We are working on Photoshop 2024 on the Mac. It works the same on Windows.

The first part is (somewhat) common knowledge, but I quickly build from there into new territory. I started with the knowledge that you can apply art effects to images if you use a 40% density with Generative Fill. I then ran with that piece of info and experimented with a number of things and came up with the tutorial you see here. Don’t worry if you don’t understand what I just said, it will be explained.

Here is an important explanation, then we will do step-by-step, but this will help people who have difficulties getting the same result.

5 Steps to Generating Art from Photos

When we generate these effects, these are the 5  steps we will take every time. It’s actually very fast and easy once you have done it a couple of times. (This is a reference, I will walk you through it). We will start with a photo.

1. Press Q for Quick mask (Takes us into Quick Mask Mode)

2. Choose a foreground color (a shade of gray)

3. Press Alt+Backspace – Option+Delete on Mac (Fills the canvas with the Foreground color)

4. Press Q key once again (Takes us out of Quick mask and turns the Mask into a selection, You won’t see the selection because it’s filled with less than 50% gray)

5. Choose Generate and enter text into the prompt

Ok, hold on one second. This is where some people got stuck.

In this tutorial, I will be setting the Brightness in the color picker to set the density (transparency of mask selection)

The number depends on a Photoshop setting.

Double click the Quick mask tool. What is your Quick Mask option set to?

If you are set to Color Indicates: Masked Areas then enter the same amount as the desired density. You know like normal. If you want 40% density, enter 40 int the B box. If you want 7%, enter 7. (If in doubt, use this option, it’s the easiest)

BUT If you are set to Color Indicates: Selected Areas then enter the opposite amount from 100 as the desired density. For example is you want 40% Density, enter 60  (100-60=40). If you want 7% density, enter 93 (100-93=7). If you wanted 20% density, enter 80 (100-80=20) (100-B=Density).
Sorry if some people got confused. Hopefully this clears it up. Not everyone has the Mask options set the same, so we have to have this explanation. So depending on your preferences, choose the correct number.

I have my preference set to Color Indicates: Selected Areas in this tutorial. So either change your preference now and follow along exactly. Or make the adjustments to the B (Brightness) number. Chart below to help.

Let’s demonstrate the effect now.

Creating a Water Color Effect to a photo with Photoshop AI

Starting with a photo

Press the Q key, or click on Quick mask to enter quick mask mode. (In quick mask you can paint or fill with color. These will be converted to selections when we exit Quick Mask).

Now we want to set the density for the mask (transparency). We want to start with 40% (The number everyone uses for this effect)

Let’s choose a Foreground color

Click on the foreground color picker

Choose the lightness from 0-100. The color doesn’t matter because masks don’t support color. Only tones.

See the 3 HSB (Hue, Saturation, Brightness) radial buttons? We will enter our number into the B field. This is the Brightness and its all we care about.

In this case I am entering 60.

This will give me 60% brightness, but 40% density because my Quick Mask mode is set to selected. (If yours is set to Masked, then just enter the desired density as the same Brightness amount (the opposite). If this doesn’t make sense, read the part above labeled “ok hold on one second…”

Click ok

Now fill the quick mask with the foreground color

Press Alt/Option+Backspace, You will see a pink color overlaying (this is the mask)

Press Q once again, or click Quick Mask to convert the mask to a selection and leave Quick mask.

You might see a warning saying selection won’t show because there is nothing selected over 50%. This is expencted. You won’t see the marching Ants selection, but the selection is active and everything is selected at 40% density (transparency).

Now, click the Generative Fill button on the taskbar

In the field, enter Watercolor into the prompt.

You will see a result, that looks like water color and takes on the shape and form of your photo.  (We are about to make it look a whole lot better.. hold on)

Ok this is where most people stop with converting photos to art in Generative fill and getting just ok results.

Lets see what I have discovered that makes it a whole lot better.

Lets start with changing the density (This isn’t even the good bit yet.. stay with me)

Much Better results with lower density

I wondered, why does everyone who does this stick to 40%? Surely there must be other numbers that work too? And sure there are, here is the result of my first discovery (more coming as this tut progresses).

Let’s do these steps again

1. Press Q for Quick mask (Takes us into Quick Mask Mode)

2. Choose a foreground color (a shade of gray)

This time we only want 7% density: enter 93% (or 7% depending on your mask setting)

3. Press Alt+Backspace – Option+Delete on Mac (Fills the canvas with the Foreground color)

4. Press Q key once again (Takes us out of Quick mask and turns the Mask into a selection, You won’t see the selection because it’s filled with less than 50% gray)

5. Choose Generate and enter watercolor into the prompt

Now look at this much better result. (This still isn’t even the good bit)

What I have discovered here is the density changes the result

The lower the density, the closer the result is to the original image. The Higher the density, the more the result will stray from the original.

THIS IS A HIDDEN AMOUNT SLIDER (The Density) But when you see why in a second, the lights will come on.

Here I performed the same 5 steps, but changed the text prompt to watercolor instead of watercolor and got an oil painted result.

We also tried it with Pencil Sketch for another art effect, this is at 30% density.

Here I used the same Pencil sketch prompt, but at 10% density. Experiment with different densities for different results.

Ok, if you stuck with me this far (And didn’t assume you already knew this and bounced) get ready for a huge payoff.

It struck me that Photoshop uses Firefly as the engine that runs Generative fill. So therefore, it seems Generative Fill must understand the Firefly settings? Let’s see.

Go to Firefly.adobe.com

Click on Text to Image

Just to get something going to populate the interface, I entered Street and Firefly generated these images. (Firefly is now on Model Version 2 and vastly improved image generation – No V2 isn’t in Photoshop yet).

Look at the settings on the right side of the interface, this is the magic.

Notice a strength slider? We are emulating this by changing the density back in Photoshop.

Scroll down and see more settings.

Let look at Movements (These are art styles or movements such as Art Deco,  Art Nouveau (predecessor to Art Deco), Steampunk, cyberpunk etc)

Ok, lets go back to Photoshop.

Open a Photo of a street.

