DV Video
DV video is an intra-frame compressed video format commonly encountered in digital tape transfers.
When a digital tape transfer copies DV data, it can preserve the tape's existing compressed representation. The computer is receiving the encoded video, audio, and associated recording data rather than necessarily making a new picture encode.
That distinction explains both the value of a copied transfer and its familiar large file rate.
What DV25 counts
Common consumer SD DV has roughly 25 Mb/s of coded video. A complete DV data stream includes more than the video payload.
| Common consumer system | Stored picture grid | Frame rate |
|---|---|---|
| 525/60 | 720 × 480 | 30000/1001, approximately 29.97 fps |
| 625/50 | 720 × 576 | 25 fps |
Selected SD forms, not every professional or HD member of the DV family.
The system's 525/625 line designation includes lines beyond the active stored picture. Its complete DV frame is organized in Digital Interface (DIF) blocks, which provide the transferable structure.
Count a complete frame
A DIF block has 80 bytes. A sequence contains 150 blocks, allocated as:
| Section | Blocks in one sequence |
|---|---|
| Header | 1 |
| Subcode | 2 |
| Video auxiliary | 3 |
| Audio | 9 |
| Video | 135 |
The total is 150. The 25-Mb/s structure has ten such sequences per 525/60 frame or twelve per 625/50 frame:
- 525/60: 80 × 150 × 10 = 120,000 bytes/frame.
- 625/50: 80 × 150 × 12 = 144,000 bytes/frame.
For the 25 fps form, that is 144,000 × 25 = 3,600,000 bytes/s, or 28.8 Mb/s, for the complete stream. One hour is 12.96 decimal GB before an additional file wrapper's overhead.
The apparent disagreement with “DV25” is thus a difference in what is being counted. The complete representation carries audio and structural/auxiliary data alongside its coded picture information.
Reconstruct the independently coded picture
DV picture compression works within a frame rather than relying on a long picture-to-picture prediction chain. This gives editing software direct access to individual recorded pictures once their data is located.
The ordinary picture coding is lossy. Copying the DV bytes avoids another video-encoding generation; it preserves the camera's encoded result rather than changing that result into an uncompressed original.
An export that decodes DV and then makes H.264 is a new representation. The two jobs can be useful for different purposes: retaining the transferred source and creating a delivery copy.
Present the source picture correctly
The reconstructed raster also needs its display interpretation. Standard-definition DV often uses nonsquare pixel presentation, and the same stored dimensions can describe different intended aspect ratios. Aspect information supplies that geometry.
Native interlaced recordings contain two field capture times in each frame. For progressive presentation, field order and the chosen reconstruction cadence determine how their motion is retained.
This is the information a playback or conversion stage needs from the transferred recording. A correctly copied DV source preserves that data for an Emby playback version whose picture shape and field motion can then be interpreted deliberately.