Skip to content

Commit

Permalink
doc updates
Browse files Browse the repository at this point in the history
  • Loading branch information
trmrsh committed May 19, 2021
1 parent 36cdb33 commit 3c3cbea
Show file tree
Hide file tree
Showing 3 changed files with 562 additions and 5 deletions.
70 changes: 65 additions & 5 deletions docs/makingmovies.rst
Expand Up @@ -8,22 +8,82 @@ Making movies from |hiper| data
*******************************

The |makemovie| command allows you to generate stills from a movie. It
is however painful to use and even more to actually compile, so this
is just to provide an easy reminder.
is a little painful to run and even more so to actually compile the
stills into a movie, so this is just to provide an easy reminder.

Making the stills
=================

TBD
|makemovie| can plot CCD images only or images plus a light curve if
you have a |reduce| log file available. I would suggest that you
either run it within an empty directory or create such a directory for
directing the output since it will create one image per
exposure. Start by creating just a few images and look at them with
your favourite PNG viewer. You will need to spend a while tweaking
things. The one thing that will change when you switch to a full run
is the X-scale on the light curve plot since that is adjusted to fit
the times of the images shown. You can adjust the apparent font size
by altering the plot width and height, at the same time as the dpi to
control the resolution. Crude, but it works. I haven't got it working
without rather a stack of "whitespace"; something for the future.

Here is a full listing of parameters I used to make a movie of the
eclipsing double white dwarf ZTF J1539+5027. In this case I was in the
directory containing the reduce log within which I had created an
empty directory "tmp". One up from the working directory was where the
raw data, run0005.fits sat.

.. code-block:: console
makemovie
source = hl
run = ../run0005
first = 2
last = 1000
trim = False
ccd = 2
bias = bias3x3_slow.hcm
flat = none
defect = none
log = run0005.log
targ = 1
comp = 2
ymin = 0.0
ymax = 0.05
ynorm = [1.0]
yoffset = [0.0]
location = e
fraction = 0.55
lpad = (0.15, 0.2, 0.1, 0.15)
cmap = viridis
width = 12.0
height = 3.0
dstore = tmp
ndigit = 4
fext = png
msub = False
iset = d
ilo = [0.0]
ihi = [100.0]
xlo = 0.0
xhi = 2048.0
ylo = 0.0
yhi = 1024.0
dpi = 200
style = dots
ms = 2.0
Making the movie from the stills
================================

I use 'ffmpeg' to make the movie with a command like::
Once you have the stills, you still need to actually make the movie. I use the command 'ffmpeg' to do so like this:

``ffmpeg -start_number 123 -i run0005_%04d.png -c:v libx264 -pix_fmt yuv420p movie.mp4 -y``

which would find any files of the form 'run0005_0340.png' starting from 'run0005_0123.png'.
which would find any files of the form 'run0005_0340.png' starting
from 'run0005_0123.png'. 'ffmpeg' has a million-and-one options and
there be ways to tweak it, but this does a reasonable job I find.



Expand Down
1 change: 1 addition & 0 deletions docs/observers.rst
Expand Up @@ -25,3 +25,4 @@ underlying principles, which contains further tips on good practice.
photometry
organisation
checklists
setup

0 comments on commit 3c3cbea

Please sign in to comment.