Flash System Operation (Long)

Started Jun 30, 2006 | Discussions
Mark VB
Senior MemberPosts: 2,912
Like?
Flash System Operation (Long)
Jun 30, 2006

Some in this forum, and more particularly in the KM DSLR forum, have talked about issues with the 7D's flash system (and to a lesser extent, the 5D). Some have asked the question whether the Alpha 100 will have an improved flash system or the same one as in the 7D/5D.

This afternoon I was talking with Phil Bradon (considered by many long-time Minolta shooters the "guru" of Minolta cameras). Although he doesn't know much about the Alpha 100, we talked a little about the 7D flash system, which convinces me that until someone does extensive tests with the Alpha 100 and a flash, there is no way to know how it will work. But there likely will be differences compared to the 7D, principally in the areas of flash metering and the lag time between the pre-flash and actual flash.

As most know, the 7D/5D (like all DSLRs) measure a pre-flash to determine the correct flash exposure. Based on that measurement, the camera cuts off the flash when it believes the proper exposure is reached. What I didn't know is that the pre-flash on the 7D when used with an external flash (5600 or 3600), is actually 14 rapid pulses. This matches the number of metering cells in the 7D's honeycomb meter pattern. There is one pulse for each metering segment, and the camera takes a reading from each segment during the pre-flash. (For the built-in flash there are only four pre-flashes, which effectively means a different flash metering algorithm is used for the built-in flash compared to an external flash.) The delay between the pre-flash and actual exposure is the time the camera needs to process the pre-flash metering to determine proper flash exposure. Thus, it is the camera, not the flash, that determines the delay between the pre-flash and actual exposure.

The Alpha 100 has a 40-segment exposure meter. The obvious question is how this new system will work for flash photography. Will it still use a 14-pulse pre-flash with external flash units, or will it use a different number of pulses, or perhaps go to a single pulse? None of this is known to any of us, I think.

The fact that the 5600HS emits a series of 14 pulses during the pre-flash also, I think, helps explain some of the exposure problems many have experienced, particularly when shooting with light modiers on the flash (such as an OmniBounce, the many Lumiquest devices, or something similar which involves bouncing the flash and/or shooting it through a diffusion surface). For the camera to perform the flash metering, there must be enough of the pre-flash pulses reflected back to the camera for it to measure. Bouncing the flash, or using a diffuser, reduces the effective power of the pre-flash like it reduces the effective output of the flash during the exposure.

However, the fact that the pre-flash is 14 separate pulses suggests to me that the level of each pulse must be weaker than I had assumed the pre-flash to be (similar to hi speed sync, which reduces the effective output of the flash to generate the series of pulsed flashes that are used in this mode). This means that when using a light modifier, or just bouncing the flash without a modifier, the effective output of the pre-flash pulse will be quite low, restricting the range over which it will provide a sufficient amount of light for the camera to measure (depending also on the specific conditions, such as ceiling height, other reflecting surfaces, the tonality of the subject, etc.).

When the camera cannot get a pre-flash reading it puts out a full power flash; that is, it maximizes the flash output since it could not determine that some lower output will provide the proper exposure. This frequently results in overexposed flash pictures.

Similarly, when the camera sees a dark subject, it also puts out more power than is actually needed because it is, generally speaking, trying to make that dark subject into a middle tone. This means that a higher flash output is required to do this, which also will have a tendency to blow out lighter toned areas of the subject. For example, guys at a wedding in dark suits, against a dark background; most of what the camera will see is black or very dark. Trying to make the subject neutral toned, white shirts or other lighter toned areas of the subject, including faces, will get overexposed and may blow-out.

So, with respect to the Alpha 100 flash system, might the 40 segment meter allow a more refined analysis of the subject that will prevent overexposed flashes by better "seeing" areas of a picture that might blow-out and reducing the flash power to preclude that? Might Sony have figured out a way to have the meter rely on only a single pre-flash pulse (or fewer pulses), reading all 40 segments at the same time, which could allow a more powerful pre-flash pulse that could minimize problems when using light modifiers or non-direct flash? Has Sony improved the camera's ability to more quickly measure the pre-flash and determine the correct flash exposure, thereby reducing the lag time between the pre-flash and actual exposure (as Nikon apparently did between the D70 and D200)?

