LUTS

Posted on by Thor 1 Comment

If you are looking for LUTs, you’re in luck. There is a website for sharing your FreshLUTs. Even though they are fresh, they are probably not as exciting as one might think.

LUTs are short for Look-Up Tables, which doesn’t sound as exciting as you were probably hoping. They are a pretty interesting process for dealing with color in high end Image and Video processing applications. Often called 3D Look-up Tables, they are used for color grading, an essential step in film production and restoration to map from one color space to another. LUTs are not to be confused with ICC profiles which aim for color accuracy, while LUTs are looking for more color quality and aesthetics.

There are a lot of LUT formats out there, it seems. In looking into this format, I have found dozens of others to investigate, but today lets look at the four available as an export from Photoshop.

Above you can see a simple screenshot for the export of different formats from Adobe Photoshop. Adobe is one of the biggest developer and supporter of the formats used in LUTs, but there are many other graphics tools which create and support LUTs. In this Photoshop export we can see four formats included in the export. Lets take a look at each of these.

ICC Profiles are well documented and available for identification in PRONOM.

filename : 'LUTs-Export-s01.icc'
filesize : 197024
modified : 2025-02-25T09:37:24-07:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'fmt/1975'
    format  : 'ICC Profile'
    version : '2'
    mime    : 'application/vnd.iccprofile'
    class   : 'Dataset'
    basis   : 'extension match icc; byte match at 8, 32'

But the other three are plain text files and still identify as such. Let us start with the CUBE format.

filename : 'LUTs-Export-s01.cube'
filesize : 884963
modified : 2025-02-25T09:37:24-07:00
errors   : 
matches  :
  - ns      : 'pronom'
    id      : 'x-fmt/111'
    format  : 'Plain Text File'
    version : 
    mime    : 'text/plain'
    class   : 
    basis   : 'text match ASCII'
    warning : 'match on text only; extension mismatch'

cat LUTs-Export-s01.cube 
#Created by: Adobe Photoshop Export Color Lookup Plugin
#Copyright: (C) Copyright 2025 ObsoleteThor
TITLE "LUT-export-s01"

#LUT size
LUT_3D_SIZE 32

#data domain
DOMAIN_MIN 0.0 0.0 0.0
DOMAIN_MAX 1.0 1.0 1.0

#LUT data points
0.000000 0.000000 0.000000

The CUBE format was first developed by IRIDAS in 2003 as a answer to ensure interoperability with other software. Adobe acquired IRIDAS in 2011 in a effort to be a leader in the color grading and enhancement market. They have published the CUBE specifications for version 1.0 in 2013.

A Cube file is a text file that defines a look-up table in the Cube format.
The Cube look-up tables store RGB values.
Advantages of the Cube format include:
  • The Cube format can describe look-up tables for a wide range of purposes, from simple gamma adjustments for display output to complex HDR image processing.
  • The format is well suited for professional digital cinema applications and for both normal range and High-Dynamic Range image processing.
  • As Cube files are text files, they are easily edited or reviewed using a text editor.
  • A Cube file can include three 1-dimensional tables or one 3-dimensional table.
  • The tables can be in a wide range of sizes.
  • Cube files are trivial to write and read.
  • All values are human-readable as they are in decimal form, and can be of high precision.
  • The input domain and output range are not limited to the range 0.0 to 1.0.

According to the specifications, a CUBE file can be a One-Dimensional Cube file or a Three-Dimensional Cube file. From the example above you can see the file is a Three-Dimensional file with the required line “LUT_3D_SIZE“. But in a One-Dimensional file, the required line is “LUT_1D_SIZE“.

cat Demo.cube
TITLE "Demo"
LUT_1D_SIZE 3
DOMAIN_MIN 0 0 0
DOMAIN_MAX 1 2 3
0 0 0
# Comments can go anywhere
0.5 1 1.5
1 1 1

Each CUBE file has one or the other and should be an easy string to look for. It is in a variable position as there can be comments before the required line and also may have a TITLE line. The TITLE and DOMAIN lines are common to every file but not required.

Now, the CUBE format is a bit different depending on the source. They all seem to have the same header, but different elements. It seems the IRIDAS Cube format is the most interoperable. The Truelight Cube format generally has the CUB extension, and the Cinespace Cube has the CSP extension, which will look at next/ You can read more about the differences on this format comparison table. This LUTCalc web site has many different types of Cube’s it can output, so there are some differences.

The other file format available in the export is a CSP. The CSP is also a plain text file, often called a cineSpace LUT file. This format come from the cineSpace software, a color management software for the film and television industry. 

cat LUTS-s01.csp 
CSPLUTV100
3D

BEGIN METADATA
#Created by: Adobe Photoshop Export Color Lookup Plugin
TITLE "LUTS"
END METADATA

2
0.0 1.0
0.0 1.0
2
0.0 1.0
0.0 1.0
2
0.0 1.0
0.0 1.0

32 32 32
0.000000 0.000000 0.000000

The CSP File Format specifications outlines header and the other two sections.:

The cineSpace LUT format contains three main sections.
Header
This section contains the LUT identifier and the LUT type, 3D or 1D.
It is made up of the first two (2) valid lines in the file. See Notes below for the definition of a valid line.

Examples
• (3D LUT) header:
CSPLUTV100
3D
• (1D LUT) header:
CSPLUTV100
1D

So there is a pretty obvious header to work with in identification. “CSPLUTV100” can be used to identify both 1D and 3D CSP files.

The other format available to export from Photoshop is 3DL. They seem to be connected to the Assimilate Inc. company and software. A specification has been posted, and it looks like there is only ASCII and not much in the way of a header.

cat LUTS-s01.3dl 
#Created by: Adobe Photoshop Export Color Lookup Plugin
#Description: LUTS
0 33 66 99 132 165 198 231 264 297 330 363 396 429 462 495 528 561 594 627 660 693 726 759 792 825 858 891 924 957 990 1023

It does not appear there is any headers or static strings to use for identification. The specification calls the format, 3DL ASCII format and that “All lines starting with ‘#’ are treated as comments.” Because of this, I don’t think positive identification can happen at this time.

For now I am just proposing 2 new file formats to PRONOM, The CUBE format And the CSP Format. Click on my GitHub submission page to see the signatures and enjoy some samples!

CD Architect

Posted on by Thor Leave a comment

Receiving electronic media from an outside source can be an adventure. Often times you find yourself sorting the valuable files and separating them from the chaff. There can be hidden files, cache files, application files, drivers, and everything in between. Determining what formats are important can sometimes be difficult, especially if you don’t know the file format of some of the files.

I was recently working on a collection of files which had been produced through some audio software. When working with audio, a WAVE file is what is usually kept as they contain the actual audio data. With these files they came with a couple other formats. One of those formats was a bunch of SFK peak files. These files are meant to be temporary as they are generated from the WAVE file to make opening of audio data faster. They are important, but can easily be regenerated. One could argue they have historical value, but also they don’t contain anything that can be used by itself, so alone they don’t have much value.

The other format found with the WAVE files have a CDP extension. These came up as unknown when using DROID. It is not a common extension so finding the name of the software which created the files wasn’t too hard. Let’s take a look at one of them.

hexdump -C tutor1.cdp | head
00000000  52 49 46 46 79 03 00 00  53 46 50 4a 66 6d 74 20  |RIFFy...SFPJfmt |
00000010  18 00 00 00 00 00 01 00  02 00 00 00 10 00 00 00  |................|
00000020  44 ac 00 00 03 00 00 00  01 00 00 00 4c 49 53 54  |D...........LIST|
00000030  88 00 00 00 66 6c 73 74  66 69 6c 65 23 00 00 00  |....flstfile#...|
00000040  44 3a 5c 53 6f 75 6e 64  73 5c 4e 65 77 20 54 75  |D:\Sounds\New Tu|
00000050  74 6f 72 20 66 69 6c 65  73 5c 53 6f 6e 67 33 2e  |tor files\Song3.|
00000060  77 61 76 00 66 69 6c 65  23 00 00 00 44 3a 5c 53  |wav.file#...D:\S|
00000070  6f 75 6e 64 73 5c 4e 65  77 20 54 75 74 6f 72 20  |ounds\New Tutor |
00000080  66 69 6c 65 73 5c 53 6f  6e 67 32 2e 77 61 76 00  |files\Song2.wav.|
00000090  66 69 6c 65 23 00 00 00  44 3a 5c 53 6f 75 6e 64  |file#...D:\Sound|

Huh, this is a RIFF file. RIFF is most commonly used as the container used for WAVE and AVI files. You can read more about the RIFF format on a previous post. The RIFF container format can be used for all sorts of things. Looking at the internals we can see a few unique list chunk’s.

Lots of references to other files, specifically WAVE files. But not a lot of actual data. That is because this format turns out to be just a project format for some software called “CD Architect“. Sonic Foundry was an audio software developer for a few years before they sold their catalog to Sony in 2003. In looking at the manual for CD Architect version 5.2, it explains the CDP Project format.

CD Architect software handles the organization of your CD using a small project file (CDP) that saves information about source file locations, edits, cuts, and insertion points. This project file is not a multimedia file, but is instead used to create the CD when editing is finished.

Looking at another CDP file from the collection, I noticed something different.

hexdump -C CDArch50a-s01.cdp | head
00000000  72 69 66 66 2e 91 cf 11  a5 d6 28 db 04 c1 00 00  |riff......(.....|
00000010  20 0a 00 00 00 00 00 00  84 38 15 b3 da 08 85 44  | ........8.....D|
00000020  b2 2a 5b 70 a1 32 15 ff  5a 2d 8f b2 0f 23 d2 11  |.*[p.2..Z-...#..|
00000030  86 af 00 c0 4f 8e db 8a  00 02 00 00 00 00 00 00  |....O...........|
00000040  78 00 00 00 00 00 04 00  11 00 00 00 44 ac 00 00  |x...........D...|
00000050  00 00 00 00 00 c0 52 40  00 00 00 00 00 00 5e 40  |......R@......^@|
00000060  00 00 00 00 00 00 00 00  04 00 04 00 40 00 00 00  |............@...|
00000070  00 00 00 00 00 00 00 00  00 00 00 00 7c 00 00 00  |............|...|
00000080  50 00 00 00 a0 00 00 00  00 00 00 00 00 00 00 00  |P...............|
00000090  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|

That’s odd, the RIFF format is always uppercase ASCII, this is lowercase. Also the important RIFF form, which was “SFPJ” in the other sample, is missing. This is not a valid RIFF format.

