• The instrument setup (Instrument -> Setup... menu) will now update the Phi position if the "Initialize Goniometer" checkbox is checked. Previously, if the position was changed outside of JDirector and this setup was performed, the position would not be updated in JDirector.

• If an axis move timeout error is detected, an attempt to resync the positions of the XYZ stage and goniometer axes will be made. This can help avoid certain situations where JDirector and CameraMan think the axes are at different positions than they actually are.

• The Error Messages dialog now lists errors with a timestamp and in reverse chronological order so that the most recent error is at the top.

• A new built-in job sample step called VerifyMount has been added which attempts to verify that the mount was successful. In particular, an attempt is made to ensure that a robot collision did not occur during the mount. If a collision may have occurred, the job is aborted. This reduces the chance of a rogue robot job that destroys all the samples by mounting them on top of each other, one after the other, or by failing to mount them in some other way.

• The Crystal View "Save Camera Image As..." pop-up menu item operation will now display an error message to the user if no suitable image writer was found based on the name of the destination file. Previously, an exception was thrown if the file had no extension, or, if no image writer was found for the extension, a zero-length file was written, and the user was not informed that no suitable writer was found.

• Increased the size of the manual diffraction center result history from 10 to 20. This is helpful for projects where there's a need to compare rasters when there might be multiple sites of interest or when regions might be rasterized again with a smaller cell size.

• Improved the diffraction center moves and syncs of the XYZ stage and Phi. In the auto-mode diffraction center, the move or sync of Phi was not interruptible, so if it hung, there was no way to stop it. This has been changed to be interruptible. In the manual-mode diffraction center, when the XYZ stage or Phi were moved or synced, it did not wait for those operations to complete. This has been corrected.

• Added support for fly-scanning. Instead of doing a step-scan for each row or column of a raster, support a fly-scan for each row or column to reduce the time it takes to rasterize a raster.

• Added support for computer cluster scoring. To speed up raster cell scoring, support scoring on a computer cluster.

• Added configuration file support. Support reading a system-wide, user, and command-line-specified configuration file. Remove the cluster-key and cluster-nodes command line options; these are now specified in the configuration file.

• Added imcadr(1) and imcadr.conf(5) man pages. Reduce the imcadr --help output to a simple usage message since the arguments are documented in imcadr(1).

In the meantime, a dead pixel mask has been incorporated in the detector server to block out this region of the image, so no additional work needs to be done from the users standpoint for processing.

A final resolution to this problem is being determined.

If there are any questions/concerns, please contact the staff.

The following command can be used in autoPROC to mask out the bad chip:

autoPROC_XdsKeyword_UNTRUSTED_RECTANGLE="798 859 2429 2526"

or use the following line in your XDS.INP file:

UNTRUSTED_RECTANGLE=798 859 2429 2526

An example image is available here: https://www.imca.aps.anl.gov/~battaile/IMCA/dummy_0001.cbf

If you are using autoPROC, this is the command line option you need to use to mask out the bad module:

• Only allow the characters "a-zA-Z_.-" in manual diffraction center "raster images folder" path elements. This avoids cryptic diffraction center errors caused by the inability of the Dectris Pilatus camserver software to handle paths containing certain characters (e.g. the space character). (The set of characters the camserver software can handle is undocumented.)

• Add NumThreadsPerProcessJob configuration option to control the number of threads to use per dataset processing job. This enables more efficient use of the available computing resources.
• Add "autoPROC_CreateGpxPictures=no" to the autoPROC "process" invocation. This keeps processing fast even if autoPROC supports and defaults to the creation of pictures of diffraction and predictions.

• Added support for a polygon raster shape (instead of only a rectangular shape) in auto-mode diffraction centering. This makes diffraction centering faster by skipping raster cells that lie outside the crystal loop and hence do not need to be scanned.

• Added support for entering X-ray beam transmission instead of attenuation for users who would prefer to think in terms of transmission. This applies to the scan editor, the manual diffraction centering configuration, and the job sample table.
• Fixed the job sample table to use correct detector distance limits. Previously, the limits were hard-coded which was wrong. Now the limits are obtained from the instrument server.
• Improved log messages. This involved fixing a case where an undefined value was printed as null, where extraneous trailing punctuation was used, and where an exception trace was unnecessarily printed for a diffraction center that terminated due to a low-score cutoff.

