Universal Scene Description

A few years ago I became obsessed with creating 3D models from physical objects. There was an app on my iPhone called 123D Catch by AutoDesk and it allowed you to take a series of photos with your iPhone camera, then combine them to create a 3D Model. This lead me down a path to eventually take a course on Photogrammetry and develop a process for capturing objects in our Museum.

Autodesk eventually discontinued the app and built the technology into their paid products. This is when we started seeing lidar introduced with handheld devices. The first one I tried was using my XBOX360 kinect sensor with the skanect software. The quality was horrible, but was fun to learn about depth sensors and structure from motion. When the iPhone finally came out with lidar sensor it was like Apple had read my mind. I love having the ability to capture objects I find into 3D models. The quality is pretty good, not as good as taking the time to capture image sets for photogrammetry and use tools like Aigisoft metashape, but apps like Scaniverse do a fantastic job. You can check out some of the models I have captured on my Sketchfab page.

With any new technology comes new file formats, and 3D formats are definitely no exception. It seems every software developer has to come up with their own proprietary format leaving the digital preservation folks scrambling to keep up. The DPC and Archivematica published a report a couple years ago and state:

“There are many challenges in preserving 3D data. As well as the complexity of the data itself, there is
a lack of interoperability between the different (often proprietary) systems that are used to create
and manipulate 3D models. Relationships to other data, software and hardware also need to be
captured and managed effectively.”

https://www.dpconline.org/docs/technology-watch-reports/2479-preserving-3d/file

With my new iPhone in hand I found myself with new file format I was unfamiliar with. Universal Scene Description is a framework to exchanging 3D data between different software developed by Pixar. The relationship between Apple and Pixar goes way back so it was no surprise the Apple iPhone has support built in for this new format and I found myself capturing and sending 3D models to others with an iPhone. The USDZ format is a ZIP package format for containing a USD 3D model and is perfect for sharing and preserving.

There is no current PRONOM signatures for identifying USD formats, so I wanted to look into creating one. This is where I ran into a problem. The current PRONOM signature syntax has no way of properly identifying the USDZ format. Let me explain.

When DROID or Siegfried is used to identify a container format such as USDZ. It will first identify the format as a ZIP file, which technically it is. This triggers the software to then refer to the container signature to see if any patterns from the files internal to the ZIP match to a known format. This is done by pointing to a specific file and a hex pattern or ascii string within the file. In the case of a USDZ the internal structure may look like this:

Listing archive: scaniverse-20210928-113055.usdz

--
Path = scaniverse-20210928-113055.usdz
Type = zip
Physical Size = 5702256

   Date      Time    Attr         Size   Compressed  Name
------------------- ----- ------------ ------------  ------------------------
2021-09-28 11:47:36 .....       297999       297999  scaniverse-20210928-113055.usdc
2021-09-28 11:47:36 .....      5403849      5403849  0/texgen_0.jpg
------------------- ----- ------------ ------------  ------------------------
2021-09-28 11:47:36            5701848      5701848  2 files

In this sample file the name of the USDZ is the same name as the internal USDC file. So the name of the USDC is variable and DROID needs a static name and path to look for patterns. The USDZ specification is clear that the only required file inside a USDZ is a USD model, anything else is ancillary and is not always going to be included. Currently the only format used is USDC, but in the future may allow a simple USD or USDA format. In addition, some of the other sample files show a very nested USDC file, making identification even more difficult.

Listing archive: Scan.usdz

--
Path = Scan.usdz
Type = zip
Physical Size = 19155195
Characteristics = Minor_Extra_ERROR

   Date      Time    Attr         Size   Compressed  Name
------------------- ----- ------------ ------------  ------------------------
2021-03-09 09:22:36 .....     19154773     19154773  /private/var/mobile/Containers/Data/Application/EFD09E66-32FB-4B08-8BED-B7E3D78FE1A8/tmp/Scan.usdc
------------------- ----- ------------ ------------  ------------------------
2021-03-09 09:22:36           19154773     19154773  1 files

The USDZ format is not the only file format which makes identification difficult through variable names and non-static patterns. An issue on GitHub has been raised to address this problem. One potential fix is to use glob patterns as suggested by the amazing Richard Lehane, creator of Siegfried. This way we could use wildcard to ignore the variable names and find any file with an extension of .USDC for example. The USDC file format has a nice 8 byte header “PXR-USDC” which is perfectly suited for identification so our container signature might look like this:

<ContainerSignature Id="1000" ContainerType="ZIP">
  <Description>USDZ 3D Package</Description>
    <Files>
	<File>
          <Path>*.usdc</Path>
              <BinarySignatures>
                  <InternalSignatureCollection>                    
	             <InternalSignature ID="300">
	                 <ByteSequence Reference="BOFoffset">
	                     <SubSequence Position="1" SubSeqMinOffset="0" SubSeqMaxOffset="0">
	                         <Sequence>50 58 52 2D 55 53 44 43</Sequence>
	                     </SubSequence>
	                 </ByteSequence>
	             </InternalSignature>
	          </InternalSignatureCollection>
             </BinarySignatures>
        </File>           
    </Files>
</ContainerSignature>

Update: I was able to get a beta version of Siegfried working with my test signature.

siegfried   : 1.11.0
scandate    : 2023-06-02T08:54:27-06:00
signature   : default.sig
created     : 2023-06-02T08:52:33-06:00
identifiers : 
  - name    : 'pronom'
    details : 'DROID_SignatureFile_V112.xml; container-signature-20230510.xml; extensions: usdz-signature-file-v1.xml; container extensions: usdz-dev1-signaturefile-20230601.xml'
---
filename : 'scaniverse-20210928-113055.usdz'
filesize : 5702256
modified : 2021-09-28T11:47:37-06:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'BYUdev/1'
    format  : 'USDZ 3D Package'
    version : 
    mime    : 'model/vnd.usdz+zip'
    class   : 
    basis   : 'extension match usdz; container name scaniverse-20210928-113055.usdc with byte match at 0, 8 (signature 1/2)'
    warning : 

I am still in the process of testing some beta versions of Siegfried in hopes of getting the glob matching to work, but still have more to do. Stay tuned!

