As many may know, the original video recordings from Apollo-11 are currently missing in action.   The Apollo-11 landing video was transmitted to earth in the form of a low resolution black and white slow scan feed multiplexed onto the telemetry data from the spacecraft.   The video was received at tracking stations where it was recorded in its raw format onto magnetic tapes.   The video was also converted, by means of a rudimentary optical scan converter to a conventional television format.  In doing so, quality was reduced.   However, the scan converted image is currently all we have of the transmission.   The slow scan reception was only seen at the tracking stations and was not relayed to Huston or to the news media.  Only the post-conversion video was relayed and recorded at Houston.

The tapes from the tracking station contain the original raw video data and could provide for higher quality masters of the video.   However, the location of the tapes is currently unknown.   What is known is that all the tracking stations sent their tapes to NASA for retention.  NASA, in turn, sent the tapes to the US National Archives where they were stored for several years.   In the late 1970’s, however, NASA requested that Apollo data be returned to the Goddard Space Center for permanent retention in their archive.   As far as anyone can tell, that is where the tapes now are.   Unfortunately, their exact location in a mountain of other tapes seems to have been forgotten.   There may also be other copies of the tapes in existence, but nobody is sure.

More info on the search for the original tapes here..

What follows is my own attempt to improve the video available from the Apollo-11 mission.  I have thus far only done about thirty seconds of footage as the process is extremely labor and time intensive.   In order to try to improve the quality of the available video I first went through the video one frame at a time, attempting to identify where a frame change on the SSTV system occurred.   In theory, there should be one original frame per roughly three frames of the recording, although this is made difficult by the fact that the two systems were not synchronized.   Another complication is that the frame rate of NTSC television is not actually 30fps, but is actually 29.97 fps.

After determining which frames held the same data from the original 10fps frame, which frames represented duplicates and which may have caught the SSTV system in the middle of a refresh, the frames of the same SSTV frame were combined to reduce noise and improve quality.  When one of the duplicate frames contained a noisy region, this was omitted in favor of the clearer of the frames.   Next, the frames were edited to remove noise and distortion by replacing the regions with data from unchanging regions of adjacent video frames.  Lossy noise-reduction filters are employed only as a last resort.

After restoration of the original 10 FPS SSTV frames, the video was processed using a superresolution algorithm, which is able to upscale the video to a higher optical resolution at far better quality than the original scan converter did.   This does not provide the same degree of definition as a true high resolution video recording does, but is the best avaliable method for scaling.   It uses data from multiple frames and motion processing to achieve high quality upconversion.

Finally the frame rate is converted to 30fps using an algorithm from the MSU Graphics and Media lab.   This method does not simply create three duplicate frames nor does it produce the additional frames by blending frame data together.   Instead, it uses a motion tracking method to produce the missing frames based on three dimensional motion vectors from the existing frame data.    This provides a considerably smoother upconversion method than conventional blend-based conversion methods.

The result:

Clarity and noise are dramatically improved, but detail is not as improved as I had hoped to achieve.   It may in a few circumstances even be diminished.    The final video has some noticable banding and the light balance is still unsatisfactory.   Ghosting remains noticable although this may be due to the phosphorus on the screen of the optical converter used in the original conversion process.

I plan on attempting to improve the video further by repeating this process with other first generation recordings of the footage (see bellow for more information) and combining the data from multiple recordings, using the “lucky region” method, which uses areas of the various recordings that provide the best image quality.

Some background info:

The video from Apollo 11 was received at three stations:

Goldstone Observatory – Located in California this was planned as the primary reception site for the transmission and had the largest, highest gain dish of any of the sites.

Honeysuckle Creek – Located in Australia, this was to be an auxiliary station for the transmission.  The video would be transmitted from Honeysuckle Creek to a control center at Sydney where it would then be uplinked to the United States via Intelsat III

Parkes Observatory – Also in Australia, Parkes had a slightly larger and higher gain antenna than Honeysuckle Creek, but it was located further to the west.   When Apollo 11 first landed, the moon was too low on the horizon for Parkes to get a good fix on the signal

The video from Parkes and Honeysuckle was transmitted to a control center at Sydney which choose which of the two feeds to uplink.   This was uplinked to Intellesat III and received at the Jamesburg Earthstation in California.   From there it was transmitted to Houston via microwave relay, as was the video feed from the Goldstone Observatory.  The Manned Spaceflight Center in Houston was thus presented with two video feeds to choose from, one from Goldstone and the other from Australia, providing the picture judged supperior by the control center at Sidney.

