toast glitch in blk fade
7 messages in this thread
I have noticed with my version 2.0 Toaster that, after blacking a tape with
black from the Toaster, when I cut in for a fade up there is a small
(approx. 1 IRE) shift in the black level-in short I see a jump on the
fades. I spoke to someone at customer service who assured me that this was
impossible. While I appreciate his confidence, I am a broadcast editor
with an engineering background, and I know what I'm seeing. I'm curious if
anybody else has noticed the same thing? I have checked all the system
timing and since it is not happening on anything other than the blk input,
it is not in whatever subroutine runs the dissolves. A puzzlement! Also,
my preview output SC is 180 degrees out of phase w/program. I know in
Toaster 1 that both PGM and PVW outputs were proc'd, and they appear to be
in version 2 as well, so I can't understand the shift. It's not crucial,
but I would like to know what I'm seeing. BTW if any NewTek design people
are listening it would be great if at some point a version came out that
would allow the user to turn off the proc circuitry, or else leave PVW
unproc'd. It would make timing a whole lot easier! All replies
appreciated. -dk
Duncan, if you are talking about using Toaster Black from the BKG
generator, it is not 7.5 ire. It is actually about 10 ire. If you save a
black framestore from ToasterPaint, you will get a black that is about
7.5
— JG
<licensed to fly>
John-Thanks for the info on the Toaster font utilities. As far as my
problems with the black go, I am taking black from an external sig
generator. It is right on at 7.5 IRE (though I didn't know about the BG
black having a level problem) but there is still a when I dissolve up from
black. As I can do a series of match-cuts to black with no shifts, I am
led to believe that it is something in the dissolve area. However, I'm not
seeing a similar hop using sources other than black. BTW, the hop is
approx
1-.5 IRE-small but noticeable particularly when dissolving to a bright
scene. Anyway, thanks for your response(s). Let me know if you have any
other ideas on the diss. problem. -dk
The jump in balck may have to do with the loading of the lines depending on
the configuration of the framestore buffers.. it might be interesting to
compare the jump with different combinations of DV1 and @ on PGM and PVW
(oops DV2). Also, a good way to get balck that is black is to use an input
with nothing on it, rather than the black BKG (10 ire) or the method of
creating the 7.5ire framestore
Duncan, just a thought on your black level problem. Have you checked to see
if the video AGC circuit on your VTR is disabled? If the AGC is left in the
on position, it will read the preroll video for it's video level, sometimes
it won't let your video signal fall below a certain threshhold. I am
assuming that you are using a deck that has a switchable agc circuit, most
industrial/broadcast decks have it.
Regards
Frank
Duncan,
I also have similar problems with the black level shift and phase shift
from frame buffer to frame buffer. The degree of glitchy-ness is board
dependant. I suspect the tolerance of the components may be the source of
the problem.
My solution? Get a new board and keep changing until it gets to your
tolerance level.
As for the 180 phase problem. It could be a hue adjustment done through
software. Or ther is a phase lock loop adjustment (trim capacitor) on the
board that will need to be tweaked. (Un-Official Procedure – Remove cabinet
CAREFULLY. Power up and configure with a STABLE sync source. Adjust trim
cap to mid position while viewing a monitor (or waveform) to obtain minimum
disturbance in the picture.
I found that it needs to be adjusted frequently and drilled a small hole in
the cabinet to get access.
Feeding Black at 7.5% into the first position is a way to avoid the glitch.
Newtek's idea of black level is 10%.
The background color called Black is about 10.5 IRE, you can get a 7.5 ire
balck by making a black screen in TPaint and saving it as a framestore
(mine is called 7.5 IRE Black). The 180 degree phase shift you notice is a
result of the Toaster and your equipment not agreeing on which field the
color frame starts on. The Toaster is always right on the RS-170a spec,
since it is digital. Analog equipment is usually tuned to be 'close
enough'. A quick way to flop the Toaster by 180 is to adjust the sc/h
phase of the source in input one. If you watch the Toaster phase while you
are doing this, it will pop over by 180 at some point. You might also try
disconnecting and reconnecting the input, hopping it will catch the frame
which your other equipment ic calling frame 1.