Embedded WAVE, thanks HP πŸ‘‹

Digital Preservation is all about identifying risks. This is done through a process which includes identification, validation, and metadata extraction. The more you know about the digital data you need to preserve over time, the more you can do to minimize those risks with the goal of making the data accessible over time.

Many formats are pretty straight forward, they are identifiable through a header and then have some binary bits or plain text that is readable by certain software. Others are more complicated. A common practice for more complex needs is to use a container. Word processing programs started out with plain text with maybe some formatting codes mixed in, then many moved to the Microsoft OLE container so you could have additional content embedded in a single file. Today file formats such as DOCX use a ZIP container, which houses all the text, images, formatting and anything else the format supports. Knowing what the format is and knowing what it may contain is important to preservation.

IM000959.JPG

I collect older digital cameras, specifically cameras with unique file formats, raw and otherwise. When I picked up a HP (Hewlett-Packard) point and shoot camera awhile back, I was initially unimpressed as it would only capture in a JPEG format and only 3 quality settings. While looking at a copy of the manual, I saw the camera was capable of capturing audio clips or voice memos for each photo taken. This can be handy when taking many photos and need a reminder about the context. This was not unique to HP, as many cameras could do this, normally a JPG was captured and the Audio would have the same name connecting the two. But when I recorded some audio on my little HP, placed the SD card in my computer, I couldn’t find the additional audio file. I also not the only one to ask about this.

There are many types of JPG files. Raw Streams, JPEG File Interchange Format (JFIF), and Exchangeable Image File Format (EXIF). Normally these formats have raster image data sprinkled with metadata. I have seen JPEG files embedded into other formats and containers, such as MP3, PDF, etc, but JPEG’s are not container formats. Or so I thought…..

View of HP Photosmart 433 folder in HP Photo & Imaging Gallery

Lets take a look at an image I took with my HP Photosmart 433. We’ll start with identification:

siegfried   : 1.10.1
scandate    : 2023-05-25T12:27:04-06:00
signature   : default.sig
created     : 2023-05-22T08:43:02-06:00
identifiers : 
  - name    : 'pronom'
    details : 'DROID_SignatureFile_V112.xml; container-signature-20230510.xml'
---
filename : 'GitHub/digicam_corpus/HP/Photosmart 433/IM000959.JPG'
filesize : 178922
modified : 2023-05-25T11:23:32-06:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'x-fmt/391'
    format  : 'Exchangeable Image File Format (Compressed)'
    version : '2.2'
    mime    : 'image/jpeg'
    class   : 'Image (Raster)'
    basis   : 'extension match jpg; byte match at [[0 16] [366 12] [178907 2]] (signature 2/2)'
    warning : 

IM000959.JPG was identified as x-fmt/391 which is a compressed Exchangeable Image File Format. version 2.2. Pretty straight forward. Next lets look at validation:

Jhove (Rel. 1.28.0, 2023-05-18)
 Date: 2023-05-25 12:35:16 MDT
 RepresentationInformation: GitHub/digicam_corpus/HP/Photosmart 433/IM000959.JPG
  ReportingModule: JPEG-hul, Rel. 1.5.4 (2023-03-16)
  LastModified: 2023-05-25 11:23:32 MDT
  Size: 178922
  Format: JPEG
  Status: Well-Formed and valid
  SignatureMatches:
   JPEG-hul
  ErrorMessage: Tag 41492 out of sequence
   ID: TIFF-HUL-2
   Offset: 606
  MIMEtype: image/jpeg
  JPEGMetadata: 
   CompressionType: Huffman coding, Baseline DCT
   Images: 
    Number: 1
    Image: 
     NisoImageMetadata: 
      FormatName: image/jpeg
      ByteOrder: big_endian
      CompressionScheme: JPEG
      ImageWidth: 640
      ImageHeight: 480
      ColorSpace: YCbCr
      DateTimeCreated: 2021-11-16T09:04:04
      ScannerManufacturer: Hewlett-Packard
      ScannerModelName: hp PhotoSmart 43x series
      DigitalCameraManufacturer: Hewlett-Packard
      DigitalCameraModelName: hp PhotoSmart 43x series
      FNumber: 4
      ................................
     Exif: 
      ExifVersion: 0220
      FlashpixVersion: 0100
      ColorSpace: sRGB
      ComponentsConfiguration: 1, 2, 3, 0
      CompressedBitsPerPixel: 1.568
      PixelXDimension: 640
      PixelYDimension: 480
      MakerNote: 0, 97, 48, 101, 114, 32, 78, 111, 116, 101, 115, 0, 0, 0, 0, 0
      DateTimeOriginal: 2021:11:16 09:04:04
      DateTimeDigitized: 2021:11:16 09:04:04
   ApplicationSegments: APP1, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2, APP2

I removed a few lines to show important parts, but we get some similar information about the format, a JPEG with EXIF version 2.2. We also learn that HP improperly ordered their tags and put Tag 41492 out of sequence, but we can ignore that for now. Looking close at the output does not give us any indication of audio formats. There is a clue when we see the mention of a Flashpix version and additional Application Segments.