But further down in the file I can see the same list chunks. Did they take RIFF format and make a proprietary version of their own? I think they may have. It seems the first example was from CD Architect version 4 and these other files are from CD Architect version 5. That complicates things. Sony stopped developing CD Architect after version 5.2d and maintained it for a few years before selling many of their titles to MAGIX Software. As far as I know there was never any new versions released. The software was very popular, as it had some really nice audio mastering features and was easy to use. Many were upset when the software was abandoned.

Creating a signature for both version 4 and version 5 CDP files will be pretty straightforward. I feel knowing what you have in a collection you are processing is the first step in making informed decisions. Wether or not you keep the project files are up for debate. Some may only want the final audio created from a CD Architect project, while others may want to see the way the audio was put together and mixed. Either way, the more you know…..

One more thing. CD Architect would default to saving a CDP project file, but could also save a “CD Image file”. This process actually would save the project to a full WAVE file with some extras baked in.

An image file is essentially a wave file with volume, crossfades, effects, mixes, and track information embedded. Burning an image file will reduce the risk of buffer underruns (especially if you have a complex project or are using a slow computer) since no audio processing is required. 

Interesting, normally when working with track information in a single WAVE file you would need a companion CUE Sheet in order to reference the track layout of the Audio CD. So I am curious how they do all of this. Lets take a look at a “CD Image”.

mediainfo CDArch52d-s02.wav
General
Complete name                            : CDArch52d-s02.wav
Format                                   : Wave
Format settings                          : PcmWaveformat
File size                                : 5.05 MiB
Duration                                 : 30 s 0 ms
Overall bit rate mode                    : Constant
Overall bit rate                         : 1 411 kb/s
Conformance errors                       : 2
 RIFF                                    : Yes
  General compliance                     : File size 5292434 is less than expected size 5292823 (offset 0x8)
 WAVE                                    : Yes
  General compliance                     : Element size 5292811 is more than maximal permitted size 5292422 (offset 0xC)

Audio
Format                                   : PCM
Format settings                          : Little / Signed
Codec ID                                 : 1
Duration                                 : 30 s 0 ms
Bit rate mode                            : Constant
Bit rate                                 : 1 411.2 kb/s
Channel(s)                               : 2 channels
Sampling rate                            : 44.1 kHz
Bit depth                                : 16 bits
Stream size                              : 5.05 MiB (100%)

Already seeing some issues with the format, but all the important bits are there. JHOVE doesn’t like them much either.

JhoveView (Rel. 1.32.0, 2024-09-12)
 Date: 2024-12-11 16:01:08 MST
 RepresentationInformation: CDArch52d-s02.wav
  ReportingModule: WAVE-hul, Rel. 1.8.3 (2024-03-05)
  LastModified: 2024-12-11 15:58:02 MST
  Size: 5292434
  Format: WAVE
  Status: Not well-formed
  SignatureMatches:
   WAVE-hul
  InfoMessage: Ignored unrecognized list type: "pqls"
   ID: WAVE-HUL-15
   Offset: 5292044
  ErrorMessage: Unexpected end of file: Bytes missing = 389
   ID: WAVE-HUL-3
   Offset: 5292434
  MIMEtype: audio/vnd.wave; codec=1
  Profile: PCMWAVEFORMAT

JHOVE is giving me two issues. The major error is the file appears truncated according to both MediaInfo and JHOVE. The InfoMessage which is less of an issue but more of a heads up that the WAVE file has an extra LIST type. “PQLS”, which was also in the CPD RIFF file we looked at earlier. So it seems by making a “CD Image” of a project embeds the project chunk data into the WAVE container. Identification is not an issue as these WAVE’s follow the standard pattern and therefore identify correctly, but one might want to be aware through further characterization these WAVE’s have some not so obvious extra data.

My attempts to find any samples from version 3 of CD Architect have failed. Until then, my proposal is to add version 4 & 5 to PRONOM with the signature on my Github page. There you will find a few samples as well.

Daisy

Posted on by Thor Leave a comment

A single file can often be self contained, having all that is needed to render itself with the correct software, but more and more often files need other files to function properly. Sometimes these groups of dependent files are within a container, such as a DOCX or ePub, but can also be found all sitting nicely in a folder. I say nicely, partly because the structure works, that is until they are treated as individual files and renamed or moved around breaking that interdependence on each other.

In the case of many Apple bundle files, they appear to be a single file when using on the MacOS, but as a folder on Windows or Linux. This can be very confusing. In other cases such as the DAISY Digital Talking Book format, it is simply a folder or disc with a few or many files within.

Current tools used to identify file formats, such as DROID, look at individual files, not groups of files to determine format. Each file within a folder may have a unique format, but when grouped with other specific formats they become something more. We will have to work on enhancing current tools if we want to avoid breaking these format types and losing their ability to render properly.

DAISY, or Digital Accessible Information System, is a type of Digital Book. The format was originally conceived in 1988 as a method to create a talking book, designed for the purpose of giving those who are visually impaired the ability to listen to books. It wasn’t until 1996, the DAISY Consortium was created in order to take the technology to those who needed it. The original version of the the DAISY format in 1994 was proprietary, but once they formed the consortium, they decided to adopt open standards for the format and in 1998, the DAISY 2.0 standard was released. You can read more on the Library of Congress Format Description page.

Lets take a look at a folder containing a DAISY 2.0 book.

ls -la "DAISY 2.02 export"
total 536
drwx------  1 tyler  staff   16384 Sep 25 22:06 .
drwx------  1 tyler  staff   16384 Sep 25 22:06 ..
-rwx------@ 1 tyler  staff    1090 Sep 25 22:05 0002.smil
-rwx------  1 tyler  staff  228413 Sep 25 22:05 aud0001.mp3
-rwx------@ 1 tyler  staff     672 Sep 25 22:05 master.smil
-rwx------  1 tyler  staff    1703 Sep 25 22:05 ncc.html

We can see three different formats in this folder. The obvious well known MP3 files and an HTML file. We also see two files with the extension SMIL.

“Synchronized Multimedia Integration Language” or SMIL is a W3C XML standard used to describe multimedia presentations. It is used in the DAISY DTB as well as other applications, but we will focus on DAISY, and it is in its third version. A SMIL file has this structure:

<?xml version="1.0"?>
<!DOCTYPE smil PUBLIC "-//W3C//DTD SMIL 1.0//EN" "http://www.w3.org/TR/REC-smil/SMIL10.dtd">
<smil>
  <head>
    <meta name="dc:title" content="Obi Project" />
    <meta name="dc:identifier" content="589c550e-303b-4c0d-9921-ae76d782fd53" />
    <meta name="ncc:generator" content="Obi v5.0.0.0 with toolkit: UrakawaSDK.core v2.0.0.0 (http://urakawa.sf.net/obi)" />
    <meta name="dc:format" content="Daisy 2.02" />
    <meta name="ncc:timeInThisSmil" content="00:00:28" />
    <layout>
      <region id="textView" />
    </layout>
  </head>
  <body>
    <ref title="Testing" src="0002.smil" id="ms_0002" />
  </body>
</smil>

A standard XML file with a link to a SMIL DTD and a root tag of <smil>. This format is recognized by PRONOM as fmt/205, although is often identified as a standard XML file. It seems the signature was created with a small offset which works with some SMIL files, but the gap between the end of the XML declaration and the start of the <smil> tag is only 20-86 bytes, not enough to allow for different character sets and full DTD URL’s. We will have to increase this gap in order to get all the SMIL files identified correctly.

With this update all the files in a DAISY 2.0 files should be identified individually, but as a set of files they make up the DAISY 2.0 format. This format requires the ncc.html file be present at the root of the folder or CD, so this file will aid in the manual identification of this format.

DAISY 3 was released in 2002 and standardized using the ANSI/NISO Z39.86 2002 name. It has been revised a couple times with the current revision being 2012. This update adds more functionality to the format with many new optional and required formats/files included in the folder. Here is a simple example:

ls -la "DAISY3 Export"
total 784
drwx------  1 tyler  staff   16384 Sep 25 22:06 .
drwx------  1 tyler  staff   16384 Sep 25 22:06 ..
-rwx------@ 1 tyler  staff     979 Sep 25 22:05 0001.smil
-rwx------  1 tyler  staff  228413 Sep 25 22:05 aud0001.mp3
-rwx------  1 tyler  staff    1014 Sep 25 22:05 navigation.ncx
-rwx------  1 tyler  staff    1881 Sep 25 22:05 package.opf
-rwx------  1 tyler  staff    7838 Nov  2  2020 tpbnarrator.res
-rwx------  1 tyler  staff  117656 Nov  2  2020 tpbnarrator_res.mp3

The SMIL format is still included, along with MP3’s, but we have some addition formats. The NCX or “Navigation Control File”, the OPF or “Package file”, and the RES or “Resource file” are a few of them. The NCX file is the first file accessed as it lays out the navigation for the whole DTB. It is also XML:

cat DAISY3 Export/navigation.ncx 
<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE ncx PUBLIC "-//NISO//DTD ncx 2005-1//EN" "http://www.daisy.org/z3986/2005/ncx-2005-1.dtd">
<ncx
	version="2005-1"
	xml:lang="en-US" xmlns="http://www.daisy.org/z3986/2005/ncx/">

This file is only recognized by DROID as a standard XML file. It probably should have unique identification like SMIL and with a root tag of <ncx>, that should be fairly easy to add.

The Package file with the extension OPF, is actually a format used by the openebook group, not to be confused by a format used by the Open Preservation Foundation 🤣. The Open Packaging Format is used and a DTB conforming to this standard must include exactly one Package File which must be a valid XML 1.0 document conforming to the OEBF Publication Structure 1.2 package.

cat DAISY3 Export/package.opf   
<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE package PUBLIC "+//ISBN 0-9673008-1-9//DTD OEB 1.2 Package//EN" "http://openebook.org/dtds/oeb-1.2/oebpkg12.dtd">
<package
	unique-identifier="uid" xmlns="http://openebook.org/namespaces/oeb-package/1.0/">
	<metadata>
		<dc-metadata xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oebpackage="http://openebook.org/namespaces/oeb-package/1.0/">
			<dc:Identifier
				id="uid">589c550e-303b-4c0d-9921-ae76d782fd53</dc:Identifier>
			<dc:Format>ANSI/NISO Z39.86-2005</dc:Format>
			<dc:Title>Obi Project</dc:Title>
			<dc:Publisher>N/A</dc:Publisher>
			<dc:Language>en-US</dc:Language>
			<dc:Creator>Creator name</dc:Creator>
			<dc:Date>2024-09-25</dc:Date>
		</dc-metadata>

The OPF format is also unknown to PRONOM and they identify as standard XML files as well. The root tag of “<package>” could be used elsewhere so the signature may need to reference the OEB package information.

