Question about framerates

[BertL]

Okay, so I was doing a debunk with video files, and ran into a question.

In the Apollo 15 mission, what framerate were the video cameras shooting at during the lunar outside walking missions stuff? I know about the 16mm 'onboard' camera that was taking 1FPS at those times, but what about the TV camera?

[dwight]

The TV camera from A12 to A17 was running at 30 fps. Apollo 11's camera ran at 10fps and was scan converted to 29.97 at the groundstations. With the colour camera, the sequential Red Blue Green frames were stored onto 3 seperate magnetic disc recorders (used for sports slo-mo) , frame buffered over 3 fields and matrixed together to form a colour signal running at 29.97 fps. Due to this process of delaying each field, you get the colour artifacts during fast movement. However you save 2/3 of bandwidth using this method.

This thread here on this site "TV signal analysis" also demonstrates how impossible it is to slow such a video format down (as the HBers claim). Further ammo in the debunk.

modified to iinclude TV analysis info

[ktesibios]

As far as I can make out, the color version of the lunar surface tv camera transmitted field-sequential video at 30 fps. Since a complete color picture required the red, green and blue frames to be stored and assembled into the final picture, you got ten complete frames per second.

This Apollo experience report about the video cameras:

ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19740002701_1974002701.pdf

is full of cool stuff for electronics geeks.

To reduce the perception of flicker, the reassembly of the three color frames into a color picture used a "rolling window" trick in converting to a commercial standard color signal. Clavius has an explanation of this:

www.clavius.org/tvqual.html

Hope that's helpful.

Incidentally, if you want to use the video frame rate as a reference for timing events in Apollo tv transmissions, video files from the 'net are probably not a good candidate for this, since there's no guarantee that the necessary frame rate conversions were done correctly.

[dwight]

ktesibios, if I can just add to clarify, the TV camera operated at 30 frames per second, that is if you watched the direct non-frame buffered feed on a BW monitor you would see realtime motion with no artifacting. It was the delay via disc that converted 10 frames of a particular colour field into the correct rate of 30 (29.97)fps. (As an aside, the tracking station monitors used gated RGB tubes to create a faux colour image prior to the disc matrixing)

Additionally most net videos are progressive scan, and need to be converted to interlaced video to be viewed correctly on a standard TV. (Discounting TVs that allow progressive scan video)

[BertL]

Okay, hypothetical stuff below.

Wouldn't it be possible to pre-record the footage at, let's say, 75FPS with a film camera, then play the pre-recorded stuff at 30FPS with a projector, and have the TV camera record the projected footage?


No, I am not an HB, but I'm just always skeptic.

[dwight]

Theorectically yes, but the video would have a very distinct look to it. In the same way you can tell the difference between video and telecined 35mm film, you _would_ without any doubt be able to pick up on this. It is easy enough to see the difference between kinescoped and 2" videotape source on the archive Apollo material. The difference between film and video would be glaringly obvious. Slowing down the video would be possible, but that would mean introducing glaring motion artifacting.

Now, the pre-recording would proclude the ability to fool the tracking stations, discuss in real time sports scores, and the ability to fool the MOCR people

[Jason Thompson]

The problem with that technique as I see it is that there are distinct differences in film and television picture appearance that would be detectable.

I know this is not a direct relation to Apollo film and TV, and this may seem long and rambling, but bear with me. I am a big fan of a British TV show called Doctor Who. It started way back in 1963. Back then the program was originally recorded on video tape. Video in Britain plays at 25fps, but the picture is divided into horizontal strips called fields that update alternately. If you start from a static image, every other field of that image updates 1/50th of a second after the start, then the other fields update 1/50th of a second later, then the first lot update again 1/50th of a second after that, and so on and so forth. Each set of fields updates once every 1/25th of a second, but they are staggered. The result is a much smoother motion of the image than can be achieved on film, where the entire picture updates at once. Some sections of episodes involving location shooting or model effecvts were shot on 25fps film and dropped into the episode during final production.

Now, for overseas sales of the episodes they were copied onto film. This simply involved projecting the video image onto a large flat monitor and pointing a film camera at it. You then got an entire episode that was originally made on video now presented as film.

In the mid to late seventies the BBC purged its archives of an awful lot of stuff, and every original videotape master of every episode of Doctor Who made between 1963 and 1971 was wiped. Some of those episodes still are missing from the archives, but those that survive do so as film prints either struck for overseas sales and retained, or else returned from overseas TV stations or private collectors. Those from 1970 and 1971 that were made in colour occasionally survive as NTSC versions of the original.

Shoot forward to the present day. Old episodes of Doctor Who are being released on DVD. Some technical wizards have developed a way to restore the film to its orginal video look by using computer software to analyse adjacent frames on the film and extrapolate an intermediate frame, then interlace it with the existing ones to return the original video look and mode of picture update. The technique is called VidFIRE. Frankly they all look superb now, and probably better than they did on their original transmission.

Now, here's where it gets relevant again. If you go through that description of events you'll find that there are several image presentation mdes described:

Original 25fps film

Original '50fps' interlaced video

25fps film transfers of original video material

25fps film transfers of original film material

VidFIREd versions of 25fps transfers of original video

NTSC transfers of original PAL video

The point is that each one of these has its own distinguishing features that allow one to identify exactly what one is watching. Film and video look distinctly different. Film transfers of video often include the faint pattern of the projection screen surface over the image. VidFIREd film cannot remove the motion blur that one gets on film, though the rest looks like video. NTSC has different colour properties than PAL.

Now, whilst the average viewer won't necessarily know exactly what type of image presentation he is looking at, there are always artefacts and characteristics that would enable someone with reasonable knowledge of motion picture production to identify how an image was produced, even without tinkering with the speed of projection. In your example, recording on 75fps film and then slowing it down would not produce motion artefacts consistent with those observed.