Since this is an image with EXIF data, lets also take a look at the output of Exiftool.

ExifTool Version Number         : 12.62
File Name                       : IM000959.JPG
Directory                       : .
File Size                       : 179 kB
File Modification Date/Time     : 2023:05:25 11:23:32-06:00
File Access Date/Time           : 2023:05:25 11:24:42-06:00
File Inode Change Date/Time     : 2023:05:25 11:24:39-06:00
File Permissions                : -rwxr-xr-x
File Type                       : JPEG
File Type Extension             : jpg
MIME Type                       : image/jpeg
Exif Byte Order                 : Little-endian (Intel, II)
Image Description               : IM000959.JPG
Make                            : Hewlett-Packard
Camera Model Name               : hp PhotoSmart 43x series
Orientation                     : Horizontal (normal)
X Resolution                    : 72
Y Resolution                    : 72
Resolution Unit                 : inches
Software                        : 1.400
Modify Date                     : 2021:11:16 09:04:04
Y Cb Cr Positioning             : Co-sited
Copyright                       : Copyright 2002-2003
Exposure Time                   : 1/29
F Number                        : 4.0
ISO                             : 100
Exif Version                    : 0220
Date/Time Original              : 2021:11:16 09:04:04
Create Date                     : 2021:11:16 09:04:04
Components Configuration        : Y, Cb, Cr, -
Compressed Bits Per Pixel       : 1.567552083
Shutter Speed Value             : 1/30
Aperture Value                  : 4.0
Exposure Compensation           : 0
Max Aperture Value              : 4.0
Subject Distance                : 1 m
Metering Mode                   : Average
Light Source                    : Unknown
Flash                           : Auto, Did not fire
Focal Length                    : 5.7 mm
Warning                         : [minor] Unrecognized MakerNotes
Flashpix Version                : 0100
Color Space                     : sRGB
Exif Image Width                : 640
Exif Image Height               : 480
Interoperability Index          : R98 - DCF basic file (sRGB)
Interoperability Version        : 0100
Digital Zoom Ratio              : 1
Subject Location                : 0
Compression                     : JPEG (old-style)
Thumbnail Offset                : 2046
Thumbnail Length                : 7112
Code Page                       : Unicode UTF-16, little endian
Used Extension Numbers          : 1, 31
Extension Name                  : Audio
Extension Class ID              : 10000100-6FC0-11D0-BD01-00609719A180
Extension Persistence           : Always Valid
Audio Stream                    : (Binary data 117820 bytes, use -b option to extract)
Image Width                     : 640
Image Height                    : 480
Encoding Process                : Baseline DCT, Huffman coding
Bits Per Sample                 : 8
Color Components                : 3
Y Cb Cr Sub Sampling            : YCbCr4:2:2 (2 1)
Aperture                        : 4.0
Image Size                      : 640x480
Megapixels                      : 0.307
Shutter Speed                   : 1/29
Thumbnail Image                 : (Binary data 7112 bytes, use -b option to extract)
Focal Length                    : 5.7 mm
Light Value                     : 8.9

Ohh, what do we have here? Exiftool mentions an audio stream. An audio stream inside the JPEG? How is this possible? The Flashpix format was originally developed by Kodak in which collaborated with HP. This was later added to the EXIF specifications. Below is an screenshot from the Exif Version 2.2 spec.

Exiftool mentioned Flashpix and additional APP2 segments. Lets take a look at the raw file in a hex editor.

Ahhh….. In one of the App2 segments we can see something familiar. A RIFF WAVE header! Lets see if we can extract the WAVE file.

exiftool -b -AudioStream IM000959.JPG > IM000959.WAV

mediainfo IM000959.WAV
General
Complete name                            : IM000959.WAV
Format                                   : Wave
Format settings                          : WaveFormatEx
File size                                : 115 KiB
Duration                                 : 10 s 681 ms
Overall bit rate mode                    : Constant
Overall bit rate                         : 88.2 kb/s

Audio
Format                                   : ADPCM
Codec ID                                 : 11
Codec ID/Hint                            : Intel
Duration                                 : 10 s 681 ms
Bit rate mode                            : Constant
Bit rate                                 : 88.2 kb/s
Channel(s)                               : 1 channel
Sampling rate                            : 22.05 kHz
Bit depth                                : 4 bits
Stream size                              : 115 KiB (100%)

MediaInfo can give us details on the embedded WAVE file, which is pretty terrible quality but is a PCM audio stream.

Embedded audio inside a raster image is not common. Most software which can render a JPEG image will most likely ignore the embedded WAVE and not even give a warning it exists. IM000959.JPG opens fine in Adobe Photoshop, but saving to a new format or making any edits will delete the WAVE file. Imagemagick also will remove the WAVE with any editing with no warning.

In order to ensure the embedded audio stream is preserved we first need to know it is there, this is where tools like exiftool can be used to extract metadata from the file and the image can be associated with having an audio stream and handled differently than any other JPEG file. More work is needed, Exiftool may mention an Audio Stream, but currently does not have the ability to pull any data from the stream.

3M Printscape

There are some file formats out there which are confusing. One such file came across my desk awhile back. This file was not identifiable with any tools I threw at it. At first I believed it to be a TIFF file variant.

You can see the TIFF header, but would not open as one, even if the extension was changed from PSC to TIF. The other hint was the phrase “3M Printscape”, I had never heard of it and there wasn’t much information available about it. It seems it was a creative product made by 3M in the early 2000’s. You could buy a package of printable cards, gift bags, etc. The problem was, there was no available software to be found. I searched on the Internet Archive, the Wayback Machine, and many other abandoned software sites. For months I searched, it wasn’t until a year later I came across one of the creative packages at thrift store. I was thrilled. That is until I was able to get the software installed.

