"There's a problem with this focus movie"

[voyager1682002]

Yes, it's really really hard to see. Thank you Mike.

Lily

[DanZ]

This CM is worth 11 points therefore it should be easy to find, but I cannot see it. I'm pretty sure the calibration track is off the screen since the red markers says it should be located near the bottom edge of the movie.

http://stardustathome.ssl.berkeley.edu/ ... 1&theY=357

Edit:
(nm... I used 200% zoom and looked again, it's there but hard to locate. Its much easier to locate if you go all the way up with the focus bar. It's kind of weird that I had to scroll up to see it. It looks like the calibration track focus was 'reversed' if you know what I mean)

There are a few like that. Just means that the track is close to the surface and/or the scope started filming from a starting point a bit closer to the aerogel. But the 11 points doesn't make sense because points should be in increments of 5 only (e.g., 5, 10, 15, 20, ...). I will report this one!

Dan

[jsmaje]

Another 70-pointer missed: 7252263V1, but the missed-movie screen has only vertical (without horizontal) red bar markers. Reviewing it I still can’t see anything at any horizontal level within that vertical column, nor any forum report of this particular PM before.
Another one with off-screen coordinates perhaps, as were a frequent complaint in phase 4? How can it possibly happen?

And no response yet to my post above about 6873177V1 & 8524895V1. I’ve seen at least another two with similar artefacts since then but, having now become allergic to them, have regretfully skipped or (misleadingly) clicked bad focus, not a solution I feel happy with.

John

[laserphil]

I thought I would give dusting one more try in the hope that the bugs from the earlier phases would be resolved. I could not have been more wrong. 6243664V1 still gave me a miss when I clicked on the lowest, in focus point . The red bars indicate that the correct point is the out of focus point at 100,450 or thereabouts. I believe that I complained about this on a number of occasions during phase 4. You have lost me!

[voyager1682002]

Another one which is not in increments of 5

5893235V1 (59 points)

[jsmaje]

Another one which is not in increments of 5

Yeah. I seem to have had non-zero-or-5-ending scores for some time. The power score list today in fact shows only four out of the first 20 that are exact multiples of 5, yet six of the next 10, and all of the last 19. Seems to be a trend there; so what's going on?

I also note that the skill rankings list has rightly been revised to feature only those who have viewed at least 50 PMs (as was always the requirement for statistical reasons). Oh well, better late than never.
John

[SmithES]

I clicked on "Bad Focus" on this movie. It definitely WAS bad focus. Yet, it was counted as a miss on a Power Movie. Happened to me on the third one back from this one this morning, too. This morning is the first time I've found a problem with Phase 5.
What's going on? Evelyn (ERSTRS)


http://stardustathome.ssl.berkeley.edu/ ... rds=?-1,-1

[caprarom]

Evelyn, this is a large HAT just above center with its "tail" trailing off the upper edge to the right. You are correct that not all power movies are in focus.

It's not as big as the one I missed today, however. Well, I didn't "miss" it, but clicked the wrong end. Then it came up again later with a different number and I got it the second time, but the rating was different. Numbers 1551608 and 9504412 look the same to me, but the ratings are 40 and 50 respectively. Curious.

[SmithES]

Thanks, Caparom, Phase 5 is most difficult of all the Phases! The fact that I had two out of focus Power Movies this morning blows me away! I'm really hoping I can remain as a duster, and will keep trying for a while longer. THANKS!
Evelyn (ERSTRS)

[jsmaje]

Hi Evelyn & caprarom,

These out-of focus PM tracks seem to have become more common despite original assurance that CM/PM’s were only pasted in properly-focussed movies. Fair enough if we had been advised of this change in policy, but I can’t recall that we were.
Meanwhile, that particular track is so obviously linear with track-like features, focussing (if not to full extent) against the trend of the background to clearly stand out.

This is nevertheless one of the many tricky issues in attempting to computationally automate relevant feature-recognition while at the same time otherwise judging a movie overall as ‘bad focus’.

John

[SmithES]

jsmaje, I really appreciate these forums when dusters answer back so quickly. I learn so much from fellow dusters. I agree (now) that the track was visible even when out of focus, but I wasn't expecting a track in an out of focus Power Movie! I remember the same thing you do---PMs were supposed to be in focus! I hadn't met an out of focus PM until today, and just a few minutes ago there was another one, and I missed it, too. Four in one day! I backtracked to see if I could find the missed track. I could not. This new development is going to take a lot of fun out of dusting!
Evelyn (ERSTRS)

[jsmaje]

This new development is going to take a lot of fun out of dusting!

