[alohadave]

attack11, I have yet to see you post any kind of reference or 'proof' that PTTL works the way you claim it does. You've posted from the flash manual snippets that do not support your argument.

"* Bounce flash messes up the distance
This is where the AF360 and AF540 will turn off the distance scale! The flash distance is no longer the camera-to-subject distance, but camera-to-ceiling-to-subject distance.
Depending on where you are this distance could be enormous. If you're outside then it's infinite "

That scale has nothing to do with the camera. It's the flash-to-subject distance that will produce a 'well exposed' image. For example, if the scale showed 5-20 feet at ISO 100, F8, then you could place the flash 10 feet from the subject and if you had the camera 40 feet away from the subject, your subject would still be well exposed. When you move the head away from 90 degrees, the flash knows that it is bouncing and the scale isn't used.

This says nothing about how PTTL works, by using a pre-flash to determine how much flash power to use. As long as the camera can see the preflash, it can try to determine the flash power to use. The camera/flash can be fooled, just like the camera meter can be in certain lighting conditions. Bounce flash is not the same as fill flash and they have different purposes, but they can both use PTTL.

[Ben_Edict]

how would you know if you've never read the manual?

so, notice that keyword "distance". it's not "that simple", and it requires the flash head be at 90d with an AF lens to transmit the distance. you can think differently if you want, but even the pttl requirements on another page in the manual state you need an AF lens for it to work properly.

this isn't guess work either. for the first few months of owning my 540 i tried bouncing with pttl in tall & average rooms, usually with white or off white ceilings. i got random results from frame to frame, so even trying to measure the pre-flash as a bounce is a gamble, and the flash has no idea what power ratio to use .. which it displays by a lack of a meter.

btw, your definition of pttl is just ttl.

Thanks for the scan. It clarifies the misunderstanding!

YOU understand "distance" as that of the flash-to-subject. That is wrong. The camera-to-subject distance is meant - and that is completely independent of the reflector's angle. So, not the least problem with P-TTL.

Ben

[newarts]

btw, your definition of pttl is just ttl.

It is *almost* the same; ie. the camera uses ttl on the pre-flash whether or not it subsequently communicates with a remote flash.

This allows the multi-segment metering system to better adjust the flash.

It implies, for example, that an optically triggered remote, single-shot flash cannot be used with an AF lens as the pre-flash fires the remote prematurely.

Dave

PS, I'm sorry there has been so much acrimony over the usage of a simple technical achronym.

[attack11]

Thanks for the scan. It clarifies the misunderstanding!

YOU understand "distance" as that of the flash-to-subject. That is wrong. The camera-to-subject distance is meant - and that is completely independent of the reflector's angle. So, not the least problem with P-TTL.

Ben

Ben, the (primary) flash for P-TTL is expected to be on the camera. So, nothing I've said is wrong, and my examples showcase how it's supposed to work. Read the url I linked.

But following the logic which we both agree to (camera to subject distance), introducing a bounce requires more power than the P-TTL algorithm is going to send. Camera, to variable height ceiling, to subject. P-TTL doesn't know the ceiling height, it's angle, anything; hence it's not meant for bouncing.

Again, the distance meter disappears when you change the flash head angle. That's the visual indicator that it will fail/random result.

[flyer]

Again, the distance meter disappears when you change the flash head angle. That's the visual indicator that it will fail/random result.

That's your interpretation. My interpretation is different, and it works. Whenever there is a highly reflective object in the picture, the result will be so and so because the camera tries to preserve highlights. So, your pictures with the mirror will never be right, wether you bounce or not. The K-7 is improved compared to other Pentax cameras when it comes to P-TTL, but it still is not perfect.

[attack11]

That's your interpretation. My interpretation is different, and it works. Whenever there is a highly reflective object in the picture, the result will be so and so because the camera tries to preserve highlights. So, your pictures with the mirror will never be right, wether you bounce or not. The K-7 is improved compared to other Pentax cameras when it comes to P-TTL, but it still is not perfect.

it's not an interpretation, it's how it works. see that meter/guide in the manual? the mirror picture (shooting into it) is proper. no clipping, proper exposure. bouncing is under exposed with clipping.

i think i'm done trying to get you guys to understand how you're misusing things.

[GeneV]

The camera's p-ttl algorithm will always want to minimize blown highlights. In both of your examples, the blown highlights will be the reflector of your flash. I think in both cases, the camera will end up telling the flash to output at its minimum flash output level. Frankly even at its minimum level, the flash reflector image will still be blown. Now, given that for both examples, the flash output the same amount of light, you will get a brighter image from a 90d flash position simply because more light is reflected back to you to light you up. With the 45d bounce, the light ends up bouncing from the mirror to the ceiling and then back to you. This also explains why both of example are underexposed.

The blown highlights algorithm is one of the most annoying things about P-TTL, and it is a reason why I use bounce (or wireless with one flash bouncing) far more often these days. White shirts are common, and too often they cause awful results in P-ttl. The underexposure problem is far less acute with bounce flash that tends to illuminate the background in a closer proportion to the foreground.

[attack11]

the whites are protected by matrix metering. in general i never shoot matrix unless it's absolutely needed (or using pttl fill).

[ma318]

Ben, the (primary) flash for P-TTL is expected to be on the camera. So, nothing I've said is wrong, and my examples showcase how it's supposed to work. Read the url I linked.
But following the logic which we both agree to (camera to subject distance), introducing a bounce requires more power than the P-TTL algorithm is going to send. Camera, to variable height ceiling, to subject. P-TTL doesn't know the ceiling height, it's angle, anything; hence it's not meant for bouncing.
Again, the distance meter disappears when you change the flash head angle. That's the visual indicator that it will fail/random result.

