Upscaling & Downscaling – Results

The upscaling and downscaling tests together make up over half of the total quality score, so a card’s performance in these two tests has a dramatic impact on the overall quality score of the solution.

Blitter Upscale Quality Tests: Jagged Edges
Test
3dfx Voodoo3
3dfx Voodoo5
ATI Radeon
Matrox
G200
Matrox
G400
NVIDIA GeForce
NVIDIA TNT2
S3 Savage 2000
Blitter 600%
Fail
Fail
Pass
Fail
Pass
Pass
Pass
Fail
Blitter 400%
Fail
Fail
Pass
Fail
Pass
Pass
Pass
Fail
Blitter 356%
Fail
Fail
Pass
Fail
Pass
Pass
Pass
Fail
Blitter 286%
Fail
Fail
Pass
Fail
Pass
Pass
Pass
Fail

ATI, NVIDIA and Matrox (only for the G400) show that they use a higher quality upscale filter as the upscaled test image was not nearly as jaggy in any of those three cases when compared to the test image on all of the cards that failed the test.

The 3dfx cards, the Matrox G200 and the Savage 2000 all seemed to do nothing more than reproduce the low resolution pixels at the higher, nonnative upscaled resolution in the test.

Blitter Upscale Quality Tests: Moiré
Test
3dfx Voodoo3
3dfx Voodoo5
ATI Radeon
Matrox
G200
Matrox
G400
NVIDIA GeForce
NVIDIA TNT2
S3 Savage 2000
Blitter 600% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 600% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 400% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 400% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 356% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 356% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 286% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Blitter 286% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass

Only the Savage 2000 with its 16-tap x 16-tap filter was able to pass this very difficult test. The rest of the cards produced an undesirable moiré pattern on the test image.

According to MadOnion, only the highest order filters with between 8 and 16 taps will be able to pass this test.

Even the ATI Radeon only features a 4-tap x 3-tap filter which prevents it from passing any of the 8 tests in this category. Luckily for the rest of the manufacturers, this test only accounts for 3.3% of the total quality score, so while S3 may have passed with flying colors, it doesn't buy the Savage 2000 much of a lead.

Overlay Upscale Quality Tests: Jagged Edges
Test
3dfx Voodoo3
3dfx Voodoo5
ATI Radeon
Matrox
G200
Matrox
G400
NVIDIA GeForce
NVIDIA TNT2
S3 Savage 2000
Blitter 600%
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Blitter 400%
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Blitter 356%
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Blitter 286%
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass

All of the solutions emerged victorious here.

Overlay Upscale Quality Tests: Moiré
Test
3dfx Voodoo3
3dfx Voodoo5
ATI Radeon
Matrox
G200
Matrox
G400
NVIDIA GeForce
NVIDIA TNT2
S3 Savage 2000
Blitter 600% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 600% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 400% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 400% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 356% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 356% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 286% Horizontal
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Blitter 286% Vertical
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail

Not even the Savage 2000 is able to pass a single part of this last upscale quality test, illustrating that there is still room to improve, even among the latest graphics accelerators.

Quality Comparison - Upscaling & Downscaling Downscaling Results
Comments Locked

0 Comments

View All Comments

Log in

Don't have an account? Sign up now