Surely not Evelyn! Remember "Expect the unexpected"?
There's fun to be found in rising to new challenges, otherwise it's just going through the same old routine.

What I do find particularly annoying are (1) carelessly-pasted fake-PMs focus-level-wise [e.g. were all the Heidelbergers really that shallow?], (2) those showing the wrong location on the missed-PM page, (3) those where the vertical or horizontal bars are missing because the 'official' coordinates turn out to be off-screen, (4) the tiniest (inevitably high-scoring) speck that in fact becomes a jpeg-compression-distorted blur indistinguishable from other jpeg random distortions or small inclusions, and (5) those really long tracks where the best-focussed section on reaching the final frame is clearly at a higher level and significantly different location than the not-yet-fully focussed blob/particle at the end (which could even be off-screen); where is one supposed to click?

The resulting temptation of (4) is to click on any little speck just in case, even though one may otherwise feel sure it’s an inclusion (was it you, Evelyn, who posted something like, "Well, if they want me to click on inclusions [despite the contrary instruction after extensive debate during early phases], then I will")?
And, re (1) & (5), to click ‘bad focus’ or simply skip.

John

PS: Fred (fjgiie), where are you when we need your words of wisdom about such matters, let alone appearing in the duster list at all?

[SmithES]

jsmaje, "Expect the unexpected?" I am of the old school. I read and obey the rules. When a rule appears to be broken (out of focus PMs, for instance) then to me it is not funny. However, I do love dusting and have been at it since August 2006, so I will try to grin and bear it. I agree with all your points, particularly 1 & 5. But, clicking on "bad focus" when it contains a TRACK, yet really IS bad focus, gets counted against me. That was what I objected to. So, my Event list grows longer and longer as I click on every little dot that has any possibility of being a TRACK.
As for your PS to fjgiie: "Fred (fjgiie), where are you when we need your words of wisdom about such matters, let alone appearing in the duster list at all?" I am guessing that Fred is dusting under another pseudonym. Are you hiding Fred?
I probably did write that comment about inclusions. I've written thousands of words over the past six years. Would like to take a lot of them back. . . .LOL. Evelyn (ERSTRS)

[DanZ]

Hi Everyone,

Sorry to be out of touch. I'm neck deep in a proposal today, but will try to address your questions and concerns very soon. Thanks for your patience!

Dan

[DanZ]

Okay, I've looked into at least one of these problems (PM values not being multiples of 5) and sure enough, there was a glitch in the computer programming. Unfortunately our head programmer is on Kwajalein Atoll in the south pacific working on the launch of NASA's NuSTAR mission, which is already enclosed in its Pegasus launch vehicle and currently sent to blast off at 8:30 AM PDT on Wednesday, June 13th. Thus, he won't be back anytime soon. Nevertheless, another tech took a look and thinks he may have solved the problem. So, please let me know if you are still finding PMs of value's that are NOT a multiple of 5 (0, 5, 10, 15, 20, 25, etc.).

Regarding PMs that are out of focus. There may have been a time when this was not the case or even a rule (though someone needs to point me to where it was stated as rule), but now that PMs can be made out of any movie where a track has been found by dusters, anything is fair game. As John pointed out, it adds a new challenge to be sure, but we want dusters to try to look for tracks anywhere they might actually pop up, even where they are not expected. Nevertheless, I will see if we can get a click on "Bad Focus" to get you out unscathed no matter what. Seems only fair. Stay tuned on that, but be patient too please (until we can get our programmer back).

If you're still finding PMs "showing the wrong location on the missed-PM page," or PMs "where the vertical or horizontal bars are missing because the 'official' coordinates turn out to be off-screen," please continue to point those out to me here and I will try to get them dealt with.

For the "tiniest (inevitably high-scoring) specks," you'll just have to live with them. They are there for statistical purposes and no one is really expected to find them. The VERY few that do (if any) add to our knowledge on duster ability, that's all.

For "those really long tracks where the best-focussed section on reaching the final frame is clearly at a higher level and significantly different location than the not-yet-fully focussed blob/particle at the end (which could even be off-screen); where is one supposed to click?" - please also continue to report those in this section of the forum if you're missing them so we can get them looked at.

Thanks again for your assistance and patience. The adventure continues to be sure!

Dan