These and other questions will need to await some tests of the camera. It is unlikely that any of the commercial sites, including this one, will undertake such an extensive testing regime of the flash system (none of them have done it before), so it likely will fall back on early purchasers of the camera to report their experiences. But, given the entry level status of the camera, will any of those people have the knowledge, experience or inclination to undertake such a testing regime? I sure hope so.

-- hide signature --

Mark Van Bergh

Gary Friedman
Regular MemberPosts: 323
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jun 30, 2006

This is, indeed, an excellent question, and in fact I was wondering the same thing -- will there be 40 pre-flashes, and will the pre-flash delay improve?

Last week, as part of my research for the upcoming ebook on the Alpha 100, I had arranged some lab time with a digital storage oscilloscope so I can actually measure what's going on and compare it to the 7D. These findings will be published in Chapter 7.

My guess, though, is that until Sony develops a sensor that possesses the same reflectivity as film, then the digital flash metering system will continue to be less reliable as compared to Minolta's superb off-the-film exposure method.

For those of you new to the Minolta flash heritage, I have written a technical article about how the flash works, and it can be viewed at http://www.FriedmanArchives.com/flash.htm .

I hope this helps!

Sincerely,
Gary Friedman

Reply   Reply with quote   Complain
clausgedved
Regular MemberPosts: 146
Like?
Re: Flash System Operation (Long)
In reply to Gary Friedman, Jun 30, 2006

Does the problems with digital sensors only go for Minolta or is the same with other brands to ?

Reply   Reply with quote   Complain
ab012_KM7D
Senior MemberPosts: 2,644
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jun 30, 2006

Mark VB wrote:

The delay between the pre-flash and actual exposure is the
time the camera needs to process the pre-flash metering to
determine proper flash exposure. Thus, it is the camera, not the
flash, that determines the delay between the pre-flash and actual
exposure.

Nice thread!

When the 40 segment was announced it also made me think of the flash changes. One thing that gives some hope - so far in the interviews with Sony Japan (in the Japanese media) one topic which comes up quite often is with regards to the the significant processing power in the a100.

For example:

-the processing speed was often a highlight in the marketing

-the instantaenous enlargement and viewing on the LCD (much faster than expected, and better than competitors

  • improved AF calculation speed has also been noted (from Sony engineers).

These give hope IMO that at least on the processing side of things Sony have made quite a bit of contribution, and hence the potential for faster processing (and from your suggestion - short preflash)

-- hide signature --

Bernard

AS rocks!
lens reviews and more on dyxum.com!

Reply   Reply with quote   Complain
Eleson
Senior MemberPosts: 1,134Gear list
Like?
more or less same for all current ones (NT)
In reply to clausgedved, Jun 30, 2006

All need to determine flash exposure before flipping up the mirror as they cannot rely on reading of the film (sensor)

-- hide signature --

Cheers
Erland

 Eleson's gear list:Eleson's gear list
Konica Minolta Maxxum 7D Sony SLT-A77 Sony 50mm F1.4 Sony DT 55-300mm F4.5-5.6 SAM Tamron SP AF 17-50mm F/2.8 XR Di II LD Aspherical (IF) +3 more
Reply   Reply with quote   Complain
matander
Regular MemberPosts: 293
Like?
Re: more or less same for all current ones (NT)
In reply to Eleson, Jun 30, 2006

Eleson wrote:

All need to determine flash exposure before flipping up the mirror
as they cannot rely on reading of the film (sensor)

Hi Brother,

Pentax D and DS (not DL) can also use TTL at lower ISO (200-400), this feature is highly regarded by Pentax íst and hopefully TTL compatibility will be kept with the upcoming D and Ds replacement cameras. K100D and K110D are the DL successor.
Mats

Reply   Reply with quote   Complain
Mark VB
Senior MemberPosts: 2,912
Like?
Re: Flash System Operation (Long)
In reply to Gary Friedman, Jun 30, 2006

Gary Friedman wrote:

This is, indeed, an excellent question, and in fact I was wondering
the same thing -- will there be 40 pre-flashes, and will the
pre-flash delay improve?

Last week, as part of my research for the upcoming ebook on the
Alpha 100, I had arranged some lab time with a digital storage
oscilloscope so I can actually measure what's going on and compare
it to the 7D. These findings will be published in Chapter 7.

My guess, though, is that until Sony develops a sensor that
possesses the same reflectivity as film, then the digital flash
metering system will continue to be less reliable as compared to
Minolta's superb off-the-film exposure method.

