jim_leonard

<em>So it seems from yours and Nello's answers to my second question that the fastest effective shutter speed might be achieved by using a normal sync. shutter speed, say 1/250th, in a completely dark room with the flash power set to 1/128th.</em><br>That is exactly correct. That's the typical manner that people do stop action photography. The flash guns aren't fast enough to stop a bullet (or even an airgun pellet as in the attached image) but anything much slower or significantly larger (movement of half an inch means a pellet is just a blur, movement of a half inch for a person would make them still quite visible ignoring the likelyhood of a person moving at 200 m/s (400mph)) comes out pretty good.<div></div>

I find that canon wireless system works fairly consistantly if it's used consistantly and assuming that none of the groups are underpowered.

 

First off, I would guess that your 420ex was running at full power most of the time, you fired your weaker flash into an umbrella and unless the umbrella is fairly close and/or you were using ISO 400 or higher you probably needed more light.

 

Secondly (and this is just an extention of the same problem that most people with flash consistancy problems with canon flashes have), all of the flash metering is basically spot metered off of the current autofocus sensor. This is especially touchy with a multiflash system because if the metered surface reflects light from group A into the camera light meter better than light from group B (due the angle of the surface for instance) then group B will get an excessive power level (and group A a similarly reduced power level) than what you would expect if the spot under the AF sensor isn't representative. This situation can be exagerated further if the area under AF spot can't see group B's flash at all (then group B will fire at full power regardless of what it'll do to the rest of the exposure).

nello's description of canon's multi flash missed a couple of flashes. Prior to the exposure, the main flash fires in a communication with the other flashes, then all of the slaves fire (one group at a time) to get metering information. All of this occurs while the mirror is still down so you can watch it happening, though it does happen fairly quickly. This is because the light meter is only in the light path with the mirror down. Then the master flash sends another communications flash to tell each of the flash groups what flash level to use. Then the mirror in the camera flips up (and the shutter opens) and everything fires. I'm not sure how the final flash sync occurs (ie is it just timed off of the preflash communications or is there another round of communication of some form). However to answer you main question, the slaves (and master) should fire the exposure flash all at the same time to within some number of microseconds.

 

 

As to question 2, you're mostly right. Ideally it would use a continuous light source but you can't do that with a flash tube so it pulses the light several thousand times per second. All of this is because when the shutter speed is higher than the sync speed, the shutter is never fully open. It's just a slit travelling across the film/sensor plane.

Note that 550ex in SE mode also powers up when the 10d powers down from timeout!

To make such a thing from scratch would be lunacy. If you don't have a well stocked metal/plastics shop it would be almost impossible to make something that wasn't extremely fragile and would take a very long time to build. If you did have a proper shop and knew how to use it, it would be a relatively trivial operation. The other possibility is that you have discarded components to work with but I'm guessing that you don't.

 

So to make anything worth having you're going to have to buy something.

 

If it's a "dedicated" flash and you want to keep all of the extra communication between the flash and your camera, then you need to purchase the rediculously expensive custom cables. You're paying the big money for the proprietary connectors. You will then need longer cables than what is provided. It's relatively trivial and inexpensive to extend the cable. Most people recommend using computer keyboard extention cables (ps/2 style) to make a very modular system. This option costs about 10 USD for the cables plus the cost of whatever dedicated flash cable you need to buy (about 40 - 50 USD) and you need to be able to solder to do it right.

 

If you don't have a dedicated flash and just need flash sync, there are fairly cheap hotshoe to pc cord adapters available in all four gender combinations (available from the likes of B&H). Get one to mount on the hotshoe of your camera and one to mount on the foot of your flash. Then use a pc sync cord of the right length between them. I was able to part this out at B&H for about 35 USD. Or if your camera has a pc sync cable attachment port already, you don't need the more expensive of the adapters bringing your cost down to about 15 USD.

Something I should also add is that different video displays are inherently sharper than others. LCD displays run at their native resolution will look very sharp, whereas CRT displays which don't have a native resolution will look less sharp. And then there are people like me who drive their CRT displays beyond their actual resolution (compare the dot pitch to the display resolution). I do it because I like things smaller on screen but I have issues when I'm creating screen viewable images because I tend to sharpen for my display (which needs more sharpening and tends to hide certain sharpening artifacts well and provides a "hardware" anti-aliasing) rather than a more typical lower resolution display.