Q for Quick mask

Choose foreground color. Choose 40% density. Either 60 or 40 into your B box depending on your Quick masks settings

Click ok

Fill with foreground color Alt/Option+Backspace

Press Q to convert to selection

Remember the movements in Firefly?

Enter Art Deco

Now our photo is converted to exactly the same scene, but in an art deco style! Mind blown!

Let’s try a different one. Steampunk.

Yes! This works beautifully too.

So, all the movements work. Lets try a different category.

Back in Firefly I see a concepts tab

Ohh futuristic looks interesting. Lets try that

I entered Futuristic in the prompt in Photoshop and Generative fill gave me exactly what I wanted.

Just for fun I opened another of my images I shot a while back

Entering Futuristic into the Prompt transformed this exact scene into a futuristic diner concept. Ok, it’s far from perfect. But start to think about the possibilites!

I encourage you to experiment with this and see what you can do. Go to Firefly for inspiration and see what you can generate.

I hope you found this tutorial helpful!

If you want to learn more about Generative Fill, check out my course here

Great to see you here at the CAFE

Colin

The post Revealing hidden features in Photoshop Generative Fill to convert photos to art and more appeared first on PhotoshopCAFE.

When you hit generate in Photoshop Generative Fill, you will see tips. I have assembled all the tips here and will demonstrate each one. I have also thrown in a couple of my own to make it more interesting. Check out the quick video, each tip is a little over 30 seconds on average. Here is my written summary.

I will show each tip and then show you an example.

1. New Languages

It supports over 100 languages now. It will be interesting to see if it generates localized content based on the language. For example, person in Japanese” 人 hito”, might be biased towards generating Japanese people. and localized content, like it does on Mid Journey.

2. Pin Taskbar

Click on the menu and choose pin, for the taskbar to stay in position. It now says put even when you relaunch. To find the taskbar choose Window>Taskbar

3. Selections matter

When you make a selection, remember it will be able to change anything within the selection area, but cannot outside the selection.

If I want to generate a t-shirt, I need to make sure I provide enough room for the shirt, this includes completely selecting what I want to remove as well as providing space for the sleeves, otherwise I’ll get a tank top. At the same time, if I don’t want the hair to change, I shouldn’t include it in the selection.

When we generate, notice the shirt is generated perfectly, but she lost half her hair.

4. The shape of the selection is very important

This one is really important, so much so, I created a whole tutorial on selections and generative fill previously 

Make a wide selection and type hat

We get a wide hat

Make the selection taller

And we get a top hat. Make the selection in the shape you are looking for. In the previous example you will see I made it in the shape of the t-shirt I wanted. This also reduces the amount of “content warnings” you get. But I’ll add another bonus tip on that at the end.

5. Change density of Generation

Changing the density of the selection will make semi transparent generations to make things appear underwater or though glass etc.

Start by double clicking Quick mask in the tool box.

Set the options to Selected areas

Click ok

Press the Q key to go into Quick Mask

Click the foreground color picker on the bottom of the tool box.

Back will give us 100% dense selection. White will make it invisible or no selection. The lighter the gray the more transparent. (It’s clear in the video at the top)

Choose black

Grab a brush and paint the area to select with black, Make it above the water. (shows red because of quick mask)

Change to a light gray and paint under the water as shown.

Press the Q key to change the Quick mask into a selection. The bottom part doesn’t show as a marching ant selection because its less than 50% opacity, but its still selected..

Type in Whale

Looks like we got 2 whales. Notice how the bottom part looks underwater because of the gray we painted into the mask.

6. Moving a generated layer

Here I have generated a truck

We want to change the position. Drag the generated layer to the other side

Notice it doesn’t blend in anymore

Press generate and it will

It now blends but we have a different truck. That’s one of the drawbacks of generative pixels.

7. Keep the same object but move it. (Bonus tip)

This tip isn’t in the Adobe tips, it’s one of my own.

Instead of hitting generate when we move an object, we need to select just the generated area outside the object.

With the truck layer active. Choose the quick selection tool

Press Select Subject

You will see the truck is selected. Refine the selection if you need to.

Choose Select>Save Selection

Name it truck.

Now we need to select all the pixels in the Generative fill layer.

Hold down the Alt/Option Key and click on the Mask in the Layers Panel. You will see the selection loads up.

Now, to remove the truck from the selection.

Choose Select>Load Selection

Choose truck under the channel

Click on Subject from selection.

Click ok

Now we have just the extra generated pixels selected.

Press generate without typing anything into the filed.

And now it blends the image into the background perfectly including shadows.

(You may think, you could just select the truck and choose inverse and generate. If you do that, it replaces the ENTIRE background, so you have to do it the way I just showed you).

8, Generate again

If you don’t like the result, press generate again to get 3 more variations.

(bonus tip) this also works after the fact, you can open the image and choose any generative layer to make more variations.

Click generate and get 3 more options each time.

9. Reduce file size

Each variation contains and image and adds to file file size.

Rollover the unneeded variations and you’ll see a trash can. (Bin for British, Kiwis and Aussies).

Click the trash can to remove the variation

Here we just kept the one we need. You can always generate more, it will remember the prompt on the generative layer.

10 to remove content

Use the lasso tool to make a selection around objects to remove.

Don’t put anything into the text field, just hit generate and it will remove the object, don’t type remove.

bingo

Note, you will often Gert faster results with the remove tool, plus it works in high resolution, doesn’t use generative credits and doesn’t need to be connected to the internet. (see Generative Fill vs Remove tool ai here)

11. Don’t micro manage Gen Fill

When removing objects, some people (too many) think if you type “remove” it will help get a better result.This isn’t the case according to Adobe documentation from day 1.

If you want a boat, just type in “boat” don’t type “add boat” it doesn’t help.

12. Re generate part of a generated image (Bonus tip)

Another of my own tips you don’t find in the Adobe documentation.

The front of this boat looks good, but the back is weird. You don’t have to generate a different boat hoping for a better result.

Just make a selection around the area you don’t like.

