[SOldBear]

Huh? The metering negotiation between body and flash should happen in milliseconds and shouldn't be noticeable. Sounds more like red-eye reduction to me.

Have you actually measured it? Or have you found the spec somewhere?

The time between the preflash strobe and the main flash is actually noticeable. Some people (my son included) react to the preflash strobe and thus likely close their eyes during the main strobe. For this reason, I don't use p-TTL mode for group photos.

In my non-scientific guesstimation, the time between the p-TTL preflash strobe and the main strobe is in the neighborhood of 1/50 sec. Note the sequence:

1. Flash fires preflash strobe.
2. Mirror raises.
3. Leading shutter fully opens.
4. Flash fires main strobe.
5. Trailing shutter closes.

How long do you think it takes for (2) and (3)?

Do this simple experiment: setup an optical slave flash, one that is "dumb" (can not ignore the p-TTL preflash strobe). Using the built-in flash to trigger the slave, take a photo in front of the mirror, with both the camera and the slave flash visible in the mirror. Check the photo to see which flash(es) fires during exposure.

---

Last edited by SOldBear; 04-10-2013 at 02:31 PM.

[BrianR]

Huh? The metering negotiation between body and flash should happen in milliseconds and shouldn't be noticeable. Sounds more like red-eye reduction to me.

This is a pretty standard complaint of all the X-ttl flash systems. Not everyone will blink fast enough, but some people are pretty consistent blinkers. It seems much more common if the subject is staring down the barrel of the flash (not surprising I suppose)- I gave up on the built in flash vs people pretty quick and cursed the lack of manual controls on the built in flash which fixes this problem (they can blink after the photo all they want).

[glanglois]

I think I missed (another) step, SOldBear.

The way I'm envisioning this sequence, steps 2 through 5 take place regardless of the type or absence of flash. The incremental duration of image-taking using P-TTL is found, then, between step 1 and step 2.

But the experiment looks good and I think I've read of others using just that technique. In that case, though, why does red-eye reduction not result in masses of closed eyes? Is it because the interval between red-eye reduction flash and main flash is long in comparison to the interval between P-TTL flash and main flash?

[SOldBear]

The way I'm envisioning this sequence, steps 2 through 5 take place regardless of the type or absence of flash. The incremental duration of image-taking using P-TTL is found, then, between step 1 and step 2.

Step 1 makes a big difference in the context of this discussion (optical slave triggered by the built-in flash), because the preflash strobe in step 1 triggers the slave. My point is: the time between when the slave is triggered and when the shutter is fully open is NOT in the range of milliseconds. This confirms Jonathan Mac's comment that "by the time the camera takes the shot the moment has already passed" (flash is not in sync).

If the slave flash is used at reduced power (say, 1/32), it may still have sufficient charge to fire again when the main flash triggers it in step 4. But the first (maybe the only) triggering during step 1 does nothing to the exposure.

why does red-eye reduction not result in masses of closed eyes? Is it because the interval between red-eye reduction flash and main flash is long in comparison to the interval between P-TTL flash and main flash?

I think because of the long duration of the red-eye reduction, people realize that something is about to happen, and try NOT to blink.

On the other hand, the preflash strobe startles people. They blink unconsciously.

---

Last edited by SOldBear; 04-10-2013 at 02:59 PM.

[Lowell Goudge]

It's not an unreasonable expectation. Twenty years ago you had around ten networking protocols. Today you've got one. An open protocol benefited everyone, software companies included.

Before anyone points it out, I got the protocols the wrong way round. As any fule kno, E-TTL is Canon's protocol, not Nikon's.

Interesting article:
Nikon Flash Interface

It was meant as a tonge in cheek comment. Nothing on cameras is interchangeable. Not lenses not grips not even image fe formats , nothing!. Standardization does not exist in cameras except memory cards but only because that is controlled by bigger market sectors. Sony tried to go it alone on memory and failed.

