Viewing or converting BIF images
Marcus E. Hennecke
I have a set of images with extension .bif which I would like to view
or convert to some more standard format. It seems to be a rather
simple file format however I can't seem to find any tool able to view
it. I have been searching the web and the newsgroups but haven't been
able to find anything of use. It does not seem to be a BIFF format nor
the byLight image format. Rather it seems to be a binary image format
but I can't find a description of this format. I have only found two
viewers claiming to read bif files. One is 20/20 which crashes the
instant I try to open one of my BIF files and the other is Nico's
viewer which does not recognize the format either. Looking at the
files with a hex editor I see that the format seems to be very simple.
It starts with the three letters "BWC" followed by a zero byte
followed by two two-byte integers which I assume are in little endian
format:
0000000 42 57 43 00 c0 0c 00 0a f0 0f 00 f0 0f 00 f0 0f
0000020 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00
...
All four files that I want to view start with this exact same sequence
although the sizes are not exactly identical. The file sizes vary
between 152069 and 183815 bytes.
Does anyone know what kind of format this is and how I can view it? It
is probably quite old as the files are from 1993. They should
represent scans of papers from the proceedings of the ICASSP '93.
Thanks in advance for any help,
Marcus
SoftSpot
The biff file format is described on their 2d-graphics page. Maybe they can
help you further in your search for a piece of software that can read those
files.
One warning though: the file is "expired" some time, I don't know if the
authors still are willing to answer questions.
"Marcus E. Hennecke" <marcus....@temic-sds.com> schreef in bericht
news:40303dde.03050...@posting.google.com...
Marcus E. Hennecke
> You might try a search at http://www.dcs.ed.ac.uk/home/mxr/gfx/
> The biff file format is described on their 2d-graphics page.
Thank you very much for your reply. In my search on the internet I
have come across this spec, but that doesn't seem to be it. BIF and
BIFF do not seem to be the same thing. The format these images are in
seems to be rather simple.
> "Marcus E. Hennecke" <marcus....@temic-sds.com> schreef in bericht
> news:40303dde.03050...@posting.google.com...
> > Hi there,
> >
> > I have a set of images with extension .bif which I would like to view
> > or convert to some more standard format. It seems to be a rather
> > simple file format however I can't seem to find any tool able to view
> > it. I have been searching the web and the newsgroups but haven't been
> > able to find anything of use. It does not seem to be a BIFF format nor
> > the byLight image format. Rather it seems to be a binary image format
> > but I can't find a description of this format.
> > [...]
SoftSpot
maybe the people at http://www.mindworkshop.com/alchemy/alchemy.html can
give the answer? In any way, a visit to their site may prove useful for you
can download their sharewareprogram "graphic workshop" for an uncrippled try
out. That program can convert almost any graphical format in almost any
other. And in case of an unknown format it also can make an "educated guess"
which in a numerous number of cases leads to a useful result.
Besides of that, the people of Mindworks are busy with the workshop and
grafical formats for a considerable time (my first contacts with the program
dates from the '80s) and maybe they have stumbled on your BIF-files already
and can help you further?
Marcus E. Hennecke
"SoftSpot" <bobob...@hotmail.com> wrote in message news:<3ebf8660$0$11154$ba62...@reader0.news.skynet.be>...
> Another hint:
> maybe the people at http://www.mindworkshop.com/alchemy/alchemy.html can
> give the answer? In any way, a visit to their site may prove useful for you
> can download their sharewareprogram "graphic workshop" for an uncrippled try
> out.
Gee, this format is quite stubborn. When I try to open one of my BIF
files or try to "Identify mystery files", Graphic Workshop crashes
immediately.
I will try to send them my question, perhaps they can help.
Thanks again!
Marcus
SoftSpot
suitable viewer here, but since I don't have any bif-file, I can't check it
out.
"Marcus E. Hennecke" <marcus....@temic-sds.com> schreef in bericht
news:40303dde.03051...@posting.google.com...
SoftSpot
I received a BIF-file from Marcus, it seemd to be in good order. Checked it
out in various viewers, convertors and painting programs running under DOS,
Win 3.X, W95, W98, Win ME, Win XP (in a grand total I checked about 30
different programs in these OS's, dating from 1988 - 2002) but only one
(SVGAshow, a DOS-program) recognized just the extension, it told it was a
320 x 240 x 256 grayscale file. But in an attempt to show it on screen it
only showed a random B&W filled graphic screen.
