GS400 lens hood?

[Dennis Wood]

You raise some interesting points Tom. It makes sense though that the cam would report a "virtual" F stop as opposed to a physical correlation with the iris blades. The stinker on the GS400 is that in auto mode, it does not deviate from 1/60s shutter. This means in bright conditions you must switch to manual and up shutter speeds, add filters, or use one of the AE program modes that actually will change the shutter speed. I discovered this very quickly shooting ski stuff...same problem on the water.

You got me thinking so I strapped on my LED headlamp, put the cam in manual, zoomed to 50% so I could watch the iris and here's what I observed.

At OPEN (F1.6) the iris is fully open (diamond shaped).

F2.8 it closes a bit.

Now from F2.8 to F8, a square ND filter (looks to be one piece, but darker part way down) starts to slowly advance across the iris. Each time you increase F/stop, the ND filter moves up slightly. The iris itself does not appear to move at all.

From F8 to F16, the ND filter is fully deployed, and I can only assume that the iris then physicaly closes. At F2.8, it is still almost fully open.

Here's a pic posted up "the guru" Guy Bruner

[Tom Hardwick]

Because the internal NDs are slivers of uncoated gelatin they have to be placed at an angle to avoid introducing lots of nasty reflections. Not only that but there are many times when you're shooting through the edge of the filter - a nasty solution but workable simply because the filter is so well out of focus.

By far the best way - for a photographer - is to have manually introduced fiilters (TRV900, DVX100, VX2100, GL/XL Canons) - that way you're in control. When you select f/8 for it's depth of field you know you're getting f/8 - modern cams shoot nearly everything at f/4.8 with greater or lesser amounts of ND to make up the difference.

The diamond shape you observed Dennis is due the the Leica lens only having a two bladed diaphragm - like the very simplest Super-8 cameras of 30 years ago. Herr Leica would be turning in his grave if he knew. The TRV900 has a beautiful (but far more expensive) 6 bladed diaphragm, and this of course is more efficient (it opens to a perfect circle at maximum aperture) and gives much nicer 6 point highlights on sparkly water.

I must say I don't like hidden ND filters, but I have to admit that they do work well, especially for the millions out there who want their cam to work indoors under a single lamp and out on the ski slopes imn blazing sunshine. In the latter regard I'm very surprised to hear that the GS400 doesn't up the shutter speed automatically. How do you know this? Are you believing the readout when you hit 'display' later? Are you also believing the f/11 readout? Big grin.

tom.

[Dennis Wood]

Tom, I'd can certainly see how the F/stop is a calculated value, particularly after looking at the ND filter's operation. What I can accept though is that the shutter speed value is not changing in auto mode. We've done experiments on shutter timing and CCD charge times corelated with NTSC field timing (done by filming a marked disc spinning at 3450 rpm) and I can tell you that when this cam indicates 1/60s it is...at least under the shoot conditions in the experiment.

Also in very bright environments the picture is blown out in auto mode. Switching to manual shows the cam at its "virtual" F16, 1/60s and zebra bars everywhere. Once the shutter speed is upped sometimes as high as 1/250s, then the zebras disappear. The difference in the footage with these speeds is fairly obvious too. I've found auto mode with the GS400 useless on open ski hills with direct sunlight.

I'm interested in your thoughts though as I really haven't considered the shutter speed to be extrapolated.

EDIT: Sorry that experiment was done at 1/500s...and the CCD charge times were indeed 1/500s.

[Georg Liigand]

I also remember GS400 not changing shutter speed automatically. VX2100, for example, has a special setting in the menu called AUTO SHUTTER which probably makes it able to change the speed automatically.

[Tom Hardwick]

Quite correct Georg, and I have it switched off at all times. For any filming of movement even doubling the shutter speed to 1/100th sec (I'm a PAL man) gives a most noticeable staccato effect, and of course spoils any chance for making fluid slow motion later. Not only that but you've only soaked a stop, not much when an ND8 can easily soak 3 stops.

Interesting to hear that the GS doesn't up the shutter speed Dennis. The TRV900 does it, but the PDX10 does indeed choose smaller apertures if the light gets too bright for the three internal NDs and f/4.8. Of course this degrades the very fine picture quality, but going the other route (aperture priority automation) gives some very strange effects.

I've seen PDX10 footage where propeller aircraft have their propellers come to a standstill as they lift into the sky and then go round in reverse. Of course this is just the propeller blades strobing with the camera's shutter, but it doesn't half look strange.

High shutter speeds can look pretty good sometimes (think battle scenes in Ryan, Brothers, etc) but in small, mega-chipped camcorders CCD smear gets to be a real problem, and the PDX10 is unusable above 1/300th sec in my view. Camcorders with bigger 1/3" chips (VX2100) are much better in this regard.


tom.

[Georg Liigand]

Nice information, Tom.

I have the auto shutter currently switched on and it works quite well. I've also noticed that with higher shutter speeds the motion is not fluid and it might be disturbing sometimes.

[Dennis Wood]

I was a bit surprised on the shutter speed behaviour in auto too. Several of the program AE modes, however, will increase shutter speed. "Portrait" mode for example attempts to maintain a low f-stop (for what little DOF there is) by upping the shutter speed.