After I installed the software in a virtual machine running Windows 98 I tried to open the PSC file but the software was looking for files with the extension STD, which is an unfortunate acronym. Turns out it stands for SureThing Document. SureThing is a software company who develops Label software. After many months of searching I thought I had found the software to render my file, but it was not meant to be.

Many months later I decided to do some more searches. That is when another copy of 3M Printscape showed up in the Internet Archive. 3M Printscape 2.0! It appears 3M decided to design their own software for version 2.0.

The preservation value of the above image is not lost on me. What took me over a year to figure out ended up being a simple pixelated image of a cardinal. Its the journey, not the destination?

From this little adventure I was able to submit two file formats to PRONOM, fmt/1275,Β fmt/1276. Also I documented the formats and linked to the software on the File Format Wiki. The 3M Printscape version 2 was also released for Macintosh, so the signature had to account for endianness, just like a TIFF file would. With the format having the string “3M Printscape” in the header, it made for an easy signature.

Hopefully, I will be the last to spend this much time on an image of a bird.

Greenstreet

During the 1980’s and 90’s, there was an explosion of software created for the PC and Macintosh. When it came to graphic design, Aldus, Adobe, Quark, Serif, and a few others were clearly the best. That didn’t stop other software developers in trying their hand with publishing design software. If you were on a budget, there were plenty of options to choose from. One of them, Timeworks Publisher, was very popular. It was released in 1987 for IBM PC and Atari with later releases for Apple II and Macintosh. The name was later changed to Pressworks. It was published by an interesting software company out of the UK called GST Software, also under the GSP name. They really enjoyed licensing their software.

Desktop Publishing software

TimeWorks Publisher may have been the first, but was definitely not the last. Pressworks was very popular so the software was sold and rebranded to many companies. In 2001 GST merged with eGames Europe as a new company, Greenstreet Software who continued to support the software. Some of which are:

  • FUJI Publisher
  • Global Software Publishing (in Europe) Pressworks, Power Publisher
  • GST Pressworks
  • 1st Press
  • IMSI TurboPublisher
  • Media Graphics Publishers Paradise Page Express
  • MicroVision Vision Publisher 4
  • NEBS PageMagic
  • PersonalSoft Publications (FranΓ§ais)
  • Pushbutton Publish
  • Softkey Publisher DOS
  • Sybex Page (Deutsch)
  • Timeworks Publisher, Publish-it, Publisher Lite, Publish-it Lite
  • VCI Pro Publisher
  • Wizardworks CompuWorks Publisher
  • Instant Home Publisher
  • Greenstreet Publisher
  • Canon Publishing Suite

All the of the software listed above could open and save to the same file format with the extension .DTP with full compatibility, also used TPL for templates. Originally the DTP file format was a single proprietary binary format which had an ascii header of “DTPI” and all seemed to end with the ascii “EODF”. Later the software was enhanced to be OLE compatible and the binary format was wrapped inside. This made it work well for moving objects in and out of the software into other OLE compatible software like Word, but is confusing to format identification software as the header is the same as a Word file. I have added the two versions of the DTP format to PRONOM to help identify them better. They are fmt/1415 and fmt/1416.

Drawing Software

In addition to the popular Desktop Publishing software, there was a companion Drawing software licensed as well. It also had many titles:

  • BHV COLOURDRAW!
  • FUJI Designer
  • Global Software Publishing (in Europe) Designworks, Power Publisher
  • GST (in North America) PressworksDraw
  • 1st Design
  • IMSI TurboDraw
  • Media Graphics Publishers Paradise Design Studio
  • MicroVision Vision Draw
  • NEBS DesignMagic
  • PersonalSoft CrΓ©ation Graphique
  • Pushbutton Design
  • VCI Pro Design
  • Wizardworks CompuWorks Designer, CompuWorks Draw
  • Canon Publishing Suite

The Draw/Design software all used the same file format as well with the extension .ART, also with full compatibility between all the titles. The TEM extension was used for templates. Not to be confused with the AOL Image format, or Asymetrix Compel Image format, or a number of other formats using the ART extension. This format also began as a single proprietary binary format with the ascii header “GST:ART” starting at offset 16. And just like the DTP format it was later wrapped in an OLE container to be more compatible. In fact, the DTP format may have embedded Art objects! This format is not in PRONOM, so lets take a closer look.

You can see from the 1stdgn.art file here, the ascii “GST:ART” string starting at byte 16. This is consistent with all the samples I have. The first 16 bytes seem to vary in each sample and probably have to do with the size of the file and dimensions of the artwork. GST:ART is unique enough and should work well for a signature.

The ART file from a later version of Draw is in the OLE file format. This container format was designed by Microsoft as a universal container to increase compatibility among software. You can see from the hex view above the file looks very similar to the DOC format used by Word. There were many software titles which used this container format, many documented here. One of the easiest ways to look inside an OLE container is to use 7-Zip. A quick listing of the file shows it is a Type = compound and includes three files. The SummaryInformation file is common among many OLE formats and can contain some metadata, but the Contents file is what we are looking for. Examining the Contents file we find it looks identical to the earlier version of the ART format. The same “GST:ART” string starting at byte 16.

A note about the Preview.dib file. It appears to be a Device-Independent Bitmap, similar to a Bitmap file, probably for a thumbnail preview.