For those of you new to the Minolta flash heritage, I have written
a technical article about how the flash works, and it can be viewed
at http://www.FriedmanArchives.com/flash.htm .

I hope this helps!

Sincerely,
Gary Friedman

Thanks Gary. Perhaps you will be the person we will be looking to for an evaluation of the Alpha 100 flash system, how it operates, and whether it has resolved some of the shortcomings of the 7D's flash system.

-- hide signature --

Mark Van Bergh

Reply   Reply with quote   Complain
Timothy S Broadley
Senior MemberPosts: 1,297
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jun 30, 2006

Mark:

There is very little doubt in my mind that Sony has tweaked the camera. I was in Japan recently and had an opportunity to play with the alpha. One thng that immediatedly leaped out at me was there was no hunting on the AF. Secondly in a series of about 10 shots, the AF did not focus on a spot where I did not expect it to be at the time of exposure. These results, as very informal as they are, seem to be consistent with David K's experiences as I understand them. Lastly I tried to get it to blown-out highlights with a some very broad range scenes I.e White light housing against Black ceiling and did not succeed.

Each of these issues occurs with my 7d. In trying the alpha, I did not get any of them to occur which, to my mind, indicates that the Sony Engineers have been involved in system development and refinment.

This is all FWIW and I know it does not speak to the flash issue you raise but if Sony did not tweak the flash, I'd be surprised. I think the alpha is going to be a pretty competent camera.

cheers
Tim

Reply   Reply with quote   Complain
Timothy S Broadley
Senior MemberPosts: 1,297
Like?
Re: Flash System Operation (Long)
In reply to Gary Friedman, Jun 30, 2006

Gary:

One thing to note is that we already know that the coating on the sensor has been changed because of the anti-dust implementation. It would not in my view be unreasonable to say that the reflectivity of the sensor has been altered as a result of that implementation alone. Now whether or not that is a good thing , I don't know but it is clear to me that Sony has messed with this camera (see other post in this thread)

cheers
Tim

Reply   Reply with quote   Complain
Biu
Biu
Veteran MemberPosts: 3,151
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jun 30, 2006

The delay from pre-flash is not due to number of pre-flashes. The pre-flash works typically in the 3-5kHz range and 14 counts are relatively fast.

It happens to all DSLRs using pre-flash. I use Canon before and the same problem. The problem is that the camera rely on the reflection from the mirror to direct the light to the exposure sensors and therefore the mirror cannot start to flip up until the pre-flashes are over. The flip of the mirror is a much slower mechanical movement and this is why a portion of the population reacts fast enough to cause the lazy eye.

Biu

Reply   Reply with quote   Complain
Mark VB
Senior MemberPosts: 2,912
Like?
Re: Flash System Operation (Long)
In reply to Biu, Jun 30, 2006

Biu wrote:

The delay from pre-flash is not due to number of pre-flashes. The
pre-flash works typically in the 3-5kHz range and 14 counts are
relatively fast.

It happens to all DSLRs using pre-flash. I use Canon before and
the same problem. The problem is that the camera rely on the
reflection from the mirror to direct the light to the exposure
sensors and therefore the mirror cannot start to flip up until the
pre-flashes are over. The flip of the mirror is a much slower
mechanical movement and this is why a portion of the population
reacts fast enough to cause the lazy eye.

True enough, but not all cameras have the same delay between the pre-flash and the actual exposure. For example, in the Nikon forums there is discussion about the time lag between pre-flash and flash exposure being much shorter on the D200 than the D70. Others have compared differences between some Canon cameras and the 7D (or 5D).

What you are explaining is the generic operating mechanism for all DSLRs, which noboby questions. But, the implementation is not the same on all cameras.

Many here and in other forums (e.g., KM DSLR) have complained that the 7D has a longer delay between pre-flash and flash than some other cameras, which leads to a greater incidence of blinking eye syndrome. While I have not found this a huge problem (but it does occasionally appear), others have and have asked whether the Alpha 100 will be the same or different. My original post was partially intended to address that question, explaining how the flash system in the 7D operates, some differences in the Alpha 100 (e.g., the 40-segment metering vs. 14-segment), and that we will have to wait for actual tests of the Alpha 100 to see just how it performs.

