TAOP: Part 4 – Light: Research – High-Speed Flash Sync
April 8, 2012
Following my last assignment my tutor suggested I investigate the effect of using high-speel flash sync – looking at the impact it has on the power output of the flash.
I googled and found this very good site and article – Concepts of Flash Photography – on the comparison and effect of high-speed flash sync versus normal flash mode. What follows is a summary of what I’ve learnt, plus some experimentation with my camera and flash gun (the Canon EX 580EXII). For anybody interested in more specific details and comparisons I’d recommend reading the full article linked to above.
- Normal flash mode emits a short burst of light. To get an even exposure the shutter speed must be slow enough so as the shutter is fully open when the flash fires – else only part of the sensor will get light from the flash, as the remainder is obscured by the shutter.
- On my camera 1/200s is the fastest shutter speed possible in Normal mode. Higher-end cameras are generally faster than this.
- High-speed flash sync releases pulses of light from when the shutter opens to when it closes. The consequence is the sensor gets an even amount of light at a shorter exposure time, but the price of this is the flash is less efficient, reducing the power and range of the flash.
- There is an immediate drop-off in efficiency when you switch from the cameras maximum Normal mode speed to speeds requiring High-speed flash sync mode.
- Once in the high-speed sync range the flash output is consistent, so should be treated the same as continuous light. If the exposure time is reduced by one stop, the ISO or aperture size would need to be changed by one stop to account for the change.
Effects of High-speed Flash Sync
This is some experimentation with my flash set to manual, and high-speed flash sync, to avoid the effects of the E-TTL taking action to adjust power output.
This shows the effects of high-speed flash sync mode with only the exposure time adjusted.
The first image shows the baseline, at the optimum setting of 1/200s. The foreground is pretty-much correctly exposed:
The second image clearly shows the drop-off in efficiency thanks to high-speed flash sync at 1/400s. Even at this point the image is clearly under-exposed:
The final image continues to illustrate the dramatic effect on flash efficiency at 1/1000s:
The drop-off in power due to high-speed flash sync is described as linear in the article, so I decided to test this myself. Here’s the base-line shot at 1/200s:
Changing to 1/400s – so one stop – and opening the aperture by one stop shows the big drop-off in power between normal and high-speed flash sync, which is expected:
Opening the aperture a further two stops to f/5.6 (so, three stops in total) gets to a similar image to the original:
Finally, decreasing the exposure time by just over one stop required an identical change to open the aperture wider:
I did open the aperture slightly too wide, hence being slightly over-exposed. But it’s close enough that I’m confident it illustrates the point. Essentially it proves the flash power output is remaining the same – so linear changes are required as though it is a continuous light source at the same level.