The video was recorded at Houston on a kinescope machine, which was considered supperior to video tape in 1969.  This is the official copy of the video in the NASA archive.   It is also from here that the final cut of the transmission was rebroadcast to the international media.    This is the feed that was seen by the US and the world, with one notable exception:  the Australian Broadcast Commission was allowed direct access to the uplink from Sydney and therefore had a more direct feed from the observatory and had the option to choose whether to use the international feed or the Australian feed independently.

Of these observatories, Goldstone had the best signal reception and the best picture was displayed on the monitors at Goldstone.   Honeysuckle Creek and Parkes both had good picture quality and signal strength, but not quite as good as Goldstone.  However the picture received in Houston from Goldstone was extremely dark.   This was due to a setting on the scan converter.  Although the picture was clear, Neil Armstrong appeared on the ladder as only a dark silhouette.

Houston initially selected the Goldstone feed for the broadcast at the start of the transmission.   However, it became clear that the video from Australia was of a higher quality.  Thus, within less than a minute, the feed was switched to the Australian video feed.    Initially this was provided by Honeysuckle Creek, but later in the mission, the feed was switched to Parkes when the moon rose high enough for Parkes to get good reception.

What we have for recordings:

• The official archive recording – A kinescope recording of the main international television feed.  This recording shows the feed after it was converted to 30 fps and relayed to Houston from the various receiving stations.   It only shows one of the feeds (the one selected for the broadcast).   Even though three different reception stations were receiving and recording the transmission, only the feed that was being used for the broadcast was recorded.   The broadcast switched between the three feeds to whichever had the best picture at the time.
• American TV news recordings – Recordings of news coverage of the broadcast as it was aired on ABC, NBC and CBS.   These recordings were made on kinesope and/or video tape.   While the American news coverage was recorded, it is unfortunate that the BBC coverage of the event has been lost.
• Australian TV coverage – A kinescope recording exists which recorded the broadcast of the moonwalk by the Australian Broadcast Commission.   The quality of the recording is poor and the image is very grainy, but portions of the broadcast have better contrast and brightness than the official archive recording from the international transmission.   The Australian coverage is extremely unique because Australia was the only country which had the ability to view the broadcast directly from Honeysuckle Creek and Parks and not only the international feed.   The Australian kinescope records the broadcast from Honeysuckle Creek before the international broadcast switched to that feed.   The international broadcast initially used the feed from Goldstone, which is extremely dark.
• Honeysuckle Creek 8mm Recording – Ed Von Renourd, a video technician at Honeysuckle Creek brought his own personal 8mm movie camera to work to record the historic events of the Apollo-11 landing.   His recording is the only known motion picture recording of the transmission as seen on the slowscan monitors at its original frame rate and resolution, before it was scan converted.   These recordings show only a few brief portions of the lunar EVA and because the camera is handheld and pointed at the screen, they are at times slightly skewed and do not keep the monitor perfectly stable in frame.   However, they also are some of the best quality images avaliable.   Other portion of Ed Von Renouard’s film show the activities at Honeysuckle Creek where technicians operate the receivers and video equipment.
• A Video Tape Recording of the Sidney Video Feed – Sidney video was the control point that received video feeds from Parks and Honeysuckle Creek which was then retransmitted to Houston and on to the international feed.   A video recording was recently discovered which shows the video from the Sidney Video station.   This video is NOT the original telemetry tape and shows video after it was scan converted.   Portions are of a higher quality than the official archive recording but it has heavy tracking noise.
• Several still Photographs – the feed was photographed before scan conversion by technicans at Honeysuckle Creek, Parks and Goldstone.   These photographs offer the best known images of the transmission, although they only capture one moment in time.    Some of these photographs were simply made by cameras pointed at the monitors but a few were made by CRT to film cameras that were mounted to a crt monitor and timed to capture a single frame of the display.   Prior to modern capture methods, it was not uncommon to use such methods to capture still images from video.

---

This entry was posted on Sunday, April 26th, 2009 at 9:04 pm and is filed under Conspiracy Theories, Culture, Good Science, History, Space. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

---

View blog reactions