Stardust@home Suggestion Thread
Moderators: Stardust@home Team, DustMods
Unless we have the support staff of people who have fluency in the other languages, we would not have the ability to assist them. I think it could be possible to get this open for the dusting in another language, but what if they had a question? Who would moderate their forums? I think the team is so busy at the moment, that this idea is simply too much for them.
It is a shame that language becomes a barrier to the science. But perhaps you could mentor them and teach them what to do and share with them the forum information. Be their translator!
It is a shame that language becomes a barrier to the science. But perhaps you could mentor them and teach them what to do and share with them the forum information. Be their translator!
From dust we come
Re: some suggestions from me:
I like this idea, as I was going to suggest perhaps adding functionality that would allow one to see who the first discoverer of a particular event was (by entering the movie ID as seen in the Events section), or by entering their login ID. Since these are recorded in the database, it shouldn't be too difficult to do a query on it for either of them.rusm wrote:some suggestions from me:
...
4. Add flag “first discoverer” in “My events”
...
-
- Posts: 19
- Joined: Thu Aug 03, 2006 2:02 pm
Drop the calibration movies,set up something where you can view only calibration movies or something like that.
So people who aren't sure what to look for can look there,and try to find particles there.
Drop the scoring and ranking,so people stop measuring.
Then the only thing to think about is finding something.
To keep everybody happy,maybe set up something when,if you start to view a new batch of movies,or you reached some kind of amount of movies,you get a notice.
Like,"Guess what you've just reached...".
You know.
Just keep the events/viewed table.
Keep us updated,it doesn't have to be much.
Cheers
So people who aren't sure what to look for can look there,and try to find particles there.
Drop the scoring and ranking,so people stop measuring.
Then the only thing to think about is finding something.
To keep everybody happy,maybe set up something when,if you start to view a new batch of movies,or you reached some kind of amount of movies,you get a notice.
Like,"Guess what you've just reached...".
You know.
Just keep the events/viewed table.
Keep us updated,it doesn't have to be much.
Cheers
a blaze in the Northern sky
Hi zuben,zuben el genubi wrote:Drop the calibration movies,set up something where you can view only calibration movies or something like that.
So people who aren't sure what to look for can look there,and try to find particles there.
Drop the scoring and ranking,so people stop measuring.
Then the only thing to think about is finding something.
To keep everybody happy,maybe set up something when,if you start to view a new batch of movies,or you reached some kind of amount of movies,you get a notice.
Like,"Guess what you've just reached...".
You know.
Just keep the events/viewed table.
Keep us updated,it doesn't have to be much.
Cheers
welcome back again!
I agree to you:
1. Drop the calibration movies (maybe limited to people, having seen enough of them)
2. Drop scoring and ranking! (My last use of this was to statistically calculate the time already spent by volunteers until now - which is impressive!)
3. Keep us updated!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
[edit]
And as for me
4. Show us more new movies! [/edit]
Cheers
peter (sturdust1)
Wir leben in einer Zeit vollkommener Mittel und verworrener Ziele. (Albert Einstein)
Suggestions for Improvement
I have a few suggestions to improve the dust search process. So far I have viewed ~600 real movies and looked at more than a few movies that were posted as being good track possibilities.
* Far too many movies show less than a few frames below the surface of the aerogel. This is critical since only subsurface frames contain data of interest. I realize that the surface of the aerogel is uneven and it is very difficult to scan a tile. My suggestion: Scan the aerogel picking the start point as is currently done, but increase the total frame count. Granted this will take longer to scan and result in larger data sets, but it will result in more movies with substantial below surface frames. For example if today 32 frames are shot, and the start point is picked such that on average there will be 16 frames above and 16 frames bleow surface level, then my suggestion says to shoot 48 frames picking the start point the same way - this will give 16 above and 32 below surface frames. I have a great concern that far too many movies are viewed with only a few subsurface frames, thus we may be missing dust data on as much as half the frames we view.
* A number of posters of particle tracks seem to not understand the importance of finding the approximate surface. Dust tracks and dust itself cannot be present above the surface of the aerogel, thus the only interesting data lies below the surface, thus understanding how to locate the surface and manipulation of the VM to examine below surface data is important. The training needs to place more emphasis on identifying the surface and the VM needs to assist that. The VM should have a optional "lock" to set the surface level and prevent VM manipulation above the "lock" point. Make the "lock" feature easy to use via the 2nd mouse button, so we don't have to stop and move off the image and click a button. The "lock" is optional so it doesn't have to be used for those who don't want to. The presence of the surface lock feature will make searching faster and emphasize how important finding the surface is. Obviously the "lock" can't be used on tilted tiles. I would observe that it is highly improbable to have a particle masquerading above surface feature in the same location as a real subsurface particle feature, thus the location of the surface is important in all searches.
Hmmm, I know I had at least one more suggestion for improving the process, but I can't remember it at present. Perhaps others Dusters will offer a few ideas.
* Far too many movies show less than a few frames below the surface of the aerogel. This is critical since only subsurface frames contain data of interest. I realize that the surface of the aerogel is uneven and it is very difficult to scan a tile. My suggestion: Scan the aerogel picking the start point as is currently done, but increase the total frame count. Granted this will take longer to scan and result in larger data sets, but it will result in more movies with substantial below surface frames. For example if today 32 frames are shot, and the start point is picked such that on average there will be 16 frames above and 16 frames bleow surface level, then my suggestion says to shoot 48 frames picking the start point the same way - this will give 16 above and 32 below surface frames. I have a great concern that far too many movies are viewed with only a few subsurface frames, thus we may be missing dust data on as much as half the frames we view.
* A number of posters of particle tracks seem to not understand the importance of finding the approximate surface. Dust tracks and dust itself cannot be present above the surface of the aerogel, thus the only interesting data lies below the surface, thus understanding how to locate the surface and manipulation of the VM to examine below surface data is important. The training needs to place more emphasis on identifying the surface and the VM needs to assist that. The VM should have a optional "lock" to set the surface level and prevent VM manipulation above the "lock" point. Make the "lock" feature easy to use via the 2nd mouse button, so we don't have to stop and move off the image and click a button. The "lock" is optional so it doesn't have to be used for those who don't want to. The presence of the surface lock feature will make searching faster and emphasize how important finding the surface is. Obviously the "lock" can't be used on tilted tiles. I would observe that it is highly improbable to have a particle masquerading above surface feature in the same location as a real subsurface particle feature, thus the location of the surface is important in all searches.
Hmmm, I know I had at least one more suggestion for improving the process, but I can't remember it at present. Perhaps others Dusters will offer a few ideas.
my suggestions
A circle or crosshair in the 'my events' movies so that I can see where I clicked on previous movies.
Feedback. When I get a calibration movie wrong, an option to replay with a circle or arrows showing where the track was.
A contrast adjustment, to expand the contrast of the image.
I think this one is more 'do able'.
Perhaps a 'double size' or 'large format' way of viewing the pictures. When i lean back in my comfy chair, it would be nice not to have to lean into the monitor to squint to see some small detail.
A circle or crosshair in the 'my events' movies so that I can see where I clicked on previous movies.
Feedback. When I get a calibration movie wrong, an option to replay with a circle or arrows showing where the track was.
A contrast adjustment, to expand the contrast of the image.
I think this one is more 'do able'.
Perhaps a 'double size' or 'large format' way of viewing the pictures. When i lean back in my comfy chair, it would be nice not to have to lean into the monitor to squint to see some small detail.
-
- Posts: 38
- Joined: Thu Aug 03, 2006 9:19 am
- Location: St. Petersburg Florida
- Contact:
Re: some suggestions from me:
Awesome suggestions I'd really like to see the "first discoverer" flag.rusm wrote:some suggestions from me:
4. Add flag “first discoverer” in “My events”
5. Would be fine to see more personal and whole project statistic (spent time, count of candidates, top candidates (w/o IDs), total movies, total checked, total bad focus, ...)
Winning isn't everything, but wanting to win is. - Vince Lombardi
-
- Posts: 38
- Joined: Thu Aug 03, 2006 9:19 am
- Location: St. Petersburg Florida
- Contact:
A thought just hit me that this mass visual inspection process could be used on the space shuttle. NASA has been having a lot of trouble with foam detaching from the orbiter's external fuel tank. I think that they take a photo survey of the foam's surface before the tank is mated to the orbiter. So by using the techniques developed by the Stardust@home team another website could be established to inspect those pics. Each small section of the external fuel tank's surface would be examined by hundreds of us space nuts thereby reducing the chances that a flaw can turn into a disaster.
Winning isn't everything, but wanting to win is. - Vince Lombardi
Warm-ups
It would be nice to have a warm-up section of calibration movies. I almost always screw something up when I first start looking at movies...my eyes/brain haven't had a chance to zero in yet. Not, mind you, that there is anything subtle about most of the calibration movies, but it would be nice to have.
/k
1. Load movies from the bottom up (I hate waiting for the important bottom few frames to load, while it's taking bandwidth loading the irrelevant top few frames!)
2. Click on the focus bar to say where the surface of the aerogel is, so that it can easily be decided which need rescanning because their focus is too shallow.
2. Click on the focus bar to say where the surface of the aerogel is, so that it can easily be decided which need rescanning because their focus is too shallow.
-
- DustMod
- Posts: 694
- Joined: Wed May 17, 2006 8:12 pm
- Location: Horsetown, USA
-
- Posts: 1
- Joined: Wed Aug 09, 2006 4:50 am
stardust physics
I would appreciate if the Stardust team wrote a short commentary on the physics of stardust and aerogel. We are told to "expect the unexpected" and it would pretty help to improve our faculty of reasoning out what is a special case and what is not. I would just like to know, at least at the most basic level, how particles react with aerogel, in order to be able to discriminate better.
Here are some questions that I've come up to when searching in the VM:
Do other particles also make a track, that is just smaller or looks different than the stardust track? Is it only a matter of velocity and size whether a particles makes a track?
Sometimes the scan is slanted, a stardust grain's track would then be slented too, right?
If there is a particle not in focus, it appears as a blured patch, whereas the track is supposed to be in focus at all level (its edge, to be precise) - does it follow that all other particles make no track at all?[/list]
Here are some questions that I've come up to when searching in the VM:
Do other particles also make a track, that is just smaller or looks different than the stardust track? Is it only a matter of velocity and size whether a particles makes a track?
Sometimes the scan is slanted, a stardust grain's track would then be slented too, right?
If there is a particle not in focus, it appears as a blured patch, whereas the track is supposed to be in focus at all level (its edge, to be precise) - does it follow that all other particles make no track at all?[/list]
-
- Posts: 1
- Joined: Thu Aug 10, 2006 2:08 pm
Image Loading
The VM loads in the images top to bottom, which means when the connection/server goes slow you're twiddling your thumbs as the images above the surface are loaded first.
I think it would be better if it was the other way round; bottom images loaded first. Then you could get on with looking at the below surfaces images straight away and it would be the more pointless, above surface images, that would be last to load.
I think it would be better if it was the other way round; bottom images loaded first. Then you could get on with looking at the below surfaces images straight away and it would be the more pointless, above surface images, that would be last to load.