Peter Werner

You are most likely looking for the Perlin Noise filter, which is located under Filter > Noise > Perlin Noise. Instead of giving you a different result every time like in Photoshop, it actually gives you parameters, which is way better. Also unlike in Photoshop, you can change your foreground and background colours interactively using the Colour panel while the filter dialog box is open. No need to cancel, change your colours and go back into the filter like in Photoshop. You can even set one of the colours to transparent! However, you should be aware of the following bug: Changing the Opacity parameter will currently lead to the preview disappearing. It will still render fine and the preview will also re-appear when changing any RGB components. Setting noise in the colour panel (hidden feature when you click the preview circle to the left of the Opacity slider) will also trigger the preview bug, and any per-colour noise you set will be ignored when rendering (the latter is intended behaviour I think). But since we are in Feature Requests: One thing I would love to see is more types of noise in addition to the standard Perlin Noise, like the "Fibers" filter in Photoshop or the "Fractal Noise" filter in After Effects, which can generate all sorts of handy procedural stuff like Voronoi ("Cell") patterns and the like.

I mean literally all lowercase as a formatting option/text style attribute. In languages such as German, a lot more words in a sentence are capitalized than in English, such as all nouns, proper names and so on. This leads to much more mixed-case text than in English. As such, it has to actually be typed as mixed case in order to appear correctly when no special capitalization settings are applied. If due to aesthetic considerations or due to the way the font was designed, the decision is made to go for all uppercase or all lowercase (and possibly back again, we all know how it is), it is incredible useful to have this as a character attribute that can be set via text styles without having to re-type, say, all headlines in a publication. I have been in that situation even for English text a few times, and I can say from experience that it also leads to errors since you are inevitably going to miss the occasional word. Also if the content is to be reused and the publication is, say, exported to HTML or ePub, you wouldn't want this kind of styling to be baked into the text, be it because of search engines, because of font differences, or because you want it to use CSS text-transform. In general, I think it is a good idea to keep formatting and content separate as much as possible. This is especially critical for Publisher once it comes out, of course. In Designer and Photo, documents will generally be less complex, so re-typing a few lines of text is, while inconvenient, is a lot less of an issue of course since it is probably feasible to do it manually. I imagine you will also be factoring out Affinity Publisher's story editor persona into a separate application at some point in the future to offer something along the lines of Adobe's InCopy and Quark's CopyDesk. In such a workflow, you would then as a designer of a larger magazine or newspaper have to tell all your editors to type out all headlines in lowercase, which then means they will be fighting auto correct and spell check, and it will possibly get messed up or at least flagged again when somebody decides to proofread the copy in story view.

Right now, an Affinity Photo document can only be RGB, Grayscale or LAB at a time. However, there are several workflows where it would be beneficial to work in more than one. Examples are: Black and white conversion – you would like to end up with a grayscale image, but work with RGB data using channel mixers, HSL and so on and be able to edit these after the fact. The final image being saved as Grayscale is so important because it ensure that it ends up being separated into the CMYK black channel, leaving CMY clean. Applying LAB mode corrections like LAB curves inside of an RGB mode retouch (could also be implemented as an implicit conversion as part of the curves adjustment layer like in PhotoLine though) Manually fine-tuning the result of CMYK conversion by manipulating the image in RGB mode (possible through soft-proofing, but requires a manual CMYK-export in the end and you have to re-do any manual adjustments that may have been made on the resulting CMYK file every time) These will be even more relevant once work files can be saved natively as TIFF retaining layers and so on since you can then actually place your working files in third party software and not have to export a bunch of separate files. Especially in the B/W case, you currently have to keep your RGB original retouch, then your BW conversion RGB file, and then a final single-channel BW exported file. Currently, there seems to be no other way than to flatten the image and convert to another mode as far as I can tell (haven't tried placing embedded documents yet though). Photoshop provides a rather clumsy workflow where you can convert your entire document into a smart object and then change the colour mode of the enclosing document. But especially if you just want to quickly apply LAB curves a bunch of times in your retouch, the nesting just becomes unusable. If you had simple adjustment layers along the lines of "Convert to LAB" and "Convert to CMYK", that would make it easy to mix the best of all worlds in one single document.

