Aftemplate

Getting used to Affinity Photo is very simple, even without any learning costs (if you are familiar with Photoshop). But APh is still missing some relatively important features. If you can bear it, it might be a good choice. If not, APh is clearly not the current choice, it still needs to be honed.

In addition, the current AP and AD brush manager is very backward, see how Photoshop CC is done

Affinity seems to use increments to speed up the save, but this will increase the file size. Usually saving as a new file will significantly reduce the file size.

I am a loyal reader of Dan Margulis. The Lab color space is very useful. It should be valued.

APh or AD only supports its internal link embedding changes, and internal updates are not updated when external files are changed (Adobe Ps supports external change responses in smart link objects)

Because the canvas adjustment is instantaneous, all pixels are processed dynamically, which is equivalent to refreshing the execution rotation every second. The rotation layer is static, allowing more performance to be used to improve quality, static only once.

Have you noticed that the quality of APH Rotate is extremely bad, if you compare Adobe. It is 15 degrees, I guess one of the reasons is because the Rotate of smaller degrees will produce lower quality. This is the truth, If you don't understand, please try to rotate 15 degrees multiple times. When the image is 90 degrees and 180 degrees, the quality is significantly improved.

This looks interesting.

I don't know if this helps...

I also noticed this problem, your suggestion is very useful, this is exactly what I want. I also noticed that the program only uses bilinear reduction canvas, as far as I know there are higher quality algorithms. But they all need gpu hardware support, including rotation at any angle, this is a hardware transformation. Because APh There has been a lack of gpu support, so shrinking the canvas and rotating the canvas provide limited performance, they should be improved in the future, after supporting gpu

Hi! Thanks to the developer for their contribution. Merry Christmas! In most cases, the performance of affinity photo is excellent. But I noticed that in some cases it is slow I used a 100 million pixels Lab/16 canvas to create twenty layers. The save of large-area modifications is extremely slow or even horrible. This is not a bug or a computer problem. Large-size (1000+) brushes are slow. Because of the lack of hardware acceleration At the same time I use the macos version, the speed after 1.7 is amazing, the filter responds in time without any lag, and everything is attributed to metal acceleration (in macos 1.7 metal) Please consider adding Lab/8 support to provide a faster preview than lab/16, especially in the current missing gpu acceleration. I will continue to use affinity photos in macos mojave (metal2 api)

Yes, I understand. I heard that you are already doing this work. How do you plan to support it in the future? By cuda or something else? Thanks.

How is it achieved? CUDA or OpenCL and OpenCV? Thanks.