<i>the image should be slightly sharper than what I want on paper? Is there a reason for that?</i><br>There are two reasons. 1) when you print, the pixels run together a bit (effectively a low radius blur). It's good to add a bit of extra sharpness to compensate for this. And 2) screen pixels (72-130ppi typical) are 3 to 4 times larger in each direction than print pixels (300-360ppi typical) so that the same levels of sharpening affect different angles of view on screen versus print and therefore look different.

A problem you're going to run into is that there are dozens of star trails right at the same level as the noise. This will seriously impair any noise removal.

Mark,

In a camera flash or other high drain devices, modern NiMH cells will provide significantly more amp hours than alkalines.

Oh, and I should have mentioned that using NiMH cells in a modern flash (the 420ex included) will cut the recycle time in half as compared to alkaline cells.

They do not draw power from the camera body.

 

As to what batteries to use in the flash you really, _REALLY_ want to use NiMH cells. First off they're rechargable, second, for high current devices like camera flashes, modern NiMH cells have higher capacities and can achieve higher drains on the battery (ie has a lower internal resistance) than alkaline cells.

 

A good place to get NiMH batteries and chargers online is http://thomasdistributing.com/

 

and yes the 420ex is a good flash and does tilt and swivel

Well, as a quick comparison... my 550ex has a duration such that a pellet fired from my inexpensive (the low end of something that isn't just a cheap plastic gun, I think it claims 580 feet / sec) will be illuminated by the flash for about a distance of 1 inch at the lowest manual flash setting. More interesting is things being shot with the pellet gun because that won't be moving as fast.

I believe Julian's problem can be mitigated by making sure none of the extra 4 pins are are making electrical contact with anything. They can cause confusion in your triggering device. A well behaved flash system will provide a few volts of sync voltage when it's ready to fire and drop that voltage for a few milliseconds (or more while it's recharging) to signal that it has fired the flash. If one of those pins is connected to the proper contact the trigger won't see the voltage drop and think that the flash hasn't finished firing.

I should add that the sync speed on the 10d is 1/200. At 1/250 you'll be using "high speed" sync and that will be far worse than using 1/200 at normal sync.

Depending on how close the candles are to each other, a thousand candles starts getting bright. And since people aren't going to be moving fast (it is a vigil after all) and you're on a tripod, you shouldn't need high shutter speeds. If you're photographing a person holding the candle then you would want a shutter speed of about 1/30 at f/2.8 and ISO 1600 (per my quick tests 5 minutes ago). This is assuming you don't want it to look like it's daylight lit. Also candles can vary drastically in brightness (ie 2 stops in either direction depending on the condition of the wick). Without knowing more information (specifically how are the people and the candles going to be organized), I can't say much more but I think you'll be fine without going to the extremes you're mentioning.

There are lots of ways... experiment and figure out what works best for you.

 

 

If you cover the four extra contacts on the st-e2 you will not get the preflash but will also fire the 420ex at full power and the 550ex at full power unless it's in manual mode.

 

For cheap you can get pc cord y adapters (and similar) and you can also get pc cord to hot shoe adapters. Plug your flash gun into the hot shoe adapters. If they're metal (or just metal over plastic like mine was) you'll likely need to mask off the 4 extra contacts but you can do it in a more permanent way on the cheap hot shoe adapters. Paint it with nail polish or take a dremel and cut away some of the offending metal. You should use the 550ex in manual mode for this but the 420ex will again only fire full power.

 

As a final solution you can get "digital" triggers which will flash after a certain number of flashes (typically syncing with the second flash but if you use the 550 and the 420 in different flash "groups" it would need to sync on the third flash which most won't do because they aren't adjustable but a "trainable" trigger might). But personally I much prefer to either work with all of my light manual or all of my light ttl. This means either just the 550 (in manual) and the monolight or just the 420 and the 550. Otherwise it is very difficult to get predictable/reproducable results.

First of all, the duration is going to be less than a millisecond regardless of the power settings and probably less than a tenth of a millisecond. Secondly the more intelligent flashguns will vary their output duration with the level of charge in it's capacitors so that the light output is consistent.

