Quote:
Originally Posted by etype2
Fair enough, you have explained low and high band to a point. I thought VTR’s had to be engineered for a given color spec. ie, Pal, SECAM, well these are formats, I understand, bad example. I think you know what I mean, but you’r saying those early modded Quads recorded full 1953 NTSC color?
That being the case and not an engineer, was it a case that early television shows were broadcast with the full 1953 NTSC color signal but the televisions from 1955 and after couldn’t reproduce the complete signal? And SMPTE-C came in about 1968 and shows were being monitored on Conrac monitors with SMPTE-C color space? I’m not understanding.
What was meant by Ed’s comments about “RCA LABS color and RCA Broadcast Color” ?
|
Details of the VTR's signal format have to be set according to the particular system's vertical and horizontal scanning rates and color subcarrier frequency. This has nothing to do with the primary colors, which are determined back at the camera before modulating onto the NTSC or PAL color subcarrier.
So, yes, these early tapes were made with TK-41 camera chains that had NTSC color.
You are correct that NTSC always used NTSC coding for the whole duration of analog broadcasting. Also that the receivers began to have non-NTSC phosphors around 1957. Also correct that SMPTE C phosphors are not NTSC. Receiver makers at first made no electrical compensation for the color of the phosphors, only for their improved efficiency (by changing the R,G,B drive levels to keep their chosen white point, which, by the way, was not NTSC, but much more cyanish).
After some years, the TV makers made adjustments to the color difference demodulators to compensate approximately for the new phosphors. SMPTE also standardized a similar electrical matrix for SMPTE C phosphor monitors that could be switched on and off. These changes were not ideal because they are applied to signals that have been gamma corrected, so some colors get overcorrected (especially bright reds) and others get undercorrected. The proper place to correct for different primary colors is at the camera before gamma correction, so the effects are linear. This is what PAL did to standardize on the new phosphors, and what programs like Photoshop do. In the case of Photoshop, since it is dealing with files that have the gamma correction baked in, they undo the gamma correction from the camera, make the color adjustment, and redo the gamma correction for the monitor.
Note that NONE of the above affects what equations are used to form the luminance and color difference signals and hence subcarrier signal in NTSC or PAL. In fact, PAL simply adopted the same equations for forming the color subcarrier as NTSC, even though the original R,G,B signals from the camera were modified for new phosphors.
So, when the composite signal (luma + subcarrier) gets to the VTR, it knows nothing about what primary colors the camera was matrixed for. It just records and reproduces what it is given.
When TK-41s began to be displaced by Plumbicon cameras, linear matrixing for different phosphors became possible (too complex and drift-prone for the all-tube TK-41s). So, camera designers modified the color-splitting optics for best efficiency and signal-to-noise, and then used a linear matrix to produce the "right" R, G, B for NTSC or PAL. But, since no color optics is perfectly matched to the human visual system, the camera's interpretation of the scene could also be fudged for best-looking pictures on the SMPTE C or EBU modern monitors. In PAL, this would be close to the theoretically correct matrix, since PAL receivers had fixed demodulator specs (borrowed from NTSC) and known modern phosphors. In NTSC, however, the TV makers had modified the circuitry for an approximate correction, and the camera makers obviously couldn't broadcast multiple versions to match multiple TV designs. So, camera designers did the next best thing, and made the cameras look best on SMPTE C monitors. BUT, the circuits following the gamma correction in the camera were still precisely NTSC.