[mmaimcal]Re: Comment in ALMA memo 489

[MORITA, Koh-Ichiro]

Dear Stephane and Mark,

In my old work in 2000, I used an MEM mosaic program
in SDE package.  For noise free cases, almost similar imaging
qualities were obtained from 64+SD and 64+ACA+SD.
But, at that time, I only used M31 model for which the ACA is
less effective in our current work.  I also used ring type
configurations for the ACA.  This type is not good one
as shown in memo 488.

However, the effect of the ACA was clear in the case including
pointing error.  So, in my previous memo (memo374),
I concentrated on image simulations for such cases.

I also remember that it was not easy to find out appropriate
weight factors for the 64-element data and the ACA data.
I sometimes changed these values and got different imaging
quality.  So I still suspect that I failed to use the
optimum one.

So, if we really need to compare results from MEM and CLEAN,
I think we should use AIPS++ package, which has both
algorithms and includes most recent knowledge.

Cheers,
Koh-Ichiro