|
Trev's Articles Third Run of AVISynth Scripts |
April 2009
Following the second run of tests, the results showed that Tests #1.1, #1.7, #1.1.1 and #12.0.2 produced the best quality output. The tests below expand on those successful scripts to see if any further enhancements can be made.
Tests ending in 'Part2' were of a different AVI clip, which is 3m 40s long. It is used to see if Audio timing is correct and verify output results from the original test compares with the output on a different video clip.
Third run Results
Items highlighted in green were confirmation of excellent results. These tests are used to confirm the final results.
|
Test Number |
Quality |
Overall Visual Rating |
|
1.1a |
Minor Combing Smooth movement Blocky Output |
|
|
1.1.1.1 |
Medium Combing Smooth movement Blocky Output |
|
|
1.1Part2 |
No Combing Smooth movement |
|
|
1.1aPart2 |
Minor Combing Smooth movement Blocky Output |
|
|
1.1.1Part2 |
No Combing Smooth movement |
|
|
1.1.1double |
Bad Combing Smooth Movement |
|
|
1.1.1.1Part2 |
Medium Combing Smooth movement Blocky Output |
|
|
1.7a |
No Combing Smooth movement |
|
|
1.7Part2 |
Minor Combing Smooth movement |
|
|
1.7aPart2 |
No Combing Smooth movement |
|
|
12.0.2.1 |
Medium Combing Smooth movement Blocky Output |
|
|
12.0.2.1Part2 |
Medium Combing Smooth movement Blocky Output |
|
|
12.0.2Part2 |
No Combing Smooth movement |
|
Quality Definitions
Jerky movement is where the picture 'stutters' on a horizontal PAN
Combing is an artifact of an Interlaced picture
Blocky is where the picture appears as large pixels with no smooth edges. Only these tests introduced these blocky artifacts. Note: The blocky effect only showed up on the final DVD, it was very difficult to see when playing the AVI file on the PC.
See the first run of tests, for examples of quality for these tests.
Blocky examples:
This first image shows how the image should look

This second image shows the image with the blocky effect. Note: this is more emphasized on a TV Screen after DVD conversion:

Test scripts and details
Test #1.1a
# PAL DV (50 fps) to NTSC DV (59.94 fps)
LoadPlugin("SmoothDeinterlacer.dll")
AVISource("PalDV2.avi")
ConvertToYUY2(interlaced=true)# Smoothdeinterlacer requires YUY2
SmoothDeinterlace(tff=false, doublerate=true)
BilinearResize(720, 480)
ConvertFPS(59.94)
SeparateFields()
SelectEvery(4, 0, 0)
weave()
|
Test #1.1Part2
Uses script from Test 1.1 against a different video clip
Test #1.1aPart2
Uses script from Test 1.1a against a different video clip
Test #1.1.1.1
# PAL DV (50 fps) to NTSC DV (59.94 fps)
LoadPlugin("SmoothDeinterlacer.dll")
AVISource("PalDV2.avi")
ConvertToYUY2(interlaced=true)# Smoothdeinterlacer requires YUY2
SmoothDeinterlace(tff=false, doublerate=true)
Lanczos4Resize(720,480)
ConvertFPS(59.94)
SeparateFields()
SelectEvery(4, 0, 0)
weave()
|
Test #1.1.1Part2
Uses script from Test 1.1.1 against a different video clip
Test #1.1.1Double
# PAL DV (50 fps) to NTSC DV (59.94 fps)
LoadPlugin("SmoothDeinterlacer.dll")
AVISource("PalDV2.avi")
ConvertToYUY2(interlaced=true)# Smoothdeinterlacer requires YUY2
SmoothDeinterlace(tff=false, doublerate=false)
Lanczos4Resize(720,480)
ConvertFPS(59.94)
SeparateFields()
SelectEvery(4, 0, 3)
weave()
|
Test #1.1.1.1Part2
Uses script from Test 1.1.1.1 against a different video clip
Test #1.7a
|
# PAL DV (50 fps) to
NTSC DV (59.94 fps) |
Test #1.7Part2
Uses script from Test 1.7 against a different video clip
Test #1.7aPart2
Uses script from Test 1.7a against a different video clip
Test #12.0.2.1
|
loadplugin("kerneldeint140.dll") |
Test #12.0.2.1Part2
Uses script from Test 12.0.2.1 against a different video clip
Test #12.0.2Part2
Uses script from Test 12.0.2 against a different video clip