Type in boat and get a better back half. Yay! (Why didn’t I think of that? Oh, wait, I did)

13. Negative reinforcement

You can direct Gen Fill by telling it what you don’t want. Say you ask for an old car, and want a run down modern car for a post apocalyptic scene, and not a classic car. Type “Old car, not classic” for example. (I wish I thought of this example instead of the one I used in the video, it’s a better idea.)

Make a selection around the hand. Type Black Glove not velvet. This way I won’t get a velvet glove for our sports theme.

Looks like I got a nice latex, and woolen glove collection in the variations.

We didn’t get back either and we got 6 fingers, perfect for the piano part in the movie Gattica. (Its not perfect, but its getting better all the time)

14. Generative Expand 

We can expand the edges of a canvas by generating pixels in the blank areas. I use this one a lot!

Choose the crop tool

Make sure the option is set to Generative Expand from the top toolbar.

With the crop tool, drag outside the image boundaries.

Press enter and Generative expand will ex[pand the background to match the image.

15. Generate a Different environment

Select your subject, using select subject.

Press Ctrl/Cmd+Shift+I to inverse the selection and select everything but the subject.

Type in something different like beach

Now we get a different background. (BTW this is what would have happened in tip 7 if we simply inverted the selection)

16. Tell Adobe what you think

When you use Generative Fill, you can rate your result and let Adobe know your thoughts.

Click the menu and choose thumbs up or down, or report or delete.

A questionnaire will appear. These are read by real people, so this is your chance to be heard by Adobe.

Here are a few more bonus tips I have for you..

Content Warning

When dealing with people mostly, you might get a false content warning. To get around that, type a period only “.” into the box and generate.

Create in High Resolution

Photoshop generates in 1024 x 1024 blocks at the moment. See my tutorial on how to get High Resolution in Generative fill. Also includes a free action.

Mixing Variations hack

See how to mix variations to get a perfect result

Bonus tip: My new Course on Generative Fill

As you may know, I made a course on Generative fill (the first and as far as I know, the only course) and everything on the course works exactly the same in the final release, except you don’t have to download the beta anymore. If you missed the discount before, I thought it would only be fair to offer it again here if you are now ready to jump in.

I’ll give you a week to use the code GENFILL2 to save 20%.  Grab it here (If you already bought the course, you don’t need to buy again, it’s unchanged, as the features are unchanged)

Check out the ai Art category here at photoshopCAFE for many more tips and tutorials on this topic.

Thanks for checking this out, I hope it helped you! Tell your friends about the CAFE.

Colin

The post Photoshop 2024 Generative Fill Tips – 16 Fast Tips and tricks appeared first on PhotoshopCAFE.

I recently made a tutorial on all the new features of Photoshop 2024. I got a lot of questions and comments on Youtube and various social medias, so I’ll address the most asked ones here. There is also a video well worth watching above (It’s more in depth).

Is Generative Fill a paid add on?

A lot of talk about this one!

Generative Credits

Adobe is introducing Generative credits. The easy way to understand this is like a phone plan or data plan. An unlimited data plan allows unlimited data, but once you pass a threshold the speed may be throttled during peak time.

If you are a paid subscriber, you will get fast processing credits. Once you pass that limit, you can still generate without extra costs, but the speed may slow down if the servers are busy.

You don’t have to pay for the credits. Each month your meter will refresh.  AKA top off. You can’t accumulate credits and roll them over.

You can see the amount of credits you will receive each month in this chart from Adobe’s website. FAQ here

This doesn’t begin until Nov 1, 2024

Main ones are
Full CC: 1000 Credits per month.
Single app (aka Photoshop) 500 per month,
Photography + Lightroom CC 250 per month. New subs after November 1 will get only 100 per month.

What if you run out of credits?

To be honest, Most people won’t get though 1000 or even 250 credits per month. But if you do, then what? You can continue to use the feature and you may experience slow downs, or even a daily cap  (We don’t know what that means yet, because it hasn’t happened, but my guess is the cap is to stop people abusing the system, like starting a generation farm and selling as stock).

If the speed throttle is bad, you have the option to buy more credits fro $4.99 per month for 100. So basically it gives you priority processing, and you go to the back of the line when you use your monthly credits. They refresh each moth like the minutes on your phone.

Generative credits do not apply to the Remove tool to other ai features in Photoshop, only Generative Fill and Generative expand. Adobe express and Firefly it also applies to Text effects and Text to image.

Is resolution still limited to 1024 x 1024

the generated resolution is 1024 x 1024 px

This means 1024 px on the longest side. If your selection is larger than 1024  high or tall, the generated pixels will be enlarged to fit the selection resulting is a softer looking image. At selections of 2or less, the resolution should match your image.

I have a tutorial here on how to get high resolution fills in your image, see the tutorial here

Where is High Resolution?

Here is my best guess. I don’t know for sure, if Adobe had told me, I won’t be able to talk about it because of NDA, but I don’t, so I’m only speculating.. but it’s an educated guess

On Adobe’s site they say future capabilities may include Higher Resolution, may incur additional Generative Credits or additional cost. Lets examine that

If you compare to Mid journey, you have the ability to generate new images, or to upscale any of the variations.the U1-4 mean create high res on demand.

When you generate high resolution, it costs 1 credit.

My guess is Adobe will allow us to generate in low resolution (like now), but in the future we may be able to choose to redraw a variation in high resolution. This is like the current practice of using Adobe Stock for free in low res with a watermark. When we are happy with the results, we license the image and it’s replaced with a clean, high resolution version.

Maybe that high res generation will cost 1 credit, or even 5 depending on the resolution. We will see.

Is it Photoshop 25 or 2024?

They are the same. Photoshop is at version 25 as an app, but all the apps in Creative Cloud 2024 are called 2024 no matter their individual app versions.

Where are my Presets, plugins etc

When you install a full dot version update, (25 vs 24.7 etc) sometimes preset and plugins might be missing or not working. (I actually had good luck with my plugins this time and they all work)

Your plugins should be in: Program Files\Adobe\Adobe Photoshop 2024\Plug-ins folder

The key is to not replace your old version until you can be sure everything is working.

When you choose update from Creative Cloud, don’t remove old versions.

You can uninstall them later if you like.

