In my previous testing I made the comment……. “Well Hypersync works for me ….. sort of!”
I felt that it was too unstable to be relied upon under pressure, but I suspected that part of it might be me, so this time round I took way more care to make copious notes during my testing to figure out what was causing the instability.
This time I had the benefit of working with a personal friend and model Emilee, who has worked with me many times before and is used to the stop-start process that goes along with trying out a new technique
As mentioned before in my previous posts regarding hypersync, my main objective was to be able to use wide open apertures around F2.8 in bright sunlight, and to use flash to fill the shadows and possibly overpower the sun. Shooting at wide open apertures with long lenses provides the shallow depth of field that makes the model pop off the blurred background. But shooting at the maximum sync speed allowed by the camera, typically around 1/250th, in bright sunlight means that you have to stop the aperture down around f8 and this unfortunately brings the background into focus.
So we started with an ambient exposure of 100 ISO, F2.8 and shutterspeeds initially in the range of 1/,1000th to 1/2,000th of a second. I shot with my Nikon D3x, Pocket Wizard Mini TT1 transmitter fitted with the Pocket Wizard AC3 zone controller to separately adjust the power of each flash head. I used one, two and three 640WS Alien Bees moonlights, powered by Vagabond batteries. On each Alien Bees flash was a Pocket Wizard Flex TT5 transceiver each set to a different group, and mounted to each of the TT5’s was a Pocket Wizard AC9 Alien Bees Adapter. The main light was fitted with the 18” Omni reflector to create more punch, and the rim lights when used were fitted with the standard 7” reflectors and 30 degree grids to kill the spill back to the camera.
In order to try and develop a repeatable lighting approach, I systematically measured the model to flash distance for each shot and kept copious notes of flash power and camera settings, so I would be able to replicate it in future. I mainly shot with the 70-200mm f2.8 lens but also used my 200-400mm F4 lens which narrows the depth of field even more when shooting at longer focal lengths.
During this test I was able to obtain better lighting consistency from shot to shot, so I suspect that part of my frustration in earlier shoots was merely a lack of familiarity. However there still remained some instability, particularly in adjusting the light output, but overall I was very encouraged by the results. Hope you like them too.
Hypersync – Constant ISO, Aperture, Flash Power and Distance, yet varying the Shutterspeed
From my previous tests I wanted to confirm my previous observation of the effect of shutterspeed on the flash exposure, something that does not happen in traditional flash photography using shutterspeeds at or below the camera’s advertised sync speed.
So with a constant ISO of 100, aperture constant at F2.8, flashes at constant power and flash to model distance kept constant, I shot my model at shutterspeeds ranging from 1/1,000th to 1/8,000th of a second in one stop increments. As you can see from the image below the relationship with flash exposure is inverse and linear. Double the shutterspeed, and all other variables kept constant, the flash exposure, as well as the ambient exposure (obviously) drops by one stop.
As a sidebar, the reason for adding the palm leaves in the foreground was to have something at the edge of the frame that would be hit by the flash so I could see how far the dreaded shutter curtain banding was intruding into the image. And you can see the flash banding start to intrude onto the palm leaves on the right hand side of each image, especially noticeable at 1/4,000th and 1/8,000th shutterspeeds.
Hypersync – Constant ISO and Aperture, increasing Flash Power with increase in Shutterspeed
So I did another experiment to confirm it, but this time, each time I increased the shutterspeed, I increased the flash power by one stop. And as you can see in the image below, while the ambient exposure decreased linearly at 1 stop increments, the flash exposure remained constant.
So this further confirms that the flash exposure is affected in exactly the same manner as the ambient exposure …. An increase in shutterspeed above the advertised sync speed will affect the flash exposure at the same rate as it affects the ambient exposure.
And of course the banding is now way more visible with the extra flash illumination. For my use, the banding is not negatively affecting the image at shutterspeeds up to 1/2,000th but as I get above that I will need to pay attention to how I frame the image if I have any objects that will receive flash illumination in the foreground. As long as I frame the image to allow for the banding I will be able to shoot with flash all the way up to 1/8,000th of a second.
Cool – now I understand it better and kind of know what I want to do next!
As mentioned before in my previous blog posts regarding hypersync, my main objective was to be able to use wide open apertures around F2.8 in bright sunlight, and use flash to fill the shadows. Shooting at wide open apertures with long lenses provides the shallow depth of field that makes the model “pop” off the blurred background. But shooting at the maximum sync speed allowed by the camera, typically around 1/250th of a second, in bright sunlight, means that you have to stop the aperture down to around f8 and this unfortunately brings the background back into focus.