Writing a signature for an OLE container format is a bit more tricky. It requires a separate signature file to go along with the regular signature xml. Basically DROID is setup to “trigger” once it discovers either a “ZIP” file or “OLE” format. If it detects one of those formats it then looks into the container signature xml for additional patterns. If it finds a match then it identifies the format, if not it reports back a generic “ZIP” or “OLE” format.

As it turns out there were two different types of OLE file types, one used “Contents” for the internal file and another which used “CONTENTS”. Since the signature is case sensitive, the container signature requires two signatures both mapped to the same PUID.

These two formats were used with quite a few software titles. Hopefully these signatures cover most of them! You can find a couple samples and my signatures on my Github.

LiveCode stack

One of the earliest hypermedia systems which predated the world wide web was called HyperCard on the Macintosh. Within minutes you could have a small application to do just about anything, calendar, address book, interactive books, games, etc. The internet archive has collected many HyperCard stacks and emulates them directly in the browser.

Riding on the success of HyperCard was another hypermedia tool called MetaCard, which later became Runtime Revolution. Today it is known as LiveCode, a cross-platform application development system. LiveCode is often used to quickly create applications which can run on many platforms including iOS. It is popular with students and higher education. The LiveCode source was opened for a time instigated by a successful kickstarter program, but closed in 2021 as the company struggled to keep paying customers.

Each LiveCode version produced unique files for each of the major versions. Currently none of the formats can be identified using preservation tools. Luckily, because the code was open-source for a time, we have details which helps us identify the formats. Let’s take a look:

#define kMCStackFileMetaCardVersionString "#!/bin/sh\n# MetaCard 2.4 stack\n
#define kMCStackFileVersionString_2_7 "REVO2700"
#define kMCStackFileVersionString_5_5 "REVO5500"
#define kMCStackFileVersionString_7_0 "REVO7000"
#define kMCStackFileVersionString_8_0 "REVO8000"
#define kMCStackFileVersionString_8_1 "REVO8100"

I took LiveCode up on their 10 day trial and was able to install software version 9.6.9 to save some samples. The software has a “Save as” option which allows you to save your code to older versions. Although one must be careful as saving to older versions may have some data loss.

The samples I was able to save had matching headers just like in the source code. The REVO string starts right at the beginning of the file making identification easy. Take a look at my GitHub page for samples and signature. Also check out the File Format Wiki Page for more information and more samples!

Open Media Framework

Awhile back I was asked to look at a file in our repository which had the extension OMF. It was not identified by DROID and didn’t appear to be in PRONOM. It didn’t take long to find quite a bit of information on the file format as it was used by many important software titles, or at least it used to. Exploring the details of this file format led me on quite the rabbit hole. You see, the OMFI format is based on a container format that once was heralded as the a better open choice over the Microsoft OLE container format growing in popularity.

OpenDoc

This all started with a multi-platform approach to an open document format started by Apple Computers in the early 1990’s called OpenDoc. It was originally an alliance between Apple, IBM, and Motorola. The idea was to have a framework any developer could use to develop software or components that would all work seamlessly together. Many developers were on board initially with many promised software titles being developed, but ultimately with much confusion surrounding the framework and Steve Jobs return to Apple in 1997, the project was scrapped.

Bento

The storage format to be used with OpenDoc was called Bento, in reference to the Japanese style of a compartmentalized container tray. Specifications were released in 1993.

There are four key ideas in the Bento format:

  • everything in the container is an object,
  • objects have persistent IDs,
  • all the metadata lives in the TOC (Table of Contents),
  • objects consist entirely of values, and
  • each value knows its own property, type, and data location.

The idea of a data model with such an organized structure was so appealing the digital preservation community there was excited to push for a Universal Preservation Format specifically for multimedia based on Bento. The idea was presented to AMIA in 1996!

Open Media Framework (OMF) Interchange

Avid Technology, a leader in audio/video editing systems, used the Bento specification to design a container format for multimedia. This allowed easy interchange of projects between many different software titles. Original specifications were published in 1994, while the 2.1 specifications released in 1997. Software titles such as Pro Tools, Cubase, Adobe Audition, Adobe Premiere, Apple Logic Pro, Apple Final Cut Pro, and many others supported the OMF format, at least for awhile. OMFI was migrated to Microsoft’s Structured Storage container format to form the core of (AAF) Advanced Authoring Format in the late 1990’s.

Identification

In order to identify an OMF file we first need to understand what is part of the OMF specifications and what is part of the Bento format. OpenDoc may not have lived very long but the Bento format held on long enough to be the structure used by a few different file formats. I am aware of the following, but there was other software being developed at the time.

Samples from each of these formats show some similar patterns. In the Bento specifications we can see:

The only version of the specifications I can find are version 1.0d5 released in 1993, but we know there was also a version 2 released later. The magic bytes are not defined in the 1.0d5 spec, but looking at the code in the Open Doc Developer Release in 1996, we can find reference to the magic bytes used in “Containr.h”.

#define MagicByteSequence "\xA4""CM""\xA5""Hdr""\xD7"

The Bento specification also defines this header information as, “Our solution to this is to define the standard Bento format to have the label at the end of the container.” Which means this byte sequence will frequently be found at the End of File. The “CM” refers to “Container Manager” and “Hdr” refers to “Header”.

Now that we have the magic bytes for the Bento container we can look at what makes the OMF file unique from others. We can find the answer in the Bento specifications.

We know that every Bento container must have a object, so in version 1.0 of the specifications on page 65 we find.

Each object must have the property OMFI:ObjID. The value of OMFI:ObjID is required and is listed in the property description for each object.

