G1 - Closed eyes problem with flash?

I have some very fast eyes in my family, this creates "lazy eyes" or "closed eyes" problem with flash, due to too long preflash and mainflash time delay.

Many years ago changed my Nikon D70 to Canon 20D system, because Nikon had terrible closed eyes problem. Last year I changed my Sony A350 system to Canon 450D for the same reason.

This delay must be less than 58 milliseconds to be fast enough.
100 milliseconds: Nikon D70
80 milliseconds: Sony A350
56 milliseconds: Canon 20D, Canon 450D

If someone visits Panasonic booth in PMA09, please ask what is the preflash and mainflash time delay of G1 and GH1? They must know it, because it is easy to measure with oscilloscope.

I should also appreciate a lot of G1 users experiences of this problem, does it exist with G1?

I do not want to change my Canon 450D system to Panasonic GH1 just to find out this very same problem
--
Antti

Antti Roine said:

This delay must be less than 58 milliseconds to be fast enough.
100 milliseconds: Nikon D70
80 milliseconds: Sony A350
56 milliseconds: Canon 20D, Canon 450D

Click to expand...

Just for interest to add to the list....

12.8 milliseconds: Olympus E-300

And that is a problem as the 12.8 mS is too fast for slave flash digital triggers, they seem to need maybe 40 mS apart to work.

I definitely never have blink problems with the E-300.

Regards............ Guy

Also curious about flash timing, has anyone tested this on the G1?

I used a hand held CRO and a simple photodiode setup to check a few cameras for that double flash timing. No G1 available so not tested.

More to the point, do G1 users get the closed eye syndrome when using flash?

Regards.......... Guy

I have no way of measuring the delay, but in the name of science, I tried the experiment of photographing my own face with the flash (and blinding myself in the process). This is using an external Targus DL-20, not the internal flash. Out of 16 pictures, I got 2 with a sort-of-sleepy look. I'd guesstimate that I get about the same percentage when taking pictures of our 4-year-old.

I vaguely recall that the chance of catching a blink by chance with a fast shutter is pretty low (something like 1 in 50 to 1 in 100, but I'm not sure), so I'd say that the preflash is slow enough to cause some chance of sleepiness with an "average" person, but not so slow as to guarantee it.

A small tip I can give you that works for me:

Tell your subjects to close their eyes, then as the pre flash starts shout 3.2.1 open!

Pretty much works every time with any camera so long as you get your timing correct and your subjects aren't hard of hearing!
Hope this helps.
--
http://www.flickr.com/photos/budvarholsten/

Once I thought I was wrong, but I was mistaken!

I use my FL-36 set to auto to overcome this problem with my e-500. No preflash in that mode.

Hi,

Many thanks for the tips

Some 5 - 10 % of the peoples has so fast eyes that they close their eyes if the preflash and mainflash time delay is longer than 58 milliseconds. This means that the eyes are closed in nearly every photo - it is amazing.

It is normal that the eyes are closed in some 2 - 5% of the flash photos.

I do not like the tricks, I just want to know:

1) Does closed eyes problem exist with G1 or GH1?
or
2) What is the preflash and mainflash time delay of G1 or GH1?
Panasonic know it but for some reason they do not give this value on brochures.

--
Antti

This is a common problem for me with the Olympus E510 and E520 DSLR's.

When the E system first came out, I never had this issue with the E1 and E300 models and still don't today when I pick one of those up out of my bag to use, with either an FL50 or FL50R flash, but Olympus changed something about the flash system in the mid and lower lines as the more recent E510 and E520 cameras have a slightly delayed exposure burst.

The Olympus flash system puts out two bursts that occur at virtually the same time when used in TTL mode. The first burst is used by the system to determine the correct exposure and the second burst actually takes the picture. With the E1 and E300 DSLR's the burst is virtually at the same time and I have NEVER, not even once, had subjects blink on me using either of those two cameras, but since moving to the more recent E510 and E520 bodies I have noticed you can actually, very briefly, see the initial burst prior to the mirror raising and taking the picture, and the delay is just enough to cause some people to blink and their eyes be closed at the moment of exposure.

I was using the E510 for an event one evening and had to resort to non-TTL auto, which does away with the pre-flash, because the main guest of honor was constantly closing her eyes whenever I would take a shot of her. Once I changed to non-TTL auto I never had the problem again.

I have also used my FL50 flash with a Panasonic FZ50 and the delay was even worse with that camera. When I used it for flash shots of people I always had to use the flash in non-TTL auto with the FZ50.

Hi Gregm61,

Many thanks for the comments. I have also considered the new Olympus E-620, because it has a high quality flexible LCD monitor and real live view. My Canon 450D live view is a joke.