I'd love to have an "all lowercase" setting in addition to "All Caps" and "Small Caps". It is not uncommon for designs to use only lowercase text, so it would be a handy feature. I have had to do this manually in InDesign so often I lost count years ago, only to occasionally then manually restore all the uppercase letters when someone changed their mind.

In addition to the current sampling settings in the Healing Brush etc., a "Current Layer and Below in Group" option would be very useful. One frequent use case would be frequency separation. I like to use two groups, one for the low frequency and one for the high pass image. I can then create additional layers in each one. For the low pass group, that's pretty simple since "Current Layer & Below" will work fine. To work with layers inside the high frequency group, however, I need to manually hide all layers below until I can clone/heal from the high frequency image only onto a new layer. That means I currently get no live preview of my results, and checking what I am doing involves manually selecting and toggling all layers below all the time. Being able to restrict sampling to the current group would solve these issues. Also, a checkbox inside the Frequency Separation command to create groups with layers in them instead of single layers would save the trouble of creating the groups, setting the high frequency layer's blend mode to normal and setting the high frequency group's blend mode to Linear Light every time, and it could even create empty layers above each base layer as well. While I am at it, it would be very nice if the Blemish Removal Tool could also get sampling options since that is really the only way one could work non-destructively with that particular tool.

It would be great to have global clipping warnings inside of the regular Photo persona. Photoshop has implemented clipping warning overlays as part of a few tools, such as Levels and Curves, but it won't for instance help you when you are trying to use the dodge and burn tools to get a pure white or pure black background and other workflows. Since I assume that this is already planned as part of the Curves and Levels improvements on the roadmap, I would suggest to make it a view option, accessible either from the View menu or in the Histogram/Scopes panels that can be used at any time, with any of the tools available instead of adding it only to two or meticulously adding it to all sorts of adjustments manually. Holding Option inside Levels/Curves and on the Luminosity slider in HSL adjustments, Black&White etc. would then temporary toggle that option to the opposite of its current setting (i.e. on when it's off and off when it's on). Video cameras often implement the overlay as a zebra pattern, and you can customize the threshold at which it shows up. That's actually a useful workflow for other areas as well. For instance, you could have your clipping warnings at 95% if you wanted to see which areas are reaching a range that while they might not be clipped, would not show detail when offset printing the image. A preference to animate a zebra pattern inside the clipping warning, like the marching ants of selections do, would also make it help stand out when the image is visually very similar to the overlay pattern.

I would love to see something like what Tandent Lightbrush supposedly does (see this article) implemented inside of Affinity Photo. It is basically like frequency separation, except that it separates illumination and surface detail instead. It could be implemented in conjunction with a blend mode that would re-combine these two passes, just like Linear Light re-combines the high-pass layer with the low-pass one for frequency separation. I don't know if the algorithm is described in a Siggraph paper or something like that, but I believe there have been several published approaches to estimating illumination and so on. I was fully expecting for something to show up in Photoshop soon after because it would be spectacularly useful for photo editing, but so far Adobe hasn't done anything in that direction yet. I'm not even sure if the original product is still around.

Just for reference, I believe the Adobe XMP SDK contains an example that reads metadata from an .indd file and might give some indication as to the basic structure of the file format.

As far as I know, Photoshop does not embed a PSD into the TIFF file, it saves data conforming to the standard and adds some proprietary tags inside the TIFF file (I believe Adobe even have a specification for the Photoshop TIFF format that they used to ship as part of the SDK). Using this approach, way layers and other data can be exchanged with other applications as much as possible while also keeping the full information that would otherwise be stored in a PSD. Curiously, the corresponding TIFF files are usually smaller than the otherwise completely equivalent PSD files. An implementation that keeps the .afphoto file embedded inside the TIFF is a usable workflow for collaborating with raw converters like Capture One and Lightroom, but proper layer support with proprietary tags would enable file exchange with Photoshop and other applications (like enabling use of the "Object Layer Options" command to toggle layers when the file is placed inside of InDesign).

