Deinterlacing converts interlaced fields into pictures suitable for a progressive screen. It can combine information from adjacent fields, interpolate missing lines or adapt its method to movement. Good processing preserves still detail while avoiding combing around moving objects. The result depends on field order, source cadence and the chosen method. Deinterlacing is a picture-processing task; it cannot recover information that the original source never recorded or fix unrelated network interruptions.
Deinterlacing and Upscaling
Deinterlacing reconstructs pictures across fields; upscaling changes their spatial dimensions.
Different methods make different compromises
A simple weave combines fields and retains vertical detail in still areas, but moving edges can show combs. A simple bob method creates a progressive picture from each field, preserving temporal updates while interpolating missing lines. Motion-adaptive methods try to choose differently for still and moving areas, with quality and processing cost varying by implementation.
Field order and cadence matter
Presenting fields in the wrong order can make movement appear to jump backwards and forwards. Film-derived interlaced material may also contain a repeated cadence that needs reconstruction. A force-on filter applied to already progressive video can soften it unnecessarily, so processing should follow the actual media rather than a general preference.
Where processing can happen
The source provider, player, graphics hardware or television may perform conversion. If upstream software has already produced progressive output, a television cannot recreate the untouched fields. Comparing two documented player modes can reveal where the difference arises, but it does not establish which component performed every earlier conversion.
An illustrative example
A football channel has smooth motion but visible combing on players’ legs. A supported automatic deinterlacing mode removes the teeth. A different forced mode reduces motion detail, so the useful choice is the one that handles the actual field structure rather than the highest-sounding label.
Practical actions
- Establish whether the incoming source is interlaced before enabling a filter.
- Test one moving scene with automatic and documented supported modes.
- Avoid stacking player and television processing changes during the same comparison.