• In auto-mode, instead of saving camera snapshots from the centering and before and after steps directly in the sample directory, they are now saved in a "camera" subdirectory. This makes the sample directory less cluttered and more uniform.
• Added support for saving ancillary files ("camera", "diff-center", etc.) with the data in auto-mode. This behavior is off by default but can be enabled via an option in the Job Control options pane. This is useful for mail-in when collecting for more than one user and want the ancillary files to be saved with the data. In this mode of operation, the recipe usually contains a path not relative to the sample image directory (i.e. user-defined absolute path or relative to the working directory) and using escape characters. This can also be used if a user wants to collect their data to a different location than the default one produced by auto-mode.
• In auto-mode, for the ancillary "camera" and "diff-center" subdirectories, if they already exist, they will now be renamed out of the way first. They are renamed to a name ending in ".old" or ".oldN" where N is an integer >= 2. This makes the latest camera and diffraction center images always appear in the "camera" and "diff-center" subdirectories. The previous behavior for "camera" was to just overwrite the images. (The "camera" subdirectory did not exist previously; it is new in this release.) The previous behavior for "diff-center" was to create a new "diff-centerN" directory where N was an integer >= 2 if "diff-center" already existed. With that approach, the path to the latest diffraction center images was not stable.
• Disabled mediaLib in order to prevent the message "Could not load mediaLib accelerator wrapper classes. Continuing in pure Java mode." from being written to the standard error stream the first time one did something involving reading an X-ray diffraction image (e.g. loading an image into the Image Viewer view).

• Fixed an issue where the log file of the autoPROC process program's output to its standard output stream was not being saved. When invoked in a terminal, process creates a log file (e.g. process.20140910_131055.log) of what it writes to its standard output stream. But when invoked as a subprocess of the aap program, it does not create this log file. Now aap creates this log file for process to achieve the expected behavior.

Aaproc (Auto Auto-Process) is a suite of programs for automatically auto-processing X-ray crystallography datasets. It includes a program, aap, for finding datasets on the file system and auto-processing them using external programs (e.g. autoPROC). The aap program writes a summary of dataset processing to a file named aap-result.html in the current working directory.

Aaproc is installed on the processing server (i.e. plankton). If we typically want to auto-process every dataset as it is collected at IMCA-CAT, Aaproc might be useful to us. An easy way to use it is to collect all datasets to their own directories somewhere below a particular directory tree. Let's say we collect all datasets under /depot17id/luke/20140725. Now simply do the following at a command shell:

$ cd /depot17id/luke/20140725 $ aap

The aap program will run indefinitely, detecting datasets as they are collected, will attempt to auto-process the datasets using autoPROC, and will write a processing summary to /depot17id/luke/20140725/aap-result.html as it goes. This will work equally well for data collected in manual or auto mode.

Documentation is available in the form of Unix manual pages. We can view the aap program manual by typing the following at a shell:

$ man aap

Similarly, we can view the aaproc.conf configuration file manual by typing the following at a shell:

$ man aaproc.conf

• Fixed a bug in auto-mode where pressing the Abort button at the bottom right of the JDirector window would not abort the currently running auto-mode jobs.
• Reclassified some auto-mode diffraction center log errors as warnings so that an error dialog is not displayed. The most common case where this is helpful is when an auto-mode diffraction center fails because the minimum score threshold is not achieved. Instead of an error dialog popping up, the error will just appear in the sample's status field.
• Reduced width of second diffraction center raster from three cells to one cell. This will avoid the possibility of the second raster walking away from the optimal Z center found by the first raster, and it will speed up diffraction centering on average because it will now rasterize just one column of cells as opposed to the usual three.

This has been resolved and the correct distance should now be written to the image header.

If you have any questions or concerns about this, please contact the staff.

Please use ONE of the following addresses:

support at imca.aps.anl.gov

or...

imca-support at anl.gov

• Fixed a bug in auto mode diffraction centering where it would abort after running for exactly 10 minutes even if it had already finished diffraction centering and was just finishing synchronizing the XYZ stage and Phi axis positions.
• For auto mode diffraction centering, a loop-find will now always be executed even if the "use stored center" option is checked.
• For auto mode diffraction centering, reduce the maximum raster size from 50x50 cells to a smaller staff-configured size. This aims to avoid really big rasters that take a long time to scan.

• Fixed a bug in the image viewer that could lead to excessive CPU utilization and a freeze when inverting the color of an image multiple times from the Contrast Adjustment tool.

• Added support for assigning a username to a puck and for a corresponding '%u' substitution character expansion in scan template paths and manual diffraction centering paths. (This is useful for staff for mail-in data collection.)

• Added support for a staff-configured auto mode diffraction center best-score cutoff. Now, if the best score of a diffraction raster is less than or equal to the cutoff, diffraction centering will abort, and the sample will be skipped. This makes the system more intelligent by not fully rasterizing and collecting data on a sample where there appears to be nothing there.