As I suggested in this thread, currently tool settings are remembered only on a per-document basis. For some tools, some people like to use different settings than the defaults (such as the Healing Brush layer sampling options in my case). This means that with the current implementation, I need to change those types of settings every time I create a new document and use a particular tool for the first time. This is obviously not ideal. I see two possible solutions to improve the workflow: 1) Add a "Make Default" button (and possibly also "Reset Settings to Default") to the tool options bar 2) Add a preference "Remember tool settings across documents" Should you choose to do both, you would obviously only need to show the button(s) if the preference setting is turned off.

That's a good point though, non-destructive layer-based cropping with a dedicated tool instead of having to mess with masks could actually be quite a nice workflow I think. Especially for screen designers and the like.

That sounds sort of like Apple's Exposée, just for layers instead of windows. I could see that being quite useful for complex composites in Affinity Photo. It might also be an interesting concept to apply to pages inside large documents, especially once Publisher comes out. Simply press a key for a bird's eye grid-view of all pages and click the one you want to go straight to it instead of messing about with a Pages panel like in InDesign all the time, which is either way too small or takes up way too much screen real estate when working on larger publications. It would also be flashy and Mac-like enough to be quite effective in marketing videos and the like.

As for the drop down usability issue, I have already asked for that elsewhere as well, so I think the developers are aware of the problem. It's not just limited to Levels, there are other places in the UI that are affected as well, such as the Channel Mixer and the Selective Colour adjustment. Double-clicking sliders to reset them is also something I'd love to see. I think the basic adjustments still need some work in general. Like its Photoshop counterpart, Levels and Curves are currently limited to Master + RGB and missing Alpha support (which Adobe has in After Effects, but not in Photoshop). And there are no output levels and no numeric input for the Gamma slider. There also isn't any clipping preview like when you hold down Option when dragging the slider in Photoshop. Pickers are also absent except for the very basic one in Curves (I'm still hoping for a white balance one in addition to Black/White/Gray point). I'm sure we'll see most of these things at some point, but I really really hope they will be in the 1.0 release since they are so essential.

TIFF is built based on chunks, meaning that the file is a collection of different data segments that are identified by some kind of ID. If a program tries to read a TIFF file, it usually reads the chunks whose IDs it recognizes and just skips all the others, and "the others" in this case means the application-specific ones. This means that once you open such a file that has proprietary chunks in another application, all the data that conforms to the standard can be read fine. The proprietary information, however, is not loaded with the document. Hence when you save that document to a TIFF file again from the second application, the proprietary information from the first software will be lost. So you can't have a TIFF file with proprietary data, open it in another application, edit it, save it, and expect to have the editable data from the original program intact. That's the big drawback of this approach. Photoshop and TIFF isn't the only combination like this by the way. PNG is chunk-based as well, and Macromedia Fireworks that fact to make PNG Fireworks' native file format (even saving vector graphics in them). Other Macromedia programs like Macromedia Dreamweaver and Flash (and after the merger also Adobe Photoshop) were then extended to support this proprietary Fireworks data in PNG files as well.

Awesome, thank you so much!!!! That blend gamma is a dream come true for me! Small thing: The slider starts at 1.0, but Gamma is an exponent, thus values between 0 and 1 should also be available. And I know I am completely insatiable, but there is another thing relating to blend gamma: Could you maybe make blurs (and possibly other convolution filters, but blurs are the main problem) perform their calculations at gamma 1.0 as well internally, or offer a "compute at gamma" option there as well? Following example situation: Fill a document with RGB( 255, 0, 0 ) and draw a medium-sized rectangle with RGB( 0, 255, 0 ) right on top of it in the center of the document and then apply a blur to the green rectangle. At Gamma 2.2, the transition area will be an unnatural looking muddy brown. Now set the blend gamma for the rectangle layer to something closer to 1.0 – now the transition will be a much better and accurate looking colour. Now try to apply the blur to the entire document instead, or just to any regular flattened photo. You are stuck with the unnatural gamma 2.2 edge artifacts there. Blurring at gamma 1.0 leads to much more natural results. While blend gamma for layers is a real power-user-only feature, I think gamma 1.0 blur would actually be a feature that can be used for advertising if you showed gamma 1.0 blur results side by side with results from Photoshop's Gaussian Blur filter, which suffers from this problem as well for Gamma 2.2 images. It might be worth considering to internally composite brush strokes at gamma 1.0 as well in the paint engine at some point in the future (or add a setting, either for brushes or somewhere in Preferences where it does not confuse casual users). I know you have more pressing issues right now, just something to maybe keep in mind for the future. Thank you so much for the blend gamma, that is huge already! EDIT: I just found this video that explains all that much better than I could, they even use the same example: https://www.youtube.com/watch?v=LKnqECcg6Gw

You would not be able to retain all information in your document since there are features in AP that Photoshop and hence the PSD format does not support. For instance, live filters would be completely stripped, or the PSD file would be an unusable mess of nested Smart Objects at the very best. Some filters, like the Affine Filter, unfortunately have no equivalent in Photoshop. That's why I suggested supporting proprietary data in layered TIF files in another thread. TIFF allows embedding proprietary information in addition to the standardized data. Photoshop does this as well, you can save as TIF instead of PSD without losing any Photoshop features or information in your document. If AP supported something like this, other software (including Lightroom, InDesign, Capture One, and more) would be able to read the files without problems (at least in flattened form), but you would still be able to retain all your original editable data. PSD as a file format lacks this kind of extensibility and you would be limited to the features supported by Photoshop.

I know, and you could also paint with a colour of your choice on a new layer above and then Cmd+Click the layer thumbnail to load the alpha as a selection when you are done. But the point is that these are all workarounds. I'm sure all these things will be resolved in due time, though. The guys at Affinity are doing a lot more than just catching up with 25 years worth of development done at one of the biggest software companies in the world, and they are doing it at an amazing pace.

Thank you, this is probably the best update yet!!! Option+Click to edit masks was actually the one single feature that prevented extensive practical use for me, and the layer isolation mode is just great, much better than the Option+Click thing in Photoshop! Now all that's really desperately missing for me are some enhancements to the colour adjustments (clipping warnings in Levels/Curves (Option+dragging sliders in Photoshop), full RGBA levels, pickers, numeric values for gamma in Levels etc.). The new live filters are also extremely handy. I really wish Liquify was supported, hope that will be added at some point. Displace would also be great, ideally linked to a channel so it can be painted interactively – I know brush strokes might not necessarily always appear where the mouse is and so on, but one can easily switch an effect off temporarily if things like these become an issue). But gone are the days of Stamp Down + Sharpen or High Pass layers on top of the document, needing to be redone after every minor edit below – that's huge!!! Two new bugs I found: Dragging on-canvas to manipulate Gaussian Blur (didn't test other filters yet) will result in strangely oscillating values. Values under about 10 are not accessible via on-canvas dragging. Under some conditions, the document view's background transparency grid seems to be replaced by a white background. I had a 50% gray layer in Soft Light mode over a few black brush strokes on transparent, and it took me some time to figure out why there suddenly was gray appearing in the background. Probably a bug related to isolation/mask editing mode. Some really small things I just noticed while playing with the beta again just now that would save an incredible amount of time in practice: Having swatches for Black, White and 50% Gray, as well as preview swatches for the current Primary and Secondary Colour settings in the Fill dialog would be helpful. Copying Photoshop's Shift+Backspace key combination in addition to Shift+F5 would be nice since the F-keys are mapped to OS functions on many Apple laptops by default. Swatches panel: The permanent None/Black/White swatches are a great idea, but an additional 50% gray swatch would be great (50% gray it is the Overlay/Soft Light/Hard Light neutral colour and thus is needed quite frequently – in Photoshop, I actually have a custom action that loads 50% gray as the foreground colour and have assigned a keyboard shortcut to that). The visibility checkboxes in the layers panel are on the right. They would make more sense positioned on the left since they would be closer to the disclosure triangles and thumbnails, hence reducing unnecessary mouse miles moving back and forth. It sounds like a small and ridiculously nitpicky thing, but these things just add up when working in a program all day. Photoshop also allows dragging over the eyeballs to show/hide multiple layers, which is faster than shift-click selecting and clicking one of the checkboxes, though the new layer isolation mode makes this a lot less problematic. Colour coding layers is also an organization feature I usually use extensively. I couldn't figure out how to apply Curves or other adjustments to a mask, either I am missing something obvious, or it's just not implemented yet The "Layer > Rasterize…" menu command