Sports Video in High Definition
Comparison of High Definition to Standard Definition for sporting events
Here are captured 2 snippets of a basketball game, with a JVC DV2000 Digital Video standard definition camcorder, and with a JVC JY-HD10 high definition camcorder.
The DV2000 is in progressive scan mode, capturing 30 frames per second, of the usual 720 by 480 pixel frames.
The HD10 is in HD mode, capturing 30 frames per second of 1280 by 720 pixels.
For sporting events, I strongly prefer progressive scan mode, which is relatively uncommon. You can see more information about it at Home Theater and HiFi. All of the files below are shot in progressive scan mode. As you play them, try a pause or slow motion. Notice the details of player actions which are visible. Also note that steady playout has a "jittery" appearance. This is a tradeoff between the feel and the image quality. I prefer to get precise pause and slow motion images. Others prefer the smooth playback, which when paused, yields scenes blurred beyond all recognition (also referred to as "film quality").
Additionally I have compressed to DVD quality MPEG2, using tmpgenc, at 9200Kbit/sec, CBR 4:3 aspect ratio (the HD material becomes letterboxed), progressive scan, motion search precision standard. These files are called MPEG2 in the table below.
I wanted to supply the orginal file as captured from each camera, but I don't know how to cut the .m2t file, so instead I used tmpgenc to create WMV9 files encoded at 20,000 Kbit/sec. They appear to me to be about the same as the raw material, visually. These files are called "original" in the table below.
You can attempt to answer two questions from these files. 1. how does the image quality comapre between DV and HD? Look at the "original" material of the same scene from each camera.
2. how does HD look when compressed down to DVD quality MPEG2? Look at the "MPEG2" material of the same scene from each camera.
3. See the advantages of progressive scan, when pausing a playback.
| few seconds of rebounding |
Captured with: |
|
| |
DV2000 |
MPEG2 here. 4Mbyte |
| |
|
original here. 9Mbyte |
| |
JY-HD10 |
MPEG2 here 3Mbyte |
| |
|
original here. 8Mbyte |
| |
|
|
| warmup shot |
|
|
| |
DV2000 |
MPEG2 here. 4Mbyte |
| |
|
original here. 9Mbyte |
| |
JY-HD10 |
MPEG2 here. 3Mbyte |
| |
|
original here 6Mbyte |
| |
|
original deshaken here 4Mbyte |
Any questions or comments, please email to Rich at richalt2 at yahoo dot com.
High Definition compression
Investigating methods of compressing HD material to practical sizes has shown excellent results using DivX. See DivX website here. I understand the DivX standard to be similar to MPEG4. Until we see a solid emergence of the HD DVD standard, and encoding standard and tools, this is a great way to save current HD material. Using DivX I have gotten 40 minutes of material to fit in a 2.3Gbyte file, compressing at 6Mbit/sec. See my results and workflow here, for creating High Definition DivX compressed video from the output of the JVC JY-HD10/GR-HD1 camcorder Transport Stream MPEG-2
Panning blur
Here are 4 shots of about 4 seconds each, where I look to explain the effect of panning blur. I believe it is mostly due to the selection of shutter speed. The series are all taken with the JY-HD10, on automatic (where I don't know the shutter speed setting, but is probably 1/30 or 1/60), then manually setting to 1/30sec, 1/60sec, and 1/250 sec. All are 720p, 30 frames/sec, then compressed to DV 480p. By the time you get to 1/250, you can see the substitution of the jittery look of the background, replacing the blurred look. Each file is MPEG2 encoded, at 9,200Kbit/sec, 480p progressive, 30frames/sec.
| ball moving down court |
JY-HD10 |
|
| |
Automatic setting |
MPEG2, 4Mbyte |
| |
Manual, 1/30 sec shutter |
MPEG2, 4Mbyte |
| |
Manual, 1/60sec shutter |
MPEG2, 5Mbyte |
| |
Manual, 1/250sec shutter |
MPEG2, 5Mbyte |
De-shaking a hand-held video
Information here.
EIA 1956 Resolution test pattern
Using the explanation from Beale Corner, http://www.bealecorner.com/trv900/respat/
I captured these two jpeg files of the EIA 1956 test pattern. My lighting was not that great, 3 65W bulbs on a white wall. I printed the test pattern on an inkjet printer, on glossy paper. Footage was shot by the JY-HD10 in HD mode. Captured, then cut a frame into Photoshop and saved as highest resolution jpeg, here,
aligned left and right edge 
aligned top and bottom edge 
The alignment at the edges is not perfect as the LCD viewfinder crops about 5% top to bottom, and about 10% left and right.
Look at "aligned left and right" image, the vertical lines (as they enter the center square) to see where the lines are no longer distinct. Thus horizontal resolution appears to be about 760.
Look at "aligned top and bottom" image, the horizontal lines to judge vertical resolution, about 510. |