[MPrince]

I don't own a Pentax flash and don't expect I ever will. I like older Nikon flashes. I have an SB-20, an SB-24, two SB-25s and an SB-26. For me, Nikon flashes are ideal for off-camera use in manual mode and also work great on-camera in manual or auto mode. That's all I need.

[pinholecam]

They already have a universal standard.
Its called 'M' mode


Being the underdog for flash, I don't think Pentax has any bargaining power in it.
Why would Canon or Nikon want to share anything with Pentax (or Oly/Panny for that matter)?

That said, starting at the lowest may be a good thing if the will is there.
Perhaps Pentax will give us Wifi communicating flash systems and controls in the near future and leap frog over the more antiquated comms protocols on Canikon.

[maxfield_photo]

But here's the thing: it works both ways. If Pentax cameras spoke E-TTL, I'd be able to mount Nikon Speedlights.

What's stopping you? Nikon still has the decency to include built in light meters in most of their flashes, or just shoot in manual. Mix and match, says I.

What's I'd really like to know is if by using Pocket Wizard Flex TT5s and Nikon flashes if I could control my lights remotely in manual mode from a Pentax camera. Maybe even use one of those new touch screen Sekonic meters. Anyone tried it?

[Jonathan Mac]

The pre-flash to main flash delay is much more than 1/50s, at least on my Metz. People thought the photo had been taken and started to relax & allow themselves to blink just as the main flash went. Getting to know manual flash is a much better option in most situations.

[top-quark]

Looking more closely at the "standard" hotshoe, different manufacturers aren't even pin compatible. The only standard bit is the larger X-sync pin. Pathetic. Bang goes that idea.

Could maybe do a blah-TTL to some-other-TTL hotshoe adapter. Just need to go back to university and study electronic engineering for a few years. So maybe not.

[Clavius]

Being the underdog for flash, I don't think Pentax has any bargaining power in it.

That's an advantage; nobody expects Pentax to market something spectacular in the flash area.

For example, I've been hoping for Pentax to release a WR WIFI flash for some time now. With a hotshoe wifi trigger for older cameras, but it can use the built in WIFI of the new cameras that are still to come...

[SOldBear]

Could maybe do a blah-TTL to some-other-TTL hotshoe adapter. Just need to go back to university and study electronic engineering for a few years. So maybe not.

It's not so simple. The adapter has to know two languages: blah-TTL and some-other-TTL.

We already see big companies (Metz, Sigma) with large research staff struggling with reverse-engineering one language.

I personally think whaterver-TTL is overrated. My 2 copies of AF540FGZ are used much less than my Sunpak 120J and Metz 54MZ.

[Neo_]

It strikes me that in all areas of technology you get popular thing, slightly less popular thing and ghettos. For example, you have iPad, Android and Blackberry Playbook.

In lighting this would be Nikon E-TTL, Canon I-TTL and every other bloody manufacturer's proprietary protocol.

I'm OK with P-TTL - I kinda understand its quirks and limitations. But I've just got a Fuji. This will be for available light shooting because I'm damned if I'm going to spend money on any of their crappy flashes. So why don't the manufacturers agree on a vendor-neutral protocol? The key word here is protocol. After all, I'd be stuffed if the PC that I'm typing on spoke a different version of TCP/IP than the server on which this site is hosted.

I know that giving up proprietary stuff is hard to do - there are egos involved, after all. And it's anathema to the very idea of vendor lock-in. But here's the thing: it works both ways. If Pentax cameras spoke E-TTL, I'd be able to mount Nikon Speedlights. I'm sure Pentax wouldn't want that. But, if I'm a Nikon user with any sort of investment in Nikon lighting, then the fact that Pentax cameras could use my existing lights with no limitations would remove a major barrier to switching brands. It would remove a barrier to professional uptake too: suddenly Pocket Wizards and other groovy third-party gadgets work with no limitations.

So, Pentax: drop P-TTL. Maybe keep it as a body option so that people's existing flashes don't suddenly become significantly less useful (although you've done that before). License E-TTL (or even reverse engineer it; what do Metz do?). There's prior art here: Pentax are one of the few manufacturers to realize that a plethora of proprietary raw formats is just rubbish and switched to vendor-neutral DNG.