As Tim and, I think, some others have noted, it would appear that there are definite changes in the camera's operating system compared to the 7D/5D. Hopefully such changes, for the better, also will apply to the flash system.

-- hide signature --

Mark Van Bergh

Reply   Reply with quote   Complain
Biu
Biu
Veteran MemberPosts: 3,151
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jun 30, 2006

This is an age long problem dated well before digital SLR. It started since film cameras using pre-flash mode. There virtually were no complaint on flagship products (such as the Canon EOS 1, Nikon F5, or Minolta Maxxum 9) because of their high speed mirror and shutter mechanism. On lower level body, people do feel the pain.

For the D200, Nikon put in a newly designed, much improved mirror assembly. This is why the delay is much shorter. Good move for Nikon. This is why Nikon emphasis this in their ad.

Biu

Reply   Reply with quote   Complain
Mark VB
Senior MemberPosts: 2,912
Like?
Re: Flash System Operation (Long)
In reply to Biu, Jun 30, 2006

Biu wrote:

This is an age long problem dated well before digital SLR. It
started since film cameras using pre-flash mode. There virtually
were no complaint on flagship products (such as the Canon EOS 1,
Nikon F5, or Minolta Maxxum 9) because of their high speed mirror
and shutter mechanism. On lower level body, people do feel the
pain.

Before DSLRs, the primary method of flash metering was off the film, during the actual flash exposure. While some systems did offer a pre-flash function, it was not the primary flash metering mechanism. This is why the problem was little discussed until the advent of DSLRs, for which "off the sensor" flash metering was not possible, and why just about all DSLRs went to the pre-flash metering system.

While I would expect that a higher speed mirror/shutter mechanism would reduce the delay between the pre-flash and actual flash, I don't think that fully explains the differences in the delay between cameras of the same brand, or between brands. But, here I am speculating, not relying on any specific evidence or information. I would love to find out just how the Nikon flash system operates - e.g., is the pre-flash a series of pulses or one burst? Does it have a different CPU and/or processing algorithm allowing it to calculate the flash exposure faster than earlier Nikon cameras and thereby reduce the pre-flash to flash delay? Are there other explanation for the shorter delay (such as a faster mirror/shutter mechanism)?

I have not been able to find any explanations of the Nikon system, at least not yet.

Finally, and in the end, I raised all of these points simply as a predicate for those asking about the Alpha 100 flash system, and whether it is the same or different than the 7D/5D (and the latter two are probably not the same, but are at least similar). There is much still to learn about the Alpha 100.

-- hide signature --

Mark Van Bergh

Reply   Reply with quote   Complain
RickD
Contributing MemberPosts: 809Gear list
Like?
Thank you and a comment
In reply to Mark VB, Jun 30, 2006

What a great informative thread. Thank you everyone for the information. On previous threads about preflash and lazy eye in the 7 and 5D, someone said that the preflash-flash delay was longer in the KM cameras because they needed a slower shutter mechanism to avoid interfering with AS. Does anyone here know if this is, in fact, a limiting step? If so, it couldn't be solved with faster processing.

On a related but different topic, I emailed tech support at Metz to ask if their M5/6 adaptors would work with the Sony alpha. They said they had not yet received a unit to test, and therefore could not comment.

-- hide signature --

RickD

 RickD's gear list:RickD's gear list
Nikon D4 Nikon AF-S Nikkor 14-24mm f/2.8G ED Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 70-200mm f/2.8G ED VR II
Reply   Reply with quote   Complain
hanugro
Senior MemberPosts: 2,339
Like?
Re: Thank you and a comment
In reply to RickD, Jul 1, 2006

Some said that mirror can not be sped up because it can interfere with AS. I think Sony need to develop a way to isolate mirror and AS mechanism or develop a way so the Off-the-Sensor flash metering is possible!

I use omnibounce and never experience lazy eyes again. I think it is because less pre-flash intensity make less people react with pre-flash. The trade off is that the GN is significantly decrease.

RickD wrote:

What a great informative thread. Thank you everyone for the
information. On previous threads about preflash and lazy eye in
the 7 and 5D, someone said that the preflash-flash delay was longer
in the KM cameras because they needed a slower shutter mechanism to
avoid interfering with AS. Does anyone here know if this is, in
fact, a limiting step? If so, it couldn't be solved with faster
processing.