The RES Resource file is also a standard XML and can be identified through its root tag of “<resources>” and resources DOCTYPE.

cat DAISY3 Export/tpbnarrator.res 
<?xml version='1.0' encoding='UTF-8'?>
<!DOCTYPE resources PUBLIC "-//NISO//DTD resource 2005-1//EN" "http://www.daisy.org/z3986/2005/resource-2005-1.dtd" []>
<resources xmlns="http://www.daisy.org/z3986/2005/resource/" version="2005-1">
  
  <!-- SKIPPABLE NCX -->
  
  <scope nsuri="http://www.daisy.org/z3986/2005/ncx/">
    <nodeSet id="ns001" select="//smilCustomTest[@bookStruct='LINE_NUMBER']">
      <resource xml:lang="en" id="r001">
        <text>Row</text>
        <audio src="tpbnarrator_res.mp3" clipBegin="0:00:02.379" clipEnd="0:00:03.416" />
      </resource>
    </nodeSet>

Now, adding these DAISY 3.0 formats will greatly increase the identification of this complex format. But we run into a problem with some of the software out there which generates these DAISY files, some of them include files not required by the format, but are included to be used by the different software. This can include some CSS files for formatting, additional XML, XSL files, DTD’s, and for DAISY files created by the PlexTalk software, additional project files.

ls -la MasterCD/AfterBuild 
total 7520
drwx------@ 1 tyler  staff    16384 Sep 24 19:34 .
drwx------@ 1 tyler  staff    16384 Sep 25 22:11 ..
-rwx------@ 1 tyler  staff     6688 Sep 25 01:32 ImdPhrInfo.imph
-rwx------@ 1 tyler  staff     3773 Sep 25 01:32 ImdTxtTabl.imtt
-rwx------@ 1 tyler  staff     1276 Sep 25 01:32 Ncc.imdn
-rwx------@ 1 tyler  staff  3716618 Sep 25 01:32 a000001.mp3
-rwx------@ 1 tyler  staff     4352 Sep 25 01:32 ncc.html
-rwx------@ 1 tyler  staff     1015 Sep 25 01:32 ptk000001.smil
-rwx------@ 1 tyler  staff      938 Sep 25 01:32 ptk000002.smil

The ncc.html file is here, indicating a DAISY 2.0 format, along with an MP3 and SMIL files, but including some additional formats.

In addition, when creating a project, four files with the extensions Ncc.imdn, ImdPhrInfo.imph, ImdTxtTabl.imtt, and METADATA.ini are automatically created. These files are called “Plextalk project files.” They store table of contents information, etc. (Plextalk project files generated by older versions of this product do not have METADATA.ini.)

http://www.plextalk.com/jp/dw_data/PRSStd/PLEX_RS_UM.html

These four files may not be crucial to the playing of the Daisy format, but they are important to the PlexTalk software.

hexdump -C ImdPhrInfo.imph | head
00000000  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
00000020  ff ff ff ff ff ff ff ff  00 00 00 00 00 00 00 00  |................|
00000030  00 00 00 00 00 00 00 00  f0 a3 0d 00 00 00 00 00  |................|
00000040  a3 06 00 00 a4 06 00 00  00 00 00 00 53 00 00 00  |............S...|
00000050  ff ff ff ff 01 00 00 00  03 00 00 00 00 00 00 00  |................|
00000060  00 00 00 00 00 00 00 00  c5 11 00 00 20 1a 00 00  |............ ...|
00000070  e5 2b 00 00 00 00 00 00  63 00 00 00 ff ff ff ff  |.+......c.......|
00000080  02 00 00 00 04 00 00 00  00 00 00 00 00 00 00 00  |................|
00000090  00 00 00 00 e5 2b 00 00  d6 0b 00 00 bb 37 00 00  |.....+.......7..|

hexdump -C ImdTxtTabl.imtt | head 
00000000  17 00 00 00 32 30 30 34  2f 30 35 2f 33 31 2f 31  |....2004/05/31/1|
00000010  36 3a 36 3a 34 37 2e 30  30 30 00 03 00 00 00 65  |6:6:47.000.....e|
00000020  6e 00 0b 00 00 00 69 73  6f 2d 38 38 35 39 2d 31  |n.....iso-8859-1|
00000030  00 0d 00 00 00 5a 3a 2f  42 6f 6f 6b 44 69 72 34  |.....Z:/BookDir4|
00000040  2f 00 0d 00 00 00 5a 3a  2f 42 6f 6f 6b 44 69 72  |/.....Z:/BookDir|
00000050  34 2f 00 0c 00 00 00 61  30 30 30 30 30 31 2e 6d  |4/.....a000001.m|
00000060  70 33 00 0c 00 00 00 61  30 30 30 30 30 31 2e 6d  |p3.....a000001.m|
*
00000980  70 33 00 08 00 00 00 48  65 61 64 69 6e 67 00 01  |p3.....Heading..|
00000990  00 00 00 00 08 00 00 00  48 65 61 64 69 6e 67 00  |........Heading.|

hexdump -C Ncc.imdn | head       
00000000  01 ff 00 ff c4 00 00 00  3c 00 00 00 2c 00 00 00  |........<...,...|
00000010  14 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000020  00 00 00 00 49 6d 64 54  78 74 54 61 62 6c 2e 69  |....ImdTxtTabl.i|
00000030  6d 74 74 00 00 00 00 00  00 00 00 00 00 00 00 00  |mtt.............|
00000040  00 00 00 00 49 6d 64 50  68 72 49 6e 66 6f 2e 69  |....ImdPhrInfo.i|
00000050  6d 70 68 00 00 00 00 00  00 00 00 00 00 00 00 00  |mph.............|
00000060  00 00 00 00 04 00 00 00  00 fa 00 00 44 ac 00 00  |............D...|
00000070  01 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000080  00 00 00 00 01 00 00 00  08 00 00 00 12 00 00 00  |................|
00000090  03 00 00 00 00 00 00 00  01 00 00 00 ff ff ff ff  |................|

I don’t have a METADATA.ini file to research, but I will be honest, these PlexTalk files will be hard to identify from their contents.

Looking at the IMPH file, there isn’t a lot of bytes which might indicate a format magic bytes. But I do see some patterns. The first 40 bytes all seem to be the same.

00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 FFFFFFFF FFFFFFFF

But making a signature from only 00 and FF might clash with other formats. It does appear that the 4 bytes FFFFFFFF occur every 40 bytes. This precision might be good enough if we repeat it a couple times.

The IMTT file is different. It appears to have information on the name, character set and all the files in the Daisy package. The first 4 bytes in my 14 samples either start with 17000000 or 18000000. Not knowing what the 17 or 18 refers to, I am hesitant to use it for identification. In between some of the data there is some consistent bytes, but at different offsets.


hexdump -C ImdTxtTabl.imtt | head
00000000  18 00 00 00 54 69 74 6c  65 00 35 39 2d 31 00 31  |....Title.59-1.1|
00000010  35 3a 35 34 3a 35 39 2e  32 36 30 00 03 00 00 00  |5:54:59.260.....|
00000020  65 6e 00 0b 00 00 00 69  73 6f 2d 38 38 35 39 2d  |en.....iso-8859-|
00000030  31 00 01 00 00 00 00 01  00 00 00 00 01 00 00 00  |1...............|
00000040  00 01 00 00 00 00 01 00  00 00 00 01 00 00 00 00  |................|
00000050  01 00 00 00 00 01 00 00  00 00 0c 00 00 00 4d 61  |..............Ma|
00000060  72 69 6f 6e 20 53 79 6d  65 00 28 00 00 00 4d 69  |rion Syme.(...Mi|
00000070  6e 75 74 65 73 20 6f 66  20 74 68 65 20 43 6f 6d  |nutes of the Com|
00000080  6d 69 74 74 65 65 20 4d  65 65 74 69 6e 67 20 32  |mittee Meeting 2|
00000090  34 30 35 30 34 00 08 00  00 00 48 65 61 64 69 6e  |40504.....Headin|

hexdump -C ImdTxtTabl.imtt | head
00000000  17 00 00 00 32 30 30 34  2f 30 35 2f 33 31 2f 31  |....2004/05/31/1|
00000010  36 3a 36 3a 34 37 2e 30  30 30 00 03 00 00 00 65  |6:6:47.000.....e|
00000020  6e 00 0b 00 00 00 69 73  6f 2d 38 38 35 39 2d 31  |n.....iso-8859-1|
00000030  00 0d 00 00 00 5a 3a 2f  42 6f 6f 6b 44 69 72 34  |.....Z:/BookDir4|
00000040  2f 00 0d 00 00 00 5a 3a  2f 42 6f 6f 6b 44 69 72  |/.....Z:/BookDir|
00000050  34 2f 00 0c 00 00 00 61  30 30 30 30 30 31 2e 6d  |4/.....a000001.m|
00000060  70 33 00 0c 00 00 00 61  30 30 30 30 30 31 2e 6d  |p3.....a000001.m|

Not sure what any of it means, but might be good enough for a signature.

Now the IMDN files might be a little easier:

hexdump -C Ncc.imdn | head
00000000  01 ff 00 ff d4 00 00 00  3c 00 00 00 2c 00 00 00  |........<...,...|
00000010  14 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000020  00 00 00 00 49 6d 64 54  78 74 54 61 62 6c 2e 69  |....ImdTxtTabl.i|
00000030  6d 74 74 00 00 00 00 00  00 00 00 00 00 00 00 00  |mtt.............|
00000040  00 00 00 00 49 6d 64 50  68 72 49 6e 66 6f 2e 69  |....ImdPhrInfo.i|
00000050  6d 70 68 00 00 00 00 00  00 00 00 00 00 00 00 00  |mph.............|
00000060  00 00 00 00 04 00 00 00  00 7d 00 00 22 56 00 00  |.........}.."V..|
00000070  01 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000080  00 00 00 00 01 00 00 00  28 00 00 00 28 00 00 00  |........(...(...|
00000090  00 00 00 00 00 00 00 00  28 00 00 00 ff ff ff ff  |........(.......|

This format directly names the two other formats. Should be easy to look for the two file names in the header. The NCC html file in Daisy 2.0 and the NCX xml file in Daisy 3.0 are directory files so it makes sense this file would do the same.

Not sure if these signatures will hold up over time, but they are a start. It would be nice if all the files we are given to preserve would have convenient static magic bytes, but alas, many do not and we have to guess.

These Daisy formats illustrate a problem in preservation that doesn’t quite have a good solution. Each of these files are individually unique and can be identified, but as a whole they represent another unique format. Tying formats together to link their interdependence on each other will be no small task, but will be necessary not only to understanding the format, but to avoid separating the files, renaming, or rearranging breaking that interdependence.

I have added the update to SMIL and new signatures for the other formats to my GitHub repository. Feel free to test and change if you find additional samples or information.