Last night I switched over to another PC where I was able to run a Linux
environment, and checked out whatever programs I have there. Also without
any succes. Although I must say that on the Linux-field I am a newby. The
only working Linux version I have is a DVD-based Knoppix version with only 3
graphic programs. All three of them gave a negative result.
Searching on the net directed me into the suspicion that the BIF-file format
might be originally an UNIX-format. So maybe there is a possibility there
which might prove more succesful?
Anyway: I am through my resources. Sorry, but I can't do anything more.
Who's next?
Nospam9212
(Marcus E. Hennecke) wrote...
>I have a set of images with extension .bif which I would like to view
>or convert to some more standard format.
Looking at "formats" page, shows "byLight" - .bif can be read by XNView.
Worth a shot.
-= Francis Yarra =-
fyarraATjunoDOTcom
http://members.aol.com/fyarra001/ads - My drywall website
http://members.aol.com/fyarra001 - My C64 website
http://members.aol.com/prsnl99 - My personal website
SoftSpot
the one bif I have. And alas: "file type can't be established"....
Thanks for the hint, but apparently the presence of a certain extension
doesn't guarantee the structure of a certain filetype...
"Nospam9212" <nospa...@aol.commune> schreef in bericht
news:20030518012157...@mb-m10.aol.com...
Marcus E. Hennecke
> On 5/7/03 11:04 AM Eastern Daylight Time, marcus....@temic-sds.com
> (Marcus E. Hennecke) wrote...
> >I have a set of images with extension .bif which I would like to view
> >or convert to some more standard format.
>
> http://www.xnview.com/
>
> Looking at "formats" page, shows "byLight" - .bif can be read by XNView.
Thank you very much for your help. I had tried this program as well as
20/20 before but as I said, this is not a "byLight" format.
The first four bytes of the file are "BWC" and a zero byte, making it
look like a C-string. I can imagine this could be an acronym for
"black and white compressed". After all, these images are supposed to
be scans of conference proceedings. At only 152kB for a whole page and
each file being a different size I suppose some kind of compression
must have been used. In my internet search I have found the following
explanation for .bif (among many others):
BIF Image Capture board bitmap image (b&w Binary Image Format)
see for example http://www.file-ext.com/b.html
Unfortunately, I haven't found a web page describing this format or a
viewer able to display it (at least none running under Windows, MacOS
or Linux).
Thanks again for your help,
Marcus
Alan Tipping
images that use a 256-colour palette. It supports multiple frames.
Go to this site -
http://www.totalwar.org/Downloads/Tools.shtml
and download the file Bif/Buf viewer (version 2.2a). Its a simple
bif viewer/editor but has nice facility that allows extracting all bif
frames to windows bmp's - and vice-versa.
Might help.
Alan
"SoftSpot" <bobob...@hotmail.com> wrote in message
news:3ec37e2f$0$11136$ba62...@reader0.news.skynet.be...
SoftSpot
Well, I tried it. And all I got (system running under XP home) was a note
that the program runs "out of memory while expanding memory stream"... And
if there's one thing I have enough of here, it is memory, of that I am
absolutely sure....
So: who or what is next?
"Alan Tipping" <alan.t...@wxs.nl> schreef in bericht
news:bbm258$2pf$1...@reader08.wxs.nl...
Alan Tipping
Sounds like the height/width and number of frames fields in the bif
have too big values (which is why you get the memory error).
How big is this "bif" file in total?
If it's not too big (less than 1 meg) E-mail me and I'll take a look
inside. Maybe I'll recognise something, maybe not.
regards
Alan
"SoftSpot" <softspo...@hotmail.com> wrote in message
news:3ededc2e$0$264$ba62...@reader0.news.skynet.be...
Marcus E. Hennecke
Wow, I don't check the newsgroups for a while and look what has
happened in the meantime:
"Alan Tipping" <alan.t...@wxs.nl> wrote in message news:<bc06am$rre$1...@reader08.wxs.nl>...