On a related but different topic, I emailed tech support at Metz to
ask if their M5/6 adaptors would work with the Sony alpha. They
said they had not yet received a unit to test, and therefore could
not comment.

-- hide signature --

RickD

Reply   Reply with quote   Complain
hanugro
Senior MemberPosts: 2,339
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jul 1, 2006

Mark VB wrote:

The fact that the 5600HS emits a series of 14 pulses during the
pre-flash also, I think, helps explain some of the exposure
problems many have experienced, particularly when shooting with
light modiers on the flash (such as an OmniBounce, the many
Lumiquest devices, or something similar which involves bouncing the
flash and/or shooting it through a diffusion surface). For the
camera to perform the flash metering, there must be enough of the
pre-flash pulses reflected back to the camera for it to measure.
Bouncing the flash, or using a diffuser, reduces the effective
power of the pre-flash like it reduces the effective output of the
flash during the exposure.

Yes for bouncing I find omnibounce is useless (always under exposure for me). I use omnibounce as direct flash to eliminate lazy eyes. If I want to bounce then I took it off since I don't get lazy eyes in bounce anyway. Bottom line is bouncing/diffuser help with lazy eyes for me.

However, the fact that the pre-flash is 14 separate pulses suggests
to me that the level of each pulse must be weaker than I had
assumed the pre-flash to be (similar to hi speed sync, which
reduces the effective output of the flash to generate the series of
pulsed flashes that are used in this mode). This means that when
using a light modifier, or just bouncing the flash without a
modifier, the effective output of the pre-flash pulse will be quite
low, restricting the range over which it will provide a sufficient
amount of light for the camera to measure (depending also on the
specific conditions, such as ceiling height, other reflecting
surfaces, the tonality of the subject, etc.).

I think this is not the case. In order for each honeycomb sensor to measure flash it needs decent light. I don't think 14 pre-flash means that the power is only 1/14 (or even 1/2) of what if only 1 pre-flash is involved because you can not direct each pre-flash to the area where each honeycomb sensor would be metering. The pre-flash would still lit the entire scene (not just one honeycomb meter).

That is also why omnibounce work in reducing lazy eyes effect because it reduce the intensity of pre-flash (in each burst and in a whole 14 of them).

Flash in general work like this, it will repeat a given burst of light until it is enough. This given burst of light is determining the GN of the flash. 5600 is more powerfull than 3600 so it can output more light in single burst.

HSS means less effective output since of those available burst you have to divide it again for HSS sync. Consider this simplistic example, supposed a given flash has capability of 10kHz (10,000 burst of given flash output a second). The maximum the flash can output in NON-HSS 1/100s exposure is 100 burst. If we use HSS with 1/1000s shutter than it become 10 burst only. And given that each area between the shutter opening need same burst of light to make it consistent then it is significantly less than 10 burst.

So, with respect to the Alpha 100 flash system, might the 40
segment meter allow a more refined analysis of the subject that
will prevent overexposed flashes by better "seeing" areas of a
picture that might blow-out and reducing the flash power to
preclude that? Might Sony have figured out a way to have the meter
rely on only a single pre-flash pulse (or fewer pulses), reading
all 40 segments at the same time, which could allow a more powerful
pre-flash pulse that could minimize problems when using light
modifiers or non-direct flash? Has Sony improved the camera's
ability to more quickly measure the pre-flash and determine the
correct flash exposure, thereby reducing the lag time between the
pre-flash and actual exposure (as Nikon apparently did between the
D70 and D200)?

More powerfull pre-flash without significantly reducing the lag would worsen the lazy eyes effect. I even propose the opposite for lazy eyes (less pre-flash intensity with faster lag time).

I think using 1 pre-flash may even be less powerfull than using multiple pre-flash because even each pre-flash in multiple pre-flash scheme still have to lit entire scene (making a 14 times pre-flash more powerfull than 1 time pre-flash).

Other manufacturer can get away with lazy eyes because they may have faster lagtime. Even 0.1s faster will count and reduce the occurance of lazy eyes. Aparently some said AS has something to do with why KM can't use faster mirror flap. So Sony also need to improve on this!

These and other questions will need to await some tests of the
camera. It is unlikely that any of the commercial sites, including
this one, will undertake such an extensive testing regime of the
flash system (none of them have done it before), so it likely will
fall back on early purchasers of the camera to report their
experiences. But, given the entry level status of the camera, will
any of those people have the knowledge, experience or inclination
to undertake such a testing regime? I sure hope so.

