MPEG to PNG Converter

How to Extract PNG Frames from MPEG Video

1. Upload Your MPEG Files: Drag and drop or click "+ Add Files" to select.mpg or.mpeg files. Batch is supported — queue several clips and extract them in one pass.
2. Pick a Quality Preset: Default is the source quality. Set the Quality Preset to Highest for archival-grade frames, or step down to High/Medium when you need smaller PNGs for thumbnails. PNG itself is lossless, so the preset controls source-decode fidelity, not file-level quantization.
3. Choose Frame Selection and Resolution (Optional): Use Specific Frame to grab a single still at a time you enter in seconds, or Multiple Screenshots to pull frames at a chosen framerate (1–30 fps) for an image sequence. Set Image Resolution by Preset Resolutions (144P through 4320P), Resolution Percentage, or custom Width/Height (aspect ratio locked).
4. Convert and Download: Click Convert. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours — no sign-up, no watermark, never shared.mpg files are uploaded over an encrypted connection and auto-deleted after a few hours unencrypted.

Why Extract PNG from MPEG?

MPEG-1 was published as ISO/IEC 11172 in August 1993 to compress VHS-quality video to about 1.5 Mbit/s for Video CDs and early digital broadcasts. MPEG-2 (ISO/IEC 13818) extended this for DVDs, ATSC/DVB broadcast, and interlaced HD. Both store frames as I/P/B groups — only I-frames are full keyframes, the rest are deltas — so extracting a usable still means decoding the GOP and rendering a clean PNG. PNG (RFC 2083, 1997; ISO/IEC 15948:2004) uses lossless DEFLATE compression with 8- or 16-bit-per-channel depth and an optional alpha channel, making it the right target when you care about every pixel.

• DVD and VCD archival — Old home-movie discs, training videos, and game cutscenes are still MPEG-2 or MPEG-1. Extracting keyframes as PNG preserves the original pixels for restoration in tools like Photoshop or GIMP.
• Broadcast and surveillance evidence — DVR exports and many CCTV systems still output.mpg containers. PNG frames keep timestamps and overlays sharp for legal review, where JPG's blocking artifacts can obscure license plates or fine detail.
• Documentation and tutorials — Frames from screen recordings or product demos kept as PNG show crisp UI text, icons, and chart lines without JPG halos around edges.
• Animation and rotoscope work — Image sequences feed straight into Krita, Blender, or After Effects. PNG's lossless pipeline avoids generational quality loss across edit passes.
• Computer vision datasets — Training pipelines (object detection, OCR, action recognition) prefer PNG so JPEG compression noise doesn't contaminate features.
• Thumbnails and posters — Pull one Specific Frame at the time you want and use it as the poster image for video players, blog posts, or a video preview for social.

MPEG vs PNG — Format Comparison

Frame Selection and Resolution Guide

Frequently Asked Questions

Will the PNG frames be sharper than the source MPEG video?

No — a PNG can't add detail that isn't in the MPEG. MPEG-1/2 uses YCbCr 4:2:0 chroma subsampling and 8-bit per channel quantization, so fine color detail is already discarded at encode time. What PNG does preserve is everything the decoder produces — no JPEG ringing around text, no additional generational loss when you edit. If your source is a 352×240 SIF MPEG-1, your PNG will be 352×240 of crisp but soft pixels; if it's a 1080i MPEG-2 broadcast, you'll get a 1920×1080 still with the interlaced fields woven in as broadcast — clean on static shots, combed on motion.

Should I extract from an I-frame or any frame?

Most decoders reconstruct any frame transparently, so you can pick any timestamp. That said, I-frames (keyframes) are the only ones encoded independently — P- and B-frames are deltas referencing neighbors. On heavily compressed MPEG-2 the difference can be visible: I-frames tend to show full detail, while a B-frame near a scene change can briefly look softer. If a specific timestamp looks worse than you expected, nudge the time by ±0.1–0.5 seconds to land on a nearby I-frame.

What's the difference between Specific Frame and Multiple Screenshots?

Specific Frame outputs a single PNG at the time-in-seconds you enter — useful for poster images. Multiple Screenshots outputs an image sequence at a chosen rate (1, 2, 5, 8, 10, 12, 15, 20, 24, 25, 30, or 50 fps) across the whole clip — useful for storyboards, animation, or feeding a directory of frames into another tool.

Why is my PNG so much bigger than the original MPEG file?

A 90-minute MPEG-2 DVD at 6 Mbit/s is about 4 GB of compressed video. Extracted as 1080p PNGs at 24 fps you'd produce ~130,000 frames at 1–5 MB each — easily 200+ GB. Two reasons: MPEG is lossy and inter-frame (a P-frame is often a few kilobytes), while PNG is lossless and per-frame. If you only need a few moments, use Specific Frame; if you need the whole sequence at smaller size, drop resolution to 50% or extract JPG frames instead.

Do PNG frames keep the alpha channel from the video?

MPEG-1 and MPEG-2 don't carry alpha — they're 4:2:0 YCbCr with no transparency channel. PNG output will be opaque RGB. If you need a transparent matte, you'll have to chroma-key or rotoscope after extraction in Photopea or similar.

Can I extract frames from a specific portion instead of the whole video?

Specific Frame handles one timestamp. For a range, run Multiple Screenshots first, then keep only the frames in the seconds you want — or pre-trim the source with Trim MPEG so only the relevant segment is processed, which is much faster than extracting and discarding.

Does interlaced MPEG-2 (DVD, broadcast) extract cleanly?

Standard-definition DVDs and many broadcasts are interlaced — each frame is actually two fields captured 1/50 or 1/60 second apart. Without deinterlacing, fast motion shows comb artifacts. Our extractor renders the decoded frame as-is, so for moving subjects you may see combing on PNG output. If you need progressive stills, run a deinterlace filter over the MPEG in a desktop editor first, then extract from that progressive intermediate.

What's the maximum file size I can upload?

The practical cap is upload size and connection speed — conversion runs on our servers. Large DVDs (4–8 GB VOB-style .mpg files) are supported; on slower connections, trim to the segment you need first. Files are deleted from our servers automatically after a few hours.

Is PNG always the best frame format?

Not always. PNG is right for screenshots, text, charts, and compositing work. For continuous-tone footage (live action, nature) where you'll publish on the web or feed a thumbnail generator, JPG at 85–95% gives 5–10× smaller files with no visible difference. PNG also has no native animation — if you want a moving preview, target MPEG to GIF or MPEG to WebP instead.