HFE

Posted on by Thor 1 Comment

Last week I had the pleasure of attending the 20th annual iPres conference on Digital Preservation in Ghent, Belgium. I enjoyed hearing from many of my respected colleagues on many aspects of preservation including one of my favorite topics, floppy disks. There was tutorials, lightning talks, and even a workshop, presented by Leontien Talboom, Elizabeth Kata, Chris Knowles, and myself. We titled the workshop “A Guide to Imaging Obscure Floppy Disk Formats“. The workshop was conceived by a mutual interest in imaging Wang 5.25in word processor disks, but expanded to include imaging of Amstrad 3in disks, 240K Brother Typewriter Disks, and Macintosh 400/800k disks.

I brought my hand soldered FluxEngine board and others brought their Greaseweazle board to show off how imaging obscure and uncommon disks can be done on a budget.

Photo of workshop taken on a Mavica Floppy Disk camera
Image taken during workshop on a Mavica FD200 Floppy Disk Camera.

During the conference we talked a bit about the different type of hardware that can be used and the difference between a disk image and flux image. There seems to be quite the exhaustive list of different types of file formats, some specific to a platform and others more generic. I recently did a blog post on the formats used by the Applesauce software, which have some unique features.

There are many disk image types which should be researched and added to PRONOM and other format description sites, but today lets take a look at a generic format used by many tools.

The HxC Floppy Emulator file format which the extension HFE is a popular format used with floppy drive emulators. There is a lot of complexity with what is included in many of these image formats, some are simply a raw sector representation of the binary data on a disk, others contain the complete flux readings from a floppy disk. The HFE format contains a little more than a raw image, including a header, a track lookup table, and the bitstreams for each track all with the purpose of emulating the physical media. The HFE format contains only a single pass over the data, where other formats may contain multiple reading of each track to get more complete data which can be helpful for damaged or purposely copy-protected disks. You can read more on Ashley’s blog, Library of Congress format description.

HFE version list

When using the HxC Floppy Emulator software, you can open and save to many different formats. The main format being their HFE native format. It comes in 5 versions. 

hexdump -C test01.hfe | head
00000000  48 58 43 50 49 43 46 45  00 53 02 00 e8 01 00 00  |HXCPICFE.S......|
00000010  07 01 01 00 ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
00000020  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|

Above is a hexdump of the main SDCard HxC Floppy Emulator file format. The format specification shows the 8 byte header “HXCPICFE”. This is a very unique pattern and should be all we need to make a robust signature for the format, but we do need to take into account the other HFE “versions” and see if they might clash or need to be identified separately.

hexdump -C test02-a2.hfe | head 
00000000  48 58 43 50 49 43 46 45  00 53 02 00 d0 03 00 00  |HXCPICFE.S......|
00000010  07 01 01 00 ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
00000020  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|

The “A2” version of the format has the same header but some different bytes further into the file. 

hexdump -C test03-rev2.hfe | head
00000000  48 58 43 50 49 43 46 45  01 53 02 00 00 00 00 00  |HXCPICFE.S......|
00000010  07 01 01 00 ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
00000020  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|

The “Rev 2” version also has the same header. But if you look at the 9th byte you can see the value changed from 00 to 01, which according to the specification, this is the revision byte. 

hexdump -C test04-rev3.hfe | head 
00000000  48 58 43 48 46 45 56 33  00 53 02 00 e8 01 00 00  |HXCHFEV3.S......|
00000010  07 01 01 00 ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
00000020  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|

With “Rev 3” we see a change in the header with “HXCHFEV3” which appears to be referred to as HFEv3.

hexdump -C test05-stream.hfe | head 
00000000  48 78 43 5f 53 74 72 65  61 6d 5f 49 6d 61 67 65  |HxC_Stream_Image|
00000010  00 00 00 00 00 00 00 00  00 18 00 00 00 02 00 00  |................|
00000020  00 1a 00 00 53 00 00 00  02 00 00 00 40 9c 00 00  |....S.......@...|
00000030  07 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000040  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|

This last format seems to be a special HxC stream image.

It seems the best option is to make three signatures to identify the three main headers. Additional software can be used to further parse the disk image. If you would like to see some sample images, you can download a bunch here. You can also take a look at my GitHub repository to see additional samples and a proposed set of signatures.

A2R / MOOF / WOZ

Posted on by Thor 1 Comment

There seems to be a never ending growing list of disk image formats. Many have features which are specific to the media and format. If you have ever imaged an older Macintosh floppy you know they are special. If you add in copy-protection which many early Apple II floppies have, and you need special drives, hardware, and a special format to store the floppy data.

When imaging special media, especially with unique media, it is best practice to image the floppies at the magnetic flux level.

Floppy disks contain magnetic fluctuations which are measured and recorded using specialized equipment. A popular method is using a Kryoflux board, floppy drive, and software. The software communicates with a custom controller board connected to a floppy drive through USB. If you are interested in the different controller boards, a good list has been compiled here.

A Kryoflux, fluxengine, greaseweazle, all can image specialized disks like a Macintosh 800k floppy, but the best controller board for them is an Applesauce setup. They are specifically designed to for the task. With that task, comes a few specialty formats.

A file format which can store flux data is a bit different than a regular disk image format. The flux data contains all the low-level recordings which can then be interpreted into disk images much like the original floppy. In the case of an Applesauce flux image, it can contain all the small nuances of the original floppy, this includes recording any copy protection or other creative methods used by software vendors throughout the years. The format used for storing this flux data is the A2R format.

A2R is in its third iteration. Let’s take a look at the basics of the format.

hexdump -C Samplev3.a2r | head
00000000  41 32 52 33 ff 0a 0d 0a  49 4e 46 4f 25 00 00 00  |A2R3....INFO%...|
00000010  01 41 70 70 6c 65 73 61  75 63 65 20 76 31 2e 38  |.Applesauce v1.8|
00000020  38 2e 35 20 20 20 20 20  20 20 20 20 20 20 20 20  |8.5             |
00000030  20 02 01 01 00 52 57 43  50 e9 49 6e 01 01 24 f4  | ....RWCP.In..$.|
00000040  00 00 00 00 00 00 00 00  00 00 00 00 00 43 01 00  |.............C..|
00000050  00 01 27 3a 25 00 91 d9  00 00 21 20 21 21 21 21  |..':%.....! !!!!|
00000060  1f 21 21 21 21 1f 24 5e  24 1f 21 21 20 21 24 5c  |.!!!!.$^$.!! !$\|
00000070  24 20 21 21 21 1f 24 5c  25 21 21 1f 21 21 23 5b  |$ !!!.$\%!!.!!#[|
00000080  25 20 21 21 21 1f 21 22  23 3f 41 3f 26 3e 43 3f  |% !!!.!"#?A?&>C?|
00000090  43 5f 41 27 3d 61 41 27  3d 61 3f 28 3e 61 3f 26  |C_A'=aA'=a?(>a?&|

hexdump -C Samplev2.a2r | head
00000000  41 32 52 32 ff 0a 0d 0a  49 4e 46 4f 24 00 00 00  |A2R2....INFO$...|
00000010  01 41 70 70 6c 65 73 61  75 63 65 20 76 31 2e 31  |.Applesauce v1.1|
00000020  2e 36 20 20 20 20 20 20  20 20 20 20 20 20 20 20  |.6              |
00000030  20 02 01 01 53 54 52 4d  75 17 5d 01 00 01 e6 da  | ...STRMu.].....|
00000040  00 00 83 a9 12 00 12 1e  11 13 1e 13 1e 13 11 1f  |................|
00000050  21 1f 11 13 1c 14 1e 30  14 20 1e 14 1e 14 1c 14  |!......0. ......|
00000060  1c 13 11 20 21 1f 11 11  0f 13 1e 14 1c 14 2e 21  |... !..........!|
00000070  13 1e 13 1e 14 1e 11 11  20 21 1f 11 11 13 1e 1f  |........ !......|
00000080  13 20 30 21 11 11 0f 13  1e 13 11 30 1f 21 20 13  |. 0!.......0.! .|
00000090  11 30 1f 14 1e 30 14 1e  11 11 11 1e 13 11 1e 14  |.0...0..........|

The A2R format uses a chunk system to store the various pieces to the format. Earlier versions used a STRM Chunk to store all the raw flux data. Version 3 changed to a RWCP Chunk to store all the raw flux data. Applesauce uses a 2-pass imaging process, doing a rapid imaging to determine where on the media surface track data exists and then a second pass that captures longer durations for processing and error correction.

Once the full raw flux data has been captured that data can be interpreted as a disk image. The Applesauce software is able to make a regular disk image, a Disk Copy 4.2 file, which are well known and identify in PRONOM as fmt/625, but can also create a couple of special disk image formats which allow for special nuances on an original disk.

The WOZ Disk Image format is an offshoot of the Applesauce project. Capturing highly accurate bit data is of no use if you don’t have a container to hold the data. The WOZ format was designed to be able to contain every possible Apple ][ disk structure and layout. It can be so accurate that even copy protected software can’t tell that it isn’t an original disk.

The WOZ format has become very popular in the Apple II community and is ideal for emulating all the old games and software titles popular in the early 1980’s. You may have guessed where the name comes from. The internet archive has a large collection of WOZ disks in their WOZ-a-Day collection. The file format of a WOZ disk image is also a chunk based format similar to the A2R format, it has two versions. Let’s take a look. 

hexdump -C WOZ 1.0/Blazing Paddles (Baudville).woz | head
00000000  57 4f 5a 31 ff 0a 0d 0a  f6 f5 92 d6 49 4e 46 4f  |WOZ1........INFO|
00000010  3c 00 00 00 01 01 00 01  01 41 70 70 6c 65 73 61  |<........Applesa|
00000020  75 63 65 20 76 30 2e 32  36 20 20 20 20 20 20 20  |uce v0.26       |
00000030  20 20 20 20 20 20 20 20  20 00 00 00 00 00 00 00  |         .......|
00000040  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000050  54 4d 41 50 a0 00 00 00  00 00 ff 01 01 01 ff 02  |TMAP............|
00000060  02 02 ff 03 03 03 ff 04  04 04 ff 05 05 05 ff 06  |................|
00000070  06 06 ff 07 07 07 ff 08  08 08 ff 09 09 09 ff 0a  |................|
00000080  0a 0a ff 0b 0b 0b ff 0c  0c 0c ff 0d 0d 0d ff 0e  |................|
00000090  0e 0e ff 0f 0f 0f ff 10  10 10 ff 11 11 11 ff 12  |................|

hexdump -C WOZ 2.0/Blazing Paddles (Baudville).woz | head
00000000  57 4f 5a 32 ff 0a 0d 0a  21 da c2 c8 49 4e 46 4f  |WOZ2....!...INFO|
00000010  3c 00 00 00 02 01 00 01  01 41 70 70 6c 65 73 61  |<........Applesa|
00000020  75 63 65 20 76 31 2e 31  20 20 20 20 20 20 20 20  |uce v1.1        |
00000030  20 20 20 20 20 20 20 20  20 01 01 20 00 00 00 00  |         .. ....|
00000040  0d 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000050  54 4d 41 50 a0 00 00 00  00 00 ff 01 01 01 ff 02  |TMAP............|
00000060  02 02 ff 03 03 03 ff 04  04 04 ff 05 05 05 ff 06  |................|
00000070  06 06 ff 07 07 07 ff 08  08 08 ff 09 09 09 ff 0a  |................|
00000080  0a 0a ff 0b 0b 0b ff 0c  0c 0c ff 0d 0d 0d ff 0e  |................|
00000090  0e 0e ff 0f 0f 0f ff 10  10 10 ff 11 11 11 ff 12  |................|

Unlike a common disk image, a WOZ image contains more than the bits on the disk, it contains a mapping of all the tracks and the associated data, this is how it can even contain copy-protection usually only possible with a physical disk. The ‘TMAP’ chunk contains a track map and the ‘TRKS’ chunk contains all the data.