The OMFI:ObjID can also be found in version 2.0 of the specification, but in addition it defines:

The OMFI:ObjID property has been renamed the OMFI:OOBJ:ObjClass property, which eliminates the concept of generic properties and makes the class model easier to understand. The name ObjClass is more descriptive because the property identifies the class of the object rather than containing an
ID number for the object.

Since both are required it seems appropriate to use those strings for identification in a PRONOM signature. You can check out the proposed signature and samples on my GitHub page.

There is so much history wrapped up in these formats and the potential they had to change how we preserve files in our archives. Luckily we have the Internet Archive WayBack machine to help us discover or remember ideas that once existed, some which may find their way back to inspire future file formats.

Sony Voice Recorder

Sony’s IC Recorders have been a popular small digital voice recorder for many consumers. The current models all use common recording formats like Linear PCM WAVE files or MP3, but it wasn’t always so. One of the first models ICD-R100 would record to the ICS audio format, which was Sony’s original sound formats used on the IC Recorders. I am still looking for samples of this format. If you do have a need to convert this format, Sony has free converter software.

The next generation of IC Recorders used a Memory Stick and therefore recorded audio to the MSV (Memory Stick Voice) format. There were actually two different types of MSV files, the first used the ADPCM codec and the next used the LPEC codec. Later IC Recorders would record to the DVF (Digital Voice Format) which also had a couple versions, one using the LPEC codec and the other the older TRC codec.

AFAIK, none of the codecs used in these file formats has been made public and these formats are not readable by tools such as MediaInfo. The only way to know details of a file and have the ability to play or convert is to use Sony software which has been discontinued and the replacement, Sound Organizer, can only recognize the LPEC codec versions of MSV and DVF. There is also a plugin for Windows Media Player available here, which is required even for Switch to work.

PRONOM currently has one signature for the LPEC versions of MSV and DVF, so lets look closer at the formats and see if we can determine what they are from the header.

The CODECs

ADPCM is an abbreviation for “Adaptive differential pulse-code modulation“. Appears to only have been used with the ICD-MS1 and possibly MS2 digital recorders.

TRC may be an abbreviation of Truespeech’s “Triple Rate CODER” or “Triple Rate Codec“, but not much info exists.

LPEC is a proprietary compression format. It is an abbreviation of “Long-term Predicated Excitation Coding“. It even had its own trademarked logo which was cancelled in 2015.

The Software

The first IC Recorders came with PCLINK software, then came with the “MemoryStick Voice Editor” software. List of compatible formats.

Digital Voice Editor came next. It could read and convert everything except “ICS” files. Click here to download the last version. Version 1 compatible formats. Version 2 compatible formats. Version 3 Compatible Formats. The software was officially retired in 2016.

The current software for managing audio files from IC Recorder is Sound Organizer. The software does open and convert some MSV/DVF files as long as they use the LPEC codec. Sound Organizer Compatible formats.

Also note, Sony made one ICD-CX series recorder which could also capture photos. It requires the Visual & Voice Player software. Audio is recorded in the DVF format.

Test Data Set

In order to explore the different formats I first needed to gather some samples. There are a few out there, but with the Digital Voice Editor 3 software, I was able to take a sample file and convert it to the many options available. You can see in the screenshot below, the different samples, their extension and the codec used. You can find my samples in GitHub here.

All MSV and DVF file have a similar pattern. The first 32 bytes have the text string “MS_VOICE SONY CORPORATION”. In between MS_VOICE and SONY, there is 4 bytes which vary slightly between the different formats. Here is a table of samples and the 4 bytes so we can see the differences.

ModelCODECEXTENSIONHex Values
ICD-Px0TRCDVF01020000
ICD-Px8TRCDVF01020000
ICD-Px7TRCDVF01020000
ICD-SXxx0LPECMSV01030000
ICD-SXx8LPECMSV01030000
ICD-SXx7LPECMSV01030000
ICD-SXx6LPECDVF01020000
ICD-SXx5LPECDVF01020000
ICD-SXx0LPECDVF01020000
ICD-MXLPECMSV01020000
ICD-BMLPECMSV01020000
ICD-STLPECDVF01020000
ICD-MS5xxLPECMSV01010000
ICD-SLPECMSV01010000
ICD-BPx50LPECDVF01010000
ICD-BP100/x20LPECDVF01010000
ICD-MS1/MS2ADPCMMSV01000000
ICD-R100/R200UnknownICS

There is an obvious pattern to the hex values as they increment 0100, 0101, 0102, and 0103. But there is some overlap between extension and codec, so probably more of a version number than specific to the codec. Currently the PRONOM signature for this format fmt/472, has the pattern for the 0102 version, but none of the others. We could simply add a variable in the signature for the different values and update the PRONOM signature so more samples would be identified. This would work well if there was a secondary characterization process to get technical metadata such as the codec and quality, but I am unaware of any tool to gather this information from the format, so I wonder if we can find any hints in the file to identify the codec so we have multiple PRONOM signatures to choose from. Also, you can see from the screenshot above that some of the LPEC formats have specific model numbers in the codec column, which could mean they may not be exactly the same. Each IC Recorder model has different quality settings and it appears, some settings may not be compatible with other models.

Looking beyond the first 16 bytes there is a lot of hex values which are unknown. A close comparison of all the samples leads me to the 4 bytes at offset 60. They seem to be the same for files with the same settings. Below is a chart of those values.

