Guides
Sume guides: the API, Formats, agents, models, media tools and pricing, each answer dated and sourced.
How to insert a video into another video
To insert a video into another video, split the main video at the insert point and play it, the new clip, then the rest, all in one render.
How to shorten a video: trim, cut, or remove pauses
Shorten a video by cutting time: trim the start or end, remove sections from the middle, or cut the pauses. Which to pick, how, and the limits.
FFmpeg trim video: cut by time, with or without re-encoding
Trim a video with FFmpeg: -ss before -i to seek, -t for the length, and -c copy to skip re-encoding, which starts the cut at the keyframe before.
Types of cuts in video editing: hard, jump, J, L, and more
The main types of cuts in video editing: hard, jump, J and L, match, cutaway, cross-cut, and smash cut. What each does, and how to build them.
Dip to black transition: fade out and in between clips
A dip to black fades one clip out to black, then fades the next in from black. How to build one when a render's fade is a crossfade, and the limits.
23.976 vs 24 fps: the difference and when it matters
23.976 fps is 24000/1001, the NTSC-film version of 24 fps: 0.1% slower, one frame in 1,001. When that gap matters, and what Sume renders do with it.
Keyframe interval explained: I-frames, GOPs, and cuts
The keyframe interval is how far apart a video's whole-picture frames sit. It decides where copy cuts and fast seeks land. How to read and set it.
How to convert frames to seconds (and back) at any fps
Seconds = frames ÷ fps, and frames = seconds × fps. How to convert frames and timecode at 24, 23.976, or 29.97 fps, and why Sume takes seconds.
What codec does MP4 use? MP4 is a container, not a codec
MP4 isn't a codec: it's a container holding video and audio compressed by separate codecs, such as H.264 and AAC. How to tell them apart.
How to check a video's codec with ffprobe or an API
Check a video's codec by probing it: ffprobe prints each stream's codec_name. For a Sume-hosted file, a free probe returns video_codec and pix_fmt.
MP4 plays audio but no video? Check the video stream
An MP4 that plays sound but no picture has a video stream the player can't decode, or none at all. Probe the codec and pixel format, then re-encode.
FFmpeg height not divisible by 2: why and how to fix it
The error means a yuv420p encode got an odd height (or width). Make both even: scale with -2 or round with trunc(ih/2)*2. How Sume rounds sizes.
yuv420p10le vs yuv420p: 8-bit vs 10-bit pixel formats
yuv420p and yuv420p10le are both planar YUV 4:2:0. yuv420p stores 8 bits per sample; yuv420p10le stores 10, little-endian, in 16-bit words.
Faststart MP4: moving the moov atom to the front
A faststart MP4 has its index, the moov atom, at the start of the file. FFmpeg moves it there with -movflags +faststart in a second pass.
FFmpeg extract audio from video: copy, MP3, or WAV
Extract audio from a video with FFmpeg: -vn drops the picture, then -c:a copy keeps the audio as it is, or an encoder writes WAV or MP3.
FFmpeg extract frames from video: one, every second, or all
Extract frames with FFmpeg: -frames:v 1 saves one image, -r 1 saves one per second, and a numbered file pattern writes every frame.
FFmpeg concatenate videos: concat demuxer vs concat filter
Concatenate videos with FFmpeg: the concat demuxer joins matching files with -c copy and no re-encode; the concat filter re-encodes mixed clips.
FFmpeg merge audio and video: -map, copy, and -shortest
Merge audio and video with FFmpeg: map the video from one file and the audio from the other, copy the video, and add -shortest to match lengths.
Convert 4:3 to 16:9 video: pillarbox, crop, or stretch
Convert 4:3 video to 16:9 three ways: pillarbox it, crop 12.5% off the top and bottom, or stretch it. With FFmpeg and the Sume API.
Cinematic bars size: bar height for 2.39:1 and 2.35:1
Cinematic bars on a 1920×1080 video are 138 px each for 2.39:1 and 131 px for 2.35:1. The formula, sizes for 720p to 4K, and how to add them.
How to find the aspect ratio of a video (and calculate it)
Find a video's aspect ratio by dividing its width and height by their greatest common divisor: 1920×1080 is 16:9. Plus the traps that mislead.
Vertical video resolution: 9:16 sizes from 720p to 4K
Vertical video resolution is 1080×1920 for 1080p, 720×1280 for 720p, and 2160×3840 for 4K. How to size any 9:16 frame, and what Sume can render.
