Skip to main content

Comparision between 10-bit Uncompressed and FFV1 Video Archive Preservation at Oxford Duplication Centre

DISCUSSING FFV1

The most common challenge in digitising moving image is the file sizes that result from the actual digitisation.  The file sizes can be huge and with that comes the increased cost of storage and maintenance for long term preservation.


Common consensus consider 10-bit uncompressed to be the preservation standard for moving image because it uses no file compression. 

It is considered the most reliable safest format for moving image preservation at the current time. 10-bit uncompressed deliver high image resolution, colour quality and sharpness whilst avoiding motion compensation and compression artefacts.




THE DOWNSIDE TO 10-BIT UNCOMPRESSED

1 hour of 10-bit uncompressed video can produce a 100gb file. To put it into perspective it would take 21 DVDs to store a video of this size.  As you can see, its a lot of data!

WHAT IS FFV1?

FFV1 is a video codec developed within FFmpeg. It is lossless, meaning that it compresses video without introducing quantization degradations. Therefore, FFV1 is a good choice for archiving and preservation. A lossless codec comparison found FFV1 to be the "most balanced", offering "relatively good speed and high compression". There are two versions of the codec supported by ffmpeg, version 1 and 3.

COMPARISON

FFV1 is a lossless codec which can store its own aspect ratio data which has started to gain in popularity as an alternative to 10-bit uncompressed.  FFV1 uses lossless compression (think of a zip drive) to store digitised moving image at reduced file sizes without any data loss.  

FFV1 is part of the  open-source FFmpeg project which has been around since 2003.  FFV1 uses entropy encoding to deliver mathematically lossless intra-frame compression.  As a result this produces substantially smaller file sizes when compared to 10-bit uncompressed digitisation.

TESTING ENCODED FILES

The testing shows that files encoded to FFV1 produces files around 1/3rd smaller than 10-bit uncompressed.  Both formats can be carried in a variety of different wrappers (container files) such as AVI (Microsoft), MOV (Apple) or MKV (open source).

The encoded video and audio streams are wrapped together in the container with other data streams that include technical metadata.  The type and variety of data that a container can hold are specific to that container format.

EXCITING TIMES BUT SOME DOWNSIDES!

Reduced files produced using FFV1 are exciting but there are some downsides.  FFV1 is open source and as such will not play on  video software on MAC and Windows, nor can FFV1 be utilised within commercially available digitisation hardware and software. It can only be converted using 'terminal command'.  This is because currently Apple, Microsoft, Adobe and Black-magic has not adopted the codec or announced plans to do so.

Any file format that does not eventually achieve widespread adoption and universal playback within the broadcasting and film-making communities has a higher risk of long term obsolescence and lack of engineering support.  

I should also mention that FFV1 version 3 is nearing release. It is now available in ffmpeg under experimental status as testing completes. The version 3 adds multithreaded encoding, which should speed up encoding by about 2.5 times. For long term digital preservation FFV1 version 3 also incorporates mandated checksums per frame. Thus if any digital damage occurs to the file in storage or transmission, an ffv1 decoder should be able to identify which frames were affected, which makes the preservationist's response much more specific than if only enabled with a whole file checksum.

Generally I'd recommend that an uncompressed codec may be the safer choice for an archive with limited digital preservation infrastructure since the technology dependencies, consequences of digital damage, and requirements for access may be less severe than with lossless codecs.


container
video-codec
audio-codec
1.
MXF
JPEG2000
PCM (uncompressed)
2.
MXF
Uncompressed
PCM (uncompressed)
3.
MOV
ProRes
PCM (uncompressed)
4.
MKV
FFV1
PCM (uncompressed)
5.
AVI
FFV1
PCM (uncompressed)



Labels:
Location: 29 Banbury Rd, Kidlington OX5 1AQ, UK

Comments

Post a Comment

Popular posts from this blog

The Repair Shop - How To Spot A Ferrotype Camera 1855-1940s

After watching The Repair Shop on BBC1 restore a beautiful and rather rare ferrotype camera I thought a blog on the process would be interesting. Not only did they repair but they managed to have the camera working, taking photographs. This was very inspirational given the age of the camera. ABOUT FERROTYPE PROCESS Ferrotypes first appeared in America in the 1850s, but didn’t become popular in Britain until the 1870s. They were still being made by while-you-wait street photographers as late as the 1950s. The ferrotype process was a variation of the collodion positive, and used a similar process to  wet plate photography . A very underexposed negative image was produced on a thin iron plate. It was blackened by painting, lacquering or enamelling, and coated with a collodion photographic emulsion. The dark background gave the resulting image the appearance of a positive. Unlike collodion positives, ferrotypes did not need mounting in a case to produce a positi...
1 comment
Read more

Onion Skin Archive Book Scanning - What is this and how do we process the pages?

CURRENT BOOK SCANNNING PROJECT.  We are currently working on a very large archive of old books that require HQ scanning to Archival TIFF images.  Once processed, these images will be prepared to PDF with OCR (optical character recognition) for a complete searchable output.   The difficulty in this order, is the books are prepared using a medium called Onion Skin Paper. Whilst we are very confident in preparing this type of medium, it is very important to be aware that there are risks with scanning, given the sometimes-fragile nature of the paper.   Tears and rips can occur, so a very gentle white glove approach is required. Equally, with the nature of onion skin, the paper is very translucent which requires a sheet of white paper to be placed under each page before scanning. This then grants a very good HQ image that we can work with.   WHAT IS ONION SKIN PAPER? Onion skin paper is a type of very light weight, almost translucent paper that ...
Post a Comment
Read more

Vellum Document Deed Scanning in Oxfordshire

One of our most interesting orders came in today in the form of a vellum property deed dating back nearly 200 years. We were asked to prepare a digital copy so our client could preserve the original. After assessment we decided on which scanner to use and scanned the deed to a high quality Archival TIFF file before applying post production tone and sharpening. The results were stunning and a wonderful image produced. Interesting fact:  To ensure you could prove a duplicate copy, some deeds had a defined curved wave cut out at the top. This meant the original's replica could be matched 100% to the original, safeguarding fraud. What is vellum? Vellum is prepared  animal skin or "membrane", typically used as  a material for writing or printing on, to produce single pages, scrolls,  codices  or books. The word is derived from the  Latin  word  vitulinum  meaning "made from calf". Typically deeds are folded and stored. Alth...
1 comment
Read more