some workaround, which aren't ideal, but there - such as generating in chunks or upscaling with topaz photo AI and then overlaying those changes to your image. But for low focus areas and smaller additions it's very powerful
One capability not mentioned is to restore blown/clipped highlight detail. Use the quick selection tool to select the blown area, expand the selection slightly, then hit generative fill without a prompt to recreate the detail in the selected blown highlight area.
That's awesome! Thank you for sharing, I didn't think of that
This is great. I am in school for graphics design, and we just doing an AI image project in my 101 Photoshop class, where you have to create a background realistic, and fill it with 10 objects generated with AI, I am enjoying this a little too much, and this video is very helpful for my photography editing. Thanks
Glad I could help mate! Best of luck with the project :) Thank you
This was amazing. I have not played around with this, but I’m going to try it tomorrow with some of your tips. Thank you.
Thank you so much!
It's so synthetic, you can't even tell anymore what's real and what isn't. I miss the days when someone showed me a photo and I was impressed by the technique and creativity. Now I'll never know who's photo is honest and which one isn't.
Thats true, but one thing: If the image is high resolution with deliberate detail, it's most likely human made as theyd never work in low resolution on purpose. But Ai will only render in windows of 1024x1024 so typically it's lower resolution or lower detail.
Not a great way but the best I know of!
One of the best intros to AI yet
Thx bud
Thank you!
Might seem like a small thing but the borders hack is very neat indeed, thanks! I find gen fill otherwise so handy for adding little props and details that would simply add to the shoot time. So adding a bracelet or knick-knack, changing a cup of coffee to a coke because it fits the scene better, generating spooky stuff for a dream sequence shot, whatever … truly game-changing.
Yeah I agree, really allows for more imaginative subtle details and changes in an instant
Thank you so much Wade, very helpful tips and tutorials! God bless you and good luck!
My pleasure! Thank you
Nice overview that covers a lot. Got me going with this; thank you.
Glad it was helpful! Thank you :)
At 0:28 you selected both eyes then put a box on the nose. How and why did you do that?
No problem, I deselected the nose so it wouldn't change and it can help keep the generation more confined to the eyes only. Hold down Alt (PC) or Option (Mac) whilst using the select tool to deselect.
@ thank you! Cool trick
SO MUCH AWESOMENESS!! Fantastic tips and tricks Wade bro Priceless tools! 🎉
Thank you mate! I really enjoyed making that video :) Been trying to think of other topics to work a similar format :)
@@WadeMcMaster GRT to hear indeed brah, GF the Greatest additions by far to PS ever!$!
Just watched again brotha and DEF think u should have a series of these on best use examples w PS GF
@@thewebstylist Note taken! Great idea. If it helps, I've worked on a video for text effects which I'll release soon enough. It's not as punch or valuepacked as this one but could be useful for many :)
I still think this could/should be a monthly series my man!
Thank you so much bro, your tutorial worked, keep going!!!
Thank you!
Thanks to your video I was able to re-edit 2 pictures for a client by removing something & adding back in part of a huge structure & it worked beautifully for me, however I do have a question. When I need to select my area to delete & let the prog find me replacements, It doesn't give me the Lasso tool to select my object, I use my brush tool and it erases my pixels, but it does work. I was curious why it wouldn't let me use the Lasso tool ? Any advice ? Thank you for your video, it helped me do what my customers wanted. 😉😀
Thank you, glad I could help! I'm not sure whats happening there sorry. Thats confusing try rasterizing or selecting different layers maybe? Sorry Im just not sure.
How do you deselect the area you want to keep?
Infusing the selection tool, hold alt (PC)/option (Mac) key and make your selection
Thanks for this video!
Omg what a resource !!!!!!!! Thanks
Thank you!
how do you get the text reflection i get random letterings
It did take a few attempts, however, Id.experiment was the the selection placement (closer or overlapping bottom of text). Sometimes you can leave the genfill field blank or type in water reflection
thanks
It's probably a stupid question... but can you use generative fill to obey colors, i.e., take a black and white line art and turn it black and gold/ white and gold, etc.?
Not really, although you can try using a semi transparent selection and try prompting that way? Here's what I mean : ruclips.net/video/ErgQD3ieUDs/видео.htmlsi=N9NVK3INyKp8Ms95
@WadeMcMaster thanks for the fast reply. I just didn't want to recreate my same "line art" in different color schemes... so I was hoping i could select and recolor it into a black/ gold color scheme like that of inverting colors etc. One last question... is this feature included with Photoshop, or is it an exlcusive offer. Dont want to buy if this isn't incl.
Everything I touch on is standard in Photoshop's latest version :)
How do u duplicate a generative fill? For example I'm trying to change some rims on a car but when I click on more than one it doesn't do a good job but if I do one at a time gives me amazing result but it won't create the same one.
Yeah that is tough, at this stage the only way I can see you getting an identical rim, would be copying the layer itself and masking it manually..I've found getting the same result difficult also. Hopefully they improve this.
