Lencarta safari flash duration

[mr bizzare]

I am looking for a bit of help, I need to know the flash duration of the safari lights at different outputs and configuration. if someone can give me this information I would be eternally grateful.

 

[Garry Edwards]

It's fairly complicated, with the flash duration varying dependent on both the power setting and the choice of output socket used.

On test, it produced what could be described as equivalent shutter speed figures (which are similar to t.1 times) What this means is that these figures are pretty much what you would get, in terms of action stopping potential, at similar shutter speed used with continuous lighting. These figures are probably more meaningful than either t.1 or t.5 times.

At full power (600Ws) from socket A the equivalent shutter speed is about 1/700th.
At minimum power (12.5Ws) from socket B the equivalent shutter speed is about 1/900th.
At various other settings, it ranged from around 1/500th - 1/1000th.

 

[alsos]

It's fairly complicated, with the flash duration varying dependent on both the power setting and the choice of output socket used.

On test, it produced what could be described as equivalent shutter speed figures (which are similar to t.1 times) What this means is that these figures are pretty much what you would get, in terms of action stopping potential, at similar shutter speed used with continuous lighting. These figures are probably more meaningful than either t.1 or t.5 times.

At full power (600Ws) from socket A the equivalent shutter speed is about 1/700th.
At minimum power (12.5Ws) from socket B the equivalent shutter speed is about 1/900th.
At various other settings, it ranged from around 1/500th - 1/1000th.

Garry might be worth clarifying those specs relate to the Classic and the Li-on is faster

 

[HoppyUK]

Garry might be worth clarifying those specs relate to the Classic and the Li-on is faster

That's the Li-on. Looks like Garry has quoted the figures from a review I did for Advanced Photographer magazine (Feb 2012 issue).

As Garry said, it's complicated and those numbers equate to the kind of action stopping potential you would get with actual shutter speeds. They are roughly close to t.1 times, but since hardly anyone ever quotes those, my best estimate is that they are similar to t.5 x2.

So when it comes to comparing brands, if you take their quoted t.5 and double the duration, you will have some idea of what to expect compared to shutter speeds, but it's not an exact science!

 

[Garry Edwards]

Yes, that's about it, Richard's review figures agree with my own findings and, IMO are probably more relevant than quoting either t.5 or t.1 times - so I nicked his

t.5 times are science-based and aren't misleading in themselves provided that people understand what they actually mean, but not everyone does understand them.

t.1 times are also science-based and are closer to real world shutter speeds than t.5 times but again they aren't perfect, and virtually no manufacturers quote them because the industry norm is to quote the more attractive-looking t.5 times and I doubt whether anyone wants to go out on a limb and quote figures that make their product look unappealing...

Richard's test method, which involves photographing a moving fan under continuous lighting at various shutter speeds, and then photographing the same subject using the flash at various power settings, provides a worthwhile comparison in that it allows us to say, for example, that the blur at X power setting is roughly the same as at Y shutter speed. It has 2 main limitations:
1. Cameras don't have enough different shutter speeds, so if you think that the amount of blur at a given power setting is somewhere between the amount of blur recorded using continuous lighting at 1/500th and 1/1000th, all you can say is that it's around say 1/700th, it's difficult to be more precise than that.
2. It's a pain, because to get meaningful results it's necessary to get the top fan blade vertical, which means taking at least 10 shots of the blade in the wrong place for every one in the right place...
Also, it has limitations caused by the moving slit of the FP shutter, which 'bends' the fan blade, which gets progressively worse as the shutter speeds get shorter.

Edit: Assuming that people understand what the figures mean, any science-based testing method, which involves using an oscilloscope to trace the way that the flash tube discharges, and the time it takes to do it, is helpful. But from looking at various websites it's pretty obvious sometimes that the figures quoted just can't be true, just as figures quoted colour temperature variations at ridiculously close tolerances just can't be true. Sometimes, I can't help wondering where these figures come from.

 

[HoppyUK]

Yes, pretty much Garry.

There's not really a problem getting incremental shutter speeds between full stop settings. I have the full sequences at one third stop increments, though they were not all published in the magazine because of space limitations.

The real problem is that making an assessment of how effective flash durations compare to shutter speeds is at least partly subjective due to the way flash actually works. You have to make a judgement on what 'it looks most similar to' in terms of shutter speeds.

In general terms, shutter speeds take a slice of time, and that's it - with sharp start and cut-off points. IGBT flash (hot-shoe guns) effectively works in much the same way, at least at lower power settings.

But studio heads are different. The total flash duration is always quite long, something like 1/100-1/200sec or thereabouts from start to complete extinction. But during that time, the brightness rises very quickly on firing to a peak, and then falls away much more gradually.

It's the duration and height of that peak that gives the action stopping. T.5 times measure peak duration (it's the length of time the flash stays above 50% brightness) but ignore the relative height. Personally, I think something like t.3 would be more realistic, but even that doesn't take into account the effect of the lower peak that occurs at lower power settings.

What we really need, and I've discussed this with Garry before, is something like a percentage peak time multiplied by the height of the peak above a certain threshold, but I don't see the industry making too many moves in that direction, because t.5 is pretty much universal and it tends to flatter their products. Not helped by the fact that some of the lesser brands appear to pluck their numbers from thin air LOL

 

[Cistron]

Yes, pretty much Garry.

