HyperSync & Beating the Shutter Curtain - Test Results

[Hermes]

Got my PocketWizard TT5's today and obviously the first thing I did after checking that they worked was try out the HyperSync with one of my studio strobes to see what it could do.

So far it's been one of the most frustrating, mind-boggling experiences I've ever had but I'm slowly beginning to understand how the technology works. I did a quick series of tests to try and figure out what the effects of changing various factors are, and I thought I'd share the results for anyone thinking of buying the new PWs or capable of offering any insight into what's going on.

The HyperSync was set to the maximum delay of -2500. I shot images at various shutter speeds, with and without FP / High Speed mode enabled in the PW transmitter, and with the strobe at full power for minimum duration and at minimum power for maximum duration. I then did the whole thing again but firing the strobe via a skyport transmitter in the TT5's hotshoe and the skyport RX transceiver to see if the Hypersync could be integrated into my existing setup and eliminate the need for a separate TT5 on each strobe.

Table of resulting images follows (sorry it's not neat and properly aligned but I didn't want to spend hours doing it).

http://img205.imageshack.us/img205/7149/sync1.jpg (http://img205.imageshack.us/i/sync1.jpg/)

[/URL] [URL=http://g.imageshack.us/img31/sync2.jpg/1/] (http://img31.imageshack.us/i/sync2.jpg/)

[Hermes]

http://img31.imageshack.us/img31/4415/sync2.jpg

[TMR Design]

Hi Hermes,

Based on those test results I would say it's not too promising. I know that many people had to play with settings a bit to get better performance. Perhaps you will have to do that as well.

[Hermes]

Rob,

From what I understand, the large black gradient at the top of the images taken at fast shutter speeds is due to the short duration of the strobe. Notice how it gets less severe at minimum power / max duration. The only obstacle that can't be dealt with is the first curtain at the bottom of the frame, and even at 1/8000 it isn't too obtrusive and could be cropped out if you were willing to frame the shot a bit taller/wider. If you're shooting outdoors where only the subject will be exposed by the flash and the background will be exposed purely by ambient light, the curtains can be worked around.

Should have mentioned that this was done with a 5D btw. With a 1 series the first curtain might be pushed back even further.

I'm guessing (and hoping) that If i get hold of a strobe with a long duration like a D-lite or a Genesis, the results will be better (the irony).

[Palladium]

a couple of issues come to mind immediately

are you use just 1 flex on your camera to trigger your strobes

or

do you have a flex on both the camera and strobe

[Palladium]

darkness gradient (shutter) at the top = too much delay

I'm not sure why you choose -2500 to run your test at

shutter at the bottom = no enough delay

you can fine tune somewhere in between to get a setting that you can work with

The camera your using can also factor into

[TMR Design]

Even if you were willing to frame and/or crop to compensate it doesn't give you much in the way of improved performance.

Yes, there is absolute irony in using a D-Lite or Genesis strobe as a workaround. I've never looked at the BXRi but would those work?

Rob,

From what I understand, the large black gradient at the top of the images taken at fast shutter speeds is due to the short duration of the strobe. Notice how it gets less severe at minimum power / max duration. The only obstacle that can't be dealt with is the first curtain at the bottom of the frame, and even at 1/8000 it isn't too obtrusive and could be cropped out if you were willing to frame the shot a bit taller/wider. If you're shooting outdoors where only the subject will be exposed by the flash and the background will be exposed purely by ambient light, the curtains can be worked around.

Should have mentioned that this was done with a 5D btw. With a 1 series the first curtain might be pushed back even further.

I'm guessing (and hoping) that If i get hold of a strobe with a long duration like a D-lite or a Genesis, the results will be better (the irony).

[Hermes]

a couple of issues come to mind immediately

are you use just 1 flex on your camera to trigger your strobes

or

do you have a flex on both the camera and strobe

For the images on the left I had a flex on both the camera and the strobe.

For the images on the right I had a flex on the camera with the hypersync still set, but with a skyport transmitter in its hotshoe and a skyport transceiver on the strobe. The hypersync delay is then routed throught the skyport (when the skyport transmitter is in the hotshoe without the flex, the frame is entirely black by 1/320)

[Hermes]

darkness gradient (shutter) at the top = too much delay

I'm not sure why you choose -2500 to run your test at

shutter at the bottom = no enough delay

you can fine tune somewhere in between to get a setting that you can work with

The camera your using can also factor into

You've got that the wrong way round. The shutter at the bottom is due to too much delay between the flash firing and the shutter opening.

