When To Worry About Telecine

There are some general guidelines you can follow to know when to analyze video to see if it would benefit from IVTC. The first rule of thumb is that telecine is only for film footage. This rules out a lot of television content, although it's common for television shows to be shot on film or even progressive digital video. If the digital video uses film's 24fps it should be treated as film. Always remember that the most reliable tool for this is your own eyes. While a computer can be fooled by images that aren't mathematically similar enough to be related, the human eye can sometimes distinguish similarities and differences that can't be caught any other way.

Identifying telecine

The best tools for finding telecine patterns and IVTCing video are DGIndex and AviSynth. Most telecined video is in MPEG-2 format, so we'll be starting with DGIndex. DGIndex, an improved version of DVD2AVI, can be used to examine a number of MPEG-2 file properties, including frame structure (progressive or interlaced) and the presence of RFF flags indicating pulldown. AviSynth can then be used to pass on the information from DGIndex (along with the video) and has filters that can help you make decisions regarding IVTC. AviSynth plugins can be used to match fields, remove combing artifacts, and decimate to recover 24fps (or at least 23.976fps) progressive frames from telecined video. This can then be used by an encoder or viewer of your choice, or even in other editing software.

Although DGIndex is powerful, particularly for getting information about your video, it's really suited for performing IVTC by itself except in limited cases. Instead it should be used as a helper application for an actual video editing program. There are a few tools capable of this, but since AviSynth is as powerful as almost anything available, and the price (free) is just as good, that's what we'll be using. It's also possible to skip DGIndex altogether, but because of the information it can give you, and pass to AviSynth as well, it should always be part of the process for MPEG-2 files.

Along with a program like VirtualDub or AVS editor like AVSEdit or AvsP, and DGIndex/DGDecode for analyzing and decoding MPEG files, using AviSynth to perform IVTC on a variety of sources is relatively easy. Most programs that can read AVI files will accept AviSynth scripts as input. For those that won't, you can also use a variety of programs (including VirtualDub) to encode to an intermediary file in an uncompressed YUV format called HuffYUV. All of these tools are free, but HuffYUV files are very large. Make sure you have close to 40GB of storage per hour of video you're encoding. The HuffYUV encoded AVI can then be substituted for your original source as input to a video encoder or other program.

Using AviSynth

You can find more information on using AviSynth in our guides Using AviSynth and AVSEdit: AviSynth For The Scripting Impaired. Those guides will give you the basics of both AviSynth, and performing IVTC with it. In addition the AVSEdit guide will show you how to use a simple script editor to write and preview scripts. Since AviSynth has a text (file) based interface while IVTC is a visually oriented operation it's important that you be able to use some other software (such as AVSEdit and/or VirtualDub) for viewing the output of AviSynth scripts. This will be necessary at several points in the process. This guide will also expand upon and update the IVTC information in the AVSEdit guide.

Using DGIndex

If you're new to DGIndex, or have a question about the instructions in this guide, you should start with our guide on Using DGIndex. Although it doesn't cover IVTC specifically, there are instructions for using various DGIndex features that will be required for this guide.