Hi Ludo,
Several thoughts:

1) the movie processing in relion-1.2 will work well on large particles, 
but not so well on small ones. (This will be improved in relion-1.3)
2) it is important the refinement is done with averages of the movies, 
so perhaps the answer to your question is yes.
3) the running average window determines how many movie frames are 
averaged together before doing the alignment against the reference. This 
is necessary because the dose in each frame is typically too low to 
align accurately. For ribosomes we have used 1-2 electron/A^2/frame and 
have used typically 5-frame averages for the alignment. As mentioned 
above: things are tougher for smaller particles, where we have used up 
to 11-frame averages. If you used a single frame (run-avg-window = 1), 
your resolution would decrease as you describe.

HTH,
S



On 01/28/2014 02:55 PM, Ludovic Renault wrote:
> Hi Sjors,
> I have a dataset collected as movie on a falcon2.
> I have done exactly what is suggested for movie processing:
> -process my integrated images in 2d/3d
> -extract subset of movies stacks.
> -resumed the 3d ref on my small subset of data with movie option 
> pointing to the 'movie' star file.
>
> but I get a degraded resolution (from 9ang to 19ang) in the end after 
> movie processing compare to using only integrated images.
> Could it be that my 'integrated' images are not exactly the averages 
> of the frames?
> Should I instead average all the frames from the movie stacks and use 
> that for initial 2d/3d classification?
>
> Also, what is the 'running average window' parameter?
> I do not understand it fully I think.
>
> Thanks!
> Ludo
>
>
>
> On Sun, Jul 21, 2013 at 8:11 PM, Sjors Scheres 
> <[log in to unmask] <mailto:[log in to unmask]>> wrote:
>
>     Sorry: hit the send button without finishing...
>
>     The important thing to be able to continue correctly with the movie
>     particles from the homogeneous subset of average particles in the
>     3d-auto-refine, is that these average particles should be the
>     average of
>     the entire movie. Otherwise, there may be trouble with
>     normalisation to
>     the correct greyscale. If you want to use only a specified set of
>     frames
>     for the movie refinement, that's still fine: just indicate so on the
>     Preprocessing GUI when extracting your movies.
>
>     Read more about movie processing on:
>     http://www2.mrc-lmb.cam.ac.uk/relion/index.php/Process_movies
>     and more about how we use relion on:
>     http://www2.mrc-lmb.cam.ac.uk/relion/index.php/Recommended_procedures
>
>     Hope that helps,
>     Sjors
>
>     > Dear Sjors,
>     >
>     > would be it possible to use the particles selected from 2D
>     classification
>     > also for movie processing?
>     >
>     > Thank you in advance,
>     > Matteo
>     >
>     >
>     > --
>     >
>     > Matteo Allegretti
>     > PhD candidate
>     >
>     > Max Planck Institut für Biophysik
>     > Max-von-Laue-Straße 3
>     > D-60438 Frankfurt am Main
>     > Office: +49-69-6303-3015
>     >
>
>
>     --
>     Sjors Scheres
>     MRC Laboratory of Molecular Biology
>     Francis Crick Avenue, Cambridge Biomedical Campus
>     Cambridge CB2 0QH, U.K.
>     tel: +44 (0)1223 267061
>     http://www2.mrc-lmb.cam.ac.uk/groups/scheres
>
>

-- 
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres