DonKC

VERY COOL! Had no idea that was there (or why LOL). Worked for me, and tested it over the original and it was pretty close (I wasn't real precise about points).

Someone in my family had pasted vintage photos in a scrapbook without my access to a flatbed scanner (to scan an entire page) or being able to remove the photo (to scan individually), so I took a photo with my iPhone (trying to be as perpendicular, straight and true as I could). I can edit on the phone with Apple's crop and perspective tools to isolate photos in the photo rather easily with repeated tweaking, and most pics turned out fine. However some photos were sloppy and taken sideways, and from different perspectives, and I wanted to try using Affinity Photo v2 thinking there would have to be tools to do that quickly. So I made a test photo to use but I have struggled for hours with what seems to be missing tools to correct these distorted photos trying to use Filters->Distort -Perspective, Mesh, Deform, etc. What I would like is movable points at each corner of the object photo within and simply move them to corners of guides I set up for resolution and size but that is frustrated at every turn. I have attached a deliberately skewed and distorted photo of a square grid in a 3.5"x3.5" vintage photo-like size that I tried to create a 3.5"x3.5" photo at 300 pixels/inch. Is there an easy way to do this in Affinity that I am missing?

After trial and error for hours, it seems (but not totally sure) that when I removed masks on the embedded files, the project could be exported without rasterization.

The 3 embeds are all .afdesign files (but may have been AI in a previous life). When I deactivate them all, there is no white background for the project (and it's a nuisance to always change the Designer document to transparent to see the difference) with transparency. When I individually edit each one and then try to export each it says whether there will be some rasterizing or nothing will be rasterized. Going through each embed, there were things like strokes not expanded or a group with one member, and I removed these random things until the test export said nothing will be rasterized. I even exported just the no-rasterize embed and the transparency is there for each. When I put anyone of these no-rasterize embeds back in the main project, it again rasterizes something, and that produces the white background. Maybe Designer always rasterizes embedded vector files (rather than track all the new objects)???

I have a project (with transparent background) with multiple objects and several multi-object files that I have placed/embedded into that project. I can edit the embedded files in the project without a problem. I can successfully export this project as a PNG and everything is as I expect with a transparent background (in other programs and online t-shirt design areas). However, when I export the project as a EPS, I do see a notice that "Some areas will be rasterized" but I ignored that because all I worked with were vectors and nothing with pixels or rasterized. When the export is finished the transparent background is gone and replaced with a white background. On further investigation on the resultant EPS file in Designer, I found that instead of vectors for the embedded files there are "images". So when I hid those image layers the white background disappeared and the background was transparent. The same thing happened when I did not export the project the embedded layers, "nothing will be rasterized" and the background remained transparent. Apparently, Designer throws away embedded file vectors and converts it to a rasterized image which produces a white background for the entire project. Is there a way to export the content of embedded files in a project that eliminates the rasterization of the files and preserves a transparent background?

Wow, lacerto, didn't know any of that and thanks for the explanation and video!

Thanks. I missed the fill mode winding. I knew where it was with Illustrator but didn't catch it here. Thanks again.

This design has only three curves on a transparent background - a rounded rectangle, and two rectangular curves (formerly rectangle objects - I think). I am unable to punch out the black rectangular curves from the white rounded rectangle. I've tried various options, e.g. expand stroke, rearrange order, etc. the minus option, the xor option, etc. Sometimes the combined results has outlines for all the curves but not transparency. Does anyone know what I am doing wrong? punchout.afdesign

Excellent point. So, I took a scanned slide 2.3mb JPEG, did some adjustments, and chose Save Flattened. That file turned out to be 7.6mb! I tried the Export with a high-quality JPEG preset and exported that to a 1.6mb file. So, yes, there's no way to know what's going on especially with a consumer slide/negative scanner. Unfortunately, as a hack, I tried to create a macro for Export -> set things up -> etc. and a warning message says export cannot be done as a macro.

I load a jpg, add adjustments, filters, etc. and just want to overwrite and save those changes back to the *same* file. I don't want to save a project file, e.g. .afphoto, or change the size, quality, name or anything else. When I press cmd-S a dialog pops up Document -> Save As (should be Save As...), Save Flattened, or Cancel, so I click Save Flattened. I would like to save that extra click and thought there would be some existing command with key shortcut, like option-cmd-S to eliminate that step. I cannot find any command that is essentially Save As Flattened. Have I missed something? Is there a hack to do this? Or is this only a feature request, e.g. Save As Flattened menu command and shortcut, at this point?

More clicks than I would like, but do it once, and I'm set, thanks!

Thank you for a clear response that explains their model and an example (and makes sense).

I appreciate the responses, so thanks for coming up with ways to address the issue, BUT the "answers" all appear to be workarounds/hacks that take a substantial amount of fiddling (mousing, clicking, dragging and dropping) and expert knowledge (macros, presets, parent/child layers, etc.) especially if there are DOZENS of photos that need similar correction(s). One might as well just open dozens of tabs, each with one photo, and paste the adjustment dozens of times. Or maybe someone would know how to make a batch macro that opens all photos in a folder, pastes an adjustment (from somewhere, e.g. preset?), exports each photo, closes the tabs. Maybe that's already possible? But that makes the users programmers and user-programmers can probably do that with Apple Automator and Quartz filters, or maybe write their own app to do that in Xcode, so why bother with Photo? As I searched this forum, I found many similar questions from users who expected that whatever adjustment was "on top" affects all the layers below. That's what most users see over and over, and of course they (and I) thought the same with exporting slices with ONE click. That paradigm was broken and hence the confusion and bewilderment. Just my two cents (for the developers).

I placed N photos in a project as layers that all needed ~the same correction (Levels) that I added on top (and not individually). I went to the Export persona -> slices -> exported the layers, and only the first had the adjustment applied. I expected ALL the layers/slices to have that applied just as if I selected/deselected the layers in Photo persona where the top layer was applied to any of the N layers. I've searched this forum and read anything to "in can't be done" to "make a macro to duplicate the adjustment to every layer", "group layers", etc. and finally "works as designed" (even though HIGHLY unintuitive). So, is there any "answer" on how to not duplicate the same adjustment for dozens of photos in a project and yet use one (or many) top-level adjustments, and export N photos each adjusted the same.

Yes, that works! I had actually spent an hour or two fooling with this until I discovered (after your post) that it would actually work but I had to crank up the radius well beyond the default 100px until it covered any existing fill (and I learned that the object had to have an existing fill).