@lapser: I also have a SX210 IS (100c). Could you attach a build for that camera, so I can try if I run into the same problems like with my G1 X?
That would be great. Thanks!
Will do. I'll have to test single shot mode in all my cameras too. Thanks for helping to find this problem. Here's the answer to the questions you brought up in the G1X thread:
Sorry, very stupid question... I figured it out. It's just the setting in the canon menu 
But it's interesting: I've always shot time lapses in single shot mode. Now I tested in the continuous mode and it works perfectly.
So what is actually the difference between these two modes (except that you loose all pictures when you run out of power in continuous mode)?
90% of my questions are "stupid". And if I don't feel stupid after figuring something out, I probably haven't really figured it out.
I'm really glad you tested it in single shot mode, since I haven't used single shot very much. I tested the script on my G1X in single shot and it brings out the problem for me too!
There are no real advantages of single shot mode with my time lapse script, except the possibility of losing pictures with an external battery. The solution to that problem isn't to use single shot mode though. It's to stop the script when the external battery gets near its low voltage cutoff.
A script just simulates pushing the shutter button to take pictures. You push the shutter yourself, without a script, to see what happens.
In single shot mode, the script pushes the shutter all the way down, and lets it up half way between pictures. The advantage of letting it up only half way is that the camera doesn't re-focus and measure exposure before each shot. All the pictures are taken with the same exposure as the first shot.
In continuous mode, the script holds the shutter all the way down for the entire time. The focus and exposure behavior is the same as single shot, but the shot rate is about twice as fast. In addition to the faster shot rate, the minimum time between shots is about .25 seconds in continuous mode, and .75 seconds in single shot. This is great for night shots, because you don't miss as many brief events, like meteors.
My CHDK build includes a shot meter function that measures the brightness of 4 different image areas after the shot, from the raw image when it's in memory. It's this raw image that was color shifted, and caused the bad meter readings and fluttering. The script uses the meter readings to set the exposure of the next shot, in 1/96 F-stop (ev) units.
In-camera real time shot metering is something that I don't think has ever been done before. It's much more accurate, and faster, than the built in camera exposure meter, especially at low light levels. The disadvantage is that you don't know the meter reading until after the shot. But if you're willing to take multiple shots, like with a time lapse, there are lots of possibilites, like optimum exposure bracketing and HDR with a minimum of shots. The multiple metering areas are quite useful too.
But the most exciting use of shot metering to me is doing smooth exposure adjustments in day to night time lapses. This is the so called "Holy Grail" of time lapse photography. My script produces pictures that can be converted to a time lapse straight out of the camera, with no post processing of the images. I don't think there's any other way to do this. The only thing comparable is "bulb ramping" where an external controller holds the remote shutter down. There's are smartphone apps coming out that control the camera through PTP and can adjust exposure between shots. I assume they use the phone as an exposure meter, but I haven't tried them. I think you still have to post process the images to get rid of the flickering.
The other major addition to CHDK in my build is the set_shot_interval(interval) function. What this does is hold the camera at the point immediately before it opens the shutter until the time from the last shot is equal to the specified interval. The time between shots is perfectly accurate. When I take time lapses with 2 different cameras at the same time, they stay perfectly in sync with each other as long as the shutter time is at least ~300 msec less than the interval. This has real potential for multi-camera synchronization without requiring usb cables connected to each camera.
[EDIT]
Another big advantage of continuous over single shot mode is that the pictures are displayed on the screen between shots, as you've probably noticed. In single shot, the screen is blank most of the time.
New build is attached now:
