Halogenuros - in HD

[onsuper8]

After much effort we've finally got the amazing Halogenuros project on the podcast in both HD and SD...

http://onsuper8.blogspot.com/2009/01/ha ... dcast.html

Enjoy!

[christoph]

hiya,
thanks for posting.

it seems to me that something went wrong with the frame rate though..
there are a lot of lot of doubled, triple and even frames that are repeated 4 times in a row. this makes for a rather stuttery playback even on computers with enough horse power to play H264 in 720p.

++ christoph

[Daniel]

Hello,

I hope you are well.

The playback rate is 24 f.p.s There will be in a few days, or next week, a final technical report with tech parameters of each shot. Part of the footage was time-stretched (under-crank), always using "replicate" mode (not interpoling). Some footage was processed through QUtility Average that smoothen the super-8 grain, by mixing frames.... Also there are natural under-cranking, shots @ 70 fps and 50 fps but edited on a 24 fps timeline. There is also footage shot @ 24fps such as the direct sound sequence.

All the best,
Daniel Henríquez Ilic

[marc]

How do I get full screen on the 301 MB version?

[mattias]

no, there's seriously something wrong with the frame rate, possibly in the encoding? it's not just extra frames, it's many extra frames, then it moves for a while, stops, starts again, no pattern that i can see, it's jerkier than a 9 fps cell phone movie.

/matt

[christoph]

glad somebody else sees this as well, i thought i was going insane ;)

the pattern is something like:
A-B-B-C-C-C-D-E-E-E-F-F-G-G-G-G-H-I-I-J-K-K...
and that even on the title sequence and fades
++ c.

[onsuper8]

To watch the full screen HD version you'll need to click the link for the H264 file in HD.

[Daniel]

Hello,

Concerning the original timeline (eQ), in some shots there is no repeated frames, and on other shots, there are repeated frames. The technical report posted by monday, have a description of a each shot.

For example on the first keying sequence, the actress footage was shot 18fps and slowed down to 24fps with replicate method. The final bird sequence was shot at 36 fps, and placed on a 24 fps timeline, this created a bit of slow-motion , natural slow motion, without repeating frames. That’s why there is no one pattern of “reaptead frames” similar for all the short film.

Then the footage was encoded to H.264 progressive scan, but this type of H.264 (used for the current internet delivery), is not an intraframe only coding. Because to meet the hosting specs (640x360 version compatible to iPod and 1280x720 compatible to Apple TV) , there are some limitations on the quality of the H.264 encoding, in particular on the maximum bit rate. This was encoded H.264 through Compressor with the highest parameters available for those destinations.

Also it might be in some cases, (probably not Cristoph or Mattias cases) computer's peroformance in relation to H.264 decoding that can also interfer.

I have been checking that H.264 footage on a quite recent iMac (released 2008) with 3GB RAM, and it's quite ok and smooth, I would say, understanding this is a huge compression ratio..... from my point of view there is some jittery motion on the opening ocean and part of the first dance sequence, that might be due to encoding scheme... but on another hand, this was firstly edited for 35mm film-out were the sensation is not exactly the same that to be monitoring that on a tv or computer screen.

Bye bye
Daniel

[Uppsala BildTeknik]

The playback is anything but smooth... at least in my computer. It sure seems to stutter a lot more than the usual 18 to 25 fps pulldown I see every day.

Something is not OK, not with the playback I get from blip.tv. And not with the playback of the SD file. I chose the SD just to make sure I wouldn´t have any problems with the palyback.

[Daniel]

Something is not OK, not with the playback I get from blip.tv.

Hello,

concerning delivery workflow, I sent the H.264 files to Giles Perkins (http://www.onsuper8.org) whom placed them for download, and the streaming is a flash compression from those very same H.264 files. In other words, the flash streaming version (of blip tv) were created from the H.264 files.

the technical report is in spanish, but the voice-off is translated to english.

also i wanted to say thank you very much to Giles Perkins and http://www.onsuper8.org for hosting the files, and thus helping in the diffusion of this research project, worldwide.

bye bye
Daniel

[mattias]

i think you misunderstand. we know all about repeated frame and we understand that it's different for different shots. in this case it varies *within* a shot though, and it's not just duplicate frames, it's like three in a row, followed by a couple of non repeated frames, then it freezes again. i suggest you redo the encoding because this is unwatchable and only makes super 8 look really really bad.

/matt

[Daniel]

Hello,

Other people over the internet have found it is ok for watching.

No other encoding will be done at this stage.

DH

[reflex]

Well, this "other people" also finds it unwatchable, which is a shame. If you're not going to fix it, remove it.

[Daniel]

....take it easy.... I will certainly not ask the footage to be removed.

[alan doyle]

great work..
the frame rate issue does seem to detract from the quality.
i would also suggest the weak link in the chain is pro 8 quality control on old film stocks,i have always found them sub standard...dirty and unsteady images..
as phil says great for that unique look.

i believe the spirit scans use a 16mm optic and do a digital zoom...so probably not complete 2k.
not as revolutionary as i would expect..the spirit scan,the new camera and the primo lens do not seem to be adding much additional quality...
i am a little confused..
cheers