Celltech/Travor RF550D Macro LED Ring Flash

[Tetragnathid]

I think I'll find it useful for macros under limited conditions, where the brief duration of a strobe flash isn't required to stop motion. As you'll see, I'm puzzled by some aspects of its behavior.

Here's the kit (got it at Amazon for US$33) and shots of the unit mounted on a Sigma 150mm f/2.8 macro and a Zuiko 50mm f/2.0 macro, both on my E5



Here are some specs:

It runs on 48 LEDs

Functions:
· Continuous light with all 48 LEDs
· Flash with all 48 LEDs
· Flash with left-side 24 LEDs
· Flash with right-side 24 LEDs

Seven power levels from 1/1 to 1/64 in steps of *½

Diffusers:
· Clear (not really a diffuser at all)
· Frosted white
· Frosted blue
· Frosted amber/orange

Seven adapter rings, lens filter thread to annular light (49, 52, 55, 58, 67, 72, and 77 mm filter threads)

The flash has a fixed duration of 1/100 seconds regardless of power level; I verified this with high-speed video (200 frames/second)***wrong; see correction in later post below***. This is, of course, different from a regular strobe, where the power is regulated via the duration rather than the intensity of the light. Clearly, 1/100 second isn't brief enough to stop a lot of the motion that one encounters with macro photography. So this is probably better for studio work than field work.

The claimed guide number is 15. It's supposed to be strictly a manual flash.

Here are photos with full, left, and right flash on the Sigma 150 mm f/2.8 macro (on an E5, ISO 200, manual exposure at 1/6 seconds, f/11, full 1/1 power, 25 cm flash-subject distance). Note that since the ring rotates freely on the lens, “left” and “right” could as easily be “top” and “bottom”



The flash exposure behaves strangely in my hands on both my E5 and E3. Of course, when the shutter speed is faster than the flash duration of 1/100 seconds, the exposure intensity falls off. But, as the shutter speed is slowed below 1/100 seconds, the exposure intensity continues to increase. The composite below illustrates this for four different shutter speeds ranging from 1/10 to 1/250 seconds. Conditions: f/11, continuous light or full 1/1 power flash, ISO 200, flash-subject distance 14cm, Zuiko 50mm f/2.0 macro lens, image a composite of out-of-the-camera ORFs. Ambient light is negligible (black photo at 1/6 seconds not shown).



Of course, the exposure behaves this way for continuous light, but I would have expected the exposure to be independent of shutter speed for durations longer than 1/100 seconds. This is puzzling, and I wonder whether there could be some strange shutter-synchronization issues. Any suggestions about the source of this behavior?

I haven't yet tried using it as a ring light for non-macro work (e.g., portraiture), but my impression is that it would be weak as a key light.

[Tetragnathid]

Of course you always catch errors after you do the post (and often when you're just climbing into bed). The manual claims a flash duration of 1/100 second, but I think I got seduced by that and made an order-of-magnitude error when I was looking at the high-speed video. 20 frames at about 200 frames per second is 1/10 second, not 1/100 second. So I think the flash duration is really longer than they say, but I have to check into this tomorrow. This would more or less explain the behavior that I saw. Now back to bed!

[Tetragnathid]

So after I slept on it I took another high-speed video, this one with an LCD stopwatch and the flash in the frame. That way I can time the flash directly from the images rather than relying on counting frames and relying on the camera and video editor to produce the number of frames per second that I expect.

The stopwatch shows 1/100 second intervals, though of course that's around the switching time for such displays and so it's a bit dicey at that fine a time resolution. But it's not hard to use it to calibrate the clip over a longer time, and what I come up with is that the flash lasts about 0.14 seconds (maybe +/- 0.01 seconds) at both full and 1/64 power.

So this indeed explains my observation that flash exposure depends on shutter speed for durations longer than 1/100 second, and the claim "Exposure time: 1/100s" in the Technical Specifications section of the ring flash's manual is extremely inaccurate.

And what about the claim of a GN of 15? The whole idea of GN generally assumes that the shutter duration is longer than the flash duration, so their figure should be applicable only for 1/7 seconds shutter or slower. So I set the shutter at 1/3 seconds, the flash at full power, the ISO at 200, and varied the aperture for a subject that was placed 27 cm from the flash. Ambient light was negligible. Under those conditions, f/22 looked about right. The calculated GN would be 0.27*22/sqrt(200/100) = 4. So I think their claimed GN figure is a bit high...EXTREMELY high if you want to use a more usual shutter speed, where you only capture a portion of the flash.

I suppose it's possible that I got a lemon. I'm going to check with the manufacturer.

[Tetragnathid]

I've been communicating with the manufacturer about the duration of the flash. They've been very cooperative about conducting the email exchange. Though there may have been some language difficulties, and I think the guy I was communicating with is more a sales/marketing than a technical person, it does appear that they claim the duration of the flash is 1/100 second. What is not clear is whether my unit is defective or whether that's an incorrect spec. If the amount of light produced in the first 10 msec were enough to fulfill their GN 15 spec, it wouldn't matter that the light continued for another 130 msec, but that's not the case on my unit.

Here's part of the final email:
1. What i mean synchronizing speed is when the shutter is fully open,the flash shine instantly. and the speed is less than 1/100 second.

2. Frankly speaking, this model of flash is a entry-level product. i'm sorry that it can't not meet your need in some respects.

3. I very appreciate your feedback. The comment will be transfered to our engineering department. Our new item of camera flash will be released next month, It will be much better than RF-550. you can give it a try if there is interest.

I thanked him for his time and told him that I wasn't going to seek a replacement/refund based on this issue. Note that they're soon going to release a better model.