BASELINE RELEASE FOR AUGUST 11, 1999 Eclipse is Version 2.2.2
Before launching into a summary of the enhancements to UMBRAPHILE in this release, I wanted to thank Edwin Aguirre for his prominent, though unexpected, mention of UMBRAPHILE in his July, 1999 Sky & Telescope article entitled "Imaging Totality. As a result, my personal e-mail inbox was FLOODED with queries, comments, questions, and annoyances. To wit: "What do you MEAN you don't have a Windoze {my edited spelling} version. EVERYONE uses Windoze!" Sorry, but not those of us who "Think Different". That said, however, I would be VERY happy to see an UMBRAPHILE clone ported to the PC world - "for the rest of them".
To the vast majority, however, who offered many very constructive ideas
(and whom I hope will become both users of the program, and eclipse-chasing
friends), I thank them whole-heartedly for their suggestions and now offer
UMBRAPHILE 2.2.2. I apologize to those to whom my replies were delayed,
or if one of their specific requests did not make it into this release.
My "real life" schedule has been extremely constraining (UMBRAPHILE is
a labor of love, but it does not pay the mortgage). I believe this
release incorporates 99% of the requested enhancements. I don't plan
on adding anything else before the August 11, 1999 eclipse, as I want to
"freeze" the current implementation to allow time to fix and test any "bugs"
should they be found. I have tested this version extensively, and
believe it
is sound. However, should you find any problems PLEASE let me
know so I can provide an update before the next eclipse.
As always, wishing you clear skies, and good seeing.
Glenn Schneider
gschneid@rtd.com
------------------------------------------------------------------------
Previous release: 2.1.1 (and 2.1.2a Beta)
What's New & Improved in 2.2.0?
1. UMBRAPHILE has been recompiled to optimize performance on G3 processors (PPC version of the program), without compromising the execution speed on PPC 601-604 machines. There are not separate G3 and PPC releases, as UMBRAPHILE PPC will execute on both, but in an optimized fashion on G3's. There is a separate release of 68K machines (UMBRAPHILE 68K) which has all of the functionality, and runs identically, to the PPC version (albeit slower) for users of older Macintosh's. I intend to continue supplying and supporting the 68K version indefinitely, so do not worry about it becoming obsolete.
2. The new G3 Powerbooks caused UMBRAPHILE to get "confused" when trying to access the serial port, and would simply not work. In previous releases UMBRAPHILE was "hard wired" to use to use serial port 1 (aka, the "modem" port), which was the serial port identifier for all pre-G3 PowerBooks. The G3 laptops sole serial port is number 2, the "printer port". As a result UMBRAPHILE was unable to set up the serial port on G3 laptops. In release 2.2.0, the user is free to select which serial port to use for the camera controller (in multi-port machines), or specify the available port for different model laptops. UMBRAPHILE will check that the selected port is actually available, and will let you know if you need to select the "other" port. You will see this in expanded dialogs which are used to set up the serial port for the controller.
3. UMBRAPHILE now provides a serial port "test" mode, where a test waveform (exposure sequence) is output through the serial port. This will allow users who are custom building camera interfaces an easier way to develop and test those interfaces without running the controller through an actual exposure sequence table. The "test" mode provides two new capabilities. First, you can specify a series of identically spaced exposures, of uniform duration, to be repeated for a specified time. This way you do not have to run the exposure sequence table, (i.e., in a "simulation" mode for the eclipse). In this mode you specify the exposure time (which is repeated for each exposure), the time between exposures, and the duration of the test. This will help you to unambiguously test the performance of the system and interface you are building to use with the UMBRAPHILE software. You will see a new TEST SERIAL PORT menu item under the UMBRAPHILE menu which allows you to do this. You can also define the communication port parameters in any way you wish. See item 4 for more details.
4. Several UMBRAPHILE users are now taking advantage of simple, direct, electronic shutter trigger provided on many cameras, rather than building an electro-mechanical interface for external solenoid activation of the shutter. Different cameras, however, expect different kinds of input signals (waveforms and voltages) to trip their shutters. This requires some experimentation and inventiveness on the part of the user to appropriately modify the control signal delivered by UMBRAPHILE and the originally suggested electro-mechanical interface. This usually can be done fairly simply in TTL logic with just a few components (e.g., inverters, triggers, flip-flops), and/or through RC filtering. Because of the diversity of interface requirements for different cameras there is not a "generic" circuit which will do this. To simplify the task for the experimentalist, however, UMBRAPHILE 2.2.0 gives you much more flexible control over the output signals delivered out of the serial port. The "default" is to use the scheme UMBRAPHILE has always used (described elsewhere). However, you can now program the serial port by specifying the 7/8 bit data word to be sent, the number of stop bits, the data parity, and the character translation mode to be used on the "Send Data" output line. You can also select full or half duplex communication (though I can't figure out why you would want to). In addition UMBRAPHILE 2.2.0 will condition the "Ready to Send" line with optional Xon/Xoff handshaking. This is useful for those working with gated signals. Note that some (not all) DIN-9 to DB-25 connectors don't implement this line. So if you plan on using it you may need to pull the signal off of the DIN-9 connector. The only thing you cannot adjust is the Baud rate. UMBRAPHILE sets up and uses a 9600 Baud output rate. This is fundamental to UMBRAPHILE setting up the exposure timing and sequencing correctly. Finally, you needn't worry about all this unless you are putting together a custom electronic camera interface.