[Palladium]

For the images on the left I had a flex on both the camera and the strobe.

...

OK - let's start using only the flex's

just looking at the first columns of the images your posted the delay is too much -

working with the hyper sync offset you should be able to dial in a much cleaner frame than what your showing - what happens when you dial the offset back to -1000

(I'm trying to find the thread awhile back where I posted some strobe tests with the mini / flex)

[RDKirk]

Should have mentioned that this was done with a 5D btw. With a 1 series the first curtain might be pushed back even further.

That's probably the worst camera to have test.

Hypersync takes advantage of the fact that many the sync speed of many cameras is actually a bit conservative--slower than necessary. For instance, the true sync speed (the highest speed at which the sensor gate is fully exposed) is actually 1/320.

The 1-series is faster because its shutter curtains cross the sensor gate at very high velocity--it's got sync speed to spare. The xxD cameras are faster because their slower consumer-grade shutters have a shorter distance to go and also have a sync speed cushion.

Hypersync takes advantage of that and re-times the sync accordingly.

Sadly, the 5D has a slow consumer-grade shutter that has a long, long way to go, so it's the slowest of them all and has zero cushion. It's just barely making the stated sync speed.

[reebar]

Thanks for doing these tests Hermes. I'm dying for the nikon versions of these new PW's to come out so I can replace my PWII's and get a hotshoe integrated into the module, hypersync, wireless ittl RF. One big thing is to push up the max sync on my d700 with the ranger A head (and optionally my skyports) so these tests are useful. I agree that it needs trimming so are you able to test a smaller offset ?

Also in the first column, you have a tt5 on camera and what did you have on the brx600, a PW2?

[TMR Design]

Thanks for doing these tests Hermes. I'm dying for the nikon versions of these new PW's to come out so I can replace my PWII's and get a hotshoe integrated into the module, hypersync, wireless ittl RF. One big thing is to push up the max sync on my d700 with the ranger A head (and optionally my skyports) so these tests are useful. I agree that it needs trimming so are you able to test a smaller offset ?

Also in the first column, you have a tt5 on camera and what did you have on the brx600, a PW2?

I'm also very curious to see how the Nikon versions work with the D700 :D

[Palladium]

see this thread and this post

http://photography-on-the.net/forum/showpost.php?p=8053321&postcount=51

[Hermes]

OK - let's start using only the flex's

just looking at the first columns of the images your posted the delay is too much -

working with the hyper sync offset you should be able to dial in a much cleaner frame than what your showing - what happens when you dial the offset back to -1000

(I'm trying to find the thread awhile back where I posted some strobe tests with the mini / flex)

That's probably the worst camera to have test.

Hypersync takes advantage of the fact that many the sync speed of many cameras is actually a bit conservative--slower than necessary. For instance, the true sync speed (the highest speed at which the sensor gate is fully exposed) is actually 1/320.

The 1-series is faster because its shutter curtains cross the sensor gate at very high velocity--it's got sync speed to spare. The xxD cameras are faster because their slower consumer-grade shutters have a shorter distance to go and also have a sync speed cushion.

Hypersync takes advantage of that and re-times the sync accordingly.

Sadly, the 5D has a slow consumer-grade shutter that has a long, long way to go, so it's the slowest of them all and has zero cushion. It's just barely making the stated sync speed.

I should point out that my goal here is to try and get the maximum possible shutter speed. If I were trying to optimise the 5d for 1/250-ish then I'd probably set less delay but frankly the gain from 1/200 to 1/250 is pitiful. I can already get 1/400-1/500 with no real shutter intrusion by fiddling with the FP settings but even that would only reduce the ambient by an extra stop and not be worth the money and space in my camera bag for the PWs.

What I'd really like to experiment with is exposing the entire frame at the top shutter speeds to overpower the sun in bright conditions without huge battery packs. The results show that I'll need a flash with a longer duration but the TT5 is at least capable of holding the first-curtain back.

