Analog Video Resolution Tests

Fullscreen test images (both 720x480 in the original), with the detail area shown. The cats image was taken from a poster by a TRV900, and the histogram expanded in Photoshop to fill the full range. The resolution target was taken from a 30-bit, 600 dpi flatbed scan and transferred from the computer via DV to the TR7000.
DV Image from Firewire

Image intensity histogram of full screen cats image, taken from Photoshop. Note that it covers the full range from 0 to 100% intensity.

Resolution detail: about 540 lines
Original DV image on tape, transferred via firewire. DV image can be transferred indefinitely (9 generations tested) with zero loss: output bits = input bits, checked with binary file comparison.
TR7000 SVideo-> TRV900

Note histogram has shifted to the left.Also, the fullscreen image shrank slightly, picking up black bars at extreme left and right (about 4 pixels at left, and 6 at right) so image is 1.3% smaller horizontally. Both are probably analog adjustments which could be tweeked in the camera.

Resolution detail: about 480 lines.
Image after analog transfer over S-Video cable. TR7000 analog output sent to TRV900 analog input (then firewire to PC). Resolution is reduced, image shrinks 1.3% horizontally, and image is about 10% darker.
AG-1980 SVHS Playback

Histogram quite similar to the direct analog transfer between DV decks. The Panasonic SVHS deck does not shift the intensity appreciably, although it does introduce a little noise.

Resolution detail: about 400 lines.
Image recorded from TRV900 via S-Video output to Panasonic AG-1980 SVHS deck, then played back to TRV900 via S-Video input and via firewire to PC. Edit switch in "normal" mode, sharpening slider at midpoint, TBC on, SP mode.
AG-1980 VHS Playback

Histogram not shifted, but lighter features are spread out (more noisy) and the transfer curve looks funny near the black level.

Resolution detail: about 230 lines.
Same as above, but using VHS tape for standard VHS record/playback mode.

XING encoder enhances contrast relative to the LSX encoder. Many more pixels went to full black.

Resolution detail: about 260 lines.
XING encoder v2.20 (352x240, 1150kbps video, 1394.4kbps incl. audio). Screen grab from the Windows Media Player on playback, rescaled to 720x480 in Photoshop. The LSX encoder has the same resolution but less saturated/contrasty color.

Note: the VideoCD test was all digital. This image was a screen capture from playback of the file before writing to CD-R, but I did later burn a VideoCD with this file, and the VCD playback looked exactly the same. This may be a "best case" for MPEG compression because the image is a still frame: bits that would otherwise encode motion or change might be devoted to better resolution and color fidelity (although I don't know if that is actually done). When viewing motion pictures on VideoCD I have seen quite noticible blocky artifacts, but I don't see any on this static frame.

You may find Stephen Dawson's comparison of RF, composite, s-video, and component playback of DVD signals on a TV of interest.

Back to camera test page.
Back to TRV900 page.
All measurements by John Beale.  Copyright © 1999 John Beale