Initializing... drag & drop files here
Supports: TS
.ts (MPEG Transport Stream) file — typical sources are PVR recordings from a TV tuner card, IPTV set-top box captures, and downloaded HLS .ts segments. Batch conversion is supported.12.500 means 12 seconds 500 ms) to grab one still, or Multiple Screenshots to extract a sequence at 1-30 fps. This is the most important choice — a 30-minute TS recording at 24 fps contains over 43,000 frames, so picking the right mode keeps the output manageable.TS (MPEG-2 Transport Stream) is a broadcast container — it carries video, audio, and Program/System Information so it can survive transmission errors and resume after corruption. That makes it ideal for over-the-air ATSC/DVB capture, IPTV recording, and HLS streaming, but useless for sharing a still image. JPEG is the universal photo format: it opens in every browser, every messaging app, every CMS, every document editor. Common reasons people pull JPEGs out of a TS file:
.ts. Extracting one frame gives you cover art for a media library like Plex or Jellyfin..ts segment in a video player is slow; converting the first I-frame to JPEG lets you scan a folder of segments visually.| Property | TS (MPEG-TS) | JPEG |
|---|---|---|
| Type | Video container | Still image |
| Codec | Usually H.264 / MPEG-2 video + AC-3 / AAC audio | DCT-based image compression (1992) |
| Designed for | Broadcast / streaming resilience | Photographic stills, web delivery |
| Typical size | 5-15 MB per minute (1080p H.264) | 100-500 KB per 1080p still |
| Universal viewing | Needs VLC, MPC-HC, or similar | Every browser, every OS, every image viewer |
| Carries audio | Yes | No |
| Frame-accurate seek | Slower (depends on GOP and I-frame interval) | N/A — single frame |
A 1 GB TS recording at 1080p contains roughly 2-4 GB of raw frame data; the right one or two frames as JPEG is usually under 1 MB total.
| Preset | Approximate JPEG quality | Best for |
|---|---|---|
| Highest / Very High | 90-95% (4:4:4 or 4:2:2 chroma) | Archival, print, evidence frames where any compression artifact matters |
| High (recommended default) | 85% (4:2:0 chroma) | Web embedding, thumbnails, social — visually lossless for most viewers, ~50-60% smaller than Q95 |
| Medium | 70-75% | Contact sheets, bulk extraction, OCR input |
| Low / Lowest | 50% or below | Visible artifacts; use only for tiny preview thumbnails |
Quality 85 with 4:2:0 chroma subsampling is the industry rule of thumb for "indistinguishable from the source at normal viewing distance" — it's what most browsers and CDNs default to.
All three are MPEG-2 Transport Stream variants. Plain .ts is the bare transport stream used for broadcast (ATSC, DVB) and HLS streaming segments. .mts and .m2ts are the AVCHD camcorder flavors — same underlying packetization, but with extra timing data (4 BDAV bytes prepended per packet on M2TS) and an AVCHD folder structure on the SD card. For frame extraction the workflow is identical; if you have a camcorder file see MTS to JPEG or M2TS to JPEG.
Pick Specific Frame and enter the time in seconds with a decimal — 45.250 means 45 seconds 250 milliseconds. Frame-accuracy depends on the TS file's GOP structure: encoders place I-frames (full frames) every 1-2 seconds and P/B-frames in between. xconvert decodes to the requested time and returns the nearest displayable frame, which is typically within 33-66 ms of your target on a 30 fps source.
Yes, with caveats. Set Multiple Screenshots, then choose a frame rate (1, 2, 5, 8, 10, 12, 15, 20, 25, 30, or 50 fps). Extracting every frame from a 30 fps source means 1,800 JPEGs per minute — useful for VFX or per-frame analysis, brutal for casual use. For thumbnails or contact sheets, 1 fps (one frame per second) is usually enough.
JPEG is 5-15× smaller than PNG for photographic content (which is what video frames are) and indistinguishable in quality at Q85+. PNG only wins for screenshots of UI, line art, or anything with hard edges and flat colors — video frames have neither. If you specifically need lossless output (graphics overlays, evidence chain-of-custody), use TS to PNG instead.
No. JPEG is a still-image format with no audio track and no subtitle support. If the TS has burned-in subtitles (open captions rendered onto the picture), they'll appear in the JPEG. If the TS has soft subtitles in a separate stream (closed captions, DVB subtitle PID), they won't be rendered onto the extracted frame.
Practical browser limits put it around 2 GB per file — roughly a 2-3 hour 1080p H.264 recording or a 30-minute MPEG-2 broadcast capture. Frame extraction itself is fast because xconvert seeks directly to your target timestamp instead of decoding the whole file. For bulk segment extraction across many TS files, upload them in a batch and run a single frame-per-file pass.
Two reasons. First, broadcast TS streams are often interlaced (1080i instead of 1080p) — xconvert deinterlaces on the fly, which can soften fine motion. Second, source bitrates for OTA and IPTV are typically 8-15 Mbps for 1080p, so the source itself isn't pristine. For sharpest stills, extract from an I-frame (most likely to land on one at multiples of 1-2 seconds) and use Quality Highest.
Yes — if you wanted a playable video rather than stills, see TS to MP4 for the most compatible output, or TS to MKV for a lossless container repackage. Use this page only when you want still images extracted from the video.