Also if you do this, make sure you turn off Auto Update, I strongly encourage this anyway, or it could update you and wipe out your presets without your knowing. It’s happened to me before.

Running the correct version of Photoshop

If you need to reinstall an older version of Photoshop you can. To get back discontinued functionality like 3D.

Click the menu in the Creative Cloud app next to the open button in Photoshop

Click other versions

A menu will appear with other versions. You can install multiple versions at once.

Where are the new tools?

You may have the new version installed. Make sure you are launching the correct version if the new features aren’t there. You might Have added 2024, but still launching 2023. Check the name next to the icon.

Where is the Remove tool?

If you are using a workspace that didn’t include the Remove tool, generative fill or other new features at the time, they will be missing from your toolbar.

Click and hold on the toolbar options, you will see a fly-out menu.

Click Edit Toolbar

Click restore defaults and done and everything should be there now. If you had a custom workspace, update that workspace, or the tool will disappear again the next time you reset the workspace.

Wrong version is opening in Lightroom

When you right-click on an image in Lightroom classic, you can choose to Edit In>Photoshop

Some people are only seeing beta and can’t get it to open in the regular Photoshop

For Mac users, go to preferences>External Editing

Change the Photoshop version

When you right-click you will see the correct version of Photoshop, You can target the beta or other versions if you prefer.

For windows users, you need to First Launch Photoshop and then when you choose Edit in, it will open in Photoshop 2024.

I hope these tips are proving useful for you.

Drop a comment and let me know!

Thanks

Colin

By the way, I heard from some people they were waiting for Generative Fill to come to the main Photoshop because they didn’t want to mess with the beta. Well, it’s here now.

As you may know, I made a course on Generative fill (the first and as far as I know, the only course) and everything on the course works exactly the same in the final release, except you don’t have to download the beta anymore. If you missed the discount before, I thought it would only be fair to offer it again here if you are now ready to jump in.

I’ll give you a week to use the code GENFILL2 to save 20%.  Grab it here (If you already bought the course, you don’t need to buy again, it’s unchanged, as the features are unchanged)

The post Updating to Photoshop 2024, Your Questions Answered appeared first on PhotoshopCAFE.

Adobe released the new update to Photoshop and we will look at the new features. Some of these features have previously been in the Photoshop beta and I’ve already done in depth tutorial in each. Where appropriate, I’ll add links to the other tutorials. This is a overview of everything that is new, in the main shipping version of Photoshop.

First: I have already been asked this question many times: “Can I now uninstall the Beta version and just work in the new update?” The answer is, it’s up to you. Both versions can co-exist on your computer. If you were just using the beta for Generative ai, sure you can get rid of the beta. New features will appear in the Photoshop beta before they make their way to the regular Photoshop, so it’s up to you if you want a steak peek or not. As always, I’ll let you know in our newsletter and make tutorials to make you aware of new features.

The floating taskbar has been tamed.

You may have noticed the taskbar that appears when you do certain things in Photoshop.

This taskbar is loaded with shortcuts to common tasks.

But sometimes it gets in the way and jumps over your image

If you click on the Menu, you can choose Pin Bar position. It will now stay wherever you put it.

I heard a lot of complaints that it didn’t remember this when you restart Photoshop and you have to keep pinning it every new session.

I’m happy to say, the bar stays put now, even when you restart.

Updating Photoshop to v25 aka Photoshop 2024

Sometimes people have trouble finding the update.

Launch your creative Cloud app, from the top taskbar on Mac, or bottom on windows. (Or search “Creative Cloud”)

On the left, click updates

Here is a step, many people miss.

Click on check for updates on the top right.

This will look for the very newest updates, often they aren’t loaded in yet.

You will see Photoshop with update next to it. You can see I already updated mine here.

NOTE: This is a new version and will replace your version 2023 if you choose that option.

Be aware, because it’s a full version update, you will have to reinstall your plugins. If this is an issue for you, keep version 2023 as well and launch that when you want to use your plugins.

Generative Fill in now in the main Photoshop version

Gen fill is now out of beta and available in the main Photoshop. It has 3 main uses. Replace, expand and remove. I’ll cover them briefly here, but choose the ai art and tools category here at PhotoshopCAFE, for lots of detailed free tutorials on these tools. Check out my full course with lesson files to learn even more. Because it’s out of beta, Generative Fill is now approved for commercial purposes.

Lets look at replace using Generative Fill ai.

Make a selection around an object to replace.

Type in what you want. Let’s do hair dryer.

Click generate

You will now see the object you described. This is not stock photography. This image has never existed before. The ai looks at all the images it has been trained on, by looking at millions of photos. It then knows what a hairdryer should look like and generates (or draws) one to match the lighting and angle of the photo. I go more in depth on my topical tutorials. Generative Fill tutorial. 

Notice the the properties panel, you can click on 3 variations.

There is also a new button in the Properties panel that allows you to see larger thumbnails.

If you click, generate again, you get 3 different variations. Generate as often as you like to get more options.

Generative Fill for expanding or outcropping

The 2nd use is to extend a canvas, this is probably the one we will use the most.

Say we want to expand this image into a panorama.

Choose the crop tool

Drag outwards with the crop tool to expand the frame.

This is very extreme, often we will do it just a little bit.

Without entering any text, click generate. (You will notice new tips as the expansion is happening). This happens on the cloud, you will need internet access to use this tool.

You will see that it generated pixels to match the existing image.

The new generated pixels aren’t factual, they are randomly created match the aesthetic. So think of this as an aesthetic way of using it, for things like design and not for location factual uses such as a hotel listings. I’ve covered all the ethics in other tutorials, so we will just stick to the tools in this one.

One of the downsides is the resolution of generated pixels is limited at this time. Here is how to hack higher resolution from Generative Fill.

Adobe has said that high resolution is coming.

Generative Expand

Before getting to the 3rd use of Generative fill, let’s look at a closely related tool. Generative expand.

Generative expand is basically Generative fill, but with fewer steps and the ability to use the sizing presets in the crop tool. I have a separate tutorial on Generative Expend here. 

Choose the crop tool and use an aspect ration preset if you like.

