View Single Post
  #3   Report Post  
Jeroni Paul
 
Posts: n/a
Default PC/AV issue ATI Radeon 7500&Mits.35MX1(new thread)

Hi

I've an ATI Radeon 7000/VE and have extensively used its video output for
multiple purposes. The results are perfect, just like a broadcasted signal,
in any TV and recorded on VCR, either cheap or high end 100Hz (never tried
those old non-standard Sonys with tint adjustment). I don't know what
differences might exist with the 7500 card but in my display properties are
some options. The TV output can be set as primary or secondary display, if
set as primary it'll force monitor output to the same refresh rate,
otherwise the monitor and TV can have different refresh but TV out will not
render overlays or some multimedia contents. There are some adjustments that
refer to color artifacts, interlacing and static/dynamic optimization that I
never played with. The startup and BIOS screen display just fine but you
have to have the TV plugged to the card at PC turn on otherwise the card
will disable TV output.
I'm in Europe and I have it set to output PAL B/G at 50Hz but also get a
good picture setting it to NTSC on a multisystem TV.

BTW I've heard ATI is the best card for TV output and NVidia is not as good.

JURB6006 expuso:
Hi;

Thanx to all who read and replied to my question (PC/AV).

To recap, ATI All in Wonder 7500 running into a Mits. 35MX1 (not sure
if it's a CS or a CK). It appears, as said that the signal is not
quite standard. The COMB filter in the Mits. really doesn't get along
well with this signal. It does do a decent job on regular TV
broadcasts though, and it has been rebuilt (which means replacing
just about every SMD lytic on it).

Now I built this system for the guy and when it was here I connected
it to my Sony KPR 36XBR which has the old analog one line COMB filter
(which has been aligned to a T on a standard NTSC signal). It
separated the Y and C just fine, but being a one liner, it had the
"hanging dots" on the first line of a block of highly saturated
color. The interesting part is that the hanging dots are stationary !
Normally they move due to interlace. Now the signal udoes/u have
frame/field interlace. I can tell on the screen, just like a normal
signal the raster lines can appear to move up or down if you let your
eyes follow. Now in the ATI video however, either the H and V
frequencies are shifted slightly or the color subcarrier is, proven
by the fact that the hanging dots do not move. The fact that the COMB
filter is not behaving well leads me to believe that it is the color
subcarrier that is shifted, but this might not be true, the Mits COMB
filter might use sweep timing to generate a clock signal, but it
doesn't matter, either way the phasing is off and it looks terrible.
Don't forget it is just fine on regular video.

Iv'e read in the past about sets that bypass the COMB filter for a
VCR, and indeed from a VCR a COMB filter only degrades the quality.
How would a set know it's a VCR rather than a video game or a
component or satellite tuner ? I guess it could be detected due to
the time base errors inherent in a VHS signal, but I haven't seen it
implemented on any schematic that I know of, but it could be buried
in logic where I would never see it.

This all could become a moot point, (but interesting) if only I could
get the Mits to use the SVHS input. Now that Iv'e read the replies
(thanx again) I wonder if the little microswitch might just be dirty
and if I plugged and unplugged the input about 50 times it might just
start working. The question now is, is the switch at the SVHS or
composite jack ? I used one of the audio cables for the composite and
can confirm that the TV is using the composite even when the SVHS
plug is inserted. This (input 1) composite has been used in the past
but to my knowledge the SVHS input has thusfar never been used.

Jumping to another reply suggesting I look in the display properties
on the PC, I looked around there and there are very few options, for
one the refresh is going to stay 60 Hz no matter what for the TV.
Certain other refresh rates for the PC monitor cause loss of the TV
display. Apparently there are some "illegal" combinations that simply
won't work when you have the desktop (not extended) on the TV. There
is one thing I might try though, I did notice a setting for 8 bit or
16 bit color. Of couse it gives me all the way to 32 bit for the PC
monitor but not the TVO. At 8 bit, my TV looked pretty good, just how
much possibility is there that the more advanced COMB filter might
benefit from the 16 bit setting? I think that would depend on if the
timing is changed. Perhaps the 8 bit setting is for gamers who would
want to devote the most resources and VRAM to the game.

I'm not particularly crazy about ATI's software. I think it goes a
bit too deep into the firmware, this evidenced by the fact that if
you set a bad combination of settings you lose the BIOS screen etc.
when booting, not just the Windows driven display. Damn video card
BIOS ought to be a little more stable than that. This card had the
same type of problems in an Athlon XP system, and I engineered the
sale of it to this guy because I was told that the ATI cards work
better with Intel. Building a P4 system it seemed a good way to go.
(nice price too)

I have downloaded and tried to install ATI's update/patch/whatever
file twice and there is an error in the software. It won't run.

Anyway, thank you all for even reading these long posts. I don't mean
to be long-winded but these problems can be complex, and I like to
include detail like what has been tried etc. The stuff about the COMB
filter not working right is more esoteric, on the practical side I
see I'll have to find out why the Mits. doesn't switch to the SVHS
input. I'm on it. (of course the Mits. is now sitting on a 4 foot
high shelf with 3 VCRs and a DVD on top, but I'll manage)

This isn't the end of my concerns though, I wonder how this signal
will record on a VCR. The CNR in the VHS system is pretty COMB
dependant. I'll let you know how that turns out next time I go there.

Next time I think I'll get the GeForce .

JURB