Does this flash during have some to do with our shutter spec? Meaning
is the t0.1 is 1/200 sec. We need to use shutter slower than 1/200
sec in order to capture all the flash power?
I have and use the DLite-4 flashes. You risk getting a dark band (due to the shutter curtain) if you shoot them at 1/250th, depending on the conditions and your camera and way of trigging etc. 1/200th usually works well.
I shoot mine with a D300, usually. The D300 has a rated sync speed of 1/250th for external flashes, but allows the built-in to be used at 1/320th under certain circumstances. What it means is that there is "some time" during which the entire CCD is exposed to light -- or in other words, the second curtain does not start its travel until the first curtain is fully open.
The spec of the camera does not say anything about how long the CCD is fully exposed. But at 1/400th, the second curtain probably start moving before the first is fully open; otherwise 1/400th could have worked as sync speed for ultra-short flash durations.
From this, we could -- as a rough approximation -- assume that the entire CCD might only be exposed around (1/250th-1/400th), i.e. 1.5 mS or
1/650th of a second at 1/250th. It also means that the first curtain needs around 2.5mS to move to the end stop, and that any shutter time needs to have 2.5mS subtracted from its duration to calculate how long the entire CCD is exposed.
This does not take into account that the active CCD area is not covering the entire opening. In fact, Marianne Oelund has actually measured this, and arrives at 1.6 mS (1/625th) for the real time the shutter is open and the CCD is exposed fully, at 1/250th:
http://forums.dpreview.com/forums/read.asp?forum=1039&message=30477133 -- and it means also that the shutter needs 2.4mS to travel across the CCD.
If the shutter speed is 1/200th, the second curtain waits another 1/1000th of a second before it starts to move. This almost doubles the time the CCD is fully exposed. At 1/160th, it adds another 1/800th, and at 1/125th another
1/500th or so. Hence, we might assume that whatever time the flash duration is longer than in the following table will be lost, and you will see a dark band at the bottom of the picture (depending on camera orientation ;-):
Speed Max flash duration (rounded a bit)
1/320th
1/1400th
1/250th 1/625th
1/200th
1/400th
1/160th 1/250th
1/125th
1/200th
1/100th 1/125th
This assume no delay in sending the sync signal. I use the Elinchrom Skyport triggers, which I've measured with a D70 with an electronic shutter to be reliable at 1/1000th sync speed (for short flash durations) with fresh batteries, dropping just a bit to somewhere around 1/800th when running low on batteries, and unreliable at 1/1250th sync speed. This means that the skyports needs just short of 1mS or 1/1000th to send the sync signal to the flash, and that amount needs to be subtracted from the max allowed flash duration.
If you use radiotriggers with a delay around 1/1000th (and I think the skyport and the PW are about the same, with the eBay "Poverty Wizards" much worse), the table would look like this:
Speed Max flash duration w/radiotriggers
1/320th No sync(!)
1/250th
1/1700th
1/200th 1/625th
1/160th
1/350th
1/125th 1/250th
1/100th
1/160th
(Hope I did all the calculations right and didn't mistype anything).
The SB800 (and most other similar flashes) have very short durations, and last longest at full power. The DLite's are shortest at full power, and will last longer at min power. At half power, the DL 2 is rated at 1/1200 and the DL 4 at 1/800.
I think Nikon specifies t0.5 timing. I also seem to remember that e.g. Metz is giving their flash duration in t0.1 timing. Probably most manufacturers give the t0.5 timing (looks best on paper), so if you don't know for sure, assume t0.5 timing... The t0.1 timing is then around 3x this, in most cases following exponential decay.
Further, as t0.5 is the time at which the flash is outputting half its intensity, we could (mistakenly!) approximate that the capacitor has then released half its energy. What that would mean in practice is that if the actual exposure time is t0.5, the lower part of your picture would be around 1 stop darker due to the 2nd shutter curtain cutting the exposure short. At t0.1, it would be far less and barely noticeable (close to 1/6th of a stop, in fact). If you accept, say, a 1/2 stop effect at most, in round numbers assume you need twice the t0.5 time.
Luckily, much more than 50% of the light is emitted at t0.5, so it would never get as bad as this.
The D-Lite 4 is rated at 1/800th (which I interpret as the t0.5 time at half power; a bit longer at lowest power). To be 100% sure you get less than 1/6 stop darkening using 1mS radio triggers, you probably need to shoot above 1/160th -- worst case. If you don't use radio triggers and shoot at 1/250th, you may actually get up to half a stop darkening at the very edge of the picture -- worst case. But I've never seen that (but using radio triggers is not recommended at 1/250th).
In practice, I don't notice any darkening when shooting at 1/200th with radio triggers, but 1/250th is not useable with the DLites. I might go to 1/160th in some cases with the DLites, to be on the safe side and still quench the ambient as much as possible.
-- Per.