What the WOZ is for the Apple II, MOOF was made for the Macintosh. You may wonder what is with the funny name, but there is a long history around “Clarus the Dogcow”. I’m sure this factoid will help you impress your friends or win at trivia night. Again, the purpose of the special format for Macintosh disks is to allow for emulating disks, even with copy protection. You can also find quite the collection of old Macintosh software in the MOOF format on the Internet Archive, even emulate your favorite game, such as Dark Castle, which I played for hours as a kid. Also a chunk based format, let’s take a look at the header.

hexdump -C Dark Castle v1.0 - Disk 1.moof | head
00000000  4d 4f 4f 46 ff 0a 0d 0a  b5 75 f9 4e 49 4e 46 4f  |MOOF.....u.NINFO|
00000010  3c 00 00 00 01 01 00 01  10 41 70 70 6c 65 73 61  |<........Applesa|
00000020  75 63 65 20 76 31 2e 37  33 20 20 20 20 20 20 20  |uce v1.73       |
00000030  20 20 20 20 20 20 20 20  20 00 13 00 00 00 00 00  |         .......|
00000040  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
00000050  54 4d 41 50 a0 00 00 00  00 ff 01 ff 02 ff 03 ff  |TMAP............|
00000060  04 ff 05 ff 06 ff 07 ff  08 ff 09 ff 0a ff 0b ff  |................|
00000070  0c ff 0d ff 0e ff 0f ff  10 ff 11 ff 12 ff 13 ff  |................|
00000080  14 ff 15 ff 16 ff 17 ff  18 ff 19 ff 1a ff 1b ff  |................|
00000090  1c ff 1d ff 1e ff 1f ff  20 ff 21 ff 22 ff 23 ff  |........ .!.".#.|

All three formats created for imaging and emulating Apple and Macintosh software are well documented and open. They are also well suited for preservation as they can contain extensive metadata in the INFO chunk which gives provenance information on the source of the files. The Applesauce software even has a camera to photograph the disk itself for archiving. All of this makes these formats great for preservation and emulation. Take a look at my proposal for a signature on my Github.

ePic

Posted on by Thor Leave a comment

Image compression has been around for awhile. It seems everyone took a crack at making better algorithms to improve quality and size. Some chose to invent new ways and others chose to use existing methods but with their own flare. Kodak tried this with their PhotoCD, but there was a couple other photo processing options that popped up in 90’s. One was Seattle FilmWorks and another was Konica PC PictureShow. Both of which used “proprietary” formats to deliver developed film on disk.

Seattle FilmWorks later called PhotoWorks, used an image format with the extension SFW and was based on BMP and JPG, but with their own twist. The same goes for the format used by Konica’s PC PictureShow.

Konica PC PictureShow Disk

If you took your film in to be developed at one of Konica’s photo labs, you could could have those images put on a diskette or later a CD-R. The disks came with software to view your photos called PC PhotoShow. The images stored on disk where in another proprietary format with the extension KQP. The KQP format was actually licensed from another company called Pegasus Imaging Corporation, later known as Accusoft. They developed their own way to compress a JPEG file which they called an ePic. An SDK called PICTools was offered for many years, but seems not to be available anymore.

ePIC (Proprietary)
  • Supports PIC format compression, replacing the JPEG Huffman encoder with the proprietary ELS entropy encoder for 15% more compression.
  • Can be losslessly converted back to JPEG format using Op_RORE.

A search on the internet for Konica KQP shows quite a few people over the years wondering what to do with their old disks and converting the old format to JPG, only to find a lack of information and available tools to do so. One such person used python to edit the file and making the file renderable as a JPG. While the method worked well for their KQP files, it might not work for all of them. Let’s look closer and understand why.

hexdump -C Sample.PIC | head
00000000  42 4d 00 00 00 00 00 00  00 00 42 04 00 00 44 00  |BM........B...D.|
00000010  00 00 34 08 00 00 24 fa  ff ff 01 00 18 00 4a 50  |..4...$.......JP|
00000020  45 47 00 00 00 00 00 00  00 00 00 00 00 00 fc 00  |EG..............|
00000030  00 00 ec 00 00 00 2c 00  00 00 18 00 00 00 00 00  |......,.........|
00000040  00 00 02 00 00 00 08 00  00 00 01 00 00 00 01 00  |................|
00000050  00 00 60 00 00 00 00 00  60 00 00 60 00 00 00 00  |..`.....`..`....|
00000060  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|

At first glance the file appears to be a Bitmap (BMP), and it does have a Bitmap header claiming to have JPEG compression, but if we look a little further into the file.

identify -verbose Sample.PIC   
identify: length and filesize do not match `Sample.PIC' @ error/bmp.c/ReadBMPImage/950.
identify: unrecognized compression `Sample.PIC' @ error/bmp.c/ReadBMPImage/1019.

hexdump -C Sample.PIC      
00000000  42 4d 00 00 00 00 00 00  00 00 42 04 00 00 44 00  |BM........B...D.|
00000010  00 00 34 08 00 00 24 fa  ff ff 01 00 18 00 4a 50  |..4...$.......JP|
00000020  45 47 00 00 00 00 00 00  00 00 00 00 00 00 fc 00  |EG..............|
00000030  00 00 ec 00 00 00 2c 00  00 00 18 00 00 00 00 00  |......,.........|
00000040  00 00 02 00 00 00 08 00  00 00 01 00 00 00 01 00  |................|
00000050  00 00 60 00 00 00 00 00  60 00 00 60 00 00 00 00  |..`.....`..`....|
00000060  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
00000400  00 00 60 00 00 00 00 00  60 00 00 60 00 00 00 00  |..`.....`..`....|
00000410  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
00000440  00 00 ff d8 ff e0 00 10  4a 46 49 46 00 01 02 02  |........JFIF....|
00000450  00 00 00 00 00 00 ff e1  00 0a 50 49 43 00 01 19  |..........PIC...|
00000460  1e 01 ff c0 00 11 08 05  dc 08 34 03 01 11 00 02  |..........4.....|

We find a JPG marker, in fact almost the whole jpg file is included, except the quantization tables for luminance and chrominance which are needed to properly display the image. This is the area the Pegasus company thought they could encode better to further compress the image. Their method was to use a new algorithm called ELS (Entropy Logarithmic-Scale). This new method was used by the PICTools software to make a Pegasus PIC file while Konica used it for their KQP format. They are identical. By choosing the luminance and chrominance values during compression, you could make a highly compressed image, but required specific software to render.

Pegasus also made use of a special custom APP marker (PIC) within the JPEG structure of the PIC/KQP and also any JPG compressed using their software. This marker which takes up around 8 bytes holds the luminance and chrominance values. Take the above sample for instance, it is compressing the image with a Luminance of 25 and a Chrominance of 30, these are integer values and in hex they would be “19” and “1E” respectively. 

hexdump -C Sample.PIC      
00000440  00 00 ff d8 ff e0 00 10  4a 46 49 46 00 01 02 02  |........JFIF....|
00000450  00 00 00 00 00 00 ff e1  00 0a 50 49 43 00 01 19  |..........PIC...|
00000460  1e 01 ff c0 00 11 08 05  dc 08 34 03 01 11 00 02  |..........4.....|
00000470  11 01 03 11 01 ff c4 00  51 00 01 00 03 01 00 00  |........Q.......|

So in theory one could strip out any part of the file before the JPG beginning of file magic bytes (FF D8 FF E0), locate the APP marker, use the values to generate the two quantization tables, insert them in the appropriate spot and save out a JPG file.

This may be the case for the first few versions of the ePic format, but later versions got more complicated. It seems a “PIC2” version replaced the earlier versions and this format is a little more complicated. 

hexdump -C Sample.KQP | head
00000000  50 49 43 32 01 08 00 00  00 64 00 01 00 b9 3e 00  |PIC2.....d....>.|
00000010  00 05 08 00 00 00 4a 50  47 45 03 00 00 00 16 24  |......JPGE.....$|
00000020  00 00 00 43 6f 6d 70 72  65 73 73 69 6f 6e 20 62  |...Compression b|
00000030  79 20 50 65 67 61 73 75  73 20 49 6d 61 67 69 6e  |y Pegasus Imagin|
00000040  67 20 43 6f 72 70 2e 06  68 3e 00 00 ff d8 ff e0  |g Corp..h>......|
00000050  00 10 4a 46 49 46 00 01  01 00 00 01 00 01 00 00  |..JFIF..........|
00000060  ff e1 00 16 50 49 43 00  03 00 00 01 00 00 00 00  |....PIC.........|
00000070  00 00 00 00 00 00 00 00  ff db 00 84 00 0f 0a 0a  |................|
00000080  0a 0a 06 0f 0a 0a 0a 0f  0f 0f 0f 14 1e 14 14 14  |................|
00000090  14 14 28 1e 1e 19 1e 2d  28 32 32 2d 28 2d 2d 32  |..(....-(22-(--2|

Instead of the Bitmap (BMP) header, a proprietary PIC2 header is used, still containing a JPG in the JFIF format along with a the PIC APP marker, but encoded in a way that the simple method of adding a quantization table may not work. With the original format the JPG and the PIC/KQP were approximately the same size, this new version significantly reduces the size of the PIC/KQP in comparison with the JPG.

The ELS compression technology used in the ePic format seems to be patented by Pegasus and Accusoft, but is not entirely hidden as the libavcodec library includes a ELS decoder. Might be a fun project to use the code to decode the PIC/KQP formats fully.