We all hope Sony will find a way to do Off-the-Sensor(OTS) TTL flash metering as it will solves ALL problems (red eye because of pre-flash and slow mirror flap, and also inconsistent flash exposure in bouncing/diffuser/etc). With OTS flash metering there will be no debate on slow mirror, flash intensity, lazy eyes, etc. It will be the perfect Minolta's flash film exposure that is the best at it times!

Reply   Reply with quote   Complain
mcl
mcl
Senior MemberPosts: 1,085
Like?
Re: Botton Line - There Are Different Levels of Compatible
In reply to Mark VB, Jul 1, 2006
Reply   Reply with quote   Complain
Mark VB
Senior MemberPosts: 2,912
Like?
Re: Flash System Operation (Long)
In reply to hanugro, Jul 1, 2006

hanugro wrote:

Yes for bouncing I find omnibounce is useless (always under
exposure for me).

This suggests a few explanations: 1) your subjects are generally light in tone and the flash meter is exposing lighter tones to be a middle tone (e.g. white that becomes gray - remember, the meter assumes all subjects are middle toned, and determines its exposure to create those middle tones); 2) you are using either too small an aperture or the distance to the subject, when combined with your aperture, is beyond the capabilities of the flash to provide a good exposure. If the flash cannot read the pre-flash it will provide a full power burst. If you are not getting overexposure with a full power burst (the problem I have encountered when the camera cannot read the pre-flash), and you are getting underexposure, then you are simply asking the flash to do more than it is capable of; 3) your flash and camera are in need of the calibration that KM was doing shortly after the release of the 7D because of certain differences between the way older 5600 flash units had their pre-flash set and the camera's expectation of the pre-flash power; 4) you have accidentally dialed in minus exposure compensation so the flash is simply underexposing because you told it to (by accident).

More info would be neede to try to figure out what the problem is, or perhaps if it is something other than mentioned above.

However, the fact that the pre-flash is 14 separate pulses suggests
to me that the level of each pulse must be weaker than I had
assumed the pre-flash to be (similar to hi speed sync, which
reduces the effective output of the flash to generate the series of
pulsed flashes that are used in this mode). (snipped)

I think this is not the case. In order for each honeycomb sensor
to measure flash it needs decent light. I don't think 14 pre-flash
means that the power is only 1/14 (or even 1/2) of what if only 1
pre-flash is involved because you can not direct each pre-flash to
the area where each honeycomb sensor would be metering. The
pre-flash would still lit the entire scene (not just one honeycomb
meter).

It is not that a pre-flash pulse (burst) is directed to any one meter segment but rather that each segment takes a reading in sequence with the pre-flash bursts (at least as it was explained to me). As for the actual power of the pre-flash, I was speculating that the use of 14 pulses during the pre-flash means that each pulse is likely weaker than a single pulse would need to be. Just like HSS flash, where the effective power of the flash is reduced because of the need to have several pulses (bursts) during the actual exposure.

That is also why omnibounce work in reducing lazy eyes effect
because it reduce the intensity of pre-flash (in each burst and in
a whole 14 of them).

Quite possibly true. Just as bouncing the flash likely also reduces the lazy eye effect. But, the reduced intensity of the pre-flash also means less light for the camera's flash meter to measure, which at times will lead the camera to think it needs to provide more flash power than it actually does resulting in overexposure.

More powerfull pre-flash without significantly reducing the lag
would worsen the lazy eyes effect. I even propose the opposite for
lazy eyes (less pre-flash intensity with faster lag time).

A weaker pre-flash will make it even more difficult for the camera flash meter to measure the pre-flash and set the proper flash exposure. There already are enough problems with flash consistency that this would be the wrong direction to go. I only rarely encounter the lazy eye effect and am more concerned with consistent flash exposures than the lazy eye effect. Since those complaining of the lazy eye effect, or at least some, seem to think this is due to the time lag, it would seem to have nothing (or little) to do with the power of the pre-flash.

I think using 1 pre-flash may even be less powerfull than using
multiple pre-flash because even each pre-flash in multiple
pre-flash scheme still have to lit entire scene (making a 14 times
pre-flash more powerfull than 1 time pre-flash).

