7/18 Cool Note From Anna on Page 3: Aerogel Schematic

Discuss your experiences with and ideas about Stardust@home here.

Moderators: Stardust@home Team, DustMods

buddhabuddha
Posts: 65
Joined: Wed May 31, 2006 2:58 pm

7/18 Cool Note From Anna on Page 3: Aerogel Schematic

Post by buddhabuddha »

As I was updating the Guess When the "We're Ready" E-mail Gets Sent thread, I thought I noticed that a post containing the aerogel schematic had gotten updated. I checked in the Updates section and, sure enough...

One more tile is now green and I believe 2 or 3 new blues ones have appeared!

http://stardustathome.ssl.berkeley.edu/ ... c.php?t=78

That makes 10 tiles that have been either scanned, processed or made into a movie.

The project mentions, "We are aiming to complete scanning and processing of 12 aerogel tiles before we launch."
Last edited by buddhabuddha on Tue Jul 18, 2006 11:59 pm, edited 5 times in total.
Aga
Posts: 19
Joined: Thu May 18, 2006 10:16 am
Location: Poland

Post by Aga »

Yeah, it's great the team keeps us updated!
dashadeaux
Posts: 21
Joined: Sat Jun 03, 2006 9:19 am
Location: USA, Peoria Arizona

Aerogel updates...

Post by dashadeaux »

A question for Anna.

I presume a tile isn't presented as green until all the focus movies for that tile are ready. This tells me that there are @15000 focus movies completed, Berkley is working on another 5000, and the target for VM start is @60000 movies completed.

In another post on this matter, it was stated that it takes Berkley one day to produce a focus movie, but this doesn't compute.

Now the question. Can you elaborate on how many focus movies Berkley produces in a day?

Best Regards,
Dash
albutterworth
Stardust@home Team
Stardust@home Team
Posts: 69
Joined: Wed May 17, 2006 12:24 pm
Location: UC Berkeley, CA
Contact:

Post by albutterworth »

Get ready for some detail :)

When the project ends we will have scanned about 5000 focus movies per aerogel tile. Right now we're getting about 4000 movies per tile, because we're leaving out the edges. We have to go back and do the edges later because they need a different focal range.


It takes about 20 hours to automatically scan one tile of aerogel and stream 4000 40-frame movies to a hard drive. We do no processing during the scan. Instead, we process the movies off-line to keep up the speed and reproducibility.

The hard drive full of raw movies is shipped to Berkeley. Here we convert each one into a stack of about 40 individual images (the focus movie) formatted ready for our Virtual Microscope browser. The process is automated, and also takes about 20 hours to produce 4000 focus movies.

The final step is to upload the 4000 focus movies to Amazon S3, from where they are served up to your web-browser. At the same time, information for each focus movie is entered into our database at Berkeley. This combined step takes about 8 hours.


As you can see, there's quite a bit of lag time between scanning and having focus movies ready for searching. That's why we want a buffer of 12 tiles (12 x 4000 movies) completely ready before we launch.


Probably more than you wanted, but I hope that answers your question.

-Anna
Stardust@home
Useful Links: | Aerogel Scanning Progress | Forum Rules |
Wolter
DustMod
Posts: 457
Joined: Mon May 22, 2006 2:23 am
Location: Enkhuizen, the Netherlands

Post by Wolter »

Indeed more then asked for. Thanks for the info.

But, as allways, an answer to a question triggers the next question... :)
I noticed that tile I015 is white whereas memory serves it was yellow before. This indicates a need for rescanning which is sometimes necessary as explained elsewhere on the site.
Can you give an estimate about the percentage of tiles which will need rescanning?

(since most tasks seems to be automated, we will need to keep you busy... :wink: )
Just dusting... Image
dashadeaux
Posts: 21
Joined: Sat Jun 03, 2006 9:19 am
Location: USA, Peoria Arizona

Post by dashadeaux »

Thank you Anna.

