Data acquisition settings for EBSD
The settings listed and discussed here are for use with Oxford Instruments HKL Channel 5 program (Schmidt and Olesen, 1989), which currently is the most commonly used software for geological applications. TSL also produces EBSD cameras and their software is called OIM. Bruker and Thermo-Noran also produce their own software and hardware. No two SEMs are exactly the same and these settings are a guide for initiating data acquisition using Oxford Instruments HKL Channel 5 program Flamenco. The specification of the camera being used to collect the data will have a large affect upon the parameters required. These settings can be used for acquiring your first indexed point and should be modified to achieve the best results for the sample being studied, the type of analysis required and the hardware setup. The settings given are from my personal experience and/or from settings published in the literature.
Number of Reflectors: increasing the number of reflectors is necessary in lower symmetry crystals to allow accurate determination of the orientation. The higher the symmetry the fewer reflectors are needed.
Number of Bands: the Channel 5 software allows the minimum and maximum number of bands to be defined, but these can be the same number.
Band Centers verses Band Edges: if the EBSP exhibits well defined band edges in the collected pattern, then band edges should be used, as this provides extra information later, but if the centre of the band is clear and the band edges are fuzzy, then band centers should be used instead.
Hough Resolution: the higher the Hough resolution, the more likely it is to be correctly indexed, but increasing the Hough resolution will increase the time per point required to index the pattern. So the value used is a tradeoff between the percentage of correctly indexed data points and the time required to collect the data. For in depth discussions of the Hough transform see Dingley and Randle, (1992) and Hough, (1962).
Number of frames noise reduction: increasing the number of frames averaged to generate each EBSP improves the signal-to-noise ratio and causes an increase in the percentage of correctly indexed grains but increases the collection time at each point.
Each time you analyse a sample you will need to optimize the data acquisition settings and the calibration for that run to ensure you achieve the maximum amount of indexing possible at the fastest analysis rate. It is important to note that when indexing only one phase the parameters can be set which result in the fastest indexing rate to correctly indexed data points. When multiple phases are being analysed, the set of conditions used has to be a compromise between those for the individual phases.
Table 1. Suggested initial parameter settings for data acquisition using HKL Channel 5’s Flamenco program
| Phase | No. of Reflectors | Minimum no. of Bands | Maximum no. of Bands | Hough resolution | No. of frames noise reduction | Reference |
|---|---|---|---|---|---|---|
| Quartz | 70-75 | 7-8 | 7-8 | 90-120 | 2-3 | (Bestmann et al., 2011; Halfpenny et al., 2006) |
| Calcite | 60-100 | 7-9 | 7-9 | 60-100 | 1-3 | (Bestmann et al., 2006) |
| Feldspar | 60-75 | 4-6 | 6-8 | 60-120 | 3-4 | (McLaren and Reddy, 2008) |
| Hornblende | 70-75 | 6-7 | 7-8 | 90-120 | 2-4 | Personal experience |
| Clinopyroxene | 70-75 | 6 | 8 | 120 | 3 | (Bascou et al., 2001) |
| Orthopyroxene | 70-80 | 6-7 | 8-9 | 90-120 | 3-4 | Personal experience |
| Garnet | 50-80 | 4-6 | 6-8 | 90-120 | 3 | (Bestmann et al., 2008) |
| Olivine | 75 | 6-7 | 7-8 | 120 | 3 | (Bystricky et al., 2006) |
| Pyrite | 60-120 | 6-7 | 8-9 | 100 | 3 | (Barrie et al., 2008) |
| Magnetite | 60-75 | 4-7 | 5-8 | 80 | 3 | (Morales et al., 2008) |
| Ilmenite | 75-80 | 6-7 | 7-8 | 70-80 | 2-3 | Personal experience |
| Zircon | 75 | 6 | 10 | 60 | 4 | (Reddy et al., 2007) |