However, Olympus do not tell anything about the preflash-mainflash delay times to be able to estimate does closed eyes problem exist with E-620...

I had Olympus OM2 and OM2n some 25 years with real TTL flash and no closed eyes problem. So the development has went in reverse direction in this respect.

--
Antti

Antti Roine said:

Hi Gregm61,

Many thanks for the comments. I have also considered the new Olympus
E-620, because it has a high quality flexible LCD monitor and real
live view. My Canon 450D live view is a joke.

Click to expand...

The E620 does look very nice.

Member said:

However, Olympus do not tell anything about the preflash-mainflash
delay times to be able to estimate does closed eyes problem exist
with E-620...

Click to expand...

I have no doubt the system is like what I now have with the E520. I will say, non-TTL Auto is very accurate with the FL50 and FL50R, and the blinking eye thing in TTL is very dependent on the subject. I often have no issues and you have to shoot TTL when using the wireless technology, but the people who are prone to blink will almost every time. Whatever the system was when the E1 and E300 were the current models, Olympus really needs to go back to it.

Member said:

I had Olympus OM2 and OM2n some 25 years with real TTL flash and no
closed eyes problem. So the development has went in reverse direction
in this respect.

Click to expand...

OM's are wonderful machines.

Gregm61 said:

When the E system first came out, I never had this issue with the E1
and E300 models

Click to expand...

As I said, my E-300 times at 12.8 mS between flashes, so no blink problems, but way too fast for digital slave flashes. So they slowed it down.

For interest, from testing other cameras (pocket sized types) back in January I came up with the following extracted from my slave flash thread at the time....
=================================
Olympus C-5060 140 mS
Olympus C-730 110 mS
Ricoh R4 130 mS

Casio V8 - odd one, three flashes, first at 0 then second at 38 mS then final at 120 mS from first one

Olympus E-300 varies from 11.5 mS to 12.5 mS [I used 12.8 mS above as that was the slowest timing found in many tests, but usually was 11.5 to 12.5 mS]

So the very quick timing of the E-300 explains the reason why I could not get the digital slaves to work, unless I switched to non-TTL to provide only one flash and set the slave to first flash trigger.

The Casio triggers digital slaves OK, and that seems to indicate that the digital slaves may totally ignore a flash that occurs within (in this case) 38mS of the initial flash.

Also in doing this test it reveals to me how I see flashes, the Oly C-5060 was a distinct "pop-pop" but the others were not so obvious that it was a double flash. My own flash detection threshold must be close to 130 mS.
=====================
Now if someone could lend me a G1 I could go test that as well.....

Regards............ Guy

Guy Parsons said:

Antti Roine said:

This delay must be less than 58 milliseconds to be fast enough.
100 milliseconds: Nikon D70
80 milliseconds: Sony A350
56 milliseconds: Canon 20D, Canon 450D

Click to expand...

Just for interest to add to the list....

12.8 milliseconds: Olympus E-300

Click to expand...

The preflash fires with the mirror down, camera takes a metering, lifts the mirror, opens the shutter, fires the main flash, closes the shutter.

The fastest sports pro DSLRs have a pre-focussed shutter lag of about 30ms. There is no way the E-300 can swing its mirror to the side fast enough to achieve 12.8ms, so that must be some other timing.

Or am I overlooking something?

Just my two oere
Erik from Sweden

Ehrik said:

The fastest sports pro DSLRs have a pre-focussed shutter lag of about
30ms.

Click to expand...

Actually I can't find anyone faster than 44ms, which makes 12.8 ms even more impossible.

Just my two oere
Erik from Sweden

Ehrik said:

Guy Parsons said:

Antti Roine said:

This delay must be less than 58 milliseconds to be fast enough.
100 milliseconds: Nikon D70
80 milliseconds: Sony A350
56 milliseconds: Canon 20D, Canon 450D

Click to expand...

Just for interest to add to the list....

12.8 milliseconds: Olympus E-300

Click to expand...

The preflash fires with the mirror down, camera takes a metering,
lifts the mirror, opens the shutter, fires the main flash, closes the
shutter.

The fastest sports pro DSLRs have a pre-focussed shutter lag of about
30ms. There is no way the E-300 can swing its mirror to the side fast
enough to achieve 12.8ms, so that must be some other timing.

Or am I overlooking something?

Click to expand...

That's definitely how it works with the E510 and E520 DSLR's....probably the E620 too. I can see the pre-flash in the finder of my E520 prior to the mirror raising, but with the E1 and E300 you do not see the pre-flash. It is substantially faster with both of those cameras. My prime test subject is my cat. The pre-flash of the E510 and E520 cause him to close his eyes for every exposure I try to take of him in TTL mode, but I can shoot him with my E1 or E300 in TTL mode and his eyes are never closed, so the E300 and E1 use a flash sequence that is substantially faster than the more recent Olympus consumer models.