• Added support for inverse wedge scans. Now, if you want to perform an inverse wedge scan, just check the "Scan in inverse wedge mode" checkbox in the scan editor. And yes, you can even do an inverse wedge scan along a vector.
• Improved internal physics constants (e.g. constants used to convert between eV and angstroms) by updating to 2010 CODATA values.
• Fixed a bug in the handling of image-taken notifications when the image file name does not match the scan template. Previously, the task that handled the image-taken notifications would exit in this case, meaning future image-taken notifications for future scans would not work. Now, an error will be generated, but future image-taken notifications for future scans will continue to work.
• Fixed a bug in the configuration file editor that prevented it from displaying an error dialog to the user in the case where the user does not have sufficient privileges to edit a configuration file.

• Fixed bug where the per-dewar "number of positions in a magazine" configuration file parameters were being ignored if the first per-dewar "number of magazines" configuration file parameter (i.e. for dewar 1) was not present.
• Fixed bug where an invalid dewar position error would occur when editing the inventory when dewar 1 has a Rigaku plate, dewar 2 has a UniPuck plate, and overriding the mounted sample to be a high magazine position from dewar 2 that does not exist in a Rigaku magazine (e.g. 16).

• Fixed bug where pressing Reset button more than once after pressing Abort button would result in JDirector indicating it was "busy" (i.e. spinning Rigaku logo) indefinitely with no way to stop it.
• Fixed bug where no useful error message was provided for CameraMan robot error codes 91-98.

After consultation, on the morning of Nov 14, staff requested the APS to re-optimize storage ring steering at our sector. This has appeared to restore the beam focus and intensity.

If there are any questions or concerns about this, or if the users on Nov 13 have noticed problems with data, please contact the staff.

• Made auto mode diffraction centering faster on average by rasterizing the smaller plane first.

• Added support for specifying the auto mode diffraction center high resolution cutoff (i.e. the high resolution cutoff to be used for diffraction centering in auto mode) in a JDirector configuration file.

• Fixed bug where for certain resizes at certain zooms, the region would incorrectly snap to a zero width or height. For example, at a zoom of 5, resizing a region of one column and multiple rows by dragging down on the bottom edge handle resulted in the region snapping to a width of zero.

• Fixed bug where if opened the diffraction center dialog, drew a region, and then closed the dialog (thus making the region disappear), clicking where the region was did not result in a mark being set.

The Hampton HT bases with the painted bar code and ALS-style bases appear to be made too small to reliably fit in the ACTOR pucks. These two base styles were responsible for two of the robot accidents this month.

We have sizing pins that we will distribute to check the size of the bases.

The following bases have shown to be problematic:

Hampton CrystalCap HT or Copper HT w/2D code (HR8-1*) Hampton CrystalCap Magnetic ALS (HR4-779) Hampton CrystalCap Copper Magnetic ALS Any other Hampton base with a painted bar code.

The following bases are recommended:

MiTeGen: any base Molecular Dimensions: any base Hampton: CrystalCap Magnetic (HR4-731, HR4-733, HR4-902) Hampton: CrystalCap Copper Magnetic Hampton: CrystalCap HT, no 2D code (HR8-094) Any base made to the SPINE specification.

1. UPDATE: 20 Nov 2013 #######

MiTeGen supplied us with 500 of their bases to test (100 of each style) and all of them passed the gauge pin test. While this is a rather small sample, we are reasonably confident that their bases will work and can recommend them for use with the ACTOR pucks.

• Added support for scan vector visualization in the crystal view. Now, if you've defined a vector in a scan in the Image Collection view, you can see a graphical overlay of that vector in the crystal view.
• Improved scan vector definition workflow. Made vector start and end point text fields in Image Collection view scan editor uneditable. On sample dismount, all vector start and end points defined in Image Collection view scans are deleted automatically. This avoids keeping meaningless vector definitions around, and it makes it less confusing to define a new vector on a subsequent sample.
• Improved auto mode job handling of a sample with a missing pin. Rather than stopping the job, JDirector now skips the sample with the missing pin and continues on to the next sample.
• Improved manual centering by clearing the mark on the crystal view after performing a manual center. Previously, after moving to the new center position and refreshing the crystal view, the mark was still displayed in the crystal view but at the old position. We considered this to be confusing. Now, the mark is cleared after performing the manual center.
• Fixed a bug where commanding an aperture retract from the crystal view could cause the UI to become unresponsive.
• Fixed a bug where commanding a backlight insert or retract from the crystal view could cause the UI to become unresponsive.
• Improved inserting and retracting the backlight from the crystal view by refreshing the camera image after every backlight insert and retract command regardless of whether JDirector thinks the backlight position has changed.
• Fixed a bug in manual diffraction center visualization where the diffraction center overlays on the crystal view were not being hidden when the diffraction center dialog was closed.
• Fixed a bug in the manual diffraction center dialog where the size of the dialog was reset every time it was closed and opened again.
• Improved the Clear Marks button on the crystal view so that it will clear the loop outline from an auto center.