[Hermes]

OK, here is a picture for those of you who want to see a shot with less delay for syncing at slower speeds.

This was taken at 1/500. There's a faint hint of a shutter curtain still visible at the bottom but nothing that would show in an outdoor shot. The delay needed was -2200, any less and the first curtain became properly visible. The gradient at the top is still there to an extent as the shutters are both moving at this speed and my strobes have short durations.

[TMR Design]

OK, here is a picture for those of you who want to see a shot with less delay for syncing at slower speeds.

This was taken at 1/500. There's a faint hint of a shutter curtain still visible at the bottom but nothing that would show in an outdoor shot. The delay needed was -2200, any less and the first curtain became properly visible. The gradient at the top is still there to an extent as the shutters are both moving at this speed and my strobes have short durations.

I suppose I could live with that degree of shutter curtain at the bottom but not the gradient at the top. I'd rather use ND's than mess around with all the gradients and vignetting.

So far I'm really not convinced that the PW's are worth it. I hope the Nikon versions are magical and work like a charm. :cool:

[Hermes]

Even if you were willing to frame and/or crop to compensate it doesn't give you much in the way of improved performance.

Yes, there is absolute irony in using a D-Lite or Genesis strobe as a workaround. I've never looked at the BXRi but would those work?

Seems the BXRi 500 has a t5 of 1/1558 which is twice as quick as the entry-level stuff.

I'll have to get my hands on some one way or another and test it out as even having a t5 longer than the shutter speed isn't a guarantee of getting even illumination - you can still get up to a stop of difference across the frame.

[Hermes]

I suppose I could live with that degree of shutter curtain at the bottom but not the gradient at the top. I'd rather use ND's than mess around with all the gradients and vignetting.

So far I'm really not convinced that the PW's are worth it. I hope the Nikon versions are magical and work like a charm. :cool:

I'm sure they'll come out smelling of roses like all Nikon gear :)

The point about the gradient at the top though is that it probably can be gotten rid of in camera of by using a strobe with a longer duration. Even the Quadra has double the duration of the 600RXs so it won't necessarily mean using unsuitable equipment.

[TMR Design]

Do you have, or can you borrow a D-Lite to test? The Genesis might be good but unfortunately it has a standard umbrella mount and no center mounting hole for an umbrella or deflector.

[Palladium]

OK, here is a picture for those of you who want to see a shot with less delay for syncing at slower speeds.

This was taken at 1/500.

5D syncs @ 1/200 so your over by a stop

[Hermes]

Do you have, or can you borrow a D-Lite to test? The Genesis might be good but unfortunately it has a standard umbrella mount and no center mounting hole for an umbrella or deflector.

Sold all mine but I can try and borrow one. The thing is though durations typically get longer as the strobes become more powerful and the D-lites are only available in 400ws. An analog 1000ws or 1200ws strobe might be more suitable if I can find one.

[TMR Design]

5D syncs @ 1/200 so your over by a stop

1 1/3 stops.

I wouldn't buy into the system unless I could get at least 2 stops more, which on my D700 and Elinchrom 600RX would be 1/640s or better. Currently I can sync a clean frame at 1/160s and I wouldn't bother unless I was still getting a clean frame.

[TMR Design]

Sold all mine but I can try and borrow one. The thing is though durations typically get longer as the strobes become more powerful and the D-lites are only available in 400ws. An analog 1000ws or 1200ws strobe might be more suitable if I can find one.

Perhaps an R1000?

[Hermes]

1 1/3 stops.

I wouldn't buy into the system unless I could get at least 2 stops more, which on my D700 and Elinchrom 600RX would be 1/640s or better. Currently I can sync a clean frame at 1/160s and I wouldn't bother unless I was still getting a clean frame.

I'd agree with that.

Thankfully I bought the TT5s for reliable ETTL which they seem to do well so the hypersync thing is just a bonus for me. Probably won't go through the trouble of using them with manual strobes UNLESS I can find a way to get a good deal of free shutter speed out of them when working outdoors. Don;t see myself ever using them in the studio. They've already misfired 8-9 times when using the sync port which is more than my sktports ever have.

The 1000r is a sports strobe designed for short durations but I think there's a non-sports version too, will have to do some searching.