You can see my previous hypersync testing approach here
So after collating my previous test data I continued my testing, this time using a live model, the adorable Latvian model Diana, who was a total pleasure to shoot with and who was patient as I experimented with various settings and lighting combinations.
Determining the ambient exposure for optimum background saturation
We shot outdoors at our pool, and I started with a simple exposure test at 100 ISO and F2.8 and the shutterspeed needed to get the background saturation I wanted was 1/2,000th of a second seen in the image below.
Hypersync testing with Nikon D3x and up to 3 Alien Bees 640WS flash heads in bright sunlight
So while the sun was out, most of my shooting was done with shutter speeds in the range of 1/1,600 – 1/3,200th second. Later on a storm cloud darkened the sky and the ambient light dropped around 4 stops so I had to slow down the shutterspeed so as not to completely underexpose the background.
I started off using the Alien Bees at full power as in my earlier testing this had yielded the best results. With all three flash heads connected I found the system quite unreliable, and it was quite a process of un-plugging, re-plugging all the connections to try and get all three units to sync and fire at full power. And most bizarely, the lower shutter release on my Nikon D3x would not allow the three units to sync at all, I had to use the upper shutter release which is a pain when shooting portrait format.
With the 3 heads at full power there was a lot of flash-to-model distance adjustment going on and it definitely killed the flow. And of course I had to allow several seconds between each shot to allow the Alien Bees to fully recharge from each full power discharge. I ended up shooting way fewer images in the 3-hour shoot than I would have in my normal glamour shooting mode. This can be a good thing and a bad thing, sometimes you see something happening in the viewfinder that you just have to capture, but the flashes have not recharged, so as I say it can be a buzz kill on a shoot with a model who moves well from pose to pose.
The images below show shutter-speeds ranging from as slow as 1/1,000th all the way up to 1/2,500th of a second. I did shoot images at 1/4,000th and at 1/8,000th of a second, but the fall-off in flash illumination is dramatic, and of course the background ends up way too dark and loses the saturation and interest that I wanted to capture.
Hypersync Testing Summary to date.
I am sure you noticed the “well, sort of…” comment in my opening line. Yes, Pocket Wizards’ Hypersync functionality definitely allowed me to shoot at shutter speeds up to 4 stops faster than the maximum camera flash sync speed without any noticeable banding. And overall I am thrilled not only with the results in my limited testing to date, but I am more excited about the possibilities to brings to our studio and our customers.
However throughout the four tests I have conducted to date, the Pocket Wizard hypersync system proved to be pretty unstable, and would give widely differing results image to image. And this occurred in all four tests conducted to date. Apparently I am not the only one experiencing this phenomenon according to my reading.
Mounted on the hotshoe of my Nikon D3x I had the Pocket Wizard MiniTT1 transmitter with the AC3 Zone Controller to adjust the flash power. On the Alien Bees I had the Pocket Wizard FlexTT5 transceivers with the AC9 Alien Bees Adapter, and I had a dummy plug inserted into the sync socket to ensure that the Flashes were not firing in slave mode. All batteries were fresh and I was powering the Alien Bees direct from the mains.
I also experimented with the AC3 controller adjusting the flash outputs. This did not go well and the output was not consistent across all the flash units despite multiple corrections, plugging and unplugging. I have no idea what was causing the instability. Something that will require some more experimenting.
At this point I am not comfortable using hypersync on a commercial shoot as it is too unreliable. But I am very excited with its potential and I am hoping that I can get the system stable enough to trust it in image-critical situations.
Maybe some other shooters can comment on their own observations………..
Back to more testing trying to get to terms with Hypersync. For a long-time shooter like myself there is a lot to unlearn and relearn with this technology. For my entire pro shooting career I always relied on the simple fact that shutterspeed has no influence on the flash component of an exposure as the flash duration was always shorter than the shutter opening.
So when shooting outdoors using flash to balance the ambient light, this meant you could set your flash-to-model distance and power level and then chase the ambient light with the shutter, slowing down or speeding up as the clouds moving overhead would change the ambient light level. And when shooting models in front of sunrises and sunsets, the same would hold true. Keep the flash-to-model distance and power constant, hold the same aperture and chase the changing lighting conditions with your shutter speed.