There's not really a problem getting incremental shutter speeds between full stop settings. I have the full sequences at one third stop increments, though they were not all published in the magazine because of space limitations.

The real problem is that making an assessment of how effective flash durations compare to shutter speeds is at least partly subjective due to the way flash actually works. You have to make a judgement on what 'it looks most similar to' in terms of shutter speeds.

In general terms, shutter speeds take a slice of time, and that's it - with sharp start and cut-off points. IGBT flash (hot-shoe guns) effectively works in much the same way, at least at lower power settings.

But studio heads are different. The total flash duration is always quite long, something like 1/100-1/200sec or thereabouts from start to complete extinction. But during that time, the brightness rises very quickly on firing to a peak, and then falls away much more gradually.

It's the duration and height of that peak that gives the action stopping. T.5 times measure peak duration (it's the length of time the flash stays above 50% brightness) but ignore the relative height. Personally, I think something like t.3 would be more realistic, but even that doesn't take into account the effect of the lower peak that occurs at lower power settings.

What we really need, and I've discussed this with Garry before, is something like a percentage peak time multiplied by the height of the peak above a certain threshold, but I don't see the industry making too many moves in that direction, because t.5 is pretty much universal and it tends to flatter their products. Not helped by the fact that some of the lesser brands appear to pluck their numbers from thin air LOL

I always thought the t-values were integrals, i.e. t.5 was 50% of the energy put out. Not correct?

 

[HoppyUK]

I always thought the t-values were integrals, i.e. t.5 was 50% of the energy put out. Not correct?

No, though I guess the two would be very similar, but not quite. See diagram on page 10 of this link http://www.broncolor.com/fileadmin/pdf/broncolor/products/System_Catalogue/BRN_SYS_2011_LOW_ENG.pdf

It shows t.5 and t.1 times clearly, and also has a somewhat stylised indication (ie exagerated) of how colour changes.

Not shown on that diagram is how IGBT flash adjusts output, which is to always fire at full power and then cut off the pulse sharply once enough light has been delivered. That effectively cuts the duration in half with every stop reduction, ending up with incredibly short times like 1/40,000 sec at lowest settings.

Studio heads are different in that for lower powers the capacitors just dump a lower output/brightness. Therefore the total flash duration doesn't change much, but the peak is much lower.

Hope that makes sense

 

[Cistron]

No, though I guess the two would be very similar, but not quite. See diagram on page 10 of this link http://www.broncolor.com/fileadmin/pdf/broncolor/products/System_Catalogue/BRN_SYS_2011_LOW_ENG.pdf

It shows t.5 and t.1 times clearly, and also has a somewhat stylised indication (ie exagerated) of how colour changes.

Not shown on that diagram is how IGBT flash adjusts output, which is to always fire at full power and then cut off the pulse sharply once enough light has been delivered. That effectively cuts the duration in half with every stop reduction, ending up with incredibly short times like 1/40,000 sec at lowest settings.

Studio heads are different in that for lower powers the capacitors just dump a lower output/brightness. Therefore the total flash duration doesn't change much, but the peak is much lower.

Hope that makes sense

Thanks for that, the diagramme makes it clear. If manufacturers stick to the t.5 figure, integrals would make more sense for comparison.

 

[Garry Edwards]

Thanks for that, the diagramme makes it clear. If manufacturers stick to the t.5 figure, integrals would make more sense for comparison.

Basically, manufacturers do stick to the t.5 figure - but as you can see from the chart, or from any other oscilloscope trace, the flash doesn't start life at full power, it takes a (short) time to reach full height, which means that integrals can't work in absolute terms.

As both Richard and I keep saying, it's complicated.
The Bron figure shows how the colour changes as the flash decays - OK, that's greatly exaggerated for the purposes of illustration, but that is what happens.

So how is the colour temperature change relevant in the real world?
Well, in two ways. You can see that the flash very quickly reaches a blue, or cold colour. From my own tests, I would say that at it's bluest colour it is something like 1000-1200K (degrees Kelvin) above 'normal' and measures at around 6000 - 6200K. As the flash tails off, it becomes warmer and the mix of warm and cold ends up with something around 5000K, which is generally accepted to be about right.

Now, if you use a IGBT controlled flash (mainly hotshoe flash but there are some others too) which give you the benefit of very short flash durations at lower power settings, a low power setting will cut off the orange/red 'tail' of the flash and the colour of the flash that's left will be decidedly blue. If that flash is the only light source then you can change the colour PP, but if it's only lighting part of the subject and other parts are lit by say the ambient light, then you'll have say one side of the face looking 'normal' and the other half looking blue...

The other thing, which is very relevant to t.5 times, is that the tail becomes more and more red as the energy reduces. Now, with some well-designed flashes it becomes very red very quickly, and soon gets almost to the infra-red part of the spectrum, which means that it becomes invisible (or nearly so) and because of this has no visible effect on the shot.

The effect of this is that although the t.5 time may appear to be fairly long, in practical terms the bit that's in the visible spectrum is less long, so can't contribute much to either the overall exposure or to movement blur.

It's because of this that some people say that the t.5 time is about twice as long as the t.1 time and that other people say that it's about 3 times as long as the t.5 time. Both views are right and both are wrong - it depends on how quickly the colour moves out of the visible spectrum, and isn't a rule that can be applied to all flash heads regardless of manufacturer/design.

I hope this makes sense - it probably won't even make sense to me when I see it on the page