A high degree of data multiplicity is important for good quality, especially for data sets intended for phasing. High multiplicity is often difficult to achieve at many synchrotrons because of the restricted geometry. One simple method to improve the data quality is not to collect a full circle but two half-circles and rotate the crystal by half the frame width before collecting the second half-circle. This way reflections become statistically more independent, improving the signal to noise ratio.
Order (sequence) of wavelengths in a MAD experiment
There are different opinions about this.
First of all, data collection in wedges (alternating the wavelength after a few degree) results in better preservation of the dispersive differences because the effects of radiation damage would be similarly spread over the data at all wavelengths. This is done at the JCSG where they usually collect in wedges of 30 degrees, I think. But of course this requires the beamline software to be set up for this.
In a paper by Ana Gonzalez  (open access), she cites older papers which show that in two-wavelength MAD experiments, the best strategy is to optimize the dispersive, rather than the anomalous differences [González, A., Pédelacq, J.-D., Sola, M., Gomis-Rüth, F. X., Coll, M., Samama, J.-P. & Benini, S. (1999). Acta Cryst. D55, 1449-1458.] and [Peterson, M. R., Harrop, S. J., McSweeney, S. M., Leonard, G. A., Thompson, A. W., Hunter, W. N. & Helliwell, J. R. (1996). J. Synchrotron Rad. 3, 24-34.]. This means that data collection at the inflection and remote wavelengths is preferred in this case.
For the remote wavelength, she suggests an energy 500 to 1000 eV above that of the peak in the case of a K edge, and 200-300 eV above the L-I edge, but further considerations are detailed in the paper. She argues that the best wavelength to start with would be the peak wavelength in order to provide optimized SAD phasing; which implies the order peak, inflection, high-remote.
George Sheldrick suggests to measure high-energy remote, then peak, then inflection, because radiation damage further increases the dispersive difference between the wavelengths. In support of this, Poul Nissen reported that for Ta6Br12 clusters (and also SeMet !) he found that SAD maps based on the high remote wavelength were cleaner than those from the peak wavelength. One possible reason for this finding is that radiation damage is higher during measurement of the peak.