Particle size
Moderators: Stardust@home Team, DustMods
-
- Posts: 65
- Joined: Sat Aug 05, 2006 6:55 pm
- Location: Oregon, USA
Particle size
It appears to me that the system is only now truly getting started.
The training tracks and the tracks we have seen were huge and
easily identified. I'm now seeing a couple of tracks that are down
in the 3-4 micron range. Actually, this is what I assumed we would
be looking for.
After all, 400 (?) real tracks out of more than a million movies will make
them a lot harder to find than those 15 micro things we've been looking
at.
Groundling
The training tracks and the tracks we have seen were huge and
easily identified. I'm now seeing a couple of tracks that are down
in the 3-4 micron range. Actually, this is what I assumed we would
be looking for.
After all, 400 (?) real tracks out of more than a million movies will make
them a lot harder to find than those 15 micro things we've been looking
at.
Groundling
-
- Posts: 20
- Joined: Mon Aug 07, 2006 1:02 pm
well..
Actually, The way I see it..
All the training movies and calibration movies are doing is helping us watch out for the tracks. Not the actual particle. A real particle will look like any other blip in the gel.
It is the track that will set it apart from other dust, hair, snot or whatever is alread in the gel.
The newer calibration movies are starting to include smaller tracks then the training and original calibration movies.
So far for me, 21 of those suckers have gotten past me and I'm kicking myself in the butt for that. Out of 3700 calibration and 10342 real movies I missed 21 calibrations. That may not seem like much but that puts my percents at:
Specificity: 99.83%
Sensitivity: 99.05%
That is not very good as that means that I could have let a few real tracks slip by.
I know that only 45 total particles are expected from the entire project. So this means that I might have missed almost 1/2 of the possible real tracks if they existed in the first 3700 real movies that I viewed.
Not very good at all.
All the training movies and calibration movies are doing is helping us watch out for the tracks. Not the actual particle. A real particle will look like any other blip in the gel.
It is the track that will set it apart from other dust, hair, snot or whatever is alread in the gel.
The newer calibration movies are starting to include smaller tracks then the training and original calibration movies.
So far for me, 21 of those suckers have gotten past me and I'm kicking myself in the butt for that. Out of 3700 calibration and 10342 real movies I missed 21 calibrations. That may not seem like much but that puts my percents at:
Specificity: 99.83%
Sensitivity: 99.05%
That is not very good as that means that I could have let a few real tracks slip by.
I know that only 45 total particles are expected from the entire project. So this means that I might have missed almost 1/2 of the possible real tracks if they existed in the first 3700 real movies that I viewed.
Not very good at all.
-
- Posts: 30
- Joined: Sun Jul 09, 2006 7:46 am
- Location: New York City
- Contact:
Good point gamalmfalyii.
Has anyone found a sideways track yet? If so, I would like to see it, because the illustration they give you in the tutorial is so general, it is kind of like describing what the Mona Lisa looks like, to a blind person, in one sentence. (e.g.: Well, it is a painting of a woman with a wierd smile.)
Has anyone found a sideways track yet? If so, I would like to see it, because the illustration they give you in the tutorial is so general, it is kind of like describing what the Mona Lisa looks like, to a blind person, in one sentence. (e.g.: Well, it is a painting of a woman with a wierd smile.)
If you can guess what video game my username references, post your answer.
-
- Posts: 146
- Joined: Thu Aug 10, 2006 7:47 am
- Contact:
Not confirmed mind you. Chances are it's not one but it's interesting anyway
http://tinyurl.com/kgpjj
http://tinyurl.com/kgpjj
That is supposed to be a track? It´s got not very much in common with the trainings....Mighty Pete wrote:Not confirmed mind you. Chances are it's not one but it's interesting anyway http://tinyurl.com/kgpjj
Lebe Dein Leben so wie Du wenn Du stirbst wünschen wirst gelebt zu haben
Sideways track example....
I have reported a sideways track example and it is currently at the "in 1st review" stage of StarDust@Home's official comments.
http://stardustathome.ssl.berkeley.edu/ ... =8153179V1
http://stardustathome.ssl.berkeley.edu/ ... =8153179V1
My other posts relating to this movie:
http://stardustathome.ssl.berkeley.edu/ ... highlight=
Ziggy Stardust...
...searching for the Spiders from Mars.
...searching for the Spiders from Mars.
Re: well..
***********************************************PocketLint wrote:Actually, The way I see it..
All the training movies and calibration movies are doing is helping us watch out for the tracks. Not the actual particle. A real particle will look like any other blip in the gel.
It is the track that will set it apart from other dust, hair, snot or whatever is alread in the gel.
The newer calibration movies are starting to include smaller tracks then the training and original calibration movies.
So far for me, 21 of those suckers have gotten past me and I'm kicking myself in the butt for that. Out of 3700 calibration and 10342 real movies I missed 21 calibrations. That may not seem like much but that puts my percents at:
Specificity: 99.83%
Sensitivity: 99.05%
That is not very good as that means that I could have let a few real tracks slip by.
I know that only 45 total particles are expected from the entire project. So this means that I might have missed almost 1/2 of the possible real tracks if they existed in the first 3700 real movies that I viewed.
Not very good at all.
I know that only 45 total particles are expected from the entire project.
??? Did you get this mixed up with a post about 45 particles per movie or micron or something like that I read,,,bcause they have no idea if there is one track or 1000 tracks!!!
Howie
Re: Sideways track example....
Now there's a clear track! Thanks for the link. I hope when the training material is updated this would be one of the examples.Ziggy wrote:I have reported a sideways track example and it is currently at the "in 1st review" stage of StarDust@Home's official comments.
http://stardustathome.ssl.berkeley.edu/ ... =8153179V1
While not stardusting, I'd rather be driving with my Alfa.
Re: Sideways track example....
Jeah, that´s a good one, I hope, I´ll find one like that, too! Congratulationspetterip wrote:Now there's a clear track! Thanks for the link. I hope when the training material is updated this would be one of the examples.Ziggy wrote:I have reported a sideways track example and it is currently at the "in 1st review" stage of StarDust@Home's official comments.http://stardustathome.ssl.berkeley.edu/ ... =8153179V1
Lebe Dein Leben so wie Du wenn Du stirbst wünschen wirst gelebt zu haben
-
- DustMod
- Posts: 694
- Joined: Wed May 17, 2006 8:12 pm
- Location: Horsetown, USA
Re: Particle size
It seems to me that according to the science info, the stardust is not expected to consist of just one particular type of dust of one particular size.Groundling wrote:...The training tracks and the tracks we have seen were huge and easily identified. I'm now seeing a couple of tracks that are down in the 3-4 micron range. Actually, this is what I assumed we would be looking for.
After all, 400 (?) real tracks out of more than a million movies will make
them a lot harder to find than those 15 micro things we've been looking
at.
Groundling
To explain previous observations(>link), it's thought to very probably vary in several materials and to have various sizes:
So, the interesting stuff might include tracks from all of the following possibilities(and more?):
- maybe some dust of ~0.01 micron size [compare this to the scale on the Focus Movies!]
maybe some of ~0.1 micron
maybe some of ~0.22 micron
maybe some of ~10 microns
maybe some of ~20 microns
and others bigger or smaller?
[On the point of possible numbers of tracks -- the official estimation(>link) was a ballpark figure of ~45 tracks ...in, I believe, (132 tiles) x (~4000 movies each) => ~500,000 movies]
Twinkle, twinkle, little dust!
How I wonder which to trust!
From stars above the world you fell!
Buried like treasure in aerogel.
How I wonder which to trust!
From stars above the world you fell!
Buried like treasure in aerogel.
Movies per tile
Hi JOC,
Additional information about number of movies per tile. I realize you were speaking approximately ~ . And that you remember this.
http://stardustathome.ssl.berkeley.edu/ ... ight=#1042
Thanks,
fjgiie
Additional information about number of movies per tile. I realize you were speaking approximately ~ . And that you remember this.
http://stardustathome.ssl.berkeley.edu/ ... ight=#1042
Thanks,
fjgiie
[Thanks for that reference, fjgiie... the later addition of an extra 1000 movies per tile for the edges will ultimately, then, make an approx total of: (132 tiles) x (4000@middle + 1000@edges) ~= 650,000 movies total]
Twinkle, twinkle, little dust!
How I wonder which to trust!
From stars above the world you fell!
Buried like treasure in aerogel.
How I wonder which to trust!
From stars above the world you fell!
Buried like treasure in aerogel.