Something that isn’t super well known is that ARRI’s much-loved ALEXA and ALEXA Mini – which continue to be industry staples – do not actually record in 4K resolution. The newer LF-format cameras do have native 4K resolutions, but before that series (and ignoring the ALEXA 65) ARRI was limited to around 3K. A few years back, ARRI did offer a solution to this problem by offering internal 4K upscaling to limit the need to do so in post.
Still, it’s always been thought that if you want to do anything significant to your footage you should save that work for post-production. Considering this, filmmaker Dave Cheung wanted to test whether it was better to save the upscaling until post or to have the ARRI ALEXA Mini handle it in-camera.
Streaming is actually what pushed everything into true 4K. Netflix has some very strict requirements for its originals and that including a 4K image sensor. The LF turned out to be the answer to the 4K requirements.
That still didn’t change the existing cameras, though ARRI did announce a firmware update for internal scaling to 4K believing that their images were of such high quality that the difference was almost imperceptible.
Are they really?
Running through all the essential formats, Dave checks out native resolution options, internal upscaling, ARRIRAW and ProRes 4444 XQ. Bringing it into Premiere Pro he does some pixel peeping to check it out.
Comparing the internal vs post upscaling there is practically no difference. And it looks great regardless. Basically, you should do what is best for your workflow. I think recording native resolutions makes the most sense unless you have to work with an ultra-fast turnaround. Post upscaling is usually going to give you more control even if the image quality is the same.
What are your thoughts on upscaling in camera vs post? Or even just about ARRI cameras these days?
[source: Dave C]