@Media Engineer, please look at FB9393169 I filed...AURemoteIO does not produce samples with builtin mic if stereo audio is configured. Please also suggest a workaround as this is a serious bug for my app and I doubt this will be fixed in iOS 15 main release.

This feature is completely broken in iOS 15 betas(including beta 5). I have filed a bug FB9393169 but there is no response so far.

Yes, I noticed it yesterday!

Ok so it is the smallest bounding box of the image in that case, that seems reasonable. So I will need to manually resize the image to fit in the original bounding box for which I will need to scale the image back. The next question that arises is how do I know the bounding box before applying a transform? I will need that to calculate the scaling factor.

Ok I used trigonometry to calculate the new extent. Formulas work! Thank you so much for clarifying.

But then I have to update EffectModel with EffectModel where ever object of type EffectModel was used. I just want to avoid touching EffectModel.

Value needs to adhere to protocol Interpolable and Codable. And Interpolable protocol includes just one function that returns interpolated value from value A to value B by a float amount. So all math types Int, Float, Double, and types such as CGRect, CGAffineTransform, UIColor that can be interpolated, Value can be that type.

I don't see these icons, how do I see them?

I am wondering why this limitation, what kind of technical reasons stop it?

This way certainly works but it is a legacy way of observing properties. I was thinking if there is modern Swift KVO based method for the same. Sorry to omit those details in my question here , I should have mentioned the details like i mentioned on stackoverflow.(https://stackoverflow.com/questions/69505085/swift-kvo-addobserver-observevalue-for-class-objects)

Thanks for the answer. I will definitely post it on Swift forums.

I want to use my custom export session class that uses AVAssetReader/Writer as I am using it in my workflow. How to correctly adapt it to ProRes videos?

Ok I see that. But when I record video using AVCaptureVideoDataOutput, I do see some elements in AVVideoCompressionPropertiesKey dictionary through recommendedSettings. I am not sure what would be the result to omitting AVVideoCompressionPropertiesKey, but is there a way to copy this dictionary from input video file for instance (I can't find a way to do this though)?

Sure I would file a bug!

Why do we need to add 0.5 offset I wonder?