2005 YU55 movie

[pfile]

this is 35 x 60s exposures with a canon 50d and 200mm lens @ f/4.

pixinsight: calibration, registration, crop, lum extraction, histogram stretches
astrometry.net: overlay grid (unfortunately it does not label the grid )
netpbm: fix fits origin problem
mencoder: make mpeg4 avi movie at 4fps

the movie is 1080p so be sure to enlarge it...

http://www.youtube.com/watch?v=X1UXZXA1S6c

[Giorgio]

Hello,
this is a difficult object, also a very fast stone! Good work! pfile
Giorgio.

[pfile]

thanks, giorgio!

[Lex]

Pfile,

Nicely done, I like this kind of films; better than Spielberg 

[christophwoehrle]

Looks nice.
Fast little giant

Chris

[pfile]

thanks lex and chris.

does anyone know if there is an automagic way to normalize the background brightness in each frame? i tried linearfit but that seemed to mess up the images. i cant remember now if i did LF before or after stretching them though.

[Lex]

Pfile,

the best way to deal with this kind of task would normally be the LinFit function.
I think this function can be applied to Lin or Non-Lin images, but to achieve a homgenic result it should always be applied to either the L ore the NL images. IMHO

Btw: I think the "fuzzy" aspect of the background area gives this little shorty a nice "scientific" touch 

[pfile]

well, that's what i thought and i have used the linear fit function before for this purpose. i'm not sure why it did not work this time.

it's all kind of a pain because ImageContainer works for some processes and not for others. working on 35 images is a big pain if the ops have to be done individually.

[oldwexi]

Hi Pfile!
 
A very difficult task, and you mananged to get an excellent viedeo!


To your questions.
The PixInsight Animation script does an Autohistogram to all images with one mouseclick.

When i make videos i let the cropped and registered images run with PI-animation using autohistogram.
In the background  i capture the video from the PC-Screen with the software "MAGIX".

Another solution for getting the the same background brightness for all images is PIxelMath.
If i want for example the background at 0.18
The Pixelmath expression could be:
(Symbols define):
B, Diffa, Diffb

PixelMath RGB/K Expression:
B = 0.18;
Diffa = median($T) - B;
Diffb = B - median($T);
iif(median($T) > B, ($T) - Diffa, ($T) + Diffb)

PixelMath works nicely with Image container.

For sure there are other solutions but this one seems to work for me.

Aloha
Gerald

[Alejandro Tombolini]

Hi

That's a very interesting information Gerald, thank you very much for sharing.

Regards, Alejandro.

[pfile]

thanks gerald. i did use the animation script while i was capturing the pictures to see if i had actually gotten anything. however, i wanted to annotate the pictures with an RA grid and also the date and time.

the problem is, even with autohistogram (which i did to all the images by hand) there were a handful that were somehow quite different and the autohistogram yielded a much dimmer picture. i'm not sure why that happened, but there were thin clouds so it could be that some of the frames are just brighter than others.

what i should do is try your pixelmath, then use imagemagick to annotate the frames and make a new movie that's got better annotations and is also not juddery. it seems that youtube had a hard time with the 4fps movie when converting to 30fps. not too surprised by that, should have tried 5fps.

[pfile]

the pixelmath worked great, thanks for the help.

[Lex]

Wow, i really admire people having the know how with PixelMath!
I only know the real minimal basics 

Btw.: Is there a docu for PM on the forum?

[zvrastil]

Btw.: Is there a docu for PM on the forum?

There's even full reference document. For some reason, it's in the Tutorial section of the PixInsight website: http://pixinsight.com/tutorials/PixelMath/en.html

Zbynek

[Lex]

Zbynek,

Thanks for this info, i'll check that out out and studdy some Pixelmath, not an easy lesson I guess