You answered the question precisely, and provided a good overview of the task at hand.

Given the time required to scan and process a single tile, and the progress to date, I comprehend the care being taken to produce a quality movie.

Looking forward to VM launch,
Jack
Sundevil
Posts: 41
Joined: Wed May 17, 2006 7:19 pm
Location: Ohio, USA

Post by Sundevil »

Can you explain why you need a buffer of more than 1 or 2 tiles? Would it really matter if the viewers finished searching all of the movies from the scanned tiles and had to wait a few days before the next one was put into the virtual microscope?

*This has nothing to do with the fact that I predicted June 17th as the start date in the other thread. :)
Nikita
DustMod
Posts: 994
Joined: Wed May 17, 2006 8:33 pm
Location: Indiana, USA

Post by Nikita »

With all the effort gone into scanning these, it suggests to me the need to be very careful as I scan to be sure that I do their efforts justice! A lot of hard work went into getting things to this point and I wouldn't want to waste their efforts by rushing and missing things, or flagging things because I wasn't careful enough.
Does anyone else feel this way?
From dust we come
spk72
Posts: 14
Joined: Fri Jun 02, 2006 8:29 am
Location: Spring Grove, IL
Contact:

Post by spk72 »

DustTrailFinder wrote: I am planning to find at least three stardust particles and naming each of them after my grandchildren. Pretty ambitious I guess, but without goals why even try to find them. Oh yes, I will definitely have fun doing this, but I'm serious about finding one for each grandchild and really think I can do it, even though I'm on a dial up connection. So in order of my finds there will be the Madison Starparticle, the Zachary Starparticle, and the Brendan Starparticle. If I only find one, then it will be the Madison-Zachary-Brendan Starparticle.
Looks like quantity rather than quality. Hope I'm wrong
Well done is better than well said.
Benjamin Franklin
Nikita
DustMod
Posts: 994
Joined: Wed May 17, 2006 8:33 pm
Location: Indiana, USA

Post by Nikita »

spk72 wrote:Looks like quantity rather than quality. Hope I'm wrong
Hey! We're just having a little fun! Please believe that almost all of us who have posted on here have expressed a serious interest in this as a science. But we also want to have some fun. That's why we posted some silly threads about t-shirts, logo's and names. If anyone finds three, that would be quality, especially since we will be ranked according to our accuracy.

I hope that you enjoy this as much as we hope to and see this not only as a scientific journey, but one of citizenship of the world. Good luck on your hunting!
From dust we come
buddhabuddha
Posts: 65
Joined: Wed May 31, 2006 2:58 pm

Post by buddhabuddha »

I'm seeing 11 tiles scanned, processed or movied now.

http://stardustathome.ssl.berkeley.edu/ ... c.php?t=78
Sundevil
Posts: 41
Joined: Wed May 17, 2006 7:19 pm
Location: Ohio, USA

Post by Sundevil »

Is there any logic to the order of tiles being scanned? Or is it that one corner was picked as a starting point, and it will branch off from there until all of them are completed?
buddhabuddha
Posts: 65
Joined: Wed May 31, 2006 2:58 pm

Post by buddhabuddha »

Aerogel schematic was updated again late today, 6/19.

Looks like we're back up to 11 tiles with 7 processed or made into a movie.
Wolter
DustMod
Posts: 457
Joined: Mon May 22, 2006 2:23 am
Location: Enkhuizen, the Netherlands

Post by Wolter »

Let's hope that at least one yellow one will turn green at the end of today. The last two that were processed apearantly needed rescanning as they turned white again.
So let's keep our fingers crossed for more green on the board ;)
Just dusting... Image
buddhabuddha
Posts: 65
Joined: Wed May 31, 2006 2:58 pm

Post by buddhabuddha »

We're at 9 now...wahooo! 3 more to go...

http://stardustathome.ssl.berkeley.edu/ ... c.php?t=78
Post Reply