> If it's not too big (less than 1 meg) E-mail me and I'll take a look
> inside. Maybe I'll recognise something, maybe not.
Thank you very much for your help! Looks like we are finally getting
somewhere.
Sorry for not replying sooner, I was away on a business trip last
week. I did check the newsgroups via web access and saw your message
and sent you a short mail with one of the BIF files. Now that I'm
back, here the longer reply:
> "SoftSpot" <softspo...@hotmail.com> wrote in message
> news:3ededc2e$0$264$ba62...@reader0.news.skynet.be...
> > Well, I tried it. And all I got (system running under XP home) was a note
> > that the program runs "out of memory while expanding memory stream"... And
> > if there's one thing I have enough of here, it is memory, of that I am
> > absolutely sure....
The same thing happened to me under Windows 2000.
> > "Alan Tipping" <alan.t...@wxs.nl> schreef in bericht
> > news:bbm258$2pf$1...@reader08.wxs.nl...
> > > Go to this site -
> > >
> > > http://www.totalwar.org/Downloads/Tools.shtml
> > >
> > > and download the file Bif/Buf viewer (version 2.2a).
The nice thing is that it can also create BIF files. I did just that
and it seems that the format it creates is almost the same I have.
Looking with a hex editor:
My own file:
0000000 42 57 43 00 c0 0c 00 0a f0 0f 00 f0 0f 00 f0 0f
0000020 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00
The file created by readbif:
0000000 00 00 43 00 c0 0c 00 0a f0 0f 00 70 0f 00 f0 0f
0000020 00 70 0f 00 f0 0f 00 70 0f 00 f0 0f 00 70 0f 00
The two files differ in the first two bytes (42 57 or "BW" in my file
vs. 00 00 in the created file). Since my files are supposed to be
scans of papers I would assume them to be black and white which seems
to make the "BW" part not unreasonable. The file I created on the
other hand was a color file (hadn't figured out yet how to create a
black&white file with readbif).
The next six bytes are identical. What follows after that is what I
assume the color table and it differs between the two files. I
downloaded the sources of readbif and I noticed that it seems to
assume a fixed length for the color table (up to the 512th byte of the
file). However, the table in my files seems to be longer than that. I
am not sure why a black&white image would require a longer color
table, but these are my observations:
My own file:
0001120 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00
0001140 f0 0f 00 f0 0f 00 f0 0f 00 f0 0f 00 fe 32 f0 ed
0001160 52 00 7f 12 c0 01 34 e0 00 00 00 25 e0 00 00 00
Here the color table (compare with the first lines above) seems to end
at octal 1153, which is the 620th byte.
The file created by readbif:
0000760 00 70 0f 00 f0 0f 00 70 0f 00 f0 0f 00 70 0f 00
0001000 00 01 00 00 00 01 00 00 00 f0 0f 00 01 00 00 00
0001020 01 00 00 00 04 00 00 00 06 00 00 ff 00 01 06 00
In this file the color table ends at octal 777 which is the 512th
byte.
Of course, the question remains what do you do with a color table
which only contains equal entries? And why does a black and white
image need a color table with three bytes per entry? Why does readbif
assume a color table with 168 entries and why do the color tables in
my BIF files contain 204 entries? How does one determine the length of
the color table anyway?
Thanks again for your help!
Marcus
Alan Tipping
In a BIF file the 1st 512 bytes is the colour palette - 2 bytes per RGB.
In a BIF colour palette the RGB bytes are truncated. R and B colours
have the last 3 bits stipped whilst G has the last 2 bits stripped. Hence
an RGB colour (3 bytes) of 125,89,230
in binary -
R = 0111 1101
G = 0101 1001
B = 1110 0110
would be stripped to -
R = 0111 1
G = 0101 10
B = 1110 0
giving a 2-byte RGB of -
0111 1010 1101 1100
Then the 2 bytes are swopped (low order 1st) to give -
1101 1100 0111 1010 .
This is why if you open a file in Readbif and look at the RGB values in the
palette then R and B are always multiples of 8 whilst G is always a multiple
of 4.
Also the 1st RGB in the colour palette is always the transparent colour (I
think!).
