What is the "ideal" t=0.5 value for HyperSync? Is 1/150th too slow?

[studio460]

<p>If you recall my previous posts regarding HyperSync (shooting above x-sync using PocketWizard triggers) tested with a Nikon D3s, I finally had an opportunity last weekend to try out the technique using my Nikon D800E instead [note that HyperSync performance is highly camera-specific]. Happily, the D800E significantly outperforms the D3s in this application--equipment used:</p>

<p>• Nikon D800E.<br /> • PocketWizard TT1/TT5 RF triggers for Nikon TTL.</p>

<p>I own several strobes, chosen mainly for their long t.5 values:</p>

<p>• Quantum Qflash Model T; t.5 = 1/300th at max. power.<br /> • Dynalite Uni400Jr.; t.5 = 1/675th a max. power.<br /> • Speedotron Force 10; t.5 = 1/850th (all power levels).</p>

<p>For these daylight-exterior tests, I used only the Speedotron Force 10, PocketWizard TT1/TT5, and a D800E. Unlike the D3s, the D800E, produced no hard shutter-curtain shadow at shutter speeds anywhere from 1/250th (x-sync), all the way through 1/8,000th (although some uneven exposure is produced, it often goes unnoticed in most real-world photography). I did experience a loss of recorded flash output, but the loss appeared to be <em>constant</em> (i.e., recorded flash loss didn't appear to increase with higher shutter speeds).</p>

<p>Now, my question: I know that Metz flashes tend to have very long t.5 values--anywhere from 1/250th to as slow as 1/150th. My question is, would 1/150th be <em>too</em> slow? Would 1/250th be better? What is the "sweet spot" among these t-value curves, vis-a-vis, your camera's shutter speed? Say you're employing a 1/4,000th shutter speed--what would be the "ideal" t.5 value? And, at 1/1,000th? At, 1/8,000th?</p>

[rodeo_joe1]

<p>Hi Ralph. I'm not sure that there's a definite answer to your question, since the t0.5 time is dependent on the pulse shape of the flash. Some flashes have their output shaped to improve the evenness of light output with time. For example: Most older Metz flashes have an inductor (choke) in series with the flashtube to give a squarer output pulse. Other flashes not using an inductor will have a more elongated exponential decay curve.</p>

<p>Also, for a given camera, as far as evenness of exposure is concerned it doesn't matter a jot what shutter speed is chosen in excess of the X-synch speed. This is because the transit time of the shutter doesn't change once the X-speed has been exceeded. For Nikon's D700 and D800 cameras that transit time is around 1/320th second.</p>

<p>To sum up. The pulse shape of the flash is more important that some spuriously measured t0.5 time.<br />The shutter speed chosen above X-synch doesn't affect the evenness of exposure as long as the flash duration exceeds the transit time. However the overall exposure definitely <em>will</em> change depending on the shutter speed chosen. Ideally, your "sweet spot" needs a flash pulse as close to square-shaped as possible, and slightly longer than the transit time of the shutter being used. If the flash has a more exponential decay, then it's duration will need to be considerably longer than the shutter transit time.</p>

<p>BTW, HyperSync is PW's stupid proprietary name for what's essentially old-fashioned FP synch. Nikon more sensibly just call it Auto FP sync.</p>

[studio460]

<p>Hey, Joe! Thanks for that detailed reply! Since you say that Metz' decay curve is more square-shaped, I would assume that Metz flashes are in fact, optimal devices for exploiting high-speed sync techniques. As tedious as they are, I suppose more careful tests are needed to gain any further conclusions.</p>

<p>I only did a quick test against a white wall the night before my daylight-exterior "test shoot" (a spec shoot for a designer), and surprisingly, the resulting illumination <em>appeared</em> to be the same from 1/1,000 all the way up though 1/8,000th. Unfortunately, during my location tests at the beach, I was is such a hurry to complete my "assignment," that I failed to record my Speedotron Force 10's output settings, which I did vary from about half-power to full-power.</p>

[ellis_vener_photography]

Since the t0.1 period for all of those flashes you list is greater than 1/300 it is no surprise that you are able to get a full

coverage of your frame ( with some falloff at the end of the exposure ) with your camera. T0.1 measures the full amount

of time the flash is emitting photographically significant light, not as Joe points out, the marketing driven use of the

misleading t0.5 at measurement which only measures the time the output is over 50% of the peak output.

 

While Nikon calls it FP, Canon calls it HSS for High Speed Sync. In either case what Nikon and Canon do with their hotshoe mount flash systems is different than what you are doing.

[studio460]

<p>Thanks for your reply, Ellis. Yes, I'm still trying to get a handle on this topic, since I still think it's incredibly useful technique.</p>

[ellis_vener_photography]

Flash duration has two ISO blessed standards for measuring flash duration, t0.5 and t0.1, I described what the measure

above. If you are interested on freezing motion , t0.1 measurement is more useful. Mind you nothing has changed in the

flash output, just what part of it is being measured.

 

Whichever standard you are applying, a longer pulse is better for what you are doing as you want the energy level to be

kept up for about 1/200th second. The delay between the start of the flash pulse and the gap between the shutter starting

to open is important. Partly that is handled by the camera's firmware programming and partly it is your choosebn sync

connection method, and partly it is the programming and hardware of the flash itself.

 

The ideas are pretty sile but the language to describe what is going on is pretty weedy.

 

I'm surprised the light level drop is consistent across the range of your exposures. I think you should try it indoors where

ambient light isn't an issue.

[studio460]

<p>Ellis said:</p>

 

<blockquote>

<p><em>I'm surprised the light level drop is consistent across the range of your exposures. I think you should try it indoors where ambient light isn't an issue.</em></p>

</blockquote>

<p>Thanks again for your comments, Ellis. Don't trust my hasty tests--these were done very informally--inside the house against a white wall. I'll have to re-test under more stringent control to make any definite conclusions.</p>

[ian_casement]

<p>With the output spread over a longer time the 1/150s t0.5 would be more desirable than 1/300s. This should provide less banding and make more cameras compatible. Unfortunately, it will influence beneficial output too, meaning, if your shutter takes 1/200s to travel across the sensor (faster than 1/150s) you are potentially losing some of your flash benefit by it bouncing off the shutter curtain.</p>

<p>As you have noted between your D800E and D3s different cameras show different results. The type of shutter mechanism, speed of the shutter curtain across the sensor and the size of the sensor will all influence this, along with the latency of the triggering device and the way its set up. I would go as far to say the influence doesn't stop there.. firmware versions in the camera and trigger firmware versions have a considerable influence too.</p>

<p>All that said, the 1/250s t0.5 of the Quantum X-Series works remarkably well across all shutter speeds without banding using the Quantum radio system on all the cameras I've tried (Nikon) this is also reported by Canon users too.</p>