I’ve been shooting some videos in the past couple of weeks and I’ve been wondering how much the source and destination formats go into making things take longer to process. Of course I knew that chugging through 4K is harder than 1080P, but I was wondering how much that is.
So today I tested. The subject was the least interesting subject. Ever.
Feel free to not watch it. Or I guess you can watch it if you want to get soothed to sleep. Your choice. I just needed a simple, reproducible, non-static, non-copyrighted thing to take a video of. The KitchenAid mixer on the counter stood in as my model for the purpose.
|Resolution||FPS||Bitrate (Mbps)||File Size (MB)||Resolution||FPS||Processing Time||File Size (MB)|
While it may seem obvious that the destination resolution is one of the keys to the deal, I would have expected that the 1080P @ 60 would be a lot slower than the 30 fps, but it’s not really that much slower. Similarly I would have thought that the downsampling of the 4K to 1080 would also be a lot more strenuous than it turns out.
To describe the setup I was using the Sony FDR AX-53 camera to capture the original footage. Each source clip was just around 1:00 (+/- .5s). From there it was imported into Adobe Premiere and exported with not modifications to the video other than the size and frame rate on the export — I didn’t retouch or grade the video in any way. The video was stored on a Drobo with five RAIDed drives so access speed shouldn’t have been an issue. This was borne out by the fact that the CPU was pegged at 100%.
Given all this so far, I would shoot at 4K 60Mbps and export to 1080@30fps.. It gives me a bit of extra margin for editing (like cropping in post but having more data) without taking up too much extra space or taking much longer to export.
Of course this doesn’t take into account the time in the editor and how log it takes to apply any effects to the video. Another consideration is 4K in 60Mbps may look worse than 1080 at 50Mbps. This is certainly another question to be answered. Next time!