README

# When doing a vendor update:
* make clean uninstall
* svn rm vendor/aravis
* svn commit -m "aravisGigE: remove aravis dir in preparation for vendor update" vendor
* change version in install.sh and uncomment http_proxy
* run install.sh
* comment out http_proxy in install.sh
* svn add vendor/aravis
* rm -rf vendor/aravis-*
* make
* update version number in documentation
* find vendor -name .svn -prune -o -type d -exec ~tmc43/bin/set-svn-ignore {} \;
* test and commit

# When adding a new camera, you need to read it's xml feature file using arv-tool
# This requires that you have both an appropriate version of arv-tool and glib in your path.
# If you're building the aravis library in $(TOP)/vendor/aravis-$(T_A), arv-tool-0.4
# will be found in $(TOP)/bin/$(T_A)/arv-tool-0.4

# Then run arv-tool-0.4 to search for available genicam cameras
# Each line gives the genicam device_name for each camera found on the current subnet
% arv-tool-0.4
Allied Vision Technologies-50-0503349286
Basler-21664550

# If you are not sure which camera is which you can use arv-tool to show features such as DeviceModelName
% arv-tool-0.4 -n "Basler-21722087" control DeviceModelName
Basler-21722087
DeviceModelName = acA1300-60gmNIR

# arv-tool can also be used to show all features on the camera
% arv-tool-0.4 -n "Basler-21722087" features

# Then run arv-tool-0.4 -n CAMERA_NAME genicam > CAMERA_NAME.xml
# to fetch the xml feature spec
# If arv-tool generates errors, check the line numbers in the error msgs
# and fix any flaws in the generated xml.
# (My first 2 Baslers both generated xml w/ 'bCustom"' instead of '"Custom"')
% arv-tool-0.4 -n "Allied Vision Technologies-50-0503349286" genicam > Basler_acA1300-30gm.xml

# Lastly, run the python script to create new db and edl files from the feature file
% aravisGigEApp/src/makeDbAndEdl.py Basler_acA1300-30gm.xml Baslter_aca1300-30gm

# Note: I found that different Basler camera models had identical xml feature files,
# but different image dimensions.
#
# For EPICS, we need the camera's max height, width, and pixels per image
# in order to allocate waveform records.
# The xml feature file doesn't specify this, so you'll need to create a MODEL_NAME.env
# file by hand to do this, using google or experimentation to determine the image size.
# Setting these values too large wastes memory and network bandwidth, but can be
# done temporarily so you can see what aravis queries when you run your ioc.