Testing
 

Home
Back Story
Details
Theory
Other Tests
Testing
Results
Conclusion
FAQ

The purpose of this test is to measure the latency between a video source and the displayed image of an HDTV.

What you need:

  • Laptop w/ video output (we used S-Video, do NOT use a vga connector) and appropriate cabling
  • Digital camera
  • Standard definition CRT television (for calibration) with the appropriate input for the laptop's video output
  • HDTV with the appropriate input for the laptop's video output
  • Timecode

Calibration

It's important that you first develop a baseline for your equipment. You also need to make a few assumptions. Firstly you need to assume that your laptop's display and video output are essentially instanteous. You also need to assume that your video cable connecting the laptop to the test display isn't going to add any noticeable amount of latency to the test.

You can test these assumptions by first performing the test on a standard definition CRT television. This will help convince you that even though there may be a slight margin of error, it's not going to be terribly significant to the test overall.

In order to calibrate against the CRT, connect the video output of the laptop to the appropriate input of the television. It is important that you use either a composite or s-video connection here instead of a VGA connector if you're TV has one as we're specifically looking to test the 480i scaling and image perfection routines. The analog CRT likely doesn't have any of these routines, but you should keep everything as similair as possible during the tests.

Try to keep the laptop in native resolution if possible. turn off the image scaling if necessary and let your laptop's LCD frame the screen in as much black as is required to keep it from scaling the image.

Once the laptop video is being properly displayed on the television, it's time to run the timecode. All  you need is 30 fps timecode, something to visually identify each frame on the displays. Get this timecode running on both displays (you'll need to make sure your laptop is set to clone your main display to the television) and set them up so that they are side-by-side.

At this point you should have your CRT and your laptop with running timecode side by side. Pull out your digital camera and set it to a fast shutter speed if possible. Make sure the room has lots of light and turn off the flash on your camera. Make sure there isn't any glare on the screens at all either. Then take a picture with both the CRT and the laptop screen in the frame.

You'll end up with a photo that shows both screens with numbers on them. In this particular test, using our assumptions above, the numbers should be the same. If the CRT is ahead of the laptop, your laptop display is really slow and you should find another laptop for testing. If the CRT is behind the laptop, either your video output is being delayed at the laptop, or the cable is bad, or the CRT tv itself is doing something funny to the image before displaying it to the screen.

Testing

Once you've got a working baseline configuration, you can proceed with actually testing an HDTV.

Repeat the same process outlined in the calibration procedure.

Connect the laptop to the test display and run some timecode. Set the laptop alongside the HDTV and take a digital photo of the two screens in the same frame.

If your HDTV has a game mode, you should turn that on to try and squeeze a bit more performance out of the television. You should also turn off any image perfection routines in the HDTV. You want to get the tv at its fastest possible speed, then you can start worrying about the image quality.

Analysing the Results

It's quite likely that you're seeing the timecode out of sync anywhere from 1 to 3 frames. At 30 frames per second, each frame that the HDTV is out of sync is 33 milliseconds.

This is the amount of video latency that the HDTV's video processor is adding to the input source.

This is HDTV Lag.

Next : Results