Anton Firszov

This is **huge** work. I mean, it can literally take engineer-weeks, and there is no guarantee for massive improvements from integer arithmetics, we can get surprised just like it happened...

Yeah, internally we would have to convert the components to 16bit just like most jpeg calculations do. This means that it might make sense to expose an `Rgba64` API and...

> port a good working progressive encoder like libjpeg-turbo Note that libjpeg-turbo has it's own memory management primitives, a bunch of C indirections, and the optimized SIMD methods are implemented...

The *semantic* of that script is the most general way of defining a progressive jpeg export. We should define an equivalent internal API to have a flexible encoder implementation. This...

> The corresponding Unity issue (opened 7/16/21) is "1351111 (Open) IL2CPP AOT Compilation Regression". > Pragmatically speaking, Unity isn't likely to resolve the behavior. @stonstad do you have access to...

You can call `differnceImage.Save(...)` to save it for manual analysis. In our test infrastructure we have a `.DebugSave(...)` extension method to save images to a dedicated artifact folder locally which...

ImageComparer intends to define a numeric similarity metric to measure difference between two images, it's primary use-case is testing. Image fingerprinting / hashing is a related, but different feature. If...

I'm generally fine sacrificing some of our libjpeg conformance in favor of performance and actual quality, but I would prefer to see some quantitative data to understand the impact better....

Related: I just noticed https://github.com/SixLabors/ImageSharp/pull/1694#issuecomment-878989715 and 2e5b0ad74da6b476f812266b193be81c63ba17f5. As far as I remember, I used ImageMagick to generate the output, which uses libjpeg-turbo internally. Does this mean that the bug is...

> There's a strong chance those images were spoilt by bad optimization. If this is the case, it should be a bug in optipng, which is very unlikely IMO, since...