Initializing... drag & drop files here
Supports: OGV
OGV is the Theora-in-Ogg video container donated by On2 to the Xiph.Org Foundation in 2002 and frozen as the Theora I bitstream in June 2004. It powered Wikipedia's video content and Firefox 3.5's HTML5 video implementation in 2009 before VP9 and WebM displaced it. The.ogv extension and video/ogg MIME type were standardised in RFC 5334 (September 2008). Theora playback has now been disabled by default in Chrome 120+, Edge 122+, and removed in Firefox 130+ — which is exactly why people convert old.ogv files: the codec no longer plays in modern browsers, but a PNG frame plays everywhere forever. Common use cases:
| Property | OGV (Ogg + Theora) | PNG |
|---|---|---|
| Type | Video container (animated, with audio) | Single image (static) |
| Codec / compression | Theora video + Vorbis/FLAC audio | DEFLATE (lossless) |
| Standardised | June 2004 (Theora I bitstream); RFC 5334 (2008) for .ogv |
ISO/IEC 15948 (2003); RFC 2083 (1997) |
| Color depth | 8 bits per channel YCbCr (4:2:0 typically) | 1, 2, 4, 8, 16-bit per channel; truecolor and palette modes |
| Transparency | No (Theora has no alpha channel) | Yes, full 8-bit alpha or 1-bit binary |
| Audio | Yes (Vorbis or FLAC) | No |
| Current browser playback | Chrome ≥120, Edge ≥122, Firefox ≥130 disabled or removed | Universal (every browser, every viewer since 1996) |
| Best for | Open-source video archives, MediaWiki content | Lossless stills, screenshots, UI graphics |
A 3-minute OGV at 720p sits around 30-60 MB; one extracted 720p PNG frame is typically 400 KB-2 MB depending on scene complexity.
| Setting | Range / Values | What it changes |
|---|---|---|
| Compression level | 1-10 (DEFLATE strength) | Visual output identical at every level; level 6-8 is the standard balance of size vs encode time. Level 10 squeezes a few extra percent at noticeably slower speed. |
| Compression speed | 1-10 | Trades encoder CPU time against compression density at the same level. |
| Colors → Original | 24-bit truecolor | Pixel-perfect — pick this unless you specifically need a smaller file. |
| Colors → By Color Reduction + Dither | 2 / 4 / 8 / 16 / 32 / 64 / 128 / 256 colors | Converts the PNG to a palette image (PNG-8). 128-256 colors with dither stays close to the original on most natural scenes; 16-32 colors is "8-bit retro" territory. |
| Quality Preset | Lowest / Low / Medium / High / Very High / Highest | Bundled preset that biases compression + sampling toward smaller or sharper output. |
| Resolution | 144p, 180p, 240p, 360p, 480p, 576p, 720p, 1080p, 1440p, 2160p, 4320p, or custom Width × Height | Down/upscale during decode. "Keep Original" preserves source pixels. |
| Mode | Output | When to use |
|---|---|---|
| Specific Frame | One PNG at the timestamp you enter (in seconds) | Pulling a poster frame, a thumbnail, or one screenshot at a known cue point |
| Multiple Screenshots — by frame count | 2-10 evenly spaced PNGs across the clip | Contact sheet, picking the best representative frame, A/B comparing scenes |
| Multiple Screenshots — by frame rate | PNGs at N fps for the full duration | Compositing, training-data extraction, frame-by-frame editing in Photopea / GIMP |
Chrome disabled Theora by default in version 120 (January 2024) and Edge followed in 122. Firefox, which carried Theora the longest, removed support in version 130. Theora was donated to Xiph.Org in 2002 and frozen in 2004, and modern browsers prefer the AV1 / VP9 / H.264 lineage. Converting representative frames to PNG (or the whole clip to MP4 or WebM) is the practical way to keep the content accessible. See caniuse OGV for the current support table.
No, because Theora has no alpha channel — every OGV frame is opaque. The output PNG will have an opaque background by default. PNG itself supports 8-bit alpha, so you can mask the background later in an image editor; the converter just can't invent transparency that isn't in the source.
OGV uses Theora inter-frame compression — only keyframes carry the full image; the rest are encoded as differences. PNG is lossless DEFLATE applied to one independent bitmap, so a single 720p frame routinely lands at 400 KB-2 MB even though the OGV averages a few KB per frame. If size matters, drop to PNG-8 ("By Color Reduction + Dither" at 128 or 256 colors) or use JPG instead — JPEG of the same frame is typically 4-10× smaller.
PNG uses DEFLATE (the same algorithm as ZIP and gzip), which is lossless at every setting. Compression level 1-10 trades CPU time for tighter DEFLATE blocks — the decoded pixels are byte-for-byte identical. Level 1 is the fastest encode, level 6 is the libpng default, and level 8-10 squeezes a few extra percent at noticeably longer encode times. Pick level 6 unless you're batch-processing thousands of files and need either speed (1-3) or maximum density (9-10).
Specific Frame is the right choice when you know the timestamp you want (e.g., 4 seconds in). Multiple Screenshots is for contact sheets, frame-by-frame editing, and "give me 5 representative stills from this clip" — set the frame count (2-10) or a screenshot fps (1-50) and the tool exports a numbered PNG sequence as a ZIP.
Yes — under "Multiple Screenshots", set the screenshot frame rate equal to the source's frame rate (most Theora video is 24, 25, or 30 fps; some MediaWiki uploads are 15 fps). You'll get one PNG per source frame as a numbered sequence. For a 30-second 30 fps clip that's 900 PNG files; expect 200 MB-1.5 GB total in the ZIP depending on scene detail.
There's no hard server cap — practical limits come from upload size and connection speed. Most users handle 500 MB-1 GB OGV files comfortably. For very long Theora archive dumps, extract a section first using a video cutter, or work with one segment at a time.
VLC captures the displayed frame at the size and color profile of your current playback window — it depends on your local Theora decoder, which Chrome and Firefox have already removed. XConvert decodes the OGV on our servers using a bundled codec, so the extraction works the same on any device and gives you control over Resolution, Color palette, Compression level, and batch frame export — none of which VLC's Snapshot shortcut offers.
PNG for anything with text, sharp edges, UI elements, or graphics — DEFLATE is lossless so screenshots stay pixel-perfect. JPG for photographic content where 5-10× smaller files matter more than perfect detail. If you need transparency support later, you must use PNG; JPG has no alpha channel. See OGV to JPG for the lossy path.