Shop OBEX P1 Docs P2 Docs Learn Events
YPbPr video generation details ? — Parallax Forums

YPbPr video generation details ?

jmgjmg Posts: 15,184
edited 2011-12-18 02:47 in Propeller 1
I'm running up a new monitor, a PL8006 Portable 8" Color LCD Monitor - and it has VGA/Composite and YPbPr options.
Composite (B&W) and VGA work, but I'm relatively new to YPbPr so I thought I'd try that.

Has anyone generated for this ?

First quick test, I simply plugged the Composite into the Green(sync) YPbPr lead and left the other RB leads floating.

Result was not quite what I expected :
I got a very good B&W display, with better integrity than the same signal, fed via Comp.Video path.
Looks to have better bandwidth, expected I guess, as the same MiniDIN connector is used for VGA, which
is rated to 800x 600.

On Comp.Video IP, some pre-shoot and post-shoot ringing is just visible, (same lead), and the Y channel shows none of that.

It looks like the RB signals are clamped on front porch, as rough test signals that are not line-gated, smear.
Signals that are low at blanking, and hi in mid picture, give clean and pure, Red or Blue backgrounds.
So it all looks promising, and colour control is there.

Questions are:
There seems to be a (deliberate) delay, on colour, where the background is delayed, and I note devices like AD725 mention
a Lumin delay line. I'd guess 200-300ns ?
As this delay is not there on VGA, is this a artifact of matching Colour TV delays ? (ie it looks deliberate)

Also, should the R,B signals idle at 0, or 50% during the clamping time ?
As this needs to add and subtract, my guess is R,B should blank to 50%, whilst Y blanks at Black Level.

I don't need to render pictures, just give some colour highlights, or toner, to a Monochrome text display.

Comments

  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2011-12-18 02:47
    You are correct that there was a deliberate delay put into the luma paths, this was to account for the extra delay experienced on the chroma paths whilst doing all the U & V matrixing and encoding onto the sub carrier. This delay line used to be a long thin wire wrapped around a former with an underlying shied to attempt a transmission line. They were usually about 1/4" dia and 3" to 4" long.

    If you have access to a signal generator that could produce sub carrier modulated Pulse & Bar then the 10T (broader) sine squared pulse should be filled with sub carrier so that the bottom edge is perfectly flat at baseline level. If the level of the sub carrier is wrong then it would be bowed up or down evenly, if the delay between the luma and croma is wrong then this shows up as a sine wave at the bottom (and obviously combinations of these two errors are possible. The narrow 2T pulse was there to demonstrate the HF impulse response and the bar would do the same for the LF components.

    Insertion Test Signals (ITS) were stuffed onto spare lines in frame blanking so that tests and adjustments could be done in real time. Other bits of ITS showed up all sorts of problems.

    The VGA signals are clamped at their "black" levels.

    http://jf.fourcadier.pagesperso-orange.fr/television/lignes_test/lignes_test_e.htm
Sign In or Register to comment.