AVS Forum banner
Status
Not open for further replies.
1 - 5 of 5 Posts

·
Registered
Joined
·
688 Posts
Discussion Starter · #1 ·
Hi all,

Has anyone compare the two as an outboard scaler? Which one is better?

I'm using dTV to scale video signal from DSS. This route gives me better picture than using the built-in from DTC-100.

hifiPJ in SGHT has a review on HiPix and he prefer HiFix scaler over DTC100.

Thanks

Thuan
 

·
Registered
Joined
·
2,379 Posts
At a minimum, dTV is much more flexible on aspect ratio, etc., than the HiPix for scaling and picture adjustments. I think the deinterlacing is better on dTV too, especially with 3/2 pulldown removal. IIRC (I sold my HiPix), the HiPix did better on noise, but of course this is card-specific.


------------------

Dan
 

·
Registered
Joined
·
1,270 Posts
The program formally known as dTV now known as dScaler does not perform the actual video scaler function. That is done in hardware by the video card. Bi-linear interpolation is the scaling algorithm used most frequently by the hardware designers.


The Radeon cards have some features built in that are meant to de-interlace video sources, but I'm not too sure how well that works since I don't have a Radeon.


When it comes to scan rate conversion, the video card does that too. All dScaler does is loot at the image capture and try to reconstruct progressive frames of video, but it does this at the rate the video is being updated which is usually 50 or 60 times a second depending on the TV standard being used. The video card will use whatever data is has in the video overlay buffer when it redraws the display, so it's up to dScaler to ensure that it provides the video data to the card fast enough to achieve full frame rate playback. The introduction of motion judder from mismatched frame rates can also be eliminated by setting the video card refresh rate to some multiple of the video frame rate. That will often be 50 (PAL), 100 (PAl x 2), 60 (NTSC), 72 (NTSC film pulldown mode), or (120 NTSC x 2 && 24p x 5).


The dScaler places the reconstructed progressive frames into the video card Overlay buffer which the card uses to overlay video into the standard windows display. The size of the video overlay buffer is set to a maximum of 720x480 I belive, so that's the maximum resolution of video overlay on a modern computer. Video in a window is autmatically scaled to the size of the window. Pretty neat hugh? In full screen mode, if you run your video card at some resolution other than 720x480, then the video card will scale the image to match your display resolution automatically. All done in hardware on the video card. So, it could be a video doubler, quad, or any custom resolution you can get your display to sync to. It's the most flexible system you can buy really. But that flexibility results in added complexity.



Of course, there is a lot more to this, but that is the basic theory of operation. The fine points of getting it set up for optimum performance on a given machine is still being worked out by the development team.




[This message has been edited by JoeFloyd (edited 05-07-2001).]
 
1 - 5 of 5 Posts
Status
Not open for further replies.
Top