Gregm61 said:

Ehrik said:

Guy Parsons said:

Just for interest to add to the list....

12.8 milliseconds: Olympus E-300

Click to expand...

The preflash fires with the mirror down, camera takes a metering,
lifts the mirror, opens the shutter, fires the main flash, closes the
shutter.

The fastest sports pro DSLRs have a pre-focussed shutter lag of about
[44ms]. There is no way the E-300 can swing its mirror to the side fast
enough to achieve 12.8ms, so that must be some other timing.

Click to expand...

That's definitely how it works with the E510 and E520
DSLR's....probably the E620 too. I can see the pre-flash in the
finder of my E520 prior to the mirror raising, but with the E1 and
E300 you do not see the pre-flash. It is substantially faster with
both of those cameras.

Click to expand...

Well the metering is normally taken off the focus screen (sensors up in the viewfinder) or on occasion the AF sensors double as (spot) meter. But in either case, the mirror has do be down for this to work. The E-300 has a pre-focussed shutter lag of 0.10s (100ms) according to IR. This is the time it takes swinging the mirror aside plus a little extra. There is no way the mirror is swung aside anywhere near 12.8ms.

The E-300 doesn't have live view, so we can rule out the imager doing the metering (that would be hard anyway).

I wonder if the pre-flash is the external flash doing auto-metering, so the E-300 isn't doing TTL metering at all?

Gregm61 said:

My prime test subject is my cat. The pre-flash
of the E510 and E520 cause him to close his eyes for every exposure I
try to take of him in TTL mode, but I can shoot him with my E1 or
E300 in TTL mode and his eyes are never closed, so the E300 and E1
use a flash sequence that is substantially faster than later consumer
models.

Click to expand...

Yes the flashes fire close together, but it can't be doing a proper TTL flash metering.

Just my two oere
Erik from Sweden

There's no doubt about it. I measured and re-measured the E-300 timing many times over and it always was that normal 11.5 mS to 12.5 mS with the odd one out to 12.8 mS as noted.

So obviously they use some other method. Maybe they were trying some form of real TTL where they were measuring flash reflectance off the front side of the sensor/filter/dustshaker bundle, or maybe off the surface of the shutter before it opens. I now think that I always used first curtain sync so next time that I dust off the test setup then I'll try second curtain sync as well to see what happens. If it's shutter reflection then the time will stretch, if it's sensor reflection then the timing will stay the same.

Aha! Just tried it now without the CRO etc, and first curtain sync looks like one flash it's so fast, second curtain sync is two distinct flashes, one at start of operation and one maybe much later just before the shutter closes. So my guess is that they measure the reflection off the shutter blades just before they open.

I also proved that by turning on a 2 second delay from mirror up to shutter fire and the first flash is always after 2 seconds, so must be measuring off the shutter blades. Presumably the later E-520 etc would fire the first flash before the mirror flops up when you have mirror pre-fire (anti-shock in the menu).

Despite the possibly odd method that the E-300 uses I never have any problems with "TTL" flash, it always delivered nice results. It's just that second trigger digital slaves didn't work due to the close timing of the flashes.

Regards.......... Guy

That's it, I remember reading about that method long ago, but it didn't occur to me writing the posts above. The shutter blades may be painted partly grey on the outside for this purpose (you may check by doing anti-shock, MLU, without a lens attached, I suppose).

So that explains the mytstery, thanks!

Guy Parsons said:

Despite the possibly odd method that the E-300 uses I never have any
problems with "TTL" flash, it always delivered nice results. It's
just that second trigger digital slaves didn't work due to the close
timing of the flashes.

Click to expand...

In principle, normal pre-flash should be more accurate, but if it's accurate enough then nothing to worry about. Maybe the slave problem was the reason they changed, but if they get "lazy eyes" instead it's an "improvement" with a drawback.

Just my two oere
Erik from Sweden

Ehrik said:

That's it, I remember reading about that method long ago, but it
didn't occur to me writing the posts above. The shutter blades may be
painted partly grey on the outside for this purpose (you may check by
doing anti-shock, MLU, without a lens attached, I suppose).

So that explains the mytstery, thanks!

Click to expand...

Yup! When looking at the E-300 shutter blades (either move the mirror by finger or use a long delay between mirror and shutter) they are a lovely shade of grey, but looks to be a paler grey than 18% grey.

The measuring method works OK, as one shot may be a room wide shot and exposure good, then next shot a small person close to the camera and to the side of the frame is exposed OK and the background gets under-exposed of course. Never seem to get blowout of subjects too near to camera whether using pop-up or FL-50 in TTL.

Regards......... Guy