DV to PNG Converter

Grab a PNG Still from a DV Tape Capture

This page pulls one frame — or a run of frames — out of a .dv camcorder capture and saves it as a PNG. It is written for anyone digitising MiniDV, DVCAM, Digital8, or DVCPRO footage who wants a picture rather than a clip: a frame for a photo book, a thumbnail, a contact sheet of a whole reel, or a reference still to edit. Two things about DV shape everything below, so they are worth knowing up front. DV is interlaced, which is why moving subjects can come out with comb lines. And DV uses non-square pixels, which is why a raw frame grab looks slightly the wrong shape until you tell it otherwise.

How to Convert DV to PNG

1. Upload Your DV File: Drag the file onto the page or click "+ Add Files". A raw .dv stream and a DV-in-AVI or DV-in-QuickTime capture all decode here, and you can queue several files to run with the same settings.
2. Choose the Frame under Frame Selection: Specific Frame is preselected with Time (seconds) set to 0, which returns the very first frame of the clip. Type any timestamp — 4.12 seeks to 4.12 seconds — or switch to Multiple Screenshots and set the Capture Rate, which runs from one frame every 0.1 seconds up to one every 10 seconds.
3. Set the Quality Preset and Image resolution (Optional): Quality Preset is preselected at Very High (Recommended); move it to Highest for a true-colour PNG (see the table below). Image resolution stays on Keep original, or you can pick Preset Resolutions, Resolution Percentage, or Width x Height in pixels or percent.
4. Convert and Download: Click "Convert" and save the PNG, or take a multi-frame run as a ZIP. No sign-up, no watermark.

The Quality Preset Decides Whether the Still Is Pixel-Exact

PNG's own compression is lossless, so a lot of guides stop there and tell you a PNG frame grab is automatically a perfect copy. On this page that is only true at one setting. The pipeline extracts your chosen frame as a lossless PNG first, then writes the final file according to the Quality Preset — and every preset except Highest routes the frame through a palette optimiser, which throws away colours before the lossless compression ever runs.

That is also why the Colors and Compression speed groups vanish from the page the moment you select Highest: both are controls for the palette optimiser, and at Highest there is no palette step to control. Compression level (1–10, preselected at 6) stays either way, because it only tunes how hard zlib works — it changes the file size, never a pixel.

On the same 720×480 frame from a real NTSC DV file, we measured the default palette-reduced PNG at roughly a quarter to a half the size of the true-colour Highest output. If the still is going into a photo book, an edit, or an archive, take the size hit and pick Highest. If it is a thumbnail or a web preview, the default is doing you a favour.

Getting the Shape Right: DV's Non-Square Pixels

A DV frame is 720 pixels wide whether the tape is NTSC or PAL, and those pixels are not square. The stored frame carries a pixel aspect ratio that tells a player how to stretch it back out; PNG has no equivalent field, so the still you download is the raw stored grid and will look narrow or wide next to how the tape plays. The fix is to type the display size into Width x Height yourself.

We probed a standard NTSC DV file to confirm the first row: the stream reports a 720 × 480 frame, a sample aspect ratio of 8:9, and a 4:3 display aspect. Leave Image resolution on Keep original and you get the 720 × 480 grid; set 640 × 480 and you get the picture as the camcorder intended it. If your footage is widescreen, check which of the two 16:9 rows applies before typing numbers.

Common Problems and How to Fix Them

• "Moving subjects have comb-like horizontal teeth" — DV is interlaced: each frame is woven from two fields captured a fraction of a second apart, and a still shows both at once. Nudge Time (seconds) a few hundredths either way to land on a moment where the subject is holding still, and the two fields line up.
• "The still looks stretched or squashed" — Non-square pixels, as above. Set Width x Height to the display size for your tape variant.
• "Reds and thin coloured lines look blocky" — That is baked into the tape. Baseline DV subsamples colour heavily: 4:1:1 in the 60 Hz variant and 4:2:0 in the 50 Hz variant, so colour detail was always coarser than brightness detail. No output format can put it back.
• "The still is soft and not HD" — DV is standard definition, full stop. 720 × 480 or 720 × 576 is roughly a third of a megapixel. Enlarging via Preset Resolutions interpolates; it does not sharpen.
• "I asked for a timestamp and got a black or empty frame" — Either the time is past the end of the clip or you landed on a dropout in the tape transfer. Nudge the timestamp earlier and try again.
• "I wanted one still and got a ZIP full of them" — Multiple Screenshots is selected. Switch back to Specific Frame for a single image.

