Upload your video
Any 16:9 MP4/MOV: show excerpt, interview, gameplay, talk. 1 minute of video = 1 minute taken from your quota.
A dumb center crop cuts faces and keeps the void. Nysos's Format Converter analyzes the frame: it tracks the face, builds a clean split when there are two people on screen, and falls back to a blurred background when the scene doesn't crop cleanly.
Any 16:9 MP4/MOV: show excerpt, interview, gameplay, talk. 1 minute of video = 1 minute taken from your quota.
Auto: Nysos picks the best framing on its own. Split: your video on top, an image or a second video below — the classic gameplay/reaction format.
1080×1920 output ready for TikTok, Reels and Shorts. Add automatic captions on the way if you want.
Within one video, a wide shot reframes on the face, a two-person shot splits, a screen recording goes blurred-background. Nysos decides shot by shot — the difference between “converted” and “watchable”.
9:16 is the full-screen format of TikTok, Reels and Shorts: 1080 pixels wide by 1920 tall. A 16:9 source at 1920×1080 cropped at full height keeps only 607 pixels of width, less than a third of the frame. Everything that happens on the edges is gone: a second guest, a shared screen, a game's interface.
Safe areas matter as much as the ratio: on all three platforms, the interface (caption, buttons, progress bar) covers the bottom and the right edge of the video. A face or a text placed there gets hidden. A converter should not just cut; it has to decide what to keep and where to put it.
Tracked reframing. Nysos detects the face frame by frame and moves the 9:16 window with it. The right choice for a single-person shot: interview, face-cam, talk, vlog.
The split. Two stacked areas in the vertical frame. Your video on top and a photo or a second video below, the gameplay + reaction format, or for a two-person podcast, each person in their own half. Nothing is cut; everyone keeps their frame.
The blurred background. The whole video stays centered, in 16:9, over a blurred backdrop that fills the rest of the frame. The fallback when nothing can be cut: screen recording, whiteboard, a very wide shot with several people.
In auto mode, the converter does not pick one setting for the whole video. It first splits the source into shots (camera or scene changes), then decides for each one: a well-framed face, tracked reframing; two faces, split; no face or a screen, blurred background. The choice is locked for the length of the shot, no switching mid-sentence.
A face is only kept if it is confirmed over several consecutive frames, so the tracker does not follow a false positive: a game character, a thumbnail, a poster in the background. On a stream, the embedded webcam is spotted and separated from the game: cam on top, action below. One video can therefore get three different treatments without you touching anything.
Resolution: reframing enlarges part of the image. A 1080p or higher source gives a sharp vertical; at 720p, the enlarged area looks soft, especially on a close-up face. If you can choose when recording, pick the highest definition.
Placement: if you shoot with vertical in mind, keep the subject near the center of the 16:9 and avoid burned-in text on the edges; it will be cut or hidden by the interface. A two-person shot benefits from some space between the people: the split then has two clean frames to cut.
Captions: add them after the conversion, not before. Captions burned into the 16:9 get reframed with the image and end up cut or outside the readable area. Nysos can generate them on the vertical, once the frame is decided.
9:16 (1080×1920) fills a phone screen: it is the native format of TikTok, Reels and Shorts, and the only one Nysos exports. 4:5 (1080×1350) is the Instagram and Facebook feed format, taller than square but framed by white in full screen. 1:1 (1080×1080) works everywhere but leaves bars on mobile.
In practice: produce the 9:16 first. A 4:5 or a 1:1 comes from trimming the top and bottom of a vertical; the reverse is impossible without inventing image.
In a classic editor, taking a video vertical means placing position keyframes on every shot to follow the subject. Over ten minutes with a shot change every eight seconds, that is 75 shots to reframe one by one, then check. A Nysos conversion takes the whole file and renders the vertical MP4, to review once before publishing.
The reframe keeps the action zone (face, subject). If the frame is too wide to cut safely, blurred-background mode keeps the whole video centered over a blurred backdrop.
Yes: the split format stacks both people in the vertical frame, each in their own half, with nothing cut.
Output is 1080×1920. Reframing necessarily enlarges part of the image: the sharper your source (1080p or more), the cleaner the vertical result.
Yes, that's exactly what the Format Converter does: one video in, the same video out in vertical. Highlight detection is a separate tool.
An MP4, MOV or WebM, or a clip already produced by Nysos. To start from a Twitch, YouTube or Kick link, use the clip maker: its clips come out directly in 9:16, with the same framing engine.
Yes: the Format Converter also takes a clip Nysos already produced, not only a file. Handy to reframe a clip as a split with your own media at the bottom.
In auto, Nysos spots the embedded webcam and separates it from the game: cam on top, action below, each at its own size. If you want something else at the bottom, an image or a second video, use the split with your own media.