One idea though - roughly cut out the first rim, place it over the new second wheel. then create a semi transparent sslection and use the same prompt - it tends to refer to the underlying image more. This video may help if you want to try that: ruclips.net/video/ErgQD3ieUDs/видео.html
Adobe is developing its software at a high speed! Looking forward to MAX in October.
Yeah 100%, I wonder what they'll announce :)
It is potentially a game changer, yet the quality of the generations needs to improve dramatically, and the quality of the resolution also.
I do agree, although some uses it does a superb job, but simpler areas. I think it'll improve greatly in time :)
Have you faced the wrong sizes of the generated background? Im doing product photos for website. I have an object that is around half meter diameter, its a roll of a garden hose. I remove background and ask photoshop to put it on standing on grass. And photoshop creates close-up background and eventually it looks like a macro shot. And my half meter diameter roll of a garden hose looks like 10 centimeters. Any advice to specify size of generated background?
Oh sorry mate I havent had that issue, but maybe you could get a photo from pexels or unsplash and blend the bottom in instead so it stands on that grass? or shot of 'lawn' instead of grass - that term may be translated as further away? i think it would be a play on words or framing photography terms if you know any that are suitable
@WadeMcMaster okay, thanks for your respond and for your advices)
help how do you select that area again time 2:02
No worries.
Ctrl click the layer (selects the shape)
Go to select - expand selection by 15px or so
Ctrl + alt click the layer (deselects the shape)
Expand selection again by 5 px or so.
If your on mac, Ctrl = command
Alt = option
@@WadeMcMaster thank you so much
Amazing! Thanks.
Thank you too!
@1:50 I swear you said "A Fucking Truck" hahahaha
Haha yeah I can hear it now
how well does it work with text? like removing letters or replacing letters? or changing font and style?
Text, not so great but it can get some results with a few attempts
Super! Looking forward to playing with some of these ideas
Have fun!
mind blown!
haha thank you!
Nice explained...Thanx
Its Superb i reqlly appreciate your efforts keep it up
Thank you so much!
It’s does a good job of removing reflections on glasses.
Never thought of that! Great example!
how can i get generative fill in my photoshop. Please anyone help me
You just need to latest version and a current Creative Cloud subscription.
Some excellent ideas, many thanks
Thank you much appreciated!
Which version of Photoshop this are?
Can you tell me the details
Photoshop beta, this video shows how to download it :) ruclips.net/video/beIs3AwNjxk/видео.html
*You showed how to make an image into a sketch or a paint, but how about backwards, making the sketch or a paint into a realistic image? Like the painnt of Tanos, I thought you were making that painto into a realistic image representation.* Maybe next time?
I didn't consider it, I'll give it a go!
Great tutorial, super helpful tips and tricks!
Thank you!
Nice breakdown! I'm curious, have you tried Stylar’s AI? I find it more intuitive than Photoshop's.
I have! Ive done two videos on Stylar :) great platform.
Oh, nice! This would be very beneficial to add to post-production on Midjourney's final work. I wish Midjourney had a dedicated standalone software with a similar generative fill feature like Photoshop. As of right now, Photoshop's AI model only uses limited stock photos.
I agree. If Midjourney was able to upscale 6x or more AND allowed this kind of editing - it'd be unstoppable.
Word on MJ St is that in-painting is not far away. 🤩🎉
Every time I try to change hair color , eye color or clothes I always get “ violates photoshop community guidelines “ and also fill eats up so much memory I have to go purge in the settings.
Really? Man I've never had that issue, sorry to hear mate. I'm not sure what's happening there
@@WadeMcMaster thanks for the response tho .
I've been playing around with Generative Fill for a while, but your techniques are eye-opening! Thanks for sharing.
One thing I hope they improve: when you told Generative Fill to change eye color or lip color, it changed the actual eyes and lips to someone else's. That's not what was asked for!
That is true, I didn't consider that until after I did the video too. However, you could try selecting only the pupil of the eye / lips precisely and get a better result? I think they'll get that done in time, being a photo editing app theyd be mad not to work on a solution like that.
another problem is it may add in a wrong background! how do we prevent that
Great ideas... Is there still resolution limitations with generative fill?
Yeah there still is unfortunately. Maybe something to address in a other video?
thanks a lot BRO you saved my life
Happy to help, thank you!
does this menu also exist in Photoshop 2023?
It does now :)
Thanks!
when I try to make a graphic design border it doesn't work. Some really scary stuff happens.
Oh really? I've always had success with that one. How many prompts have you tried?
Wow! Great 👏🏻 Thank you
No problem. Thank you!
Where is Generative fill in Photshop 2023 Release, not Beta? Ho to open it?
It's not, it's only in the beta version sorry mate
This seems like it's finally going to remove my need for a thumbnail editor. I can just come up with thumbnail ideas and generate them myself now. Thanks for tge tips!
I assume this is all paid though right and not free?
It does require a paid creative cloud subscription to download Photoshop beta. You can sign up for their online image generator and fill for free now though. More info on getting Photoshop beta is in this video if it helps: ruclips.net/video/beIs3AwNjxk/видео.html
Great vid thanks.
Thank you!
Sos grande papaaa espero la nueva version
Thank you!