The 1st 2 bytes in your file ("BW") appear to be incorrect. If some scanning
software has inserted these 2 bytes that would throw off the colour palette
and
the header (which immediately follows the palette) which is probably why you
get the "out of memory" (the height/width and number of frames fields in the
header would all be in the wrong positions).
Try taking a copy of your file and stripping off the 1st 2 bytes. Does
Readbif.exe
read it now?
Can't remember much about the header info but I'll look into this. If you
look at the
site that you got "Readbif.exe" from, the source code for this program is
also
provided and that may help you decipher the header info.
regards,
Alan
"Marcus E. Hennecke" <marcus....@temic-sds.com> wrote in message
news:40303dde.03061...@posting.google.com...
SoftSpot
I tried what you advised: strip off the first two bytes of the bif-file
(used a hex-file editor to achieve that and checked out both files, original
and stripped with the hex-editor to check if all had been done correctly)
and then try to open it with readbif, but it all resulted in the same: out
of memory error.
Any more hints on what may be done?
SoftSpot
Alan Tipping
Yeah - I just tried it and it does'nt work :(
Marcus sent me an image of the "presumed" BIF file and
I'm playing around with it.
To my knowledge there is only 1 BIF file format; that supports
both single and multi-frames. Eg: LBM files have several formats
that are indicated in the header (ILBM , PBM etc:) and different
compression algorithms are used depending on the format - a real
pain in the jacksie.
Therefore if Marcus's files are really BIFs we should be able to
puzzle this out.
One of the main things to consider is that most other
graphics file formats that use a 256-colour palette have
an order of HEADER/PALETTE/BODY.
BIFs have an order of PALETTE/HEADER/BODY which
can be confusing when viewing a BIF via a hex editor.
Very interesting problem. Always good to be able to crack a
problem that is not bog-standard as such. I'll keep with it - but
there is no guarentee Marcus's files truly are BIF's.
regards,
Alan
"SoftSpot" <softspo...@hotmail.com> wrote in message
news:3ef31127$0$308$ba62...@reader0.news.skynet.be...
SoftSpot
several filetypes, but I do have a long experience in handling graphics. It
seems to me that Marcus' BIF-file is a B&W file, and that would mean that it
has a different structure than the files you know.
It would be very convenient if Marcus would be able to give us a hint or a
clue to the program which generated the files, since they seem to be of age
already. Unfortunately it seems he can't... I guess they originate from a
DOS-program rather than anything else, that's why I started my search with
all the DOS-graphical viewers, convertors and painting programs I know and
have.
Marcus E. Hennecke
"SoftSpot" <softspo...@hotmail.com> wrote in message news:<3ef34273$0$1039$ba62...@reader1.news.skynet.be>...
> You fetch my thoughts perfectly. I'm not such a crack on the definitions of
> several filetypes, but I do have a long experience in handling graphics. It
> seems to me that Marcus' BIF-file is a B&W file, and that would mean that it
> has a different structure than the files you know.
> It would be very convenient if Marcus would be able to give us a hint or a
> clue to the program which generated the files, since they seem to be of age
> already.
Well, these are papers from the 1993 International Conference on
Acoustics, Speech and Signal Processing (ICASSP). It was the first
time they handed out proceedings on CD which is why they gave them
away to the attendees for free. I already contacted the IEEE Signal
Processing society but they weren't much help (giving me lectures
about copyright issues instead).
Anyway, the CD does contain executables to look at the proceedings.
However, if one starts them, the monitor goes blank and they leave the
graphics card in a rather unusable state. Took me a while to get it
back to working again. Of course, I could try to send you the
executables but the problem is that they also load all kinds of
indices so that one can more easily browse the papers (or rather:
images). I couldn't really identify the one executable doing the
actual displaying (if there is one). What I can try to do is send you
just the files which I think might be relevant for the executables to
work (can't try it out, though). But caution: It might mess up your
graphics card.
Thanks very much for your help!
Marcus
SoftSpot
to get the stuff going.
As for the risk of "messing up my graphic card", err, well, I'm willing to
take that risk.
(maybe I'll manage to get my old PC back in shape and running, then I can
use that as an experimental station: if that one messes up, there's no harm
done) ;-)
SoftSpot
SoftSpot