Alej

Here's a video of how it works, I just do some brushes, not quite as snappy the first one but acceptable however hitting 'alt' and color picking is not workable at all. The lag should be very visible on the captured video. This is also a tiny document test to take doc size out of the equation, it was a 446 by 96 px small canvas, just one layer. Also, just alt clicking doesn't show the color picker, it has to be dragged for it to appear, the first appearance is actually the slowest of all (not quite visible on the video as it just appears after some frozen time). Any workarounds in the meantime would be greatly appreciated. VerySlowColorPicker.mp4

So, sadly it was about a condescending comment making fun of people, which I was willing to give it the benefit of the doubt. What an ignorant comment and person then, hope it's reported and banned. @Lupearn Some people, for example, would go and prepare command line based massive Substance Designer files that do a lot of stuff on folders regarding channel packing, optimizations, resizing, material generation based on alphas (or smoothness to roughness conversion depending on engine), etc times 2000. And sometimes a quick test without having to go through the whole complicated pipeline just yet as a quick test makes a lot sense... and there's no real need to convert, it can still be done on PS, a company has no problem paying $20 per seat for a PS license, this is more about principle and wanting to find a different venue. Your comment is so condescending, short sighted and narrow. Compare that to the civilized and insanely helpful member here @kirk23 who has shared macros, workarounds for the "zero alpha pre-multiplies the RGB channels, etc", thanks a lot for the efforts. I will continue to use Affinity Photo on my personal time and professional time as much as possible, the community is ultimately very helpful (beyond a few rotten apples) and the devs are making great strides overall

No need for a direct format conversion, affinity can load and export psd files directly. And the format’s features support are kept as up to date as possible in my experience.

Exactly, the whole issue here is, whether the AP way is superior, is that people REALLY want AP to be a replacement, to make it work, for that to happen it has to have as less friction as possible. We aren’t bashing Serif devs, we are incredibly happy that they are doing what they do with performance and keeping things in check, is just that it’s so darn close in some things but not quite that it’s frustrating. If I want to get rid of a certain other photo program I have to be able to do the same in a similar enough way, at least at the beginning, no macros, no splitting, no preview woes… I jumped into this question thread right away because I opened a TGA file that had an alpha channel, that I wanted to invert, paint some stuff on it, Levels on a selection mask, resave. A 2min turnaround max for a dirty test before deciding on larger scale production decisions which turned in hours and hours of head-scratching, reading forums, etc… what happened? I had to do it back on known grounds and it took me 1hr 2mins instead of just 2mins and felt beyond frustrating, but that’s how much I wanted it to work.

Received a PSD files, it happens to have box rectangles with uniform rounded corners, it might be related to that? Also gradient fills properties show up. Let me know if these are actually bugs I should be mentioning here? or is it to be expected for the time being while compatibility is being worked out? Latest AP, 1.10.14 RectangleBugs.psd

Alright, I think that it is a beginner's mistake regarding blending modes, I have confirmed in other softwares that the 'Fill' opacity seems to not do anything either regarding the global blending opacity of effects and also the color/patter/inner-outer-glows/etc are not affected by the layer's blending mode either. So the 'workaround' (group and apply blending mode to that) seems to be the right way. Sorry for that. Here's the video in case, 1.10.4 appeared today for me. AP_10_4_FillOpacity_BlendingModes_bugs-Low720p_MP4_nosound.mp4

Ok, doesn't appear just yet on the Mac App Store. Will give it a test drive afternoon to see. Also, my bad, this has been a pretty rush test as it isn't an important task but felt weird compared to how I do it elsewhere, the blending modes themselves for color overlay don't seem to work either. Workaround: had to put everything under a group and use the blending modes there instead, actually having the blending mode on a layer with it's Layer Effects doesn't do the blending with the final pixel value.

Hello! thanks for the quick followup. 1.10.35514 File attached. But starting to fear that I might be doing something non-clever. Updating to today's version.FillOpacityPotentialBug.afphoto

Posting a video, it is self explanatory, can't seem to find why the slider does nothing on any of the layers shown. That said, if I create a new Text layer for example, the opacity slider of that one will work. Not clear why this current example seems to do anything. (Blending mode do work but can't modulate it). EDIT: my bad, was selecting the 'brushes' opacity for the layer's opacity. However the fill's opacity doesn't seem to work. Screen Recording 2021-10-29 at 11.23.53 AM.mov

Hopefully this is not against forum rules but another known app (it starts with “Pixel” and something about “mator Pro”) has added quite the Siri Shortcuts support for Monterey. It works with a reduced subset because they have quite the one shot ML actions (remove background, separate subject from background, etc etc). Maybe something useful could be done too with Affinity apps, especially now that the desktop has support for Shortcuts now, but pretty sure there are more pressing matters.

Hello! I'm having the same issue, iMac 2020, 5700XT Big Sur. It isn't TERRIBLY slow but it shows a considerable hiccup/lag. Can make a video of it if it is still being investigated.

Woah, I especially like the scene objects highlighting (because I hadn’t seen it before like this). Pretty sure the Serif guys are working hard for this but without neglecting the potentially more important imminent few fixes and QoL tweaks. To be honest, I have used these similarly related AI things a total of zero times in other applications... so I wonder how useful it can really be (if others can chime in), however, the Depth Map and scene labeling though, that looks interesting.

That’s an interesting workaround thanks for the share. Just as a caveat for anybody doing confidencial or company contract work: besides the potential several GBs of data per conversion round trip there’s still the security/privacy issue. They mention they delete it after an hour of the conversion but still... might not be the service the problem but anybody outside or around it. I would not use it except for personal testing purposes and even then the PSD file sent can have a lot of personal information attached to it.

Thanks a lot for this, after several years I found your breakdown of which tools works without anti-alias and the overview quite informative. I have been checking and following the Affinity Suite of tools and I’m using it more and more. Some time ago I didn’t even understand what “Pixel Persona” meant, had only Affinity Photo at that time. Now that I have all of them (both Mac and iPad) it makes more and more sense. What do you thing about the existing isometric view capabilities? Maybe that could help? Also regarding sticking to pixel persona only, could it be useful to quickly sketch with the curve tools, rasterize and then pixelize or clean on top? In any case, cheers.

Solid take, in my experience the photoshop way is a work around to be honest for forcing .TGA files to have the alpha channel (since the exporting is treated completely different if it is .PNG or .TGA there). The main issue with this though is that sometimes, especially video games, control over the colors where the alpha is zero (or “fully” transparent) is required (premultiplied alpha blending workflows for example). At this point I think it’s better to use something else, .PNG and transparency of the layers (or globally making one like rasterizing to mask with a layer at the very top) works perfectly fine for UI-like assets. For intricate R-G-B-A handling I’m just using substance designer when possible in the end (which also has another set of constraints, like textures there are always power of 2, sometimes a positive, sometimes a drawback).