How to upscale video to 4K: scale factors and step order
To upscale video to 4K, enlarge it by 2160 ÷ its height: ×2 from 1080p, ×3 from 720p. How to do it with FFmpeg or Sume, and why to upscale last.
Video freezes but audio continues: causes and fixes
If a video freezes but the audio continues in every player, the file holds a still frame. How to check its streams, and the three Sume render causes.
How to combine photos and videos into one video
Combine photos and videos into one video on a single timeline: each photo held for a set time, each clip cut to length, one soundtrack under it all.
What is text-based video editing? How transcript cuts work
Text-based video editing cuts a video by deleting words from its transcript. How word timings become a cut list, and how to build it with Sume's API.
AI image generator with a transparent background: 2 steps
Make an AI image with a transparent background in two steps: generate it on a plain backdrop, then remove the background to get a PNG with alpha.
Can a JPG be transparent? No: make a transparent PNG instead
No. A JPG has no alpha channel, so every pixel is opaque. Cut the subject out and save it as a PNG, a format that can store transparent pixels.
AI sticker generator: from a prompt or photo to a PNG cutout
Make stickers with AI: generate one flat design with a bold outline on a plain background, size it for print, then cut it out to a transparent PNG.
AI logo generator: PNG concepts from a prompt, not SVG
An AI logo generator drafts logo concepts from a prompt. On Sume they come back as PNG, JPEG or WebP, not SVG; cut one out for a transparent PNG.
How to make a Slack emoji with AI (under 128 KB)
Slack says square images under 128 KB with transparent backgrounds work best. Generate one, cut it out to a PNG, shrink it, then click Add Emoji.
AI icon generator: app icon art at 1024 × 1024 from a prompt
Generate app icon art from a prompt: one centered symbol, no text, several options per call, at 1024 × 1024. Then resize, cut out and finish it.
AI T-shirt design generator: art on a transparent PNG
Generate T-shirt art on a plain background, upscale it to your printer's pixel size, then cut it out so the design sits on a transparent PNG.
AI poster generator: print sizes and room for your title
Generate AI poster art at your poster's shape with room for the title, upscale it for a 300 DPI print, then add the words in a layout tool.
AI flyer generator: make the art, then type the details
Generate flyer art at US Letter or A4 shape with empty panels, then type the headline, date, prices, and address yourself so you can check each one.
AI infographic generator: draw the layout, add the numbers
An image model can draw an infographic's layout, icons, and colors, but not reliable numbers. Generate a tall frame, then add the data yourself.
AI invitation generator: card art with room for the details
Generate invitation art at card shape, such as 1600 × 2240 for a 5 × 7-inch card at 320 DPI, then type the names, date, and address yourself.
AI greeting card generator: a card front from your photo
Make a greeting card with AI: restyle your own photo as the card's front at the card's shape, then write the message inside yourself.
AI packaging design generator: concepts, labels, and limits
Use AI for packaging and label concepts: describe the pack, send your logo as a reference, compare renders, then build the print file on the dieline.
AI tattoo generator: design ideas and a preview on your skin
Ask an AI for tattoo designs as black linework on white, preview one on a photo of your arm by sending both images, then take it to a tattoo artist.
AI comic generator: consistent characters, panel by panel
Make an AI comic one panel at a time: reuse one reference image per character, give each panel its shape, and letter the speech bubbles yourself.
AI children's book illustrations: same character, every page
Settle the main character in one picture, reuse it as a reference on every page, and generate each page at the printed size with room for text.
AI headshot generator: a studio headshot from your photos
Send clear photos of yourself as references, describe a studio look, and keep the result that still looks like you. With LinkedIn's photo specs.
How to colorize a black and white photo with AI
AI can colorize a black-and-white photo: send the scan as a reference with a colorize prompt. The colors are the model's guess, so keep the original.
Restore old photos with AI: repair damage, then upscale
To restore an old photo with AI, send the scan with a prompt listing the damage to fix, then upscale it up to 4x. Compare the result with the scan.
Virtual staging AI: furnish an empty room from one photo
Virtual staging with AI: send the empty-room photo with a prompt naming the style and what must not change, then check the walls and windows.
How to change the color of an object in an image with AI
Send the photo to an AI image model, name the object and its new color, and list what must stay. One request per color, checked against the original.
AI hairstyle changer: try a new cut or color on your photo
Send a clear photo of yourself to an AI image model, describe the haircut or color, and say your face stays. Compare several versions of the preview.