Skip to main content

Shepherd writeup
draft-ietf-cellar-ffv1

# Summary

The document shepherd is Peter Bubestinger-Steindl.
Ben Campbell is the responsible Area Director.

This document describes FFV1, a lossless video encoding format. The design of
FFV1 considers the storage of image characteristics, data fixity, and the
optimized use of encoding time and storage requirements. FFV1 is designed to
support a wide range of lossless video applications such as long-term
audiovisual preservation, scientific imaging, screen recording, and other video
encoding scenarios that seek to avoid the generational loss of lossy video
encodings.  This document defines version 0, 1 and 3 of FFV1.

The working group has chosen Informational as the requested publication type.


# Review and Consensus

Since FFV1 (version 0,1,3) already exists and is in production use for many
years already, there was a clear consensus for these versions. Although the
number of people interested in, and working with this format is growing, the
document itself was written and reviewed by a rather small group.

This small group however, consists of a team of experts in audiovisual format
encodings, where most of them have actively been working with FFV1 since its
early days. The knowhow and expertise in this group is a good mix between AV,
codec implementation, real world lossless video use cases, as well as mathematics.

FFV1 was mainly described for use in an audiovisual preservation context, which
is currently still its major use case. Its reference implementation is in FFmpeg
(which is also currently the most widely used one). Additional implementations
were written according to an earlier version of this document, by one of its
authors, Jérôme Martinez in "[MediaConch](https://mediaarea.net/MediaConch)" and
"[RAWCooked](https://mediaarea.net/RAWcooked)".
MediaConch was designed as conformance checker for validating FFV1 files
according to its specs.

I have personally tested and verified all three: the reference implementation (see:
http://download.das-werkstatt.com/pb/mthk/ffv1_stats/latest/), as well as
MediaConch and RAWcooked.

I found the document very well written and easy and clear to understand, and was
actually a bit surprised that it is so comprehensible, considering that it is
describing a video encoding format. I have not double-checked the math behind it
though.


## Addressed Comments

On April 29th 2020, Murray S. Kucherawy posted comments on version -13. Those
comments were mainly remarks and suggestions about wording, formatting and a few
clarification improvements.

These comments were addressed and fixed in several patches and sorted out
completely by June 23rd 2020.



# Intellectual Property

Each author has confirmed conformance with BCP 78/79 to the best of their
knowledge.

During the PREFORMA project, which was looking for "good quality standardised
file formats for preserving data content in the long term" (Quote:
http://www.preforma-project.eu/), Professor Björn Lundell from the Swedish
University of Skövde (http://www.preforma-project.eu/skovde.html), was to check
licensing and patent situations of evaluated formats.

He did not discover any IPR claims on parts/algorithms of FFV1.
The current known consensus about this is, that is is highly unlikely that there
are IPR claims on FFV1. Especially since it is based on 40 year old algorithms.



# Other Points

  * There is a downref to
     [RFC 4732 (Internet Denial-of-Service Considerations)](https://tools.ietf.org/html/rfc4732)
  * The downref is currently not listed yet in
     [IETF's Downref Registry](https://trac.ietf.org/trac/iesg/wiki/DownrefRegistry)




Back