ATI AIW 9800 Pro worst capture card ever?

[veedeo]

I spent the last two weekends resurrecting my old video capturing PC from 2004 since I could not find a decent modern capturing device. Good news, I got it up and running!! And tried two ATI AIW AGP capture cards, the 9000 and 9800 Pro.

Here is the exact same frame captured using the two cards on the same PC via SVideo. The 9800 Pro is just atrocious! It has hideous diagonal noise lines throughout the entire video!

ATI AIW 9800 Pro:


ATI AIW 9000:


This post is just a warning to others to avoid the 9800 Pro like the bubonic plague.

[lordsmurf]

There's nothing wrong with the 9800 card, or any of the other ATI All In Wonder cards, for that matter.

What you're seeing there is simply an issue of spacing between cards and shielding. Your computer has electrical and/or FM interference inside the case. This topic was recently discussed in relation to onboard sound vs PCI sound cards, and how noise can affect both types of cards. The same applies to video cards.

"Leaky" power supplies can also cause this. What you're seeing is "herringbone noise".

[smartel]

is there anyway way to test this ?
and in case , find the cause and maybe the cure about this "herringbone noise" thing ?

[lordsmurf]

Remove all PCI cards, put the soundcard on the farthest PCI slow away from the AGP slot.
Then see how it performs.

Based off what you see, it's either solved, or other issues must be looked at.

[smartel]

review all the connections and cables . keep away from transformers , can't be stronger than the weakest link

[jagabo]

What you're seeing there is simply an issue of spacing between cards and shielding.

Which means the card has insufficient shielding.

[lordsmurf]

Well .... yes and no.
Under normal condition, shielding is sufficient.
When it comes into contact with unusually noisy environment, it's not enough.

[veedeo]

Remove all PCI cards, put the soundcard on the farthest PCI slow away from the AGP slot.
Then see how it performs.

Based off what you see, it's either solved, or other issues must be looked at.

I actually don't have any PCI cards in the computer, I'm using onboard audio for now.

If this is a shielding problem, then it only exists on the 9800, not the 9000 since I'm not seeing the herringbone noise when I use that one.

One thing to note is the 9800 requires it's own power cord, where as the 9000 does not. Maybe that has something to do with it.

Either way, I'm sticking with the 9000.

[lordsmurf]

Maybe that has something to do with it.

Yes. Power noise is "leaking" from your power supply. That's all it is.
The card is likely fine, your computer not so much.
But using the other card is a good workaround for you.

[pippas]

It shouldn't be beyond the talent of decent pcb designers to design a card that can reject any external power supply 'noise' inputs.

The fact that the alternative model the OP describes does not suffer from the same problem, suggests a pretty weak design for this particular model.
Perhaps not the 'worst'capture card ever', but sounds like a pretty poor design layout?

[themaster1]

My guess: the s-video plug on your vcr is loose and/or bad cable.

[lordsmurf]

My guess: the s-video plug on your vcr is loose and/or bad cable.

That's quite common, too.
Worth a try, although assuming the same cable was used on the 9000, probably not the issue.

[veedeo]

The SVideo cables were not loose and yes, the same cables were used. I have the VCR plugged into a Panasonic ES-15 via SVideo and then the ES-15 connected to the purple ATI cables via SVideo. All nice and tight.

Last night I put in a PCI sound card and the 9000 still behaves well with no herringbone noise.

[veedeo]

I searched the forums for other causes of this problem and found this post:

https://forum.videohelp.com/threads/162064-ATI-AIW-9700-Pro-noisy-video-fix-%289800-AIW...ld-look-too%29

"Ty in a knot" blames the problem on a poorly designed board layout. He had to modify the board to fix it. He called it "harring bone" noise, which I assume is the same as herringbone noise.

[pippas]

I rest my case!! (post #10)

[lordsmurf]

Nice find from almost 7 years ago.
Try that modification, and see how it works from there.

Sounds like simple board correction worked, when this problem is present.