cgardner
Offline
• • • • •
Upload & Sell: Off
|
The 580exII has a thermal protection override which shuts down the unit to cool without warning if overheated. If you had been using it heavily that might have been what happened.
Batteries might also be the culprit. I've had a similar problem with my 580ex which I traced to one of the four AA's being faulty. The batteries would appear to charge OK and run the display, but fail under the load of recharging the capacitors.
Other 580exII problems have been caused by poor / intermittent contact between the flash and the hot shoe due to the weather seal, metal foot, redesign of the pins, etc. Also check the hot shoe of the camera which can sometime loosen.
As for the IR? This is from Chuck Westfall, of Canon USA from a 2009 e-mail exchange I had with him about the role of IR in the Canon flashes.
The sensitivity of the wireless metering sensor on the front of a 550EX, 580EX, 580EX II or ST-E2 isn't stated in any of the documentation I've seen. Like the European web site says, the EX Speedlites with master unit capability emit visible light preflashes through their main flashtubes, so IR is not involved for wireless flash communication with those products. The ST-E2 also uses a conventional flashtube and reflector to emit its wireless signals, but the reflector is covered by a translucent red panel. If you're in a darkened room, you can see the preflash from an ST-E2 (as well as the AF Assist beam, but that of course is a separate issue), so for that reason, the wireless signals are not completely in the IR range. However, the peak wavelength for the wireless communication signal from the ST-E2 is 750 nanometers, which is at the low end of the IR spectrum. On the other hand, the AF assist beam for the ST-E2, as well as for all other EX Speedlites, has a peak wavelength of 695 nanometers, which is still within the visible light spectrum.
So apart from the ST-E2 which dates back to the 420ex, all the Canon devices signal with visible coded pre-flash light bursts from the flash tube using this sequence in ETTL mode...
http://super.nova.org/TP/CanonETTL.jpg
To the extent any IR only filtration occurs it would be due to the lens over the slave sensor which is the smaller dark lens on top of the bigger red one which is the AF assist and has nothing to do with the flash signaling.
Canon flash metering has evolved from TTL (film bodies) to E-TTL (bodies < 2004), to E-TTLII (bodies > 2004) and the EX flashes are made backwards compatible with all three modes an the original IR only ST-E2 master device.
Canon film cameras metered off the film surface during the actual exposure using what it called "TTL" metering. A problem of the 580exII is that it sometimes switches by itself to TTL mode and firing at full power is related to that legacy metering system. Losing contact with the hot shoe make the 580exII think its been instructed to switch to the legacy TTL film body mode. The flash is fired, but command to stop the flash never comes with a digital body so the flash fires at full power.
Digital sensors required moving the metering to the viewfinder before the actual exposure and using pre-flash to meter for flash. As noted in the diagram all the flash metering — which includes a baseline ambient reading separate from the one controlling shutter/aperture — all the pre-flashes and metering occur after the shutter is fully pressed. That can create exposure problems in some scenes if you meter the shutter/aperture with half-press of the shutter button then recompose in P, Tv and Av. A solution is to move AF lock to * and only half-press to lock AE and fully-press the shutter to meter for flash after recomposing.
Why was IR filtration used on the flash tube based output ST-E2 then abandoned on subsequent Master devices? My guess is that when the pre-flash metering technology was being developed the timing of pre-flash commands was causing a blink reflex in the subjects. But by the time the 550ex was issued the problem had been solved. The pre-flashes are only a few bytes in length and occur immediately before the main flash. From behind the camera the pre-flashes and main flash blend together with the main flash. Many continue to think the commands must be IR because they don't see any pre-flashes and confuse the red AF assist lens with the flash signaling. Adding to the confusion is the fact the ST-E2 remains popular as an invisible AF assist lamp for shooting ambient only in very low light.
Canon flash continues to get beat-up over the misunderstood IR thing. The negative feedback and bashing started back when the ST-E2 was the only way to trigger a slave. The problem with the ST-E2 isn't the fact it uses IR but that it is small and underpowered for the task which limits its range to about 10 feet — you can spit further.
Due to the combination of the ST-E2 range, that it and the 420ex were ETTL only and the dismal performance of the original AF point biased ETTL flash metering Canon flash did really suck in bodies prior to the 20D and 5D and the more powerful 550ex and 580ex Master flashes. One of the main reasons I held off buying into a Canon DSLR system until 2004 were the reports I read of how dreadful flash metering was on the early Canon models D30, D60, 10D.
The 20D and 5D where among the first ETTL-II models which use zone evaluation for metering flash. I bought a 20D and a pair of 580ex flashes. I skipped the ST-E2 because I'd been reading all the negative feedback and been using a pair of Vivitar flashes on bracket for fill and stand as keep since the early 70s and wanted a similar configuration with Canon. I was one of the first to sort out how the ETTL-II flash metering actually works in practical terms via testing and shared my results in my tutorials at http://photo.nova.org/ Chuck Westfall thought enough of them to recommend them in one his monthly PhotoNotes columns a few years back.
Signaling range depends on the footprint of the Master and its intensity. If you want maximum signaling range with the Canon optical system, logically you'd want to use the most powerful flash Canon makes as Master. By avoiding the ST-E2, using an ETTL-II body and the 580ex as the Master I dodged most of the problems that plagued the Canon flash system prior to 2004. I still use my pair of 580ex flashes without any radio triggers and have no problems with range because I don't used modifiers which block the sensor on the slave.
IR or not? It really doesn't affect performance one way or the other. Yes line of sight is needed outdoors and the range is more limited (25') but indoors the sensor window on the slave just needs to see the pre-flashes from the Master which can be bounced off ceilings and walls indoors at distances up to around 50'. The biggest shortcoming of the Canon flash system, apart from the lame underpowered ST-E2, is the fixed position of the slave sensor. For reliable signaling the front of the slave must point to the Master with the flash head rotated separately towards the subject. For years I've been suggesting the slave sensor be modified to make it plug-in module that could be positioned outside a modifier. That's what I did with my Wien Peanut triggers with Vivitars — plug them into the end of a PC cord. Being able to mount the slave sensor outside a SB would solve many of the signaling problems. But Canon has focused instead on things like adding waterproofing and thermal nanny mode instead of addressing the shortcomings of the system. I thought the 580exII was huge step backwards from the 580ex which is in all respects a very good performer. There are rumors of new Wi-Fi based flashes under development so we can only hope the next generation will be better.
That's not to say radio triggers aren't useful, I just can't justify the expense because I don't find them necessary for the type of shooting I do.
|