• When performing diffraction center scoring with Labelit, changed to compute the score as a function of the good Bragg candidates total. This fixes a bug introduced in version 0.18.0 (released in March 2013) where the score was computed as a function of the spot total when it should have been computed as a function of the good Bragg candidates total. JDirector uses Imcadr for diffraction centering, so this bug has existed in JDirector since March as well and is now fixed.

• Increased the auto mode diffraction center step XYZ stage initialize timeout from 10 seconds to 3 minutes to ensure it won't bogusly time out during an inopportune NoMachine NX session disconnect or connect.

• Added an expansion of the raster images folder path in the diffraction center dialog to make it easy to see what the path will expand to. This also makes it consistent with the scan template expansion in the scan editor.

• Made Phi jog control display the position and jog magnitude with a number of decimal places appropriate for the precision of the Phi axis.
• Added a no-spaces-in-path validation rule for scan templates.
• Improved robot error messages.
• Improved communication with the diffraction center subprocess when it first starts up ensuring it has enough time to start up before needing to respond to commands.
• Improved how the diffraction center dialog is opened and closed to try to avoid a difficult to reproduce system resource leak.
• Added support for vector data collection (a.k.a. collect along a vector).
• Improved scan validation error messages.
• Made the scan template of recipe editor scans default to the sample image directory.
• Fixed a stack overflow error when saving or restoring a schedule.
• Added support for the PILATUS_1.2 CBF image header format.
• Fixed the scan exposure time validator to not allow negative exposure times.
• Improved the scan editor to align controls better and be easier to read and navigate.
• Added an expansion of the scan template below the scan template in the scan editor to make it easy to see what the scan template will expand to.

• Made manual diffraction centering stop if the subprocess exits or hangs.
• Made manual diffraction centering responsive to being stopped in more situations, and improved how it stops the subprocess to decrease the chance of leaving the beamline in an inconsistent state.
• Improved how the diffraction center step in auto mode executes the diffraction centering subprocess, and added capturing of the standard error output of the subprocess if a failure occurred.
• Made the diffraction center step in auto mode interruptible. Previously, no steps were interruptible; a job could only be stopped after a step completed. Now a job can be stopped in the middle of a diffraction center step. This is helpful if a diffraction center step hangs. (Of course this should never happen, but if it does, at least now there is a good way to stop it.)

• Added diffraction center support for choosing the analysis package (e.g. LABELIT or d*TREK) used for analyzing (including scoring) diffraction center images.
• Added diffraction center support for configuring the following analysis parameters: high and low resolution cutoff, minimum spot area, and minimum signal height.
• Made diffraction center raster cells selectable on the result view, and added a detail view showing analysis properties (e.g. spot total, good Bragg candidate total, etc.) for the selected cell.
• Added diffraction center raster result analysis property drop-down enabling selection of what property is displayed in the raster result view. Previously, a score was displayed. Now the displayed value can be the score, spot total, good Bragg candidates total, or total integrated signal. The score is not exactly the same as before. Before, the score was a scaled score between 0 and 100, but now it is not scaled.
• Made the generated HTML result file display scores as they appear in the GUI. That is, they are the unscaled scores as displayed in the GUI, not scaled scores between 0 and 100.
• Changed diffraction center configuration to indicate "None" for an optional value that has been unchecked rather than showing the value grayed out. This better communicates that no value is being specified. For example, if detector distance is unchecked, it now shows "None" for its value, indicating that the detector distance will not be moved.
• Improved diffraction center configuration descriptive text, labels, layout, and window resize behavior.
• Improved diffraction center subprocess communication when subprocess terminates prematurely.

Limited testing at 94% attenuation has shown that when the beam is moved to the edge of the camera region at 5x magnification, the beam will be returned to the predefined origin.

Please keep in mind that at high attenuation, the reported beam position will not be accurate and the beam may be moved away from the defined origin. This can be tested by turning off feedback, moving to the required attenuation and observing if the reported position deviates significantly from that reported with standard attenuation (94%). If this occurs, leave the positional feedback off until returning to a lower level of attenuation.

If problems are noticed with beam drift or any malfunction of the positional feedback system, please notify the staff.

Continued consultation with Instrumentation Technologies and Dectris has suggested that the profile change was enough that the XBPM needs to be re-calibrated. A quick test of this on Feb 18 showed this to be the case.

Recalibration will be attempted on Feb 23, after which the beam feedback should be functional.