For in-app preview, it uses Settings > Interpolation. For export, it uses Export > Video > Interpolation.
Pardon the newbie question, but I see a menu-bar “Settings” menu and not an “Export” menu–how do I get to this option? (I did find an Interpolation setting in the Export pane, having pressed Advanced, after Deinterlacer, but I presume it’s only for the deinterlacing; at any rate, I set it and saw no result when I exported again.)
What I’m going for is what you’ll see in the lowest-right scaled inset below (which I scaled from the base image with Photoshop & nearest-neighbor). What I’m getting is the inset Shotcut scaled in the video, just “behind” it. (Some might think I’m splitting hairs, but the high-frequency (sharp) edges on the nearest-neighbor scaled pixels prevent the eye from thinking they need to re-focus; even though it’s scaled antialiased text, it’s more comfortable; clearer to my eye.)
P.S. Is there an easy way to put, say, a red border or drop shadow around the inset to help people feel that it’s not a part of the base image?
My experience with making video of fine text, is best to avoid the situation as much as possible. I find it’s best to have the application or web browser text be as large as possible for a video that still allows you to show all of the information you want to show.
I’ve found that recording in 4k (~5k CBR bit rate) 30fps , for a 1080p video provides the best detail, which I scale down the source to the Video Mode of 1080p provides a superior image. I’ve only worked in Progressive mode. I use OBS to record browser windows & games. For a free program there is a lot of control on how you want your recording to be.
I will try to post examples in the next few days to a week here.
You can make up your own image through a graphic editor, or you can use Shotcut to create a color, then mask the color around.
To get a color into Shotcut, click on Open Other, then Color.
This is a rather crude example, images taken from a smaller resolution then scaled up.
I left the red border box short on purpose.
Been using OBS on a 4k+ screen for just a few months, but it’s mostly working fine (after days dialing it in: had to disable my AMD graphics, then re-enable to see anything but a black screen in OBS–hit on that by accident, but that’s another story).
Encouraged by your results I captured at the full 3840x2560, and had Windows scale 300% so the Java Frames were a good size. (Java is not dpi-aware, so it’s just duplicating pixels, but that’s yet another story and I’m okay with duped pixels so long as they aren’t smeared by some over-zealous “keep it pretty” scaling method…)
Got the captured FLV into Shotcut, but Shotcut is running the video in a very choppy way–I presume because it’s such a large image. Any suggestions on optimizing its execution?