Well, all of this worked really well when using flash with shutter speeds equal to or slower than the camera’s sync speed.
Now with the advent of HSS and Hypersync we are tricking the camera and getting the flash to sync at shutter speeds well above the camera’s advertised sync speeds.
We know that shutterspeed will affect the ambient light exposure, but with these elevated shutter speeds, the shutterspeed is now also affecting the flash illumination on the model as well, something entirely new in my shooting experience, and it is a new aspect to shooting that I have to understand and relearn.
So this round of trials is to try and understand how much the shutterspeed will affect the flash component of the exposure.
Once again I started shooting in late afternoon in full sunlight. The equipment setup is as follows
Nikon D3x with Nikon 70-200mm f2.8
Trigger is the Pocket Wizard Mini TT1 with the AC3 Zone Controller
Receiver is the Pocket Wizard TT5 with the AC9 Alien Bees Adapter
Alien Bees 640WS flash heads, all bare heads with standard reflector, no modifiers
My previous tests with Hypersync and the Alien Bees illustrated that very useable full frame flash exposure is possible up to 1/1250th of a second with this setup. I say “full frame” as with shutter speeds above 1/1,250th there was a slightly noticeable intrusion of the inevitable shutter curtain shadow banding that crept into the frame more as the shutter speed increased, making the one edge of the frame unusable.
However if you fully understand the position of the banding and are prepared to compose your image accordingly it is possible to get decent images all the way up to 1/8000th of a second, but the flash exposure is being diminished several stops by the fast shutterspeed.
Ambient Light Testing
As before, I established the desired ambient light exposure set the ISO at a constant 100, kept the Aperture at a constant F2.8 and varied the shutterspeed. Unfortunately a large cloud moved over the set as I did this so the exposures recorded are about 2 stops less than full sunlight, but the images nevertheless clearly illustrate the effect of increasing the shutterspeed on the ambient light exposure.
Single Alien Bees Main Light FULL Power
Next I put a single bare head Alien Bee 640WS light approx. 8ft from the model, slightly off-center of the camera position, on the same side as the ambient light. I did the test with the Alien Bees at full power keeping the aperture constant at F2.8 and the ISO constant at 100. Starting from 1/500th second I increased the aperture for each image, allowing the flash time to fully recharge from the full discharge each exposure. I took the shutterspeed all the way up to 1/8,000th as I wanted to see how far the shutter shadow banding intruded into the full frame with the camera in the vertical portrait position (all my previous tests were in the landscape format)
As you can see from the images, the banding only becomes noticeable at around 1/2,000th sec, and only intrudes about 10% of the frame at 1/8,000th sec. The image is fully useable without any cropping all the way up to 1/2,000th sec, but this was not the goal of the test.
What we can see is that increasing shutterspeed from 1/500th to 1/1,000th reduces both the ambient and the flash exposure by 1 stop, and there is a similar correlation from 1/1,000th to 1/2000th , from 1/2,000th to 1/4,000th and from 1/4,000th to 1/8,000th.
So many photographers have reported widely diverging results with hypersync depending on the equipment used, so while the flash exposure appears to have a linear relationship with shutterspeed, I am hesitant to just accept this.
Single Alien Bees Main Light HALF Power
I did the same test again with the Alien Bees at ½ power (controlled from the AC3 control) and the same relationship appeared to hold true but there was not enough flash illumination to make the images at the higher shutterspeed useable.
Three Alien Bees – One Main Light , Two Rim Lights, all FULL Power
This time I added two additional Alien Bees as rim lights about 10ft from the model. Because of my findings in the previous tests I limited the shutterspeed to 1/2,000th of a second. Once again the relationship of shutterspeed to flash illumination appears to be linear.
Basically, with this setup it appears that if you double the shutterspeed you will decrease the flash illumination by one stop, which ultimately means that when you end up with an underexposed subject, you will need to shorten the flash to model distance, or add additional flash units to maintain the shutterspeed.
But as you can see the results are very promising, and obviously more testing is needed prior to relying on this technique in a commercial shoot. And knowing how the banding will affect your image will allow you to use really fast shutterspeeds, as shown in this image shot at 1/4,000th second, 5 full stops faster than the maximum camera sync speed.
The next test will be with a live model using elevated shutter speeds, and getting comfortable with the affect of shutterspeed on flash exposure