5. UMBRAPHILE now allows you to save and restore all of the various settings in the dialogs it uses. This way you do not have to manually set up a lot of parameters each time you start UMBRAPHILE anew. You will see new SAVE SETTINGS and GET SETTINGS items under the UMBRAPHILE menu. When you SAVE SETTINGS all of the settings in place for all of the dialogs are saved to a file of your choosing, which can be restored (via GET SETTINGS) any time you wish (except when the controller is actually running).
6. UMBRAPHILE will now allow you to SAVE the content of the various text windows (line circumstances, local circumstances, exposure time table) to a plain ASCII Text file. Just select SAVE WINDOW from the UMBRAPHILE menu, then select the name of the window whose content you wish to save.
7. If you build an exposure sequence table (RUN CONTROLLER), but decide
not to execute it by selecting CANCEL in the window in which you set a
UT offset prior to execution, the exposure table window will now be left
on the screen (rather than taken down). This will allow you to save
the table to a file without executing it. You can still dismiss the
table by clicking on the close box.
What's New & Improved in 2.2.1?
A few bugs squashed, and some minor enhancements...
1. UMBRAPHILE now CORRECTLY handles setting the Universal Time derived from the system clock with an offset. Previously if the system clock time was earlier then the desired target time, a negative offset did not produce the correct result. Now, if you enter an offset in the UT = System Clock +/- [hhmmss.f] dialog it will be correct.
2. Somewhere along the way the UMBRAPHILE controller would not build exposure sequence tables for other than 36 exposure sequences. This has been fixed and exposure sequences for ANY even number of exposures can be built.
3. A few users complained that the audio alarm which sounded at 10 seconds before second contact was too long. It now will beep the speaker only 10 times.
4. An audio alarm (single beep) has been added to the serial port test
routine, every time an exposure is triggered.
What's New & Improved in 2.2.2?
1. UMBRAPHILE 2.2.2 now allows you to specify a minimum exposure time for the series of evenly spaced exposures during totality, rather than a hard-coded default of 0.01 seconds. This is specified through a new tunable parameter, MINEXP, in the UMBRAPHILE data file. UMBRAPHILE 2.2.2 *REQUIRES* this parameter to appear in the Camera Controller section of the data file to run the Controller. Older data files which do not have this parameter, must be edited to add this parameter. A new data file template, for the August 11, 1999 eclipse is included with the software. When building an exposure sequence table, UMBRAPHILE will define a series of NUMEXP/2 - DRNUM uniformly spaced exposures from Contact 2 to mid-eclipse. Exposure times will be logarithmically ramped from the minimum exposure time to the maximum exposure time, such that each successive exposure is longer by a constant multiplicative factor (the logarithmic base) then the previous one. The sequence is repeated in reverse order from mid-eclipse to third contact.
2. In version 2.2.1 and earlier the first exposure after the CII diamond ring was not actually the default minimum of 0.01 seconds, but was the ramp sequence logarithmic base times 0.01 seconds. This has been fixed with the generalization of a user definable MINEXP (item 1).
3. UMBRAPHILE now checks for consistency between the controller parameters DRNUM, DRSEQII and DRSEQIII, and gracefully reports an error if an inconsistency is found.
4 . Item (2) in 2.2.1 was not quite correct, in that UMBRAPHILE would handle any sequence of even number exposures of 36 or less, but not greater than 36. In version 2.2.2 exposure sequence tables are not limited to 36 exposures.
5. The exposure sequence table, for sequences > 36 exposures, is now
presented in a scrollable window. Scrolling can be done at any time
EXCEPT when an exposure is actually being taken. Scrolling while
the exposure sequence table is being executed is allowed, but the white/black
annotation inversion of executing exposures may not be correct. The
annotation will resynchronize after the scrolling has stopped.
Acknowledgements
I want to thank to Dan McGlaun, Terry Falcon, Bert Hartig, Sue Preston, and Heinrich Steinmetz for their suggestions, comments, and beta testing leading to this release. Have a great eclipse!
Glenn Schneider
sofTouch APpLications
http://rtd.com/~gschneid/sofTouch.html