Frequently Asked Questions

Is the PNG a pixel-exact copy of the DV frame?

Only on the Highest preset. PNG's compression is lossless, but the default Very High (Recommended) preset reduces the frame to an optimised colour palette before that lossless compression runs — smaller file, not an identical one. Select Highest and the frame is written as true-colour PNG with no palette step, and the still is an exact copy of the decoded frame. That is the setting to use for archives, print, or anything you plan to edit further.

Which frame do I get if I do not change anything?

The first one. Frame Selection opens on Specific Frame with Time (seconds) set to 0, so an untouched conversion returns frame zero of the clip. On a tape transfer that is often a moment of black, a lens cap, or a colour bar leader, so it is usually worth typing a real timestamp before converting.

Why does my still look horizontally squashed compared to the video?

Because DV pixels are not square and PNG cannot record that. An NTSC 4:3 tape stores a 720 × 480 grid with an 8:9 pixel aspect, meaning the picture is meant to be displayed 640 pixels wide; a PAL 4:3 tape stores 720 × 576 with a 16:15 pixel aspect and is meant to be shown 768 wide. A player reads that ratio from the video stream and corrects on the fly. A PNG has no such field, so it shows the stored grid. Set Width x Height to the display size from the table above and the still matches what you see on playback.

Why are there comb-like lines across anything that moves?

Interlacing. Every DV frame is built from two fields scanned a fraction of a second apart — on a CRT they were drawn in sequence, but a still image shows both woven together, so a moving subject appears in two slightly different places at once. There is no setting here that deinterlaces a single grab; the practical fix is to choose a low-motion instant. A held shot, a pause between gestures, or a static scene will grab cleanly, because the two fields barely differ.

What resolution will the PNG be, and can I get HD out of it?

Standard definition, and no. The tape holds 720 × 480 (NTSC) or 720 × 576 (PAL) — around 0.35 megapixels — and no conversion invents detail that was never recorded. Image resolution stays on Keep original by default, which is the honest setting. Choosing a larger Preset Resolution enlarges the pixel grid by interpolation, which makes a bigger file with the same amount of real detail in it.

Can I get a series of stills instead of just one?

Yes. Switch Frame Selection to Multiple Screenshots and set the Capture Rate, which is preselected at one frame per second and ranges from one every 0.1 seconds (ten per second) down to one every 10 seconds. Each sampled frame comes back as its own PNG, delivered together as a ZIP. It is the quickest way to build a contact sheet of a reel or to find the one moment where nobody blinked.

Why did the Colors and Compression speed controls disappear?

You selected the Highest quality preset. Both of those groups configure the palette optimiser — Colors sets the palette size (256 down to 2, preselected at 128) and whether to dither, and Compression speed sets how much effort the optimiser spends. At Highest there is no palette step at all, so the page hides controls that would have no effect. Compression level stays visible because it tunes the lossless zlib stage, which still runs.

Should I save a DV frame as PNG, JPG, or TIF?

PNG at the Highest preset is the right default for a still you might edit or keep: exact pixels, alpha support, and every browser and editor opens it. Choose DV to JPG when the still is going into an email, a message, or a web page and a much smaller lossy file is worth it. Choose DV to TIF when the destination is a print workflow or a preservation archive that expects TIFF — bearing in mind that browsers other than Safari will not display a TIFF at all. If what you actually want is the whole moving clip rather than a still, use DV to MP4.

What happens to my DV file after the conversion?

It is uploaded over an encrypted connection, processed on our servers, and deleted automatically a few hours later — no sign-up, no watermark, never shared or made public. Plan for the transfer, though: a DV capture runs at a fixed 3.6 MB per second of tape, so a full hour is around 13 GB. There is no per-file cap, but the whole file has to arrive before a frame can be pulled from it, so if you only want a still from one scene it is worth shortening the capture before you upload it. To split a reel you have already sent into workable pieces, the video cutter handles that separately.