 

However in 1 millisecond something traveling 20 mph will move .3 inches and in .1 millisecond will move .03 inches. What level of detail do you need to resolve in your images?

Tom:

 

Regarding my aperture comments, I'm not suggesting that changing the aperture will change whether there will be a flash sync problem, just that it will be made more apparent in the same way that dust on the sensor is much more apparent at very small apertures, ie just as the shadow of the dust is clearer, the shadow of the shutter will be clearer.

 

As to the duration of the flash, all I'm saying really is that you should test everything well. Who's to say a studio flash doesn't have a similar characteristic to the 430EZ that was tested. And no I wasn't refering to the high speed sync setting on the flash, I was refering to the section where the author popped the flash once a second for 8 seconds. As I said above read the section "Dependence of Flash Output on Charge Time" on that page. Basically what the article alludes to is that one second isn't enough to completely recharge the capacitors but they're charged enough to fire the flash at the level requested so the duration of the flash was increased as the voltage in the capacitor dropped to keep the total light output constant. Thus flash duration on a 430EZ is not just variable dependent on the output level but also can vary by a factor of 3 or 4 AT THE SAME OUTPUT LEVEL.

If my thinking is correct, the smaller the aperture, the more pronounced the edge of this darker band will be. Also packs will have different timing characteristics based on the output level and may even have different timings characteristics based on other things (look at the section Dependence of Flash Output on Charge Time on this page: http://eosdoc.com/manuals/notes/discharge/ ). In other words thorough testing with a new system of any kind is a good thing.

As far as the ste-2 "defying the laws of physics", I find that so long as the flash sensors (on the front of the flashguns) are pointed in the general direction of the subject and ste-2 is pointing in the general direction of the subject, (point the front of the flashgun at your subject and then swivel/tilt the head to where you need it. Leave the ste-2 on the hotshoe since your camera is pointed at the subject) everything works just fine due to the IR reflection off of the subject. Things tend to work unpredictably if I try and point the IR sensor on the flashes at the ste-2 itself because of the somewhat limited cone of illumination from the ste-2. Once I figured this out, I've never had a problem with my flashguns not triggering. This also means that you can use an ste-2 with a 550ex on a stroboframe rather than connecting the two with a cable.

Rob, if you read the article it suggests using two of them as optical slaves triggered by on camera flash. It then suggests putting them 3-5 feet from the subject (using distances to get different flash ratios) on either side. So I do believe it will achieve hugely better results than what you'll get with any on camera flash but it still won't compare to a couple of cheap slaves aimed at (two foot square) pieces of foam core (for a simple and cheap example).

While it is annoying and not a good long term solution, you may be able to use flash exposure lock (FEL) to your benefit. That triggers the preflash when you press the exposure lock button and I believe that so long as you shoot the picture within some short delay (16 seconds?) it won't fire the preflash when you trip the shutter button.

I don't know what you are working with so my first suggestion may not work. Basically up the film speed / sensor sensitivity to the point that you can bounce the flash. On my canon 10d I am more than happy to shoot at ISO 800 and do longer distance bounces and prefer ISO 1600 with bounced flash to anything with direct flash. My second suggestion would be to just not worry about it for candid shots but for portraits aim the flash backwards and to the side a bit and have someone holding a piece of white foamcore or poster board for you to bounce the flash off of. It's cheap and very effective albeit somewhat unwieldy.

You mention that you have the digital rebel / 300d. That camera does not have flash exposure compensation (which is a horrible omission on canon's part). I believe that the 550ex will give you flash exposure compensation using the controls on the flash (I know that the flash offers it but since I've never tested it I don't know if it will work with the 300d) where the 420 doesn't give you any control at all (the 420 has no controls other than on/off and slave setup).

As the previous poster said, you need to set the 550 to manual (and as a standalone flash rather than master or slave while we're at it). If you only want to use the flash to trigger the strobes and don't want the light of the 550, point the 550 backwards and set the power to 1/128. If you do want the light of the 550, set it as you would want it. If you don't set the flash to manual, then you would be confusing the ettl also and getting nothing good out of it.