I think your understanding of how the distance is used by the p-ttl algorithm is not quite correct. You believe that p-ttl must use this camera-to-subject distance to calculate the actual final flash output level. You believe that, without this distance data, p-ttl would not be able to figure out the correct final flash output level.

P-TTL will attempt to calculate a proper flash exposure with or without the distance data from the camera. It does it by using the camera's meter to measure the exposure from the pre-flash. Based on the pre-flash exposure, the camera will make appropriate adjustment for the final flash level. (This is the reason why even non-auto focus "Pentax-A" lens would work with P-TTL.) In any case, p-ttl (or any ttl from any vendors) could still be fooled by very bright objects in the frame. For very tricky lighting situation or some special effect, you might be better off with manual camera exposure and manual flash output control or use flash compensation.

[newarts]

I just demonstrated to my satisfaction that with just the on-board flash the flash parameters are set by the pre-flash. This happens with or without an AF lens (I tried both AF and A lenses.)

The proof is that with an optically slaved dumb auxiliary flash, the shot is underexposed compared to a shot with no auxiliary flash firing in synchronization with the camera's preflash.

Dave

[kyc]

notice the second bullet point on the right side. that's you why need a) an AF lens and b) the distance to the subject (90d head angle).



you can bounce P-TTL all you like.. it's gonna be random.

Count me in the camp that says you can bounce P-TTL properly. If you want an actual official reference from Pentax, let's go straight to the flash manuals. For the AF360FGZ we have

Page 53: Bounce Flash. With the AF360FGZ you can tilt the flash head... this function is effective when photographing in P-TTL or TTL mode...

Page 35: Wireless Mode (P-TTL Photography). How does P-TTL wireless work with an external flash (with on-camera flash serving as control only, i.e., not firing) if only the subject to lens distance matters? If I put the flash close to the subject you'd need a different illumination than if the flash was further away?



As for requiring autofocus to work, I believe many are finding that as long as you have an A-lens that P-TTL works. Perhaps for simplicity (not needing to explain between A-lens and other manual focus lenses) Pentax describes P-TTL as requiring AF lenses.

Anyway, we have clear definitive statement from Pentax that P-TTL works when bounced and wireless (bounced or direct).

--KYC

[kyc]

Ben, the (primary) flash for P-TTL is expected to be on the camera.

I don't think this is a requirement. You can do wireless P-TTL with on-camera controller (doesn't fire) and off-camera main flash. It's in the manual.

--KYC

[egordon99]

actually, instead of leaving it up to interpertation i offer you these 2 images.

iso100, f8.0, da* 16-50, my bathroom: back to back within seconds so ambient light shouldn't matter.

why are the histrograms, and exposures so different if you can bounce pttl. there's clipping, so it's not the room shape.

90d


45d


if pttl worked with a bounce, then i should a) not be clipping and b) have a properly exposed frame taking into account the light source is in frame.

I think what happened in that test is the flash was firing at 1/1 power to TRY and expose for f/8 ISO100, but 1/1 was not powerful enough. Bouncing eats up a lot of light. Try the test again but with ISO400, f/2.8.

Many folks have said you're wrong, so I'll be one more voice saying "you're wrong"

Canon's E-TTL, Nikon's iTTL, and Pentax's P-TTL all pretty much work the same way. And they all work when the flash is bounced. I probably shoot more than 10,000 E-TTL'd shots a year, and when I shot Pentax it worked about the same.
x-TTL WORKS WHEN BOUNCING!

The only part where the manual is wrong is in stating you need an autofocus lens. You need a lens which transmits the aperture to the body, so an A lens would work as well.

Here's how the x-TTL systems work -
1)Flash sends a brief pre-flash at a known power level (pretty low power)
2)Light goes wherever it's going to go for the shot, be it direct, or bounced on the wall BEHIND you
3)Metering sensor in the body reads the light reflecting BACK from the subject
4)Since the body knows the f-stop of the lens, and the ISO, it can calculate WHAT power to fire the main flash for proper exposure. The metering sensor doesn't CARE ONE BIT how the light "arrives" at the subject, it just sees the light reflecting back to the sensor.

[egordon99]

to prove what? i already know how to use pttl, the point i'm trying to get across is if you're not aiming the flash where the lens is pointing you're not going to get what you want (every time). pttl exposures don't need exposure compensation in raw conversion, etc; it's supposed to provide "perfect" fill.

i can provide you examples of how pttl will compensate for a light source in frame if you want.


these were shot at the same location, within feet of each other but there isn't a light source in one. how'd the camera react? different shutter speeds and flash power.

left is 1/60, right is 1/160; iso 400, f2.8, da* 50-135. camera was in AV so i could literally run around, point and fire without a concern about exposure. i like to diffuse with a fong white cap because it's large.


that's a pic of me at a flickr meet shooting that 400mm on the other page.

Do you really understand how FLASH PHOTOGRAPHY (not P-TTL specifically) works? Shutter speed has NO bearing AT ALL on the flash exposure (as long as you're within the X-sync range of whatever you body is capable of)

[egordon99]

Again, the distance meter disappears when you change the flash head angle. That's the visual indicator that it will fail/random result.

Nope, that's the visual indicator that the body/flash has NO idea where you're going to send the flash light. You could send it to the wall behind you, the ceiling, etc.....

With the flash straight on, it's an easy calculation (assuming one knows the guide number) for the minimum range (what would give you proper exposure at the minumum power level, I think it's 1/128) and the maximum range (1/1 power) for the given ISO, and f-stop.