You recognize that in HSS mode, with several bursts during the exposure, the effective power of the flash is reduced to give the same amount of light during each burst while the shutter is open. It would need to be the same with a multiple-burst pre-flash; each burst has to be of the same power, and multiple bursts likely means that each burst is weaker than if the system used only one pre-flash burst.

Other manufacturer can get away with lazy eyes because they may
have faster lagtime. Even 0.1s faster will count and reduce the
occurance of lazy eyes. Aparently some said AS has something to do
with why KM can't use faster mirror flap. So Sony also need to
improve on this!

See above. Assuming the lazy eye effect is related to the lagtime between pre-flash and flash, the solution is to change the lagtime, not reduce the power of the pre-flash.

Yes, Minolta (KM) engineers said they had to use a low vibration mirror/shutter assembly in order to work with the AS system. However, that generally accounts for the somewhat slower maximum flash sync speed of the 7D/5D/Alpha 100, of 1/125 with AS/SSS on and 1/160 with AS/SSS off. However, with studio strobes and the 7D, you actually can go up to 1/200 shutter speed and get proper flash sync (although that's not in the instruction book).

We all hope Sony will find a way to do Off-the-Sensor(OTS) TTL
flash metering as it will solves ALL problems (red eye because of
pre-flash and slow mirror flap, and also inconsistent flash
exposure in bouncing/diffuser/etc). With OTS flash metering there
will be no debate on slow mirror, flash intensity, lazy eyes, etc.
It will be the perfect Minolta's flash film exposure that is the
best at it times!

No problem there.

-- hide signature --

Mark Van Bergh

Reply   Reply with quote   Complain
hanugro
Senior MemberPosts: 2,339
Like?
Re: Flash System Operation (Long)
In reply to Mark VB, Jul 2, 2006

Mark VB wrote:

hanugro wrote:

Yes for bouncing I find omnibounce is useless (always under
exposure for me).

More info would be neede to try to figure out what the problem is,
or perhaps if it is something other than mentioned above.

It is just the flash GN is significantly reduces that the flash light can't illuminate the object enough. I up the ISO to 400 and ussually is OK! But I think your argument about inconsistent exposure and low pre-flash intensity because of difusser may be valid.

However, the fact that the pre-flash is 14 separate pulses suggests
to me that the level of each pulse must be weaker than I had
assumed the pre-flash to be (similar to hi speed sync, which
reduces the effective output of the flash to generate the series of
pulsed flashes that are used in this mode). (snipped)

I think this is not the case. In order for each honeycomb sensor
to measure flash it needs decent light. I don't think 14 pre-flash
means that the power is only 1/14 (or even 1/2) of what if only 1
pre-flash is involved because you can not direct each pre-flash to
the area where each honeycomb sensor would be metering. The
pre-flash would still lit the entire scene (not just one honeycomb
meter).

It is not that a pre-flash pulse (burst) is directed to any one
meter segment but rather that each segment takes a reading in
sequence with the pre-flash bursts (at least as it was explained to
me). As for the actual power of the pre-flash, I was speculating
that the use of 14 pulses during the pre-flash means that each
pulse is likely weaker than a single pulse would need to be. Just
like HSS flash, where the effective power of the flash is reduced
because of the need to have several pulses (bursts) during the
actual exposure.

It is exactly my point. If each one of those 14 pre-flash pulse also lit the entire scene then it is impossible to make it weaker just because only 1 honeycomb matrix read them. If that is the case then the honeycomb sensor won't read the reflective light properly.

Consider this, each of that 14 pre-flash pulses has power required so that the reflected ligth can be read by 1 honeycomb matrix (but it still has to lit entire scene anyway). Then the same amount of light produce by 1 pulse would be enough to be read by all 14 honeycombs (because, again, the flash will lit the entire scene anyway). With this in mind then if we make all 14 honeycombs to read the light at the same time then 1-pulse pre-flash system would be 1/14 the intensity of the 14-pulse pre-flash system. We then can double this 1-pulse pre-flash system and still only require 1/7 the intensity of 14-pulse pre-flash system.

That is also why omnibounce work in reducing lazy eyes effect
because it reduce the intensity of pre-flash (in each burst and in
a whole 14 of them).

Quite possibly true. Just as bouncing the flash likely also
reduces the lazy eye effect. But, the reduced intensity of the
pre-flash also means less light for the camera's flash meter to
measure, which at times will lead the camera to think it needs to
provide more flash power than it