In the meantime, a signature identifying the two versions should be added to PRONOM. Check out my proposal on my GitHub. If you need to convert your KQP or PIC files back to JPG here are a few links:

Konica PC PictureShow Version 4 (PIC2)

Accusoft PICTools Apollo Demo (Windows 7 Compatible)

Konica PC PictureShow for Macintosh

FASTA & FASTQ

Posted on by Thor 2 Comments

There seems to be a never ending source of file formats out there. Documenting past obsolete formats, one would assume a point at which there are no more to find, but in reality more are re-discovered everyday by the Digital Preservation community. When it comes to more modern formats, it seems more are invented everyday, too many to keep up with identification. Document one, 10 more pop up, it seems never-ending. Such is the case for scientific formats, including sequencing formats.

I was speaking with a colleague from another institution the other day and a file format was mentioned I hadn’t heard of before. It seems many of their scientific data was stored in a format called FASTA “Fast A” (“fast-aye”). This format specifically stores DNA sequence data and is used quite a bit, it seems. I was even more surprised the next day when I went to process some new submissions for our repository only to find one submission contained three FASTA files. I love researching file formats, but sometimes in order to understand the format structure you have to know something about the content as well. Let’s explore the FASTA and FASTQ file formats. If you would like to take a peek at the Human Genome in FASTA, go here.

Both the FASTA and FASTQ formats are text based and have a simple structure. Identification of each of these should be pretty simple, but to avoid conflicts with other formats, the signature might have to be more complex.

The FASTA format is well documented as many in the scientific community use it. Basically the format starts with the greater than “>” character followed by a description, a new line character, then the sequence. For example:

>MCHU - Calmodulin - Human, rabbit, bovine, rat, and chicken
MADQLTEEQIAEFKEAFSLFDKDGDGTITTKELGTVMRSLGQNPTEAELQDMINEVDADGNGTID
FPEFLTMMARKMKDTDSEEEIREAFRVFDKDGNGYISAAELRHVMTNLGEKLTDEEVDEMIREA
DIDGDGQVNYEEFVQMMTAK*

Pretty straight forward, but so much of the format can be variable, a simple signature would clash with too many other formats. There are some rules with what characters can be used in the sequence so it might be possible to limit the signature to only allow certain characters. At first I thought it might only be able to contain the standard characters representative of adenine (A), cytosine (C), guanine (G), and thymine (T), but as it turns out the FASTA format can contain Nucleic Acid Code’s and Amino Acid Code’s. These codes allow more than the four I was expecting, but do limit what can be represented.

Take the NCBI Sequence Viewer for a spin and download some data as FASTA.

The FASTQ format adds more structure and is more limiting, but also presents some challenges. Here is a sample of its structure:

@SEQ_ID
GATTTGGGGTTCAAAGCAGTATCGATCAAATAGTAAATCCATTTGTTCAACTCACAGTTT
+
!''*((((***+))%%%++)(%%%%).1***-+*''))**55CCF>>>>>>CCCCCCC65

Instead of a greater than symbol, the FASTQ format uses an “@” symbol followed by an identifier. The identifier can be basically anything and as long as needed. There is a newline character followed by the DNA sequence, which is only the four characters I have heard before. It can contain an A, C, G, T, or N. The “N” can represent an unidentified nucleotide or indicate that the software was unable to make a basecall. A newline character again and the “+” symbol. This is place before the fourth line with is a quality score and is the same number of characters as the sequence.

See what I mean when you have to learn about the context of a format in order to make a proper signature!

One of the problems I am left with is how to determine how many of the sequence characters to use in the signature to not have any conflicts. Too few and it might conflict with another format or simple text file. Too many and the signature gets complicated and may exclude a short sequence file. As far as I can tell there is no set minimum or maximum for the sequence. Not sure what the genome for Pinus Taeda would look like in FASTA with 22.18 billion base pairs. The other problem is often times these formats are compressed into a GZIP file, so they need to be extracted before identification.

These two formats are just a couple of the many sequencing formats being used in the bioinformatics community. I am sure others will pop up in the future. Until then, I have with the help of others put together a signature which seems to work well for the samples and data sets we have access to. Take a look at my GitHub for the signature proposal. If you find any issues, let me know!

SDIF

Posted on by Thor Leave a comment

I have used and have researched a lot of audio editing software. Some are very simple and straightforward, others are feature rich and take some time to learn. While looking in a format, I came across some Audio software which nothing like I have used before. At first I was confused, I figured it would be simple to open a certain file format and play the audio. Not so fast.

Max is software which proudly says it is an, “infinitely flexible space to create your own interactive software”. Created by Cycling ’74 software, Max has been around for awhile, being developed in the mid 1980’s. It allows the user to make “patches” stringing around components and effects to accomplish an infinite amount of options and outcomes.

The software produces simple project files and patch files, but hey are just JSON data, at least in the latest version. But when working with audio files the software can save to a number of formats.

One of the options is a format called “SDIF”, which stands for “Sound Description Interchange Format“. SDIF was jointly developed by IRCAM and CNMAT, with proposals starting back in the mid-1990’s. Originally written as a Spectral Description, it was later changed to refer to a Sound Description.

The Specification states the general idea was to “store information related to signal processing and specifically of sound, in files, according to a common format to all data types. Thus, it is possible to store results or parameters of analyses, syntheses…” So not exactly the same as a simple WAVE file you can open and edit, this format was meant to store signal data for analysis.

Each SDIF file consists of a header and then an overall a succession of frames, not unlike chunks in the IFF/AIFF/RIFF formats, ordered in time. Each frame matrix declares a “Type” which can be a combination of many options. Lets take a look at a SDIF file:

hexdump -C test.sdif | head
00000000  53 44 49 46 00 00 00 08  00 00 00 03 00 00 00 01  |SDIF............|
00000010  31 54 52 43 00 00 00 20  00 00 00 00 00 00 00 00  |1TRC... ........|
00000020  00 00 00 01 00 00 00 01  31 54 52 43 00 00 00 04  |........1TRC....|
00000030  00 00 00 00 00 00 00 04  31 54 52 43 00 00 00 c0  |........1TRC....|
00000040  3f 74 7a e1 40 00 00 00  00 00 00 01 00 00 00 01  |?tz.@...........|
00000050  31 54 52 43 00 00 00 04  00 00 00 0a 00 00 00 04  |1TRC............|
00000060  3f 80 00 00 45 95 35 c3  00 00 00 00 00 00 00 00  |?...E.5.........|
00000070  40 00 00 00 46 06 e2 14  00 00 00 00 00 00 00 00  |@...F...........|
00000080  40 40 00 00 45 3b 42 3d  00 00 00 00 00 00 00 00  |@@..E;B=........|
00000090  40 80 00 00 43 5d 94 7b  00 00 00 00 00 00 00 00  |@...C].{........|

This test file has the opening frame “SDIF“, to identify it as an SDIF, then a reference to the type “1TRC“. I would try and explain a Matrix 1TRC Sinusoidal Track, but I have no idea what it means. Something, something sine wave, etc. Someone much smarter than me can make use of this format. Here are a couple examples of SDIF with other frame types.

hexdump -C angry_cat.part.sdif| head
00000000  53 44 49 46 00 00 00 08  00 00 00 03 00 00 00 01  |SDIF............|
00000010  31 4e 56 54 00 00 00 88  ff ef ff ff ff ff ff ff  |1NVT............|
00000020  ff ff ff fd 00 00 00 01  31 4e 56 54 00 00 03 01  |........1NVT....|
00000030  00 00 00 61 00 00 00 01  53 74 72 65 61 6d 49 44  |...a....StreamID|
00000040  09 30 0a 44 61 74 65 09  54 68 75 5f 41 75 67 5f  |.0.Date.Thu_Aug_|
00000050  5f 33 5f 32 31 2e 33 32  2e 34 35 5f 32 30 30 30  |_3_21.32.45_2000|
00000060  5f 0a 54 61 62 6c 65 4e  61 6d 65 09 53 69 6e 75  |_.TableName.Sinu|
00000070  73 6f 69 64 61 6c 54 72  61 63 6b 73 0a 57 72 69  |soidalTracks.Wri|
00000080  74 74 65 6e 42 79 09 50  6d 5f 56 65 72 73 69 6f  |ttenBy.Pm_Versio|
00000090  6e 5f 31 2e 32 2e 32 0a  00 00 00 00 00 00 00 00  |n_1.2.2.........|

hexdump -C cymbalum-c4.res.sdif| head 
00000000  53 44 49 46 00 00 00 08  00 00 00 03 00 00 00 01  |SDIF............|
00000010  31 52 45 53 00 00 0d 20  00 00 00 00 00 00 00 00  |1RES... ........|
00000020  00 00 00 04 00 00 00 01  31 52 45 53 00 00 00 04  |........1RES....|
00000030  00 00 00 d0 00 00 00 04  42 49 27 7a 39 59 fc ab  |........BI'z9Y..|
00000040  3d 35 06 c9 00 00 00 00  42 6e 68 68 39 63 99 b1  |=5......Bnhh9c..|
00000050  3e 25 f7 c0 00 00 00 00  42 c6 02 bb 39 8c 31 79  |>%......B...9.1y|
00000060  3f bb 7e 6e 00 00 00 00  43 01 82 96 3a 1d 36 44  |?.~n....C...:.6D|
00000070  3e d9 21 12 00 00 00 00  43 07 35 f0 3a 20 6f 6e  |>.!.....C.5.: on|
00000080  3f 02 32 7f 00 00 00 00  43 30 84 0b 39 97 f9 1b  |?.2.....C0..9...|
00000090  3e c6 43 c7 00 00 00 00  43 4d e4 e4 39 88 14 90  |>.C.....CM..9...|

Unfortunately, the common tools I use to explore AV formats don’t seem to work on this format. MediaInfo, FFProbe, Exiftool, all give me unknown file warnings. So I had to compile the SDIF software in order to get some details.

querysdif angry_cat.part.sdif 
Header info of file angry_cat.part.sdif:

Format version: 3
Types version:  1

Ascii chunks of file angry_cat.part.sdif:

1NVT
{
StreamID	0;
Date	Thu_Aug__3_21.32.45_2000_;
TableName	SinusoidalTracks;
WrittenBy	Pm_Version_1.2.2;
}

Data in file angry_cat.part.sdif (9504872 bytes):
 1933 1TRC frames in stream 0 between time 0.000000 and 5.794875 containing
    1933 1TRC matrices with  45 --400 rows,   4 --  4 columns

An interesting thing is that a SDIF file can be in text form as well.