I didn't realise Generative Fill was only for the subscription version of Photoshop :( I've just purchased the 2023 version for £130 and now it wants me to pay £20 month for the beta subscription version so I can at least have access to Generative AI. I was so excited to use it too :(
I had no idea either! Oh man, I'm really sorry to hear about that :/
What’s the app or website called?
It's Adobe Photoshop beta, you will need to pay for Adobe creative cloud and then download Photoshop beta. More info in this video: ruclips.net/video/beIs3AwNjxk/видео.html
Great
A year later and still a gold mine of PG GF tips brother!! A number of these I still haven’t thought to use so great reminders just now #Priceless
Thank you as always man! I'm glad its stlil helpful :)
lol he turned the rock in to jay uso😂 man those guys are definitely family.
hahaha AI is expanding our knowledge
TY for another great video. Exactly what I need to learn ATM.
Q: I'm working on a few projects that involve a lot of image processing. My PS skills are pretty average. I feel like I'm drowning.
What do think about using online services like Fiverr?
Also, a video on file management for creating videos from stills would be really handy. It's not uncommon to end up with thousands of images, audio files, movie clips, animation clips, PS files etc - all in various formats and stages of processing. What's the best way to keep track of it all?
Hey mate, no problem I'm glad you like it!
Fiverr can be good for basics, but I struggled to find high quality Photoshop users on there. They may be there and require some extra searching or maybe what they do might suit, depending on what you need. Either way i search deep.
In regards to video management, I go with sub folders upon sub folders.
Usually:
- folder (date or month)
- project folder (name of project)
- then more sub folders like:
-- editor files
-- images and clips (break it up further if needed)
-- video output for exporting the whole or portions of the video
I hope that helps man :)
Also if you need more, break up each folder even more. OR have a naming system on the file like
_INTRO- me talking.mp4
_INTRO- Screen cap footage for broll
Etc etc
@@WadeMcMaster Thanks for that. I'm currently doing a similar thing with filing. I find Bridge quite handy too. I thought there might be some you-beaut method I'm missing out on.
@@WadeMcMaster 🤓👍
No worries man!
Nice ❤
"Discover unlimited possibilities for creativity with Photoshop’s This Fill AI-discover 33 innovative ways to enhance, edit, and alter your images effortlessly."
Great!
Thank you
Wow! Useful tips and tricks that I would like to experiment with. I am certainly not a big fan as everything seems artificial these days.
There is a lot more of it, I do think it'll make human made imagery more sought after though. There's definitely way more demand for art and design than there is artists so it's an interesting time.
YAY I'm "LIKE" #1000 :D
This is a good video.
But what you have shown us here is the result of using Photoshop tools, not how to use them.
We, the beginners who constitute the vast majority of your channel’s followers, hope that you will provide us with a video that explains in detail how to use the various Photoshop tools related to artificial intelligence and Generative Fill.
Thank you!
Appreciate it, thank you. I have a bunch of those videos on my channel.if you want to check them out!
Check this Playlist for ai: ruclips.net/p/PLsaZXe5Zf2eGljO-3t0zX7jiVqhKBVSKh&si=aEupB7lcqp-L-fgT
This playlist for general Photoshop tutorials: ruclips.net/p/PLsaZXe5Zf2eFcA9htb5xFkhB70-gICXEZ&si=BMjLwTztpIIfRMgv
Nice video, but too fast for beginners like me 😊
I understand for sure! Thank you, I also have this video which slows the pace a bit: ruclips.net/video/beIs3AwNjxk/видео.html
it's a big advantage here if you can speak english, not like mine does.. it's really hard for me to use this gigantic tool perfectly with the german language here. grits often come out here.
True, I've never considered that. Hopefully they are improving their language model as this would be severely limited ona global scale.
The Post Office told me to sod off when i asked about sending a pint to U
They don't know how to do their job properly then. Bloody disappointing. Just pop a cup in an envelope, ya know?
Haha, thank you very much for the Super thanks mate :) Great appreciated!
@@WadeMcMaster You're top drawer, pal. Great knowledge, good pace of delivery and patient enough for old farts like me. Cheers. The Ai stuff is what draws me in
Thank you for the positive feedback :) I'm glad I'm able to break things down in a digestible way. Much appreciated mate :)
Class, you are. 🌟
anonymous bald gentlemen got me rolling ngl
haha, thanks mate. I've been slowly trying to inject some dry humor in as I get more comfortable being my stupid self in these videos. I feel like dry suits my straight forward delivery too.
plot twist: his thumbnail is made by ai
👋
Try your best, you can find 1001 ways instead of 33 ways. 😁
This is true. 1001%
6:20 your scary me
Haha sorry mate, that ai gets creepy sometimes
And unfortunately it’s still useless due to resolution limitations…
There's some workaround, which aren't ideal, but there - such as generating in chunks or upscaling with topaz photo AI and then overlaying those changes to your image.
But for low focus areas and smaller additions it's very powerful
We can easily change gender now 😂😂...
Well yeah, thats possible! lol
Finally
Sos grande papaaa espero la nueva version