Do you really think that would work?? you're assuming that Pentax and Nikon cameras are electronically identical?? serious?
This is pure sci-fi.
Cameras are computers and no way an algorithm designed for the way Nikon cameras communicate and react to the electronics in their flashes and their camera metering would work on another camera designed in a completely different way, firmwares are proprietary, closed source and not standard, so this is not going to happen.
As far as I'm concerned, I agree with David Hobby, TTL, whatever it's called, will never work perfectly, simply because there is no "correct" exposure in the flash world, the only correct exposure is the one where light does what you want it do do, and a camera will never be able to read your brain.
So instead of letting an engineer in Japan 30 years ago decide what your flash exposure should be, turn to M mode and live happy.

[top-quark]

Do you really think that would work?? you're assuming that Pentax and Nikon cameras are electronically identical?? serious?
This is pure sci-fi.
Cameras are computers and no way an algorithm designed for the way Nikon cameras communicate and react to the electronics in their flashes and their camera metering would work on another camera designed in a completely different way, firmwares are proprietary, closed source and not standard, so this is not going to happen.
As far as I'm concerned, I agree with David Hobby, TTL, whatever it's called, will never work perfectly, simply because there is no "correct" exposure in the flash world, the only correct exposure is the one where light does what you want it do do, and a camera will never be able to read your brain.
So instead of letting an engineer in Japan 30 years ago decide what your flash exposure should be, turn to M mode and live happy.

Dude, I know that the internals and firmware of a Pentax and Nikon camera are completely different. Same applies to my laptop and a Sparc server somewhere on the Internet. This doesn't stop the server responding sensibly when my laptop issues the command "GET /index.html HTTP/1.1". That's because they use the same hardware and software-independent protocols to communicate. And that's because sometime in the past twenty years or so, lots of hardware and software companies gave up their proprietary ones.

Camera / flash communication isn't fundamentally different. One set of signals from the camera says "Aperture is F5.6", another set from the flash says "I estimate 1/16 power is enough". These are just bits-on-the-wire (or bits over the air) and there is no good reason why the aperture information cannot be indicated with exactly the same series of logic pulses independent of the hardware being used.

As to whether the camera calculating flash power for you is a good thing in-and-of itself is another question. Overall, I'd say yes. When I want a bit of fill because I'm on a picnic in the bright sunshine and my kid's doing something really cute, she's not going to hang around while I fire off a few test shots. I simply want it to work.

[IchabodCrane]

Dude, I know that the internals and firmware of a Pentax and Nikon camera are completely different. Same applies to my laptop and a Sparc server somewhere on the Internet. This doesn't stop the server responding sensibly when my laptop issues the command "GET /index.html HTTP/1.1". That's because they use the same hardware and software-independent protocols to communicate. And that's because sometime in the past twenty years or so, lots of hardware and software companies gave up their proprietary ones.

Camera / flash communication isn't fundamentally different. One set of signals from the camera says "Aperture is F5.6", another set from the flash says "I estimate 1/16 power is enough". These are just bits-on-the-wire (or bits over the air) and there is no good reason why the aperture information cannot be indicated with exactly the same series of logic pulses independent of the hardware being used.

As to whether the camera calculating flash power for you is a good thing in-and-of itself is another question. Overall, I'd say yes. When I want a bit of fill because I'm on a picnic in the bright sunshine and my kid's doing something really cute, she's not going to hang around while I fire off a few test shots. I simply want it to work.

Every camera manufacturer uses different logic to drive their cameras and that's before we even think about flash photography. Adding flash, you create an interaction between camera and flash unit that's going to be independent among manufacturers simply because each camera is different. Therefore, trying to drive to a common flash protocol may necessitate a corresponding drive to a common camera logic, too. At that point, what's the difference between Canon, Nikon, Pentax, etc.?