sdiftotext test.sdif 
SDIF


SDFC

1TRC	1	1	0
  1TRC	0x0004	0	4

1TRC	1	1	0.005
  1TRC	0x0004	10	4
	1	4774.72	0	0
	2	8632.52	0	0
	3	2996.14	0	0
	4	221.58	0	0
	5	1943.02	0	0
	6	123.951	0	0
	7	6705.04	0	0
	8	4304.97	0	0
	9	3554.29	0	0
	10	23.7822	0	0

1TRC	1	1	0.01
  1TRC	0x0004	10	4
	1	4774.72	0.0353114	2.06098
	2	8632.52	0.00442518	0.68795
	3	2996.14	0.0238517	-1.42295
	4	221.58	0.0089712	-2.44141
	5	1943.02	0.00768914	2.64629
	6	123.951	0.0397061	-0.17527
	7	6705.04	0.0245643	-0.168753
	8	4304.97	0.00894803	1.45553
	9	3554.29	0.0265175	2.57231
	10	23.7822	0.0419019	-2.17731

1TRC	1	1	0.2
  1TRC	0x0004	10	4
	1	2284.56	0.02781	2.47054
	2	4222.62	0.0151738	1.55309
	3	31.1554	0.00421461	-0.657285
	4	310.99	0.0122306	1.25794
	5	215.192	0.0174093	1.25468
	6	6253.69	0.000894192	2.21334
	7	8533.32	0.0296167	2.07209
	8	8044.77	0.0423002	2.54088
	9	6087.45	0.0264733	-2.05523
	10	7052.7	0.0287347	0.426339

1TRC	1	1	0.205
  1TRC	0x0004	10	4
	1	2284.56	0	0
	2	4222.62	0	0
	3	31.1554	0	0
	4	310.99	0	0
	5	215.192	0	0
	6	6253.69	0	0
	7	8533.32	0	0
	8	8044.77	0	0
	9	6087.45	0	0
	10	7052.7	0	0

1TRC	1	1	0.21
  1TRC	0x0004	0	4

ENDC
ENDF

An interesting format for sure. But wait, there is more!

My initial interest in this format was when I was given access to a set of MUBU files. I was unclear on how there were created at first and it took me down a long path of learning about SDIF and the Max software from Cycling ’74 and IRCAM. MUBU turns out to be a toolbox for Max which adds more analysis features.

MUBU stands for MUlti-BUffer, which helps overcome some limitations. It is actually a container using the SDIF standard. Lets take a look.

hexdump -C test.mubu | head
00000000  53 44 49 46 00 00 00 08  00 00 00 03 00 00 00 01  |SDIF............|
00000010  31 4e 56 54 00 00 00 78  ff ef ff ff ff ff ff ff  |1NVT...x........|
00000020  ff ff ff fd 00 00 00 01  31 4e 56 54 00 00 03 01  |........1NVT....|
00000030  00 00 00 53 00 00 00 01  4d 75 42 75 2e 43 6f 6e  |...S....MuBu.Con|
00000040  74 61 69 6e 65 72 2e 4e  75 6d 54 72 61 63 6b 73  |tainer.NumTracks|
00000050  09 31 0a 4d 75 42 75 2e  43 6f 6e 74 61 69 6e 65  |.1.MuBu.Containe|
00000060  72 2e 56 65 72 73 69 6f  6e 09 31 2e 35 0a 4d 75  |r.Version.1.5.Mu|
00000070  42 75 2e 43 6f 6e 74 61  69 6e 65 72 2e 4e 75 6d  |Bu.Container.Num|
00000080  42 75 66 66 65 72 73 09  31 0a 00 00 00 00 00 00  |Buffers.1.......|
00000090  31 4e 56 54 00 00 00 38  ff ef ff ff ff ff ff ff  |1NVT...8........|

A MUBU file has the same SDIF frame header, but also include a “1NVT” frame, which is a Name Value Table. This is where the MUBU container is referenced. The MuBu file has its own structure:

If I query the MuBu file like I did the SDIF, I get the following:

querysdif test.mubu
Header info of file test.mubu:

Format version: 3
Types version:  1

Ascii chunks of file test.mubu:

1NVT
{
MuBu.Container.NumTracks	1;
MuBu.Container.Version	1.5;
MuBu.Container.NumBuffers	1;
}
1NVT
{
MuBu.Buffer.Index	0;
}
1NVT
{
MuBu.Track.MxRows	2;
AudioFile	1;
MuBu.Track.NonNumType	0;
MuBu.Track.MaxSize	93515;
meta_ISFT	Lavf60.16.100;
MuBu.Track.Name	mytrack;
MuBu.Track.BufferIndex	0;
MuBu.Track.SampleRate	48000;
FileName	Wilhelm_Scream.wav;
MuBu.Track.MxVarRows	0;
MuBu.Track.MxCols	1;
meta_MetaDataSource	WAV;
MuBu.Track.EndTime	1623.5;
FilePath	/;
MuBu.Track.SampleOffset	0;
MuBu.Track.TimeTags	0;
MuBu.Track.Size	77929;
MuBu.Track.Index	0;
}

1TYP
{
  1MTD	M000	{unnamed}
  1FTD	M000
	{
	  M000	Track-0-MatrixData;
	}
}

Data in file test.mubu (3741392 bytes):
77929 M000 frames in stream 0 between time 0.000000 and 1.623500 containing
   77929 M000 matrices with   2 --  2 rows,   1 --  1 columns

The MuBu file contains one audio track and one buffer. This is a simple test file, but MuBu files can be quite large with multiple tracks.

Working with the Max software or OpenMusic is not something I found to be easy to understand. I am sure if I was more musically inclined and with a little practice I could make some of this work. For the time being, a signature to identify a SDIF and MUBU will have to do. Check out the GitHub for my proposed signature and a couple examples.

Sibelius

Posted on by Thor Leave a comment

Music notation software is among the earliest software for desktop computers. SCORE in 1987, Finale came around in 1988, Capella in 1992, and Sibelius in 1993. Many others came and went during this time. Music notation software was so much more than the typical word processing or desktop publishing system. Specialized fonts were needed to display the music notation and there are many other variables for different instruments allowing individuals and others the ability to create complicated compositions in an inexpensive way.

Sibelius [SI] + [BAY] + [LEE] + [UHS] was originally developed for the Acorn system in 1986, then released on Windows and Macintosh in 1998-99. The software became very popular and in 2006 was purchased by the software giant AVID. The software was used enough to get a preservation assessment by the British Library in 2017 and draft status format description by the Library of Congress, written by the amazing Ashley!

Both reviews of the format emphasize the proprietary nature of the file format which has been used since the early versions. Aside from the early Acorn release, the Windows and Macintosh versions used a binary format with the SIB extension. They are actually quite easy to identify.

hexdump -C Sibelius-s01.sib | head
00000000  0f 53 49 42 45 4c 49 55  53 00 00 40 00 02 00 a4  |.SIBELIUS..@....|
00000010  f1 ed 00 00 00 30 00 00  00 02 00 00 00 01 00 00  |.....0..........|
00000020  00 2a 00 00 00 00 00 00  00 00 0f 53 49 42 45 4c  |.*.........SIBEL|
00000030  49 55 53 00 00 40 00 02  00 00 00 00 00 00 00 3a  |IUS..@.........:|
00000040  00 00 00 00 38 a1 28 06  b3 d2 2f 66 03 04 16 4e  |....8.(.../f...N|
00000050  5f 5c 8d f3 95 27 3e f1  2a 1b 68 de 08 81 e8 9a  |_\...'>.*.h.....|
00000060  ea 1c bf dd 54 0e 92 8d  4d be e3 34 ed 42 78 36  |....T...M..4.Bx6|
00000070  d2 e1 67 7b 8d f7 98 6a  3a 70 c4 8b 0b 08 7b 26  |..g{...j:p....{&|
00000080  f9 45 00 00 00 00 48 71  7c 4c 98 df 0b 38 7d 9d  |.E....Hq|L...8}.|
00000090  2a 2d 84 9c a4 39 0f 4d  da a2 cc 97 ad 3d b0 55  |*-...9.M.....=.U|

This is exactly how PRONOM and other identification methods determine they are Sibelius files. PRONOM has assigned the format fmt/696 and is looking for the hexadecimal bytes 0F534942454C495553.

The problem with this identification method is that all the Sibelius files are identified as such, regardless of version. As mentioned by Ashley, version of the software used is highly important as new features were added all the time making backwards compatibility difficult. Add in the fact that there were different releases for each version which would limit these features even more and I can see how a musician could get very frustrated. If you created a score in Sibelius 5 and tried to open in Sibelius 5 Student version, you may find your composition lacking in many ways. The only way to avoid compatibility issues is to always open in the latest “Ultimate” version. Sibelius Ultimate can open all versions of the SIB format back to version 2. The software even has an export feature which allows you to export back to a previous version stripping what is necessary to ensure compatibility.

Sibelius export to previous version

For those with a bunch of SIB files in their archives, how would you know which software version created the file? Well lets take a closer look at the bytes and see if we can find some patterns. Let it be known, I am not reverse engineering the format, just looking for patterns which will allow for proper identification!

I am not the first person to ask this question, many others want to know the versions of their SIB files. Thankfully others have found some clues on which bytes hold the version information. It seems we can determine the version based on 4 bytes shortly after the SIBELIUS string. Specifically bytes 10-13. 

hexdump -C Sibelius2-s01.sib | head
00000000  0f 53 49 42 45 4c 49 55  53 00 00 08 00 22 00 47  |.SIBELIUS....".G|
00000010  98 4c 00 00 00 3a 00 00  00 00 4e 81 49 34 41 2c  |.L...:....N.I4A,|
00000020  fa 76 62 f9 71 53 a9 93  0f 54 1e 20 6c 63 61 4d  |.vb.qS...T. lcaM|
00000030  f7 b2 b0 a7 5d bd 82 3a  0d 86 02 8b f2 89 d2 a0  |....]..:........|
00000040  83 1f 8d e0 37 1b ed 1c  6a 8b 82 08 4b 6d 64 60  |....7...j...Kmd`|
00000050  71 59 e8 aa ef b1 3c df  5c 25 0a 9f 66 50 69 de  |qY....<.\%..fPi.|
00000060  2a d3 4e 2a cd 97 88 06  67 5f 50 64 0f 8f 86 2b  |*.N*....g_Pd...+|
00000070  08 0d 3f f7 80 26 e0 63  f6 7d 4e f8 e7 c0 3f fc  |..?..&.c.}N...?.|
00000080  7a 77 ea b3 4a b9 30 59  13 47 6e 09 0a 0b ae 3c  |zw..J.0Y.Gn....<|
00000090  c1 93 85 f6 41 f8 58 22  4b 92 35 3f b2 f5 3f 9d  |....A.X"K.5?..?.|

From what others have gathered and updating it with more recent versions I have come up with a list.

