hppsmtools 0.99
hppsmtools is a program for communicating with the HP PhotoSmart and Konica Q-EZ digital cameras. I've written it because I don't (and don't want to) have Windows and wanted to access the camera anyway. Quite possibly it will work for you as well.
In the meantime, of the order of 40 people have provided feedback about using this program, and most of them were happy with it. I, on the other hand, was not when I got myself a notebook with a somewhat flaky serial port and so I gave the program an overhaul. Since I also had an chance to glimpse at the protocol docs, this program should be much better behaved than its predecessors. In particular, doing a ^C should be safe in most cases.
In January 2000, Jeff Schaller <schaller+psmget@southwind.net> did some good things to this program. Among other things, he made it work with the PhotoSmart C30 and made the source look so pretty that I'm no longer exactly ashamed of it. Later, he convinced me to split the binaries and do a big cleanup, and in the end did most of the work. What you see now is the result of that effort.
In December 2002, Brian Goines provided man pages for the four programs. Thanks, Brian.
Installation
You may want to change PREFIX (and ADDIT_DEFS if you are on a libc5 system) in the Makefile, and/or config.h (see below) before compiling.
Doing make will hopefully produce the binaries psmget, psmdir, psmdel, and psmsho, and make install will put them into $(PREFIX)/bin, the man pages will go to $(PREFIX)/man/man1. That's it, nothing else gets dumped into your file system.
Psmget does not have any configuration files. However, if the environment variables PSM_BPS and/or PSM_PORT exist, it will use the values in there as the bitrate and the port used to talk to the camera. Built-in defaults (defined in config.h in DEFAULT_PORT and DEFAULT_BITRATE, in case you want to change that) are 115200 and /dev/ttyS0, respectively. Legal values for PSM_BPS are 19200, 38400, 57600, and 115200, PSM_PORT has to contain the full path to the device.
Usage
The following options are available in all four programs
-p <port> \ These two override the settings -r <bitrate> / in PSM_PORT and PSM_BPS
-t sync the camera's clock with the host computer's -h print a short help
In addition, psmget and psmdir understand
-n <string> sets a prefix for the names of the files written
(For psmdir, the idea is that you can do
psmdir -n blabla > index.txt and then comment the
images with the proper names filled in)
psmget and psmdel understand
-i the number in the parameter(s) are image ids as opposed to
image numbers. Basically, you can ignore this option
unless you use psmsho, when you have to use -i with the
number psmsho returns.
psmget understands
-b get thumBnails instead of full size images. The file names
are then *.tbn.jpg. If you have a better idea, tell me...
psmsho understands
-f don't switch off the flash (once you switched off the flash,
it will remain off until you switch off the camera)
psmdir understands
-w don't show directories, give some not so terribly interesting
numbers about the camera
psmget {<number>}|all reads the numbered picture and puts it to stdout
or to files <prefix><image id>.jpg if -n was given. Thus,
psmget 1 >test.jpg
reads the first picture to the file test.jpg, whereas
psmget -n test 1 2
writes pictures 1 and 2 into the files test01.jpg and test02.jpg.
psmget -n today all
reads all pictures in the camera.
psmdel {<number>}|all deletes the given pictures
psmdir prints a directory of the camera with dates (in the rather odd form the camera gives them) and the image quality.
psmsho instructs the camera to take a picture. By default, the flash is turned off by this command. You may try to do some kind of a webcam with this command, I won't bet how long your flash card will take this, though. psmsho returns the image id of the picture just taken (or -1 on failure, whatever your shell makes of that)
You probably don't want to read the following two paragraphs.
There is a catch -- when you communicate with psmget, you'll use image numbers, counting the images from 1 to the number of images consecutively. The camera, on the other hand, deals with image ids, numbers that are usually identical with the image numbers but may differ when you have deleted images.
Now, the hppsmtools try to hide these image ids from you, but psmsho returns an image id (because I couldn't find an easy way to find an image number from an image id), and therefore the dreaded -i option exists. Also, just to confuse you, the image ids are displayed in the download messages generated by psmget, and of course they are in the exif headers.
Your probably want to forget the last two paragraphs.
Examples
Get all images in the cameras, naming them mypic1.jpg through mypicnn.jpg:psmget -n mypic all
Get image 15 and write it to file gronk.jpg:psmget 15 > gronk.jpg or psmget -n gronk 15 (actually, the second form will write to gronk15.jpg)
Take a picture and write it to test.jpg (in (t)csh):
psmsho; psmget -i $status > test.jpg
the same, only delete the image after the fact
psmsho; set tmpvar=$status; psmget -i $tmpvar; psmdel -i $tmpvar
(notice the i)
Look at the directory of the flash card and set the time while being at it: psmdir -t
Look at the directory with the names that the a psmget -n would store
the pics in filled in:
psmdir -n mypic
If it doesn't work
Check the permissions on the device the camera is attached to and all the usual things. If you may read and write to it, it's serious and you should really contact me. Which you should do anyway, or else this program will always only do what I need and will always be the mess it is now.
If it works
Please contact me anyway.
Non-Legalese
Since it looks like this program is almost serious, I decided to go for a serious license: the GPL. If you want to lift the exif stuff, do so, I won't sue you even if it ends up in a closed-source thing. Otherwise, read COPYING.
I did not re-engineer or disassemble programs from either HP or Konica to write this program (quite obviously). All the information that went into the program come entirely from watching the camera and the software HP bundles with the camera talk to each other (and from a glimpse I was allowed to take at the HP's docs).
To-Do
Probably still a lot. I'll have to look into portability issues. Wishes may be directed to me and will probably be encouraging. Bug reports, too. If you send in bug reports, core dumps or debug logs (set MASTERDEBUG in psmget.c to 1, recompile, catch stderr into a file, and send it to me), even strace logs will be helpful.
The compression ratios for the C-30 are very weird indeed. I doubt that this is correct, and if so, than HP's (or whomever's) engineers have messed up royally. Any tips on where the errors are is greatly appreciated.
Updates
Check out http://unimut.fsk.uni-heidelberg.de/demi/photosmart
Contacting me
Write an E-mail to
msdemlei@users.sf.net
Dec 2002