Post by thread:video subcarrier recovery (was Re: Crystal Tutoria

"Peter L. Peres" <spam_OUTplpTakeThisOuTACTCOM.CO.IL> wrote: > incorrect for Australia and places which have PAL because the color > decoder crystal is phase modulated at +/-90 degrees every other line, and > this phase modulation is a pain to remove for other uses. On alternate lines the phase of the burst relative to the color subcarrier +/- 135 degrees. The color subcarrier is continuous, and can be recovered from the burst using suitable circuitry. This would be a pain if you had to do it in a discrete implementation, but there are plenty of chips available that will do it for you. In NTSC, the subcarrier frequency is 455/2 times the line rate, so the subcarrier phase differs by 180 degress on successive lines. Since there are 525 lines in a complete frame, this also means that alternate frames have alternate subcarrier phase, for a two-frame (four-field) color frame sequence. In PAL, the subcarrier frequency is 1135/4 of the line rate, plus an additional 25 Hz. This results in the subcarrier phase advancing 90.576 degrees per line. Since the 25 Hz offset adds up to a complete subcarrier cycle per frame, the total subcarrier advance per frame is 90 degrees ((90.576 * 625) modulo 360), so there are four frames (8 fields) per color-frame sequence. Reference: _A Technical Introduction to Digital Video_, by Charles Poynton

On Thu, 29 Oct 1998, Eric Smith wrote: > "Peter L. Peres" <.....plpKILLspam@spam@ACTCOM.CO.IL> wrote: > > incorrect for Australia and places which have PAL because the color > > decoder crystal is phase modulated at +/-90 degrees every other line, and > > this phase modulation is a pain to remove for other uses. > > On alternate lines the phase of the burst relative to the color subcarrier > +/- 135 degrees. The color subcarrier is continuous, and can be recovered > from the burst using suitable circuitry. This would be a pain if you had > to do it in a discrete implementation, but there are plenty of chips > available that will do it for you. The color subcarrier is quadrature modulated with the color information, so it is continuous (if there is color in the image) but its phase is not. Excepting for the burst and some test signals sent in certain invisible TV lines the color carrier proper is never sent over the air in PAL. As far as I know, there are no chips made for mass market TV demodulators that recover the color subcarrier proper. The only shortcut is the double frequency TDA Philips chip series that may or may not correct for color framing. The burst is indeed at +135 and -135 degrees, alternated every line, but the difference between +135 and -135 is exactly 90 degrees and the crystal in ordinary PAL decodes is direct-locked to the burst (at 0 or 180 degrees usually) on every line, which means that it jumps 90 degrees in phase forward and backward 15625 times per second, if we neglect the color framing jumps. Peter

"Peter L. Peres" <plpKILLspamACTCOM.CO.IL> wrote: > The color subcarrier is quadrature modulated with the color information, > so it is continuous (if there is color in the image) but its phase is not. No, the PAL color subcarrier *IS* continuous phase. It is the color *burst* that isn't continuous phase. Since the horizontal line time is not a multiple of the subcarrier phase, the subcarrier phase *relative to the line timing* varies from one line to another. As I described previously, the subcarrier phase for each successive line advances 90.576 degrees (not the 90 degrees that is usually given in simplified explanations). > The burst is indeed at +135 and -135 degrees, alternated every line, but > the difference between +135 and -135 is exactly 90 degrees and the crystal Yes, but as stated above, burst phase is relative to the subcarrier phase, which is continuous in time but advances relative to each scan line. If you looked at the same point within two consecutive scan lines (e.g., the exact midpoint of the color burst), there would NOT be a 90 degree phase difference between those two points: line subcarrier phase color burst phase color burst phase relative to line relative to subcarrier phase relative to line 1 0.000 +135 135.000 2 90.576 -135 315.576 3 181.152 +135 316.152 4 271.728 -135 136.728 5 2.304 +135 137.304 6 92.880 -135 317.880 etc. Eric

On Thu, 29 Oct 1998, Eric Smith wrote: > "Peter L. Peres" <.....plpKILLspam.....ACTCOM.CO.IL> wrote: > > The color subcarrier is quadrature modulated with the color information, > > so it is continuous (if there is color in the image) but its phase is not. > > No, the PAL color subcarrier *IS* continuous phase. It is the color *burst* > that isn't continuous phase. Yes, you are right, I did not express myself accurately. However, for extraction purposes it still is a pain in the butt. > Since the horizontal line time is not a multiple of the subcarrier phase, the > subcarrier phase *relative to the line timing* varies from one line to > another. As I described previously, the subcarrier phase for each successive > line advances 90.576 degrees (not the 90 degrees that is usually given > in simplified explanations). Ok, if it advances in jumps, not continuously. Otherwise there would be color aberration on the horizontal scan line, to the tune of 0.5 degrees per line (roughly). This aberration does not exist in physically functioning systems today. The 0.576 degrees per line add up to 180 degrees over the 312 lines of a frame and this is the beginning of the color framing sequence. After 4 frames it starts over. {Quote hidden}> > The burst is indeed at +135 and -135 degrees, alternated every line, but > > the difference between +135 and -135 is exactly 90 degrees and the crystal > > Yes, but as stated above, burst phase is relative to the subcarrier phase, > which is continuous in time but advances relative to each scan line. If you > looked at the same point within two consecutive scan lines (e.g., the exact > midpoint of the color burst), there would NOT be a 90 degree phase difference > between those two points: > > line subcarrier phase color burst phase color burst phase > relative to line relative to subcarrier phase relative to line > > 1 0.000 +135 135.000 > 2 90.576 -135 315.576 > 3 181.152 +135 316.152 > 4 271.728 -135 136.728 > 5 2.304 +135 137.304 > 6 92.880 -135 317.880 > etc. All my experience with broadcast and cameras and SSGs and all that suggests that the master color subcarrier oscillator in the transmitter does indeed run at a stable frequency, but it has a slave that is synced to it during the BG pulse and that is used to prepare the actual color information for each line. (The slave is synced to the appropriate phase of the 4 phase master signal). This is to avoid the 0.5 degree per line color shift which is very un-German (the PAL standard was elaborated in Germany). Peter