|
Post by LunarOrbit on Aug 13, 2010 0:46:11 GMT -4
It'll be on their website by tomorrow probably. And I was watching American Chopper too...
|
|
|
Post by dwight on Aug 13, 2010 6:38:58 GMT -4
Glad you all enjoyed it. There were a few moments where I started a sentence and realized I was about to throw in a german word instead of the english, and reaqlized the sentence structure was all wrong for the english word. Ahh the joys of live TV.
|
|
|
Post by Obviousman on Aug 13, 2010 8:11:38 GMT -4
Hey - how am I going to get my copy signed by the world famous author?
|
|
|
Post by Ginnie on Aug 13, 2010 11:13:43 GMT -4
Hey, how come on the SpaceVidcast from the ISS there isn't any stars?
|
|
|
Post by ka9q on Aug 14, 2010 5:02:54 GMT -4
Dwight, what's that large launcher model over your left shoulder? A Chinese Long March?
At first I thought it was one of the middle Ariane models, but those don't look like Latin characters.
|
|
|
Post by ka9q on Aug 14, 2010 5:29:48 GMT -4
Oh, yet another question for you Dwight. Do any videotapes survive from the Doppler buffer recorders shown on page 103 of your book? Since they were regular US broadcast standard 2" quad tapes, not the special wideband telemetry tapes used to record the Apollo 11 slowscan signal, maybe there was less pressure to recycle them. If any of those tapes do survive, we could now do a much better job of displaying them in color without the losses inherent in the original sequential-to-NTSC conversion.
I asked Bill Wood at Goldstone this question a while ago. He doesn't think any survived, but I wonder if maybe you heard Stan talk about them.
|
|
|
Post by drewid on Aug 14, 2010 7:28:21 GMT -4
Whereabouts are you in Germany Dwight?
|
|
|
Post by dwight on Aug 14, 2010 13:24:49 GMT -4
Hi everyone,
The rocket is a Chinese Szhenzou with the following national flags on the side, an Australian, Chinese and American. I dont recall the exact rocket type though. On my bookshelf I have a V2, a paper model Saturn 1B 1:144 scale, Saturn V 1:144 scale, the lego LM, a 1:32 scale CSM, the Shuttle 1:144 scale, the Arianne 5 in 1:48 scale, a lunar globe, the Gemini and Mercury capsules in 1:36 scale and a CSM/LM 1:96 scale,
I dont know if those 2" tapes exist. I doubt it though. There are color sep negatives of the GCTA material. Be interesting to see what can be done with those.
I am based in Köln. Or Cologne for english speakers.
|
|
|
Post by drewid on Aug 14, 2010 14:28:26 GMT -4
I'm now intrigued by the possibility of reconstructing a proper 30fps B&W stream from the colour footage. It would need to be as original and clean as possible to stand _any_ chance of working, Oh and I'm in working in Frankfurt for a few months so we're practically neighbours
|
|
|
Post by dwight on Aug 14, 2010 19:28:09 GMT -4
I have raw sequencial footage from the ASTP in flight press conference and the B&W image flickers badly due to the different luminance values for each Red Blue and Green filter.
Let me know when you get to Frankfurt drewid.
|
|
|
Post by drewid on Aug 15, 2010 5:23:42 GMT -4
I have raw sequencial footage from the ASTP in flight press conference and the B&W image flickers badly due to the different luminance values for each Red Blue and Green filter. Let me know when you get to Frankfurt drewid. Yeah but the luminance/contrast should be correctable, That's the easy bit. Here's what I'm thinking, the incoming and outgoing fields are unique, the held fields aren't. It ought to be possible to come up with a filter (or series of filters) to compare adjacent frames and diff out the held fields leaving just the split fields. Then correct for the luminance differences afterwards. I've not really thought it right through so I haven't got particular maths in mind yet. I really need to get my hands on a few frames to try some different possible sequences. I'll be in Frankfurt again from tomorrow
|
|
|
Post by ka9q on Aug 15, 2010 7:54:04 GMT -4
I'm now intrigued by the possibility of reconstructing a proper 30fps B&W stream from the colour footage. It would need to be as original and clean as possible to stand _any_ chance of working, Great minds think alike! As Dwight says, the hard part would be correcting the flickering due to the big differences in luminance. However, for scenes like the over-discussed John Young jump salute from Apollo 16, the scene is almost entirely monochrome; about the only color are the US flag and the red stripes on Young's suit. So it should indeed be possible to adjust the levels of the red, green and blue fields to produce something like a B&W video at the full resolution of the camera. Besides recovering information lost in NTSC's I&Q channel bandwidth limiting, going back to the original sequential tapes would let us recover the full vertical resolution of the TV camera. It was a 525-line camera with interlacing, but the scan converter system essentially treated it as though it were only a 262 line camera operating at 60 Hz. It would have been possible to get 525 line resolution by exposing a full frame in each color rather than only one field, but that would have made the color motion artifacts even worse.
|
|
|
Post by ka9q on Aug 15, 2010 8:41:18 GMT -4
I have raw sequencial footage from the ASTP in flight press conference Well, that's better than nothing! How hard would it be to put that on the web using as little compression as possible? I'm not sure how MPEG-4 or even -2 would handle the rapid changes between sequential fields. As I recall, MPEG-1 doesn't do inter-frame coding so it might be a good choice, as would motion JPEG, i.e., encoding each field as a separate JPEG with no inter-frame coding.
|
|
|
Post by dwight on Aug 15, 2010 11:03:19 GMT -4
I'm just making it now. Will post a link when it is uploaded. UPDATE dl.dropbox.com/u/4960558/astp.avifor the mjpeg section of the press conference (it is only a section as the whole file would have been way too big). Can you please let me know when you have downloaded it as I need the space on my server.
|
|
|
Post by ka9q on Aug 15, 2010 17:20:11 GMT -4
Here's what I'm thinking, the incoming and outgoing fields are unique, the held fields aren't. It ought to be possible to come up with a filter (or series of filters) to compare adjacent frames and diff out the held fields leaving just the split fields. Are you talking about reversing the scan conversion process and recovering the original field sequential frames? That occurred to me too a while ago. It should be possible, just by following the rules by which they were put together, to disassemble the NTSC frames by converting them to RGB and then selecting the one color that has changed since the last field. Then if you could adjust for the luminance variations to remove the flickering, you could reassemble the fields into frames and regain the original 525 lines of vertical resolution in the camera. The only real drawback is that you wouldn't be able to undo the effects of bandwidth filtering in the NTSC encoder and recover the full original resolution of the individual color fields. NTSC uses a matrix to convert from RGB to YIQ (luminance and two color components I and Q). It transmits the Y component in the place of the original B&W component with full bandwidth, and it puts the I and Q components on a quadrature modulated subcarrier at 3.58 MHz. Before modulation the I and Q signals are bandpass filtered. From memory, I is filtered to something like 1.6 MHz and Q to only 600 kHz.
It could be that the original sequential signal from the camera was also filtered to fit the Apollo communication system; I'll have to find out.
|
|