Sword & Helm

Sword & Helm

Description

An oriental sword & helm on a grassy hill
This scene was primarily created as a particle grass experiment. The sword and helm have been part of the scene since the beginning. The lousy tree in the background was added to fix the visual (aesthetic) weight after a last-minute aspect ratio change. I\'m planning to apply heavy focal blur in post, which will hopefully hide the tree.

@Janus: I\'m slightly concerned about the particle settings... if the animated grass (done via IPO texture translation) is a problem then please cancel the session -- without the animated particles this is essentially a static image.

Thanks to everyone for helping out!

Message boards : Comments and discussion : 780

Author Message
Syn
Send message
Joined: 13 Mar 07
Posts: 4
Credit: 88,366
RAC: 0
Message 7970 - Posted: 26 Mar 2008, 22:42:43 UTC
Last modified: 26 Mar 2008, 22:59:09 UTC

Very, very, small Work Units.

One single WU takes about 2-5 Minutes. Fun to crunch :D

Stuart
Send message
Joined: 30 Dec 06
Posts: 3
Credit: 10,326
RAC: 0
Message 7975 - Posted: 27 Mar 2008, 18:07:16 UTC
Last modified: 27 Mar 2008, 18:07:36 UTC

out of curiousity what takes longer to crunch, lots of small WU\'s or bigger WU\'s but less of them?

AC
Project donor
Avatar
Send message
Joined: 30 Sep 07
Posts: 121
Credit: 143,874
RAC: 0
Message 7976 - Posted: 27 Mar 2008, 18:46:37 UTC
Last modified: 27 Mar 2008, 18:51:04 UTC

I dunno which gives the best performance (I suspect there\'s a sweet spot where parts are just big enough to offset network and database overheads, but small enough to be distributed very evenly), but either way I didn\'t intend to submit such small WUs. I upped it to approximately 100x100 parts at the last minute... wish I\'d gone with 32x32 or so instead.

I did notice that CPUsec/sec dropped sharply, but I think it\'s artificially high at the moment because of \"saved up\" time from 726. Parts/min (but not frames/hr) was the highest that it\'s been in months though (not a big surprise).

Oh, and for the record: my power went out in mid-transfer last night and CATS seems to have recovered beautifully. Kudos to the dev(s).

loki
Send message
Joined: 7 Feb 07
Posts: 97
Credit: 158,266
RAC: 0
Message 7977 - Posted: 28 Mar 2008, 1:35:23 UTC

The video had those weird horizontal lines in it I think it happened in a previous session too with the large frame size.

loki
Send message
Joined: 7 Feb 07
Posts: 97
Credit: 158,266
RAC: 0
Message 7978 - Posted: 28 Mar 2008, 1:38:10 UTC

Please show a postprocessed work... I\'d be interested to see.

AC
Project donor
Avatar
Send message
Joined: 30 Sep 07
Posts: 121
Credit: 143,874
RAC: 0
Message 7982 - Posted: 28 Mar 2008, 22:59:33 UTC - in response to Message 7976.
Last modified: 28 Mar 2008, 23:00:46 UTC

... CATS seems to have recovered beautifully. ...

Actually, strike that -- I was short 12 frames. I may have killed CATS prematurely though, so the kudos remain.

those weird horizontal lines in it I think it happened in a previous session too with the large frame size

Yep, I looked for it but couldn\'t find it, but I remember the session you\'re talking about. Blue particles, right?

Please show a postprocessed work... I\'d be interested to see.

Will do. VirtualDub doesn\'t have any trouble with the big frames, but unless there\'s some real need, I\'ll probably post a pretty compressed 1280x720 version. I\'m planning on running it on dual monitors, else I\'d have rendered it at that res anyway.

loki
Send message
Joined: 7 Feb 07
Posts: 97
Credit: 158,266
RAC: 0
Message 7984 - Posted: 29 Mar 2008, 4:10:20 UTC

It was session 422, particles in motion... 2880x1620.

Will do. VirtualDub doesn\'t have any trouble with the big frames, but unless there\'s some real need, I\'ll probably post a pretty compressed 1280x720 version. I\'m planning on running it on dual monitors, else I\'d have rendered it at that res anyway.


Yeah, half resolution should be fine.

loki
Send message
Joined: 7 Feb 07
Posts: 97
Credit: 158,266
RAC: 0
Message 7998 - Posted: 1 Apr 2008, 1:35:54 UTC

sweet post-processing video!

Profile DoctorNow
Project donor
Avatar
Send message
Joined: 11 Apr 05
Posts: 392
Credit: 2,166,535
RAC: 23
Message 8016 - Posted: 3 Apr 2008, 7:22:14 UTC

That\'s pretty good, I love that one. :)

However, the first download over the normal link did fail and gave me errors.
And it took me 2 mirrors to download the full file, the first one seems to be incomplete.
____________
Life is Science, and Science rules. To the universe and beyond
Proud member of BOINC@Heidelberg
My BOINC-Stats

AC
Project donor
Avatar
Send message
Joined: 30 Sep 07
Posts: 121
Credit: 143,874
RAC: 0
Message 8020 - Posted: 3 Apr 2008, 13:31:07 UTC - in response to Message 8016.
Last modified: 3 Apr 2008, 13:39:49 UTC

Hrmm.... Which file was it that failed? Was it a Pre- or Post-processed video?

Since I\'ve got the original copies of the postprocessed, I was able to verify the MD5s and download them back from the server. They were bitwise identical to the originals. I also got the expected MD5s for the high and low quality preprocessed (didn\'t try q85). Of course that only shows that at least one mirror does have the right files.

Oh, and hey thanks for the positive comments, Doc & Loki!

Profile DoctorNow
Project donor
Avatar
Send message
Joined: 11 Apr 05
Posts: 392
Credit: 2,166,535
RAC: 23
Message 8021 - Posted: 3 Apr 2008, 15:22:53 UTC - in response to Message 8020.

Hrmm.... Which file was it that failed? Was it a Pre- or Post-processed video?

I\'ve downloaded the postprocessed Helm001T_POST_720x404_30FPS.avi.
There in the menu under \"Download\" it gives only errors and it seems the file is completely shown as ASCII then.
Then I tried mirror 1-link, gave me only a file about 2 MB or so.
Link of Mirror 4 finally worked and gave me the complete file.
Don\'t know if the mirros are fixed now, but the \"Download\"-link is still broken.
____________
Life is Science, and Science rules. To the universe and beyond
Proud member of BOINC@Heidelberg
My BOINC-Stats

Profile Janus
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 16 Jun 04
Posts: 4461
Credit: 2,094,806
RAC: 0
Message 8022 - Posted: 3 Apr 2008, 16:20:54 UTC - in response to Message 8021.
Last modified: 3 Apr 2008, 16:22:49 UTC

Could you name the failing mirrors? - the numbers are merely a reference to which order you should try them in, it changes every time someone downloads files.

Profile DoctorNow
Project donor
Avatar
Send message
Joined: 11 Apr 05
Posts: 392
Credit: 2,166,535
RAC: 23
Message 8023 - Posted: 3 Apr 2008, 17:03:37 UTC - in response to Message 8022.

Could you name the failing mirrors?

Okay, sorry.
Mirror 1 was FastUSMirror3.
As I\'ve only tried this and the next one succeeded, I don\'t know for Mirror 2 (Yoda) and Mirror 3 (Hong Kong). ;-)
____________
Life is Science, and Science rules. To the universe and beyond
Proud member of BOINC@Heidelberg
My BOINC-Stats


Post to thread

Message boards : Comments and discussion : 780