ExtensionCODECQualityOffset 60
DVFTRCHQ00300001
DVFTRCSP00350001
DVFTRCLP00370001
DVFLPEC (ICD-BP-100/x20)SP00150001
DVFLPEC (ICD-BP-100/x20)LP00190001
DVFLPECSP002A0001
DVFLPECLP002C0001
MSVLPEC (ICD-BM/MX/SXx7/SXx8/SXxx0)SP004A0001
MSVLPEC (ICD-BM/MX/SXx7/SXx8/SXxx0)LP004C0001
MSV/DVFLPEC (ICD-SXx7/SXx8/SXxx0)STHQ00200002
MSV/DVFLPEC (ICD-SXx7/SXx8/SXxx0)ST00240002
MSVADPCMSP00050001
MSVADPCMLP00090001

Just to be sure this value at offset 60 was indeed an indication of codec and quality I manually switch out the 4 bytes from a LPEC ST file for a TRC HQ file. Sure enough, the software now saw the file as a TRC HQ audio file, even though the original is a Stereo file.

There is a very good chance this is not all the options. I only have one physical recorder which only records in Mono. But this gives us a really good idea of how to tell the difference between files. Below are the patterns I am submitting to PRONOM.

MSV ADPCM

4D535F564F494345{4}01000000534F4E5920434F52504F524154494F4E{28}00(05|09)0001

DVF TRC

4D535F564F494345{4}01020000534F4E5920434F52504F524154494F4E{28}00(30|35|37)0001

MSV/DVF LPEC

4D535F564F494345{4}01(01|02|03)0000534F4E5920434F52504F524154494F4E{28}00(15|19|20|24|2A|2C|4A|4C)00(01|02)

Perhaps we can alter the existing PRONOM signature for fmt/472 to catch all we may miss to:

4D535F564F494345{8}534F4E5920434F52504F524154494F4E6D73766C637374772E73706900000000

This is one example of a file format which has a proprietary component which was never released from the vendor. When the vendor stopped supporting the software to open and read these formats, the risk increased for long-term preservation. It would be really nice when a vendor discontinues a technology, which was used by consumers, they would make the documentation for the format openly available. If you know more about the format, please reach out or if you have samples which don’t match the patterns mentioned here.

Early Microsoft Excel

The first version of Microsoft Excel was released on Macintosh in 1985. Before that there was MultiPlan.

MultiPlan version 4 and Excel version 2 used the well known and documented BIFF format. Before BIFF2 the formats are a bit of a mystery. AFAIK, Microsoft never released any documentation on the file format used for Excel version 1 and MultiPlan 1 -3, they emphasized using the SYLK format for interchange. To make matters worse, there were upwards of 100 different versions of the early MultiPlan, ported for dozens of different systems. Some of them are discussed on the TRS-80 website.

Or you can take MultiPlan 1.06 for a spin over at PCjs!

Needless to say documenting and finding a pattern which could be used to identify the early versions of MultiPlan and Excel 1 are difficult. These versions are missing from the PRONOM registry, but hopefully with enough samples, some patterns can be found to confidently identify formats from the early days of spreadsheets!

Marco Pontello’s TrID identifier software has signatures for the early Multiplan and Excel formats. His software scans for patterns in samples and finds commonalities between them. So the more samples he can scan the more accurate the identification can be.

Currently the signatures are as follows.

Microsoft Excel for Mac Spreadsheet (v1.x)
		<Pattern>
			<Bytes>532700</Bytes>
			<ASCII> S '</ASCII>
			<Pos>0</Pos>
		</Pattern>
		<Pattern>
			<Bytes>AB27000000000000000203</Bytes>
			<ASCII> . '</ASCII>
			<Pos>4</Pos>
		</Pattern>
Multiplan for Mac spreadsheet (v1.x)
		<Pattern>
			<Bytes>11AB000013E8000000000000</Bytes>
			<ASCII> . . . . . . . . . . . .</ASCII>
			<Pos>0</Pos>
		</Pattern>
Multiplan spreadsheet (v1.x)
		<Pattern>
			<Bytes>0CE9000008AB08001F0016000200</Bytes>
			<Pos>0</Pos>
		</Pattern>
Multiplan spreadsheet (v1.0x)
		<Pattern>
			<Bytes>08E700</Bytes>
			<Pos>0</Pos>
		</Pattern>
		<Pattern>
			<Bytes>0100</Bytes>
			<Pos>6</Pos>
		</Pattern>
		<Pattern>
			<Bytes>000000</Bytes>
			<Pos>11</Pos>
		</Pattern>
Multiplan spreadsheet (v2.x)
		<Pattern>
			<Bytes>0CEC000008AB08001F001A000300</Bytes>
			<Pos>0</Pos>
		</Pattern>
Multiplan for Xenix spreadsheet (v2.x)
		<Pattern>
			<Bytes>0AEC000008AB08001F001A000300</Bytes>
			<Pos>0</Pos>
		</Pattern>
Multiplan spreadsheet (v3.x)
		<Pattern>
			<Bytes>0CED000008AB08001F001A000000</Bytes>
			<Pos>0</Pos>
		</Pattern>

There seems to be some patterns between versions, but then also some major differences. Without a specification or an understanding of the system the samples were created on, it is hard to identify these formats with certainty. There could be hex values which are the same for the samples we have but different for others, headers can often have values indicating dates or length of the file, so finding variations in files is key to a good signature.

Keep an eye on my GitHub PRONOM Research folder as I add more samples and prepare a signature for PRONOM.

Adobe Illustrator and PDF

Adobe Illustrator is a power design tool. Originally released in 1987 for the Macintosh, it has been the vector design tool of choice for many professionals.

Originally the Adobe Illustrator Format (AI) was based on postscript. With each file having a postscript header. This all changed with Illustrator version 9 moving to PDF as its core, released in the year 2000.

Illustrator 8 vs Illustrator 9 header.