Let’s make a square with 1:1

Drag with the crop tool. Position the image within the frame.

Click generate

The crop tool will crop away the sides and generate pixels at the top and bottom to fill the square.

This is game changing for social media, and changing image aspect ratio. Previously, we had to do a lot of cloning and content aware fill to do the same thing.

Generative Fill for removing object

The 3rd main use for Gen fill, is removing things from a photo.

Make a selection around the car.

Click generate

The car is gone and background has been created to fill the hole.

Generative Fill is amazing, but there are things it can’t so, or do well, and we address that next. First more info about the changes.

Generative Fill as paid add on

Generative Fill is entering a paid phase in its life. Don’t freak out yet though, let me tell you what’s happening.

All this ai stuff takes an incredible amount of super-computing power. The processing is done in large and expensive data centers. So it’s not free for Adobe to run, in fact it’s very expensive. So here is what Adobe has announced:

Starting November 1 2023 there will be a limit on generations. For Full Creative Cloud members, it’s 1000 a month, for Photoshop and photography program, it’s 500 a month. Each generation counts as 1. Honestly, from using a paid mid-journey account for a year, most people won’t get anywhere near the 500 or 1000. But if you do hit that allotment, you will still be able to generate, but it will be slower. You can buy more fast credits if you want.

Credits refresh each month and you can’t accumulate credits, it just restarts. From the lounge on Adobe’s website, it sounds like the higher resolution generation will count as more than 1 credit each. No more is known at this time.

Only Generative Fill and Generative expand count as credits. The remove tool and other ai tools don’t use credits.

Remove tool

If you want to remove things from your photos, the ai powered remove tool is a better option on many cases. I compared the remove tool ai with Generative fill here. 

Choose the remove tool

Paint over the plant.

The plane disappears and is filled with natural background.

New way of using the remove tool

This is a brand new behavior. You can circle the object you want to remove, much like you would with magic eraser on the Google Pixel phone.

And it disappears.

Advantages of the remove tool over Generative fill include:

1. Often faster
2. On device, so no internet needed
3. Works in full resolution
4. No ai Gen credits used

New Gradient tool

The new gradient tool features re editable and non destructive gradients, I made a tutorial here on the new Photoshop gradient tool

For the sake of this example, let’s do lighting on a mask. I made another tut here, and it’s popular!

Start with a wall

Lets make it nigh-time

Add an adjustment layer

Choose color lookup

Click Load 3D LUT on the properties panel

Grab moonlight

Now it’s night.

To add a spotlight, choose black to white as fairground and background

Choose gradient tool (make sure it says Gradient, and not Classic Gradient)

Grab black to white and make it a radial gradient.

Drag to add a spotlight effect

Drag the dot on the side to change the shape.

You can add more than 1 light, see the video at the top

Some people wills, we could do that in Lightroom and Camera RAW. You would be right, but remember, this is a fully functional gradient tool that is capable of so much more, I’m just using it on a mask here to demonstrate.

Ok, that’s a wrap up of the new features in a nutshell. Check out all my other in-depth tutorials to explore there features more.

Thanks

Colin

By the way, I heard from some people they were waiting for Generative Fill to come to the main Photoshop because they didn’t want to mess with the beta. Well, it’s here now.

As you may know, I made a course on Generative fill (the first and as far as I know, the only course) and everything on the course works exactly the same in the final release, except you don’t have to download the beta anymore. If you missed the discount before, I thought it would only be fair to offer it again here if you are now ready to jump in.

I’ll give you a week to use the code GENFILL2 to save 20%.  Grab it here (If you already bought the course, you don’t need to buy again, it’s unchanged, as the features are unchanged)

The post Photoshop 2024 New Features appeared first on PhotoshopCAFE.

In this Lightroom / Camera RAW tutorial, we will learn an easy way to add edge lights and background lights to separate an image from the background.

This is an effective way to add some pop to a photo. In this case, we are using a person. The technique will work for any object and also works well in product photography and even photography for ebay.

In the Lightroom Classic develop module, or in Camera RAW (they are both the same and this tutorial applies to both Adobe apps.)

Click on create mask

Choose Radial Gradient

Drag out the gradient to add a spotlight effect on the background.

Adjust the Exposure to brighten the light.

Change the temp to add color to the light. In this case we are making it more blue (cool light).

Let’s add a second light on the other side. (This wills show you how to use both gradient types).

Choose Add, and Linear Gradient

Drag out the gradient to add some light that would be originating just above the model and off to the side.

Because we are adding to the mask, notice it takes on the settings of the previous gradient. If you want different settings, create a new mask, rather than adding.

Let make sure the lighting only affects the background. We will customize the lighting on the model. The exclude our person (Julia, is the model’s name), choose Subtract and select subject.

The model will be unaffected by the lights now.

Lets add edge light to the model

Click Create New Mask

Choose Subject

You will see our model is masked.

Now we want to use the mask as a stencil, so we can add the rim light, but have the edges protected.

Hold down Alt/Option in Lightroom (Shift in ACR) and you will notice add/subject now says Intersect

Choose interest with Brush

Intersect will only add a mask where the 2 masks overlap. This is magical when adding an edge/rim light, because we don’t have to worry about painting outside the model’s mask.

Paint around the edge

Brighten the exposure

Notice it now looks like a back/side light is on the model, to match the spill on the background. This light wouldn’t be placed in front of the model, as it’s acting like a hair light, (but larger and lower), kicker, or separation light, whatever you decide to call it. (I say this because I always get comments on the lighting and I have lit enough people in the studio to know how lights look).

Add more light on the other side to match the radial gradient we created in step 1).

I hope you found this simple tutorial useful and that you can adapt it to your own projects and go even deeper with it. Let me know in the comments.

I just want to mention, this is a small taste f the kinds of things I teach in my brand new Masking in Lightroom and Camera RAW course. (And no, it isn’t just for portrait photography, a lot of it is landscape as well and all lessons work on any type of photography).

Thanks for checking it out

Colin

Masks in Lightroom and Camera RAW (in Photoshop)

This new 3+ hour set of 26 tutorials will help you master the use of the new Masks in Lightroom or /and Camera RAW. Get it here

The post Add Edge Lights in Camera RAW and Lightroom, with ai Masks appeared first on PhotoshopCAFE.

Here is a simple way to get rid of edge halos on mask edges in Lightroom or Camera RAW. We are in the develop module of Lightroom Classic, but this tutorial will work exactly the same for Adobe Camera Raw.

Let’s lighten the foreground and gazebo without changing the sky and water.

Click the new mask button

Choose the ai Subject

A mask will automatically be added to the subject. Sometimes, it will get the whole thing and other times, it will get a partial selection as shown.

We need to add the grassy foreground

In the Masks panel, click Add

Choose Objects

You will see a brush and a selection option in the mask panel. Choose brush.

Paint over the area you want to select. It doesn’t have to be perfect, just enough to let Lightroom/ACR know what you want it to select.

The ai will do its thing and make the selection.

Adjust the Exposure and Shadows to brighten the foreground

Notice we have a light fringe on the transition between the ocean and the foliage. This is called a halo, and it’s common.

Lets fix it

Inside the mask, choose subtract

Choose Brush

Set the feather and flow to 100.

Turn on Auto Mask

If you look at the cursor, you will see a minus sign. This means the brush will remove.

The way auto mask works is to keep the center inside the areas you want to select. We will take the area between the 2 rings and use that to retouch our image. Watch the video at top to see it in action.

Begin to paint

Notice just the area between the 2 rings is going over our foliage, and it’s working nicely!

Here is the result. Some things can be really simple, but still effective.

I hope you enjoyed this tutorial and found it useful!

Thanks

Colin

This is just a little taste from my brand-new course on Masking in Lightroom and Camera RAW.

If you want to master these very important image editing tools, check out the course. 

Masks in Lightroom and Camera RAW (in Photoshop)

This new 3+ hour set of 26 tutorials will help you master the use of the new Masks in Lightroom or /and Camera RAW. Get it here

Get the course here

The post Easy fast way to fix edges of masks in Lightroom and Camera RAW appeared first on PhotoshopCAFE.

Adobe has released another new tool in the Generative ai family. This time, Generative Expand, makes it easier and quick to expand a photo or change its size or shape. See how to get the most out of it here.

How to Change the Shape of a photo and Fill

At this time, this feature is only in Photoshop beta. Here is a short tutorial on how to get the Photoshop beta.

Here is one of my photos. I shot it horizontal, but I want to make it vertical or square, but I don’t want to ruin the shot by cropping anything out. Generative Expand will help.

Choose the Crop tool

You will notice you can choose a ratio from the floating toolbar, or the top toolbar).

We want square so Ill choose 1:1

Drag the crop tool to the crop you want, it will be locked to a square.

You can also reposition the photo within the square.

Press generate

Photoshop will generate the area in the sky and water to expand the photo.

You have 3 variations to choose from the in properties panel.

A couple of notes on this.

First of all, it is the same as generative fill and uses the same algorithm. In fact it is generative fill, but it’s saving you a few steps in making the selections. See how this works in Generative Fill

Because it’s Generative Fill, it still has the resolution limitation of 1024×1024. For posting on social media, this may be fine for some images. Here is how to get higher resolution in Photoshop Generative Fill

Let’s look at other things we can do With Generative Expand.

Crop an image to a specific size

We can crop a photo to an exact size if we want.

Choose the crop tool, make sure Generative Expand is selected. If you don’t see it, make sure you have launched Photoshop beta.

Choose WxHxResolution

Let’s make this new image 4 inches by 5 inches at Print resolution (300 ppi (dpi once printed))

Enter 4in, 5in and 300 for resolution.

Now you will see the crop shape. When you apply this, the selected area will be resized to 4″ x5″ and any transparent areas will be filled by ai generated pixels.

Tip: Tap the x key to toggle between wide and tall.

When you are happy with your crop, apply it by clicking the checkmark at the top, or the Enter/Return key.

And here we have a beautiful crop.

Telling Photoshop what to fill with using prompts

So far, we have filled the edges based on the existing photo, and this will be enough for many people. This is how I will be mainly using it.

However, Generative Expand is capable of more, and I’ll demonstrate here.

Let’s expand this photo, with very specific things, maybe we are making a CD cover, or a Tunecore cover for Spotify/Apple Music. (Tunecore/Apple Music uses 3000 x3000px btw)

Choose the crop tool and expand the image

Let make it fill the top with a mountain scene

Type in Mountains and press Generate

Here we go. Notice it respects the depth of field.This also lets us get away with the lower resolution generation.

Let’s take it further

Choose the crop tool and expand the bottom.

This time, type in lake in the prompt

This is what we get, I love that it even creates the reflection.

I hope you enjoyed this tutorial – I have a full course On Generative Fill if you want to learn ai generation in Photoshop

Drop a comment and let me know if this was helpful.

Browse the other free photoshop tutorials here at the cafe

Download 100s of dollars worth of Photoshop adds on for free here at the Vault.

Also follow me on Instagram, Tik Tik, Threads and Twitter.  for shorts and reels.

See you next time!

Colin

The post How to Resize a photo to any size and fill the overcrop with Generative Expand in Photoshop appeared first on PhotoshopCAFE.

Because we are using Generative Fill, it’s only found in the Photoshop beta at the moment.

 Here is a short tutorial on how to get the Photoshop beta.

Most video are horizontal. Sometimes you want to change a horizontal video to vertical or to square for social media such as Instagram reel, Youtube shorts or tie Tok. Or maybe you just want to recompose a video and need to extend the edges. Here is how.

On the left is before, on the right is after. Its generative fill, but this is working on a video in Photoshop.

Choose file>Open and choose your video, it will open in Photoshop and the timeline will appear. (Not all video formats are supported, but .mov H.264 is the most commonly used and supported format.)

For more on Photoshop and video check out this tutorial 

If you look in the layers panel, you will see a video group. Everything in this video group will be on a single layer on the timeline. You will see why you need to know this soon.

Lets extend the canvas.

Choose the crop tool

Drag out the crop to extend the top or bottom. Tip: Find an edge that doesn’t have movement, or shoot a photo that only has movement in the middle. This technique won’t work properly on a video that has edge to edge movement.

Press enter to apply the crop. We have a 1:1 square crop here.

Grab the rectangular marquee selection tool

Drag over the blank area, making sure to overlap the existing pixels a little bit.

Choose Generative fill

Enter text for what you want to fill the blank layer with and then press generate

Because we are in the Video group in the layers panel, you will notice the generated layer (in purple) is added at the end of the timeline, rather than stacked on top of the video

Grab the Generated layer and drag it above the Video group in the Layers panel.

Drag the clip to the beginning of the timeline.

Drag the end of the generated layer in the timeline, so it’s the same duration as the video.

You can see its now extended.

Choose your favorite out of the variations (How to use Generative Fill in Photoshop)

Press the spacebar for the video to play. Notice how convincing this is. We have essentially created a Cinemagraph

Exporting the video

Once you have created this, you can save it as a PSD and everything will be as it was when you reopen the file. To share or distribute the result, it needs to be encoded as a video.

In Photoshop choose File>Export>Render Video

Use the default format or H.264

If you use the high quality preset, the video will render using the current dimensions.

You can choose different options from presets if you want to share the video on a specific platform, as it contains YouTube presets and more.

To learn more about video in Photoshop, I have created 2 courses

Video In Photoshop (Also available as a book from Peachpit press, authored by myself, Colin Smith)

Making Movies In photoshop

Check out the video above to see this in action.

I hope you found this tutorial helpful, browse 100s more here at the CAFE

See ya,

Colin

Also follow me on Instagram, Tik Tik and Twitter.  for shorts and reels.

My new course on Generative Fill in Photoshop: https://photoshopcafe.com/shop/generative-fill-in-photoshop-introduction-course/

The post How to Extend a Video in Photoshop with Generative Fill, ai Cinemagraph appeared first on PhotoshopCAFE.

In this tutorial, we will look at how Object Selection differs from Select subject. Then we will do a comparison between the two tools and I’ll provide some tips and a workflow to get the best results. Check out the video to get more details.

How Object Finder and Select Subject differ

For either of the tools, you will want to choose the Object Selection Tool first.

At the top, choose the drop down menu and select Cloud. This will always give a better result than Device. (You can permanently change it in preferences).

Click on Select Subject and a Marching Ants selection will appear around the main subject(s).

In this case it selected the three woman.

Ctrl/Cmd+D to deselect.

Let’s try the Object finder. With the Object Selection tool still active, make sure Object Finder is on

As you roll lover selectable objects in the photo, they will appear highlighted.

To select them, click and the selection will be created.

The main difference from Select Subject, is the ability to select different objects individually and isolate them. In the past, the quality of the selection was inferior to the selection quality of the Select subject. As you will see, this is no longer the case.

Using Object Selection, manually

Another function of the Object selection, is the ability to tell Photoshop what objects you want to select, these can be large or tiny, simple or complex.

Turn off Object Finder and use rectangle mode.

Draw a box around the object to select.

The ai will make its best selection around the object.

Combination selections

Make a box around part of an object, the box doesn’t even have to be accurate, just good enough to tell Photoshop what you want to select.

The selection will be made.

To add to a selection, hold down the Shift key and make a box around another object

Both are selected.

What if we want to add the shape in between?

For odd shaped selections, choose the Lasso mode.

Hold down Shift and make a selection

Notice it added to the selection nicely. We need to remove the empty area inside the shape.

Hold down Alt/Option to remove from the selection.

Drag inside the shape to remove the center.

And the ai does it’s thing.

If I add a layer mask, you can see how the selection looks. Not bad, for nat making any modifications.

Select Subject vs Object Finder

Ok, the moment of truth. Let’s compare the quality of selections between the 2 ai powered out selection tools.

We will start with Select subject and a subject that has a lot of hair. Yes, it’s a simple background, but the point here is to compare the two tools and the quality of the edges, not compete in a difficult selection challenge.

Here is the selection. I won’t make any refinements or modifications in this test, just a single click only. At the end of the tutorial, I’ll show how to refine the edges.

Lets click the layer mask button to mask this Select subject.

Onto Object finder.

Make sure Object Finder is on.

Click to select

I added a mask to this too. and put the 2 different selections next to each other for you to look at. We have Object Finder on the left and Select Subject on the right.

We will look at different colored backgrounds. The first thing that struck me, is this tool is now usable, the selection isn’t bad at all. Its not perfect, but remember, we haven’t worked it yet and all selection tools, just give you a starting place.

closer up, you can see things like the eyelash.

Here it is on white

and on black, I notice the Object finder (on the left) seems to do a better job with the fine details on the hair.

Here is a different selection. Select subject is doing better with the smooth details on the back of the skater, however its missing part of the board (all of these are easy fixes).

Object Finder Special features

Lets look at some of the settings

If you click this icon, you can show all selectable objects at once.

Each of the highlighted areas can be clicked on to automatically load the selections.

Click the gear to see the settings (these are all covered in the video)

The main ones are the ability to change the way the overlay is displayed.

Refresh is useful when you are compositing. As you work on the image, Photoshop will constantly analyze and update the overlays. If you choose Manual Refresh, the analysis will stop. Click the rounded arrow to manually update (refresh) when you are ready.

Refine the selection

Whichever tool you use, you will still have to refine the selection for production-level selections. I’ll briefly show you the main steps here. I have many tutorials here at the CAFE that show how to refine the edges of selections and they are all still valid. The ai tools, add more tools to make quick selections, but they don’t replace the other tools, which in many cases, can still yield better results with a little more effort.

5 ways to fix selection edges in Photoshop

Make the selection with Object Finder (or any selection tool).

Click on Select and Mask

Now we enter the select and Mask workspace. (Maybe I should do a deeper tutorial on this tool and explain all the options sometime?)

The main settings are here.

First, move transparency to 100%, so you can clearly see what you are working on.

If you have hard edges that are a little uneven, try these settings.

Radius to 1px.

Tiny amount of feather and contrast. Only use smooth if there are objections jaggies, using smooth makes it almost impossible to get good hair edges, but you could work in 2 passes. Beyond the scope of this tutorial.

His jacket it looking really good, but the hair on his temples are semi-transparent.

grab the refine brush from the left panel.

paint over the hair edges

This is a great tool for hair, fur, smoke and glass.

Choose decontaminate colors to remove edge halos

The best way to output is New Layer with Layer mask.

And the Final Cut out. These are good tools to remove the background in a photo in Photoshop.

I hope you found this tutorial useful. Tune in for a new tutorial each week and browse the hundreds of tutorials here at the CAFE (at least bookmark, as PhotoshopCAFE will be a valuable resource for you, as it is for millions of other users.)

Thanks

Colin

Also follow me on Instagram, Tik Tik, Threads and Twitter.  for shorts and reels.

My new course on Generative Fill in Photoshop: https://photoshopcafe.com/shop/generative-fill-in-photoshop-introduction-course/

The post Object selection tool in Photoshop got an update. How to make quality ai selections appeared first on PhotoshopCAFE.

With all the excitement around Generative Fill in Photoshop, its kind of overshadowed another amazing new ai tool. The remove tool, is equally an amazing, and maybe even more useful? Watch the video, or read on to see.

I made a tutorial on the Remove Tool when it was in Photoshop beta. Now its in the shipping version.

I have made a course on Generative Fill in Photoshop. Check it out here

Before I explain the differences in the tools, lets put them through their paces. The video shows the settings and variations of this basic theme I used.

At this time, Remove tool in in the main Photoshop version, generative fill is only in the Photoshop beta. Here is a short tutorial on how to get the Photoshop beta.

The Remove Tool

For the Remove tool, I created a new layer and then painted on the objects I wanted to remove.Creating anew layer gives non destructive use. Make sure to turn on Sample all layers for it to work. The remove tool in in Photoshop beta and also Photoshop 24.5 or newer.

The remove tool, is nested under the spot healing brush. If it doesn’t show in the toolbar if you are using a workspace preset. Click the 3 dot menu on the toolbar and choose edit toolbar to show the tool if its missing.

Make a new Layer

Paint over the object you want to remove

It removed the object and rebuilds the pixels behind the object based on the surrounding pixels and ai. Much better than Content aware fill, which is not ai.

I also removed the shadow in the video.

Meet Generative Fill

Generative fill is only in Photoshop beta right now. It only works on the cloud, so an internet connection is required. Generative fill is capable of creating objects and extending a canvas, In this comparison, we are only going to use it for object removal. I have plenty of other tutorials on other usage of Generative Fill.

For this tutorial, I used the lasso tool to make selections, and then chose, Generative Fill

Generate to fill with Generative fill, which uses a machine learning model than has been training on millions of images from Adobe stock.

It fills the area the car used to be with relevant pixels.

Removing a large object with repetitive background.

Lets remove this cabin.

Starting with Generative Fill.

Make a selection with the lasso tool.

Choose Generative Fill>Generate, leaving the prompt blank.

Tip: According the Adobe documentation, adding words like remove, don’t work. If you want to remove something, leave the text prompt empty. “Currently, the prompts used for the description are supported in English only. For best results, use descriptive nouns and adjectives in text prompts instead of instructive prompts like, “fill the area” or “create a scene“.”

Here you can see it didi a decent job.

Let’s grab the remove tool

Paint over the building and release

We also get a decent result. Although with this, we can see some repetition in the trees.

On close examination, the fill with Remove is the same resolution as the original image. It also works faster than Generative Fill.

If you look at Generative Fill results up close, they are soft and pixelated. That is because Gen-Fill uses a maximum resolution of 1024×1024 px to fill a selected area. This will change in the future. Here is a tip to get better resolution.

Removing a Difficult object from a picture

In this shot, we want to remove the tower.

With the Remove tool, paint over the tower first.

Then remove the power lines, it works well.

The moment of truth. How will it fare, while painting over an area of semi transparency, such as this veil.

It works wonderfully!

Ok, let’s try Generative Fill

Select the tower.

Hit Generative Fill>Generate

Oops, not so well here. None of the 3 variations are any better.

Lets try selecting the area through the veil.

Not too bad,

Note, if the entire tower was selected in a single selection, Generative fill worked much better. If you select part of an object, it tries to extend it rather than replace it.

You can’t type in remove, as it doesn’t understand those types of instructions, I tried sky, but also got weird results.

Removing a person, with a complex background

Starting With Gen Fill. Make a selection with the lasso.

And this did a fantastic job.

Same thing with Remove tool.

A very muddled result in the background.

Applying the Remove tool a few more times over the background changed it into a more acceptable result, although not as good as gen-fill on this one.

Removing a face

Let remove a person from this group.

Make a selection with lasso tool

Choose Generative Fill>Generate

On all 3 variations, Gen-Fill added a new face instead of removing the person entirely. On a second attempt, it did the same thing.

Let’s try the remove tool

I had to use it on smaller areas to rouch up the result, but it was very easy to remove the person.

The Conclusion

It looks like each tool excels in different circumstances and types of images.

The remove tool works very well on smaller areas and simpler backgrounds, although it can handle more complex backgrounds, you need to work it a few times to massage the results.

The remove tool is faster because it’s installed locally and thus doesn’t require an internet connection either.

The results are in high resolution and they are ready for prime time, because they are in the shipping version of Photoshop.

The Generative Fill can actually generate objects and thus works well on complex backgrounds, but remember its still low resolution and not yet for commercial purposes (beta).

The very best solution, is to use both together. I often start with Generative Fill and then use the remove tool to manually tweak the results.

I hope you found this tutorial useful, definitely check out the video if you haven’t, there is so much more o be gleaned than this brief writing.

Check out my new course on Generative Fill in Photoshop, Check it out here.

Also follow me on Instagram, Tik Tik and Twitter.  for shorts and reels.

Great to see you here at the CAFE

Happy Photoshopping

Colin

The post Generative Fill vs Remove tool ai in Photoshop, when to use each tool appeared first on PhotoshopCAFE.