The Milky Way Project Talk

a different way to "draw bubbles"

  • chairstar by chairstar

    Since I've found so many areas of interest - not just IDRC's but other things like green knots, which do not have a round or square shape, I'd like to see if there is a way of allowing us to ACTUALLY drawing "lines" around irregularly shaped areas, as opposed to being constrained to the classic square and round shape that we are to use now.

    Perhaps incorporating a drawing software, allowing free-hand lines could be put in.

    What do the rest of you fellow MWP fans think? Please just add your comments / ideas here. No need to PM me.

    Thanks all,



  • norfolkharryuk by norfolkharryuk

    I'd like to see more detailed drawing tools for the same reason, however the computers that read our inputs need a neat tolerance to pick out where we agree or not - maybe a polygon tool rather than freehand lines - or a grid / box system. Nice thought though!


  • JLConawayII by JLConawayII

    At a basic level, this really doesn't seem necessary. A rectangle is usually sufficient to mark a spot on an image where an object exists; I don't believe defining its exact shape is relevant to our particular task. On the other hand, I recently had an image with a very complex, sprawling IRDC covering half the image (posted in the "pretty pictures" thread) where I had to use half a dozen overlapping rectangles to suitably indicate the locations of the clouds. A more precise outline tool would have been handy in this particular situation. Overall I think the tools we have now are probably sufficient for what we're doing.


  • Helice by Helice

    I agree with JLC, but I think if we had just a rotate command like for drawing bubbles on the squares that would make them flexible enough without demanding much extra software being inserted. I often capture quite a bit of non-interesting ideas because I can't rotate by 90 degrees.


  • chairstar by chairstar

    HEY Helice,

    That IS a good idea too! I haven't seemed to notice how my boxes DO overlap other things that I'd ordinarily NOT mark. So rotating the boxes is a good idea.

    I'd also (now that I think of it), like to ask all of you out there, do you have as much difficulty in rotating the bubble rings? It seems that the "tabs" (like rotate, lengthen, etc) are so close together, I need to be very precise where I put my curser.

    An example would be trying to draw the bubble ring onto an area slightly skewed off perpendicular. There's <>_ and ) , so close together that it makes manipulating the curser rather difficult for me.

    Anyone else have a comment on this?


  • suelaine by suelaine

    I have found that the round tool has been quite precise in outlining the shape of a fairly regular shaped nebulus. The enlarging and reducing, the rotating of the tool makes a fairly accurate outline of the perimeters. The square boxes are a blunt tool that give little accuracy when we are identifying very small triggered stars for instance. I just stack the boxes or call them star clusters. As for the dark nebulae, I just stack the boxes like a stair case. A horizontal tool would be good.


  • skendrew by skendrew scientist, admin

    Thanks for your feedback. We spent some time figuring out the best way to draw things with the interface, and we really wanted a freehand tool. The problem is that it's very difficult merging a lot of freehand drawings into a consistent "average", which we need to produce the catalogue. So we decided to stick with the ellipse tool for the bubbles, and the cruder box shapes for the smaller objects and infrared dark clouds. For Phase 2, which we launched in March 2012, we're showing some of the small bubbles that were drawn with simple boxes, in close-up, so we can collect more precise drawing of them for better classifications.