I have been playing with my camera to try to determine why the ISO 200-1600 video modes are so dark, and I think I've figured something out. It appears Sanyo has a significant error in the manual. I think the reported video ISO values are actually for photographs, and vise versa.

Here is what I've done so far:

Using my dSLR, I composed a shot on all manual that matched the camera settings as presented on the display... ISO 200, Shutter speed .5 and shot it. The exposure was much too dark. I then shot the image using ISO 800 and the exposures were a match... not perfect exposure but matching, and just slightly over-exposed.

I then took video and it was exceedingly dark... as dark as my original ISO 200 image from my dSLR. IT TOO WAS A MATCH!

I have performed this test twice, and am certain Sanyo's posted ISO 200-1600 is for the camera, while the ISO 50-400 is for video. It would explain why the camera takes dark video (supposedly at ISO 1600) but bright pictures (at ISO 400).

Can someone confirm similar function on their HD1? If you don't have a dSLR, just try shooting a few images and video in a dark location... somewhere just a little too dark to capture video. If the manual is correct, the digital picture should be darker than the video.

Ok, I'm a n00b and could be all wrong, but imho:

Stills and video can't be compared like that. Video has an exposure time of 1/30 (or if shooting HR: 1/60) of a second so it will always be darker than a still shot taken with a 1/2 exposure time regardless of ISO settings... but like I said I could be completely wrong :) I would suggest doing the still at 1/30 to do a valid comparison of the ISO settings between stills and video. I'm willing to be educated here if I'm wrong tho :)

I can't believe I didn't think of film's frame rate constraints... I know better than that.

On further review:

The image's EXIF data matches what the manual reports, and when I use my dSLR at the same settings as EXIF, I get a properly exposed picture.

I set up a more useful test with a dynamic subject, my cupboard. It has a vast array of colors and shades. I then matched everything... F-Stop, shutter speed, ISO, lighting, etc using EXIF data. At a fleeting glance, the image from the dSLR contains a vast amount more detail, and looks brighter. After sampling sections in Photoshop, I find that the light areas in the photo are of equal brightness (within 2%)... it's only the dark areas that differ. The MD1 displays nothing in the shadow areas (all black), whereas my dSLR has clearly defined objects. For example, I can read the writing on the cereal box in the dSLR image; a box that doesn't even show on the pic from the MD1.

I'm thinking there must be some post processing going on in my dSLR to pull those low-light details out... and of course a more capable sensor. Interestingly, if I apply the right set of filters to the HD1 image while in Photoshop, I get a similar amount of (much noisier) detail in those shadow areas. My cereal box appears, and I can almost read it.