Devil_Inside

Since there is no proper way to define precise selection size, the suggested way is to select an area with a selection marquee and scale and move that selection marquee using the transform panel. There is an issue with this approach however, as it seems the selection marquee scaling has some weird sampling and the resulted selection has blurry edges and is actually of a different size than what is specified in the transform panel. This behavior is amplified if the initial selection marquee is very small and is scaled to a very large size, as shown in the following example. Steps to reproduce: 1. Open a 2048 image and select a small area in the middle of the image using Rectangular Marquee Selection tool 2. Change X, Y, W, H of the selection to 0,0,2048,2048 3. Copy selection to clipboard and go to File -> New From Clipboard You can clearly see that the resulted pasted image has weird blurry edges. In a lot of cases when you copy and paste a scaled selection, the resulted pasted layer also has a weird increase in size and it differs from the selection that you specified and copied to clipboard. When the difference between the initial selection size and the final scaled selection is small, the issue is not so apparent but it is still there and is very frustrating. It's a "gamebreaking" issue when you need to precisely re-arrange parts of a flat image. You copy one area to the clipboard, you paste and the resulted pasted layer has the wrong size and blurred edges and is impossible to align and position to a new spot. It is mind-boggling that such a basic feature is so broken.

Quickest way to reproduce 100%: 1. Open Affinity Photo 2. Create a new document 3. Select Brush tool 4. Ctrl+Alt+Shift+S 5. Cancel

Thanks for the update, Chris! I hope this will be reconsidered!

So there is no plan to change it? Or it just hasn't happened yet?

@Chris B Any updates on this topic?

Wow, just wow, this thread... I have selected the area I want to keep and crop everything else. It just boggles my mind that there is no simple option to do that.

If you don't want to change the current default behavior, it would be greatly appreciated if some of these values could be exposed in the settings, so everyone could adjust them to their liking. Thanks!

Stumbled upon this issue myself. The 15% increment is indeed a really weird decision. 1. I can't switch quickly between 2-3 different brush sizes. If I start with a 64px brush and I go a couple steps up, I'm unable to get back to 64px no matter what I try. It goes 64 -> 73.6 -> 84.6 -> 71.9 -> 61.2 2. Why is there a 57.3px size for a raster brush? It would make a lot more sense to round these values to whole numbers. 3. 1px increment for the Pixel Tool is too small when using larger brushes. This issue is very noticeable when you want to use the [ ] shortcuts to quickly switch between a couple different brush sizes. IMHO the best way to handle this, is to have a predefined range of brush sizes to go through, smaller steps for smaller values, larger steps for larger values. There is another issue with the Liquify Persona: Regular Photo Persona brush size slider is exponential, it goes 0-100 for the first half of the slider, and then up to 4k for the second half of the slider. This makes it really convenient when working with smaller brushes. In Liquify Persona, brush size slider is linear, with smaller brush sizes being almost inaccessible due to not only being a small % of 4k, but also due to snapping points on the slider at 32, 64, 128, etc. I haven't been using Liquify Persona much, but is it that common to use 2k - 4k brushes for liquify effects? I never wok with images larger than 4k (rarely 4k), so a 2k - 4k brush seems excessive to me. The performance when using large brushes is also horrible. I was wondering if it would make more sense to use an exponential slider for Liquify Persona as well? P.S. Similarly to brush sizes, current "CTRL + Scroll" zoom increments also make no sense since you always end up at some weird 147.4% zoom values. A predefined range or just absolute step values would be so much better.

1. Create a constrained group of several shapes. 2. Convert it to symbol 3. Create a new instance of that symbol and scale it to see the effect of constraints working 4. Scale one of the shapes inside the first instance and notice that the second instance instantly breaks. Video with reproduction steps: https://streamable.com/4xqv8 Affinity Designer: ver 1.7.3.481

+1

Thanks for the quick reply. Yes, I figured now that "sync off" is the solution to my issue, however I hope that this behavior will be changed in the future. It feels that all the operation that are done to the actual instance (and not it's internal objects) should be applied to that instance only. They shouldn't propagate to the symbol and other instances. Mask is one example. Layer effects is another example, then goes the layer opacity and even layer visibility. IMHO all these make no sense whatsoever.

I have 2 instances of 2 different symbols in my scene (4 objects total). I applied 2 different masks to the 2 instances of the first symbol and it worked as I expected, each symbol instance was masked appropriately by its mask. When I tried to mask the 2 instances of the second symbol, however, whenever I apply the mask to the first instance, the other instance immediately inherits the mask, so I'm unable to set a different mask for each instance. I tried to test this in a new empty document, and I couldn't make it work the way the first symbol worked (separate masks for each instance). This is quite confusing and counter-intuitive. I think the mask shouldn't be part of the symbol.

Thanks a lot for a detailed answer, Carl. I agree that a proper perspective distortion would be the ideal tool for this, but since it's not there, I thought to cheat with simple text on a path. For the small icon I'm doing I used the method you described, and while not perfect, it's good enough. Thanks again!

Yeah, perspective distortion could be a useful feature, however I don't think I need that kind of precision for my example. It seemed that a text on a path would be a much simpler and quicker way to do it, if the option to keep characters vertically aligned was there.

Yes, that's an example of why I need an option to have them vertical. Hopefully @carl123 has a workaround.