Even though AI files begin with a PDF header, there is much more to them which makes them a unique file format. So as Dov Isaacs put it, “PDF files are not Adobe Illustrator files and vice versa”.

Working in digital preservation the need to identify a file format is vital to the process. It is also important to identify when the format changes over time in order to properly maintain that file. Adobe Illustrator files created in version 8 or earlier are substantially different than those created in version 9 and greater and will need different software to render properly.

This is where identification tools come in handy.

Using the “File” command in a CLI we get:

Illustrator9v8-s04.ai: PostScript document text conforming DSC level 3.0

Illustrator9-s04.ai: PDF document, version 1.4, 1 pages

While partly true, we need more specific identification if we want to properly preserve these file in the long term. Enter PRONOM, which is a file identification registry based on signatures to identify file formats. Using a tool like DROID, Siegfried with the PRONOM registry we can get better identification.

siegfried   : 1.10.0
scandate    : 2023-04-07T11:59:18-06:00
signature   : default.sig
created     : 2023-03-23T15:09:43Z
identifiers : 
  - name    : 'pronom'
    details : 'DROID_SignatureFile_V111.xml; container-signature-20230307.xml'
---
filename : 'Illustrator9-s04.ai'
filesize : 77829
modified : 2023-04-07T11:04:53-06:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'fmt/558'
    format  : 'Adobe Illustrator'
    version : '9.0'
    mime    : 'application/postscript'
    class   : 'Image (Vector)'
    basis   : 'extension match ai; byte match at [[0 8] [1536 557]]'
    warning : 
---
filename : 'Illustrator9v8-s04.ai'
filesize : 323748
modified : 2023-04-07T11:05:11-06:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'fmt/557'
    format  : 'Adobe Illustrator'
    version : '8.0'
    mime    : 'application/postscript'
    class   : 'Image (Vector)'
    basis   : 'extension match ai; byte match at 0, 673'
    warning : 

This identification is possible because of signatures built for the file format specific to each version. The file format wiki has a list of the current signatures for the Illustrator format. The problem is, the last signature added to PRONOM was for version 16 (CS6). Since then there have been more changes to the format.

If we attempt an identification of a Illustrator file created with current 2023 software we get this result.

filename : 'Illustrator2023-s01.ai'
filesize : 1195445
modified : 2023-02-16T12:29:16-07:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'fmt/20'
    format  : 'Acrobat PDF 1.6 - Portable Document Format'
    version : '1.6'
    mime    : 'application/pdf'
    class   : 'Page Description'
    basis   : 'byte match at [[0 8] [1195439 5]]'
    warning : 'extension mismatch'

Although technically correct, as the Illustrator file has a PDF 1.6 header, identification needs to know this is an Illustrator file. So if we create a new signature by adding the following hexadecimal pattern:

255044462D312E36*3C696C6C7573747261746F723A547970653E446F63756D656E743C2F696C6C7573747261746F723A547970653E*252150532D41646F62652D332E30*254149355F46696C65466F726D6174203134

filename : 'Illustrator2023-s01.ai'
filesize : 1195445
modified : 2023-02-16T12:29:16-07:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'BYUDev/3'
    format  : 'Adobe Illustrator CC 2020'
    version : '24.2+'
    mime    : 'application/postscript'
    class   : 
    basis   : 'extension match ai; byte match at [[0 8] [8766 45] [45347 348]]'
    warning : 

Lets break down the hexadecimal pattern. The “*” is a wildcard indicating there is 0 to many bytes in between.

255044462D312E36 translates to: %PDF-1.6

3C696C6C7573747261746F723A547970653E446F63756D656E743C2F696C6C7573747261746F723A547970653E translates to: <illustrator:Type>Document</illustrator:Type>

252150532D41646F62652D332E30 translates to: %!PS-Adobe-3.0

254149355F46696C65466F726D6174203134 translates to: %AI5_FileFormat 14

Identification is based first on the PDF Header, then some XMP metadata indicating this is an Illustrator document, then the Postscript header, then finally the version identifier. Each Illustrator since version 5 has a file format version, when Adobe switched from the CS labels to CC, they stuck with version 13 until 2020, when the format was changed to version 14. With one catch, when Illustrator version 24 (2020) was first released it was format version 14, but still had the PDF 1.5 header. This was changed in version 24.2 to a PDF 1.6 header which added a bigger Canvas size.

In the current PRONOM signatures going back to version 9, there was some offsets assumed for the space between the PDF header, postscript header, and version number. I have found through many samples there are quite a few which are outside those offsets, especially as the size of the AI file gets larger. Therefore I am suggesting the “*” wildcard between all segments.

One area that still needs a bit more research is with Illustrator versions 9-12 (CS2). These do not include the XMP metadata indicating they are Illustrator Documents, so they will more often get misidentified as PDF. I did find, however, AI files have a string “/AIPrivateData”, while saved as PDF, they have “/AIPDFPrivateData”. So signature will have this added to distinguish.

Another anomaly is some samples I found on the Illustrator 9 CD-ROM. Illustrator 9 was released in June of 2000, but many of these files were created in February of 2000, they have a PDF 1.4 header but have a format version 4, which is what version 8 uses. So these files were probably created with an early build of Illustrator 9 and format was incremented to 5 in the public release.

You can see my submission suggestion on my GitHub page along with the PRONOM Signature and sample files. There is still a couple tweaks I need to make, but let me know what you think.

Note: All Illustrator files and PDF’s saved with Illustrator compatibility checked which include a section of the file called “AI Private Data”, this is where all the Illustrator data lives. It includes a “creator” version and a “container” version which could also be used to identify an Illustrator files version.