#PAR&TargaNTSC Distortion
7 messages in this thread
Hello to all,
We 've had our new PAR\TCBIV for about a week now and it is fantastic!
It is installed in a Gateway2000 4DX2-66V, 32meg RAM, w\ Micrpolis 2217A.
My NTSC monitors are a Magnavox combination VCR/13" screen, and a Philips 31"
TV.
Both are around 1 year old. The 13" mantains a very high resolution and is
comparable to our new Sony Trinitron 31". The Philips resolution basically
stinks but if I can achieve an acceptable result on it the rest is cake & pie.
The problem – our images are distorting when viewed through the PAR,
Almost full screen at the bottom of the TV screen and narrower at the
top of the screen. This narrowing phenomena only happens on the
right side of the TV screens. The right side of the image is about
1/2 inch from the edge of the screen at top and the bottom on the
higher resolution screens and about an 1\8 of an inch from the right
side at the bottom of the screen and 1\2 inch at the top on the lower
res screen. Also, when there is a moderate to extreme horizontal
light transition in the image from the PAR, the verticle edges of the
screen are distorted inward and continue to the top of the screen from
the line where the light changes from dark to lighter. This happens
on both verticle sides of the TV monitors. The greater the light
transition from dark to light the greater the distortion. This
distortion is reproduced on all the monitors as well as the tape
recording. A real life example is this- We are animating a
manufacturing process. The area where the machines are located is
illuminated by floresent lights hung from the ceiling about 3\4 of the
floor to ceiling highth distance from the floor. At the factory the
work area is bright while the area above the lights is dark. We
would like to reproduce this effect. Reducing the light transition
solves extreme distotion problems but the diagonal "squinching" at
the top of the TV monitor is always there.
The Targa board produces these undesired affects to a lessor degree,( not
as squinched or distoted by light diferences ), but are not within the viewing
area when overscan is on. The Targa card is installed in a different computer,
same brand and configuration as stated above.
I understand Genlock in theory, but in practice I'm not sure I have
everything connected correctly or my settings right. The PAR genlock is set to
internal, the horizontal position is set to .069us, and the subcarrier is set to
90deg, the default. I have experimented with these settings, does'nt change
anything in respect this problem. Have also disconnected the TCBIV but this has
no effect either way. The monitors are connected to the NTSC\Composite port on
the PAR. I have checked the dB output and it is 7dB, have attenuated this to
1dB and amplified to 10 dB, no change.
I could probably live with this development if I could produce the overscan
results on the PAR as with the Targa. I havn't found that option available on
the PAR if there is one. Anybody know if it is?
Anyway, does anyone know what causes this sqinching and distortion? Is
there a solution? I've thought about having the TVs calibrated but the
distortion occures with any recording device that we use so I havn't done that.
Thanks for putting up with this long winded explanation.
Randy Jackson
ComWorks Communications, Inc.
Lexington, Kentucky
Randy,
What you are encountering is very peculiar indeed, if one discounts
the monitors as the culprit. I have worked many years with Targa's
and have not seen this phenomena. I can also attest to the fact that
on my PAR I do not encounter this problem in the least.
I think from your description that the distortion is as a result of
your monitors. The only other thing I can think is that you have a
strong magnetic field somewhere, maybe caused by a VCR on a monitor
or ?
I work with a broadcast monitor to check both Composite and CAV outputs. I have
an EVO-9650 for VTR previews which is connected to both Composite from the
Targa and SVHS from the PAR. Looks very good indeed.
So I suspect the monitors and or a magnetic field somewhere in proximity to the
Monitors. Problems like this can drive you nuts, I wish I had a better answer.
-JE
Thanks John,
I enjoy reading your responses in this forum and appreciate your
quick response. The problem is solved and I will paste the solution
below. It was solved in the 3DS section under" Render Switching". I
thought these two problems were separate and probably are but
rendering to NULL let 3DS3 render to the full 752×480 res and pushed
all the distortion out of view. YEAH!!!
<Thank You David. Rendering to NULL did the trick. It took care of
the problem I posted to the Hware\Opsys\Video section also. We
design and install CATV systems and have run into some pretty strange
magnetic induction effects and I was hoping that this wasn't one of
them. (See-PAR NTSC Distortion in the Hware\OpSys\Video section.)>
Randy,
Glad there was a simple fix. From your description it sounded like only certain
portions of the screen where distorted. Thanks to David for seeing the obvious.
<g>
>> some pretty strange magnetic induction effects and I was hoping that this
wasn't one of them. <<
Me too, glad it wasn't. 🙂
-JE
>> Also, when there is a moderate to extreme horizontal light transition in the
image from the PAR, the verticle edges of the screen are distorted inward and
continue to the top of the screen from the line where the light changes from
dark to lighter. This happens on both verticle sides of the TV monitors. The
greater the light transition from dark to light the greater the distortion.
The distortion you are seeing at the top edge of the picture could be caused by
over driven video. In a simple setup like the PAR or targa+ I would start by
looking at the termination resistance. In a composite signal the sync would
also be over driven. Be sure it's 75 ohms. An unusually high resistance could
cause this distortion to occur. Other causes are erratic horizontal sync, but
unlikely given the simplicity of your setup.
Bending of the rectangular portions of the video raster are usually sync
related. If you had a WFM, check out the front and back porches of the signal,
the blanking and integrety of the burst. There might be some mis adjustment
causing video/burst to interfere with the horizontal sync. Just moving your
aspect ratio to overscan the problem is like putting a blindfold on. It does
not cure the sync problem which will magnify when you dupe your work. In
editing we use monitors designed to expose even the slightest problems to
correct them early on. Good luck.
Don,
Thank you very much. We have made some adjustments to the wiring
confiuration and things have improved considerably. As humiliating as this is
I'm going to post what we were doing here for all to see.
We were using a Y-type RCA adaptor cord to split the signal to the two
monitors. ( large red palm print now on foreheads). The resistance was
40.1ohms. I know better than to do this type of thing. There is a pattern
forming here. As we get more involved with the "finer" points of professional
animation, watching the sun rise, deadlines, marketing, materials and lighting
techniques etc… we seem to be forgetting the fundamentals.
Thanks again for the very excellent help.
Randy
Very good Randy. I know the scene very well. Just this morning I got a
panicked call from a local video post house that has been without an engineer
for two years now. Budget cuts, mean and lean business philosophy stuff. They
had all sorts of complaints lately about audio buzzes and hum, video that looks
sick, and graphics that gets lost off the screen. I must have removed about a
$100 worth of radio shack Y adapters and 93 ohm terminators from their system.
Now their VO talent doesn't sound like hees talking with a clothespin on his
nose. <G> I've got about two days of work there to get everything back in
alignment. Trouble is their editor, and two cameramen have discovered the
H-phase and sc phase trim pots through the tiny holes in their TBC's and have
worn them out tinkering. That should be good for a half day of TBC repair too.