VersionHex 10-13
Sibelius 1.200 00 00 0E
Sibelius 2.x00 08 xx xx 
Sibelius 3.x00 0A xx xx 
Sibelius 4.x00 1B xx xx 
Sibelius 5.000 2D 00 03 
Sibelius 5.100 2D 00 0D 
Sibelius 5.2.x – 5.400 2D 00 10 
Sibelius 6.0.x00 36 00 01 
Sibelius 6.100 36 00 17 
Sibelius 6.200 36 00 1E 
Sibelius 7.000 39 00 0C 
Sibelius 7.0.1 – 7.0.200 39 00 0E 
Sibelius 7.0.300 39 00 13 
Sibelius 7.1.000 39 00 15 
Sibelius 7.1.2 – 7.1.300 39 00 16 
Sibelius 7.5.x00 3D 00 0E 
Sibelius 8.0.0 – 8.0.100 3D 00 10 
Sibelius 8.1.x00 3E 00 00 
Sibelius 8.200 3E 00 01 
Sibelius 8.300 3E 00 02 
Sibelius 8.4.x00 3E 00 06 
Sibelius 8.5.x00 3E 00 07 
Sibelius 8.6.x, 8.7.0, 8.7.100 3F 00 00 
Sibelius 8.7.2, 2018.1, 2018.4.x, 2018.5, 2018.6, 2018.700 3F 00 01 
Sibelius 2018.11, 2018.1200 3F 00 02
Sibelius 2019.100 3F 00 04
Sibelius 2019.4.x, 2019.5, 2019.7, 2019.900 3F 00 06
Sibelius 2019.1200 3F 00 07
Sibelius 8.6-2019.1200 3F 00 0A
Sibelius 2020.100 3F 00 0B
Sibelius 2020.3, 2020.600 40 00 01
Sibelius 2020.900 40 00 02
Sibelius 2022.500 40 00 03
Sibelius 2022.1100 41 00 02
Sibelius 2022.1200 42 00 00
Sibelius 2023.300 42 00 01
Sibelius 2023.800 43 00 07
Sibelius 2024.3.100 44 00 01

That is a lot of versions and I feel there may be some gaps that still need to be identified. It appears that the first two bytes are the major version and the second set of bytes is the minor version. Although it looks like a few major version bytes span across a few software versions. With this chart, one could be very specific in identifying which Sibelius version wrote the file, but for archiving purposes it seems we can group many of these capturing just the major version. The export screenshot above seems to have broken down significant changes and grouped similar formats together, the biggest being 8.6 through 2019.12. A comparison of “student” and “first” formats don’t have any obvious bytes which indicate as such, so for now they are all lumped together.

There is one other similar format which needs to be mentioned. Sibelius Scorch was a product made to share scores online. This has been replaced with Sibelius Cloud Publishing, but for awhile was the best way to share a score with others in a way that protected the original. I have no idea how they were made, but sites like scorestreet.net and sibeliusmusic.com were sites you could upload your score to for sharing. Some SCO files appear to have a PDF embedded within them for proper printing.

hexdump -C smd_h_0000000000097761.sco | head
00000000  0f 43 43 53 43 4f 52 43  48 00 00 36 00 1e 00 c0  |.CCSCORCH..6....|
00000010  d4 55 00 00 00 30 00 00  00 01 00 00 00 01 00 00  |.U...0..........|
00000020  00 22 0f 43 43 53 43 4f  52 43 48 00 00 36 00 1e  |.".CCSCORCH..6..|
00000030  00 00 00 00 00 00 00 3a  00 00 00 00 03 56 11 b9  |.......:.....V..|
00000040  70 dc fe 90 50 48 30 df  eb 39 88 23 8e 88 78 bf  |p...PH0..9.#..x.|
00000050  da ab ab 5b e2 13 98 89  66 eb 94 67 8d 16 00 00  |...[....f..g....|
00000060  00 00 cf 6f 0c 67 85 ec  57 90 e5 c1 ea 8a eb 9f  |...o.g..W.......|
00000070  c8 13 d2 1d 75 bd a5 9f  eb b9 ef 1d 25 79 45 2c  |....u.......%yE,|
00000080  05 bb 74 41 e8 8f 27 6a  01 07 d0 f5 3b 17 ce 87  |..tA..'j....;...|
00000090  7b c2 82 d9 41 6b 82 2f  d8 b8 17 32 fa d3 59 05  |{...Ak./...2..Y.|

I am not sure the best way to handle all the different versions within the PRONOM registry. I went ahead and made a few signatures based on the export dialog of Sibelius 2024. Even with combining a few together, it leaves us with 17 new PUID’s. Maybe further discussion can refine these down a bit more? Regardless, each file can be associated with a specific Sibelius version, making it easier to open and migrate if needed without fear of opening in the wrong version. Take a look at some samples and my signatures on my GitHub page and let me know if there is a better way.

Shorten

Posted on by Thor Leave a comment

I was recently going through some of my old CD-R’s and came across this 11 year old fun memory.

I remember going to this 2003 Toad the Wet Sprocket concert in Salt Lake City with some friends, I had seen this band perform before, but this was the first time I was able to get a recording of the show. Normally having a recording of a concert of a well known band was a little shady, but for some bands, they not only allow recording of their live concerts, but they encourage it. There has been a few bands over the years who have this philosophy, one most have heard of is the Grateful Dead, because of all the tape trading, the band’s numerous concerts will live on forever.

The scene of recording concerts is still alive and well, and if you are into recording and sharing it is expected you share in a lossless audio format. The world of lossless audio is definitely in the minority of all those who listen to music on the daily. Most of us have been placated with the infinite playlists on services like Apple Music, Spotify, and Amazon Music. Most probably don’t care about owning music anymore, but for the few who consider themselves Audiophiles, having a lossless audio file is the only choice.

When it comes to formats, there are a few lossless formats to choose from, they all come with some advantages as well as some downsides. WAV files contain the full PCM audio stream, and while internet bandwidth today can handle full uncompressed audio, it can still be beneficial to use some compression for archiving or sharing over the web.

The most common lossless format today is the Free Lossless Audio Codec or FLAC, but there are also quite a few who like the Apple Lossless Audio Codec. Both offer many advantages, especially with metadata, cuesheets, and can contain cover album art. But many years ago another lossless format was most often used with bootleg recordings and audio sharing.

Shorten was one of the first lossless formats, developed by Tony Robinson in 1993 for SoftSound. It could cut the size in half of a typical 16-bit WAV file. It achieved this by using Huffman coding, kinda the same way a JPEG works, by reducing the frequency of how often patterns occur. Today FLAC and ALAC have replaced this format and offer improved features and support. Many audio players have dropped support for shorten making it difficult to use this old format.

The Shorten format uses the .SHN extension. It is one of the formats listed on the Library of Congress Sustainability of Digital Formats with the ID fdd000199, although a couple links don’t appear to work as it hasn’t been updated since 2011. Support was ended for this format and many of the links found on various websites are for broken, usually referencing the etree wiki. Much of which is archived on the Internet Archive.

Let’s take a look at the what makes up a lossless compressed SHN file. A quick look at a sample header:

hexdump -C test.shn | head
00000000  61 6a 6b 67 02 fb b1 70  09 f9 25 59 52 a4 d1 a8  |ajkg...p..%YR...|
00000010  dd cf 85 5a 01 57 a0 d5  a8 b6 6b 6d d2 41 10 80  |...Z.W....km.A..|
00000020  40 20 10 18 04 0a 01 44  d6 40 20 11 0d 8c 0a 01  |@ .....D.@ .....|
00000030  04 80 44 20 16 4b 0d d2  c3 b8 f8 55 a0 11 80 59  |..D .K.....U...Y|
00000040  98 56 1d b1 79 51 9f 39  f1 12 d2 d3 75 5c cd 08  |.V..yQ.9....u\..|
00000050  06 25 68 6b 52 5e 9f 4c  39 cd c1 32 c4 0d a9 b7  |.%hkR^.L9..2....|
00000060  69 34 56 f0 96 fa 46 89  a2 6e 8c ba d5 d0 58 de  |i4V...F..n....X.|
00000070  f5 44 5b aa 61 82 c7 85  88 37 d6 ee cb ab 4e 44  |.D[.a....7....ND|
00000080  91 19 b7 38 d4 20 ae 98  98 d1 2c 4a 4e 88 dd 3e  |...8. ....,JN..>|
00000090  36 68 1b 59 a8 7d 84 23  76 0a 84 21 a1 cd 80 8e  |6h.Y.}.#v..!....|

The first four bytes seem to be consistent among my samples. It makes me wonder if the ascii values have something to do with the author, Anthony (Tony) J. Robinson. In the source code for the shorten software, the file shorten.h defines the ascii “ajkg” as the magic header for the SHN format. Also found in current ffmpeg code. Although the tools don’t have much to say about them.

mediainfo test.shn 
General
Complete name                            : test.shn
Format                                   : Shorten
Format version                           : 2
File size                                : 3.17 MiB

Audio
Format                                   : Shorten
Compression mode                         : Lossless

ffprobe -i test.shn
Input #0, shn, from 'test.shn':
  Duration: N/A, start: 0.000000, bitrate: N/A
  Stream #0:0: Audio: shorten, 44100 Hz, 2 channels, s16p

Using the older SHNTOOL, we can get more information. 

shntool info test.shn  
-------------------------------------------------------------------------------
File name:                    test.shn
Handled by:                   shn format module
Length:                       0:32.23
WAVE format:                  0x0001 (Microsoft PCM)
Channels:                     2
Bits/sample:                  16
Samples/sec:                  44100
Average bytes/sec:            176400
Rate (calculated):            176400
Block align:                  4
Header size:                  44 bytes
Data size:                    5697720 bytes
Chunk size:                   5697756 bytes
Total size (chunk size + 8):  5697764 bytes
Actual file size:             3325489
File is compressed:           yes
Compression ratio:            0.5836
CD-quality properties:
  CD quality:                 yes
  Cut on sector boundary:     no
  Sector misalignment:        1176 bytes
  Long enough to be burned:   yes
WAVE properties:
  Non-canonical header:       no
  Extra RIFF chunks:          no
Possible problems:
  File contains ID3v2 tag:    no
  Data chunk block-aligned:   yes
  Inconsistent header:        no
  File probably truncated:    unknown
  Junk appended to file:      unknown
  Odd data size has pad byte: n/a
Extra shn-specific info:
  Seekable:                   yes

Many Shorten Audio Files are found out there in archives and file sharing sites, so even though the format isn’t used to create new files, it will still be around for awhile. My GitHub has my signature proposal and a couple of samples.