has an ellipsis in the name, but shows no UI when clicked (no dialogue box or similar). The Dodge Brush icon actually looks very similar to a magnifying glass (zoom tool), which might confuse some users. Also, naming the tools in this group something like "Darken Brush", "Lighten Brush", and "Saturation Brush" might make more sense, as these Photoshop-style names are relics of the darkroom era and not familiar even to many photographers these days. There apparently is no Divide blend mode I think I mentioned this somewhere else before already, but in case I didn't: Curves and Channel Mixer use a popup for channel selection. This requires an unnecessary additional click every time (just like it unfortunately does in Photoshop) compared to individual buttons. Also, access to the Alpha channel in these adjustments would be very helpful. The new Colour Chords are great, but it would be nice if the menu actually showed a preview of the resulting swatches next to the label inside the corresponding menu item. The current implementation requires some guesswork on the user's part as to what the results will be. Especially with the new Chords option, the ability to select and move/duplicate/drag to different document/delete/copy/paste/group multiple swatches at once would be useful The Brushes panel and many others that scroll don't show scrollbars by default when a touchpad is active in the system. That means they cannot be scrolled with a Wacom pen unless the user briefly touches the touchpad first to make the scrollbars appear. I think there is a setting in System Preferences to always show scrollbars, but it would be nice if all panels in AP simply showed scrollbars by default if a tablet is detected (like the layers panel does). An alternative would be if it allowed grabbing the list with the pen and dragging the contents like lists on iOS touch screen devices. It would be nice if the options bar for brush-based tools had a quick access button that shows the Brushes panel, just like the "More" button shows the brush editor, or a popup that shows the contents of that panel, like the Gradient tool does with swatches. A "Hide Selection" type option would be very nice Modifier keys for the Marquee tool to access the add/subtract modes etc. would be useful. Right now, it is also not possible to subtract from a selection when the second shape (the one to be added/subtracted from the first) starts inside the area of the first selection since the software will automatically switch to a move mode when moving over pixels that are already selected. The Elliptical Marquee tool should have Antialiasing turned on by default It would be useful if the Transform, Mesh Warp and Perspective tools had a switch that switches between transforming the actual layer and transforming the selection (similar to Photoshop's "Transform Selection" command). Right now, one workaround is to enter QuickMask mode, but that obviously does not work when the user is already inside Quick Mask mode. The colour used for Quick Mask seems like is not currently customizable, which can be an issue (hard to see) depending on the colours used in the document The crop tool does not seem to currently snap to document boundaries Sorry for the incredibly lengthy post with so much random stuff, hope this is still useable feedback for you! Thanks again for another amazing update, the software is really coming along nicely!

It is similar to Photoshop's "Blend If" feature, except that the user interface is a lot better. It basically allows you to make a layer partially transparent based on its luminosity or that of the layers below. For instance, you can use this to set up a layer so it only affects the shadows of your image, or you can quickly make a white background transparent for a test comp (you'll probably still have to unpremultiply somehow in most cases). You could, say, limit the effect of a saturation adjustment to the shadows by adjusting the curve for the underlaying layers ("Underlying Composition Ranges"). It is also great for exposure blending and so on.

Occasionally, you do want to flatten images when you don't need your layers any more, but still want to continue building on top of what you already did. Exporting and re-importing of course would be cumbersome for that. An even more essential feature, though, is Photoshop's "Stamp Down" command (option+click on "Flatten Image", or Cmd+Option+Shift+E), which gives you a flattened copy as a layer on top of your existing layer stack. Most of the use cases for that are rooted in bad decisions on Adobe's part and could be avoided (such as having to create a flattened copy to be able to sharpen an image for output after retouching it, or needing a flattened image in order to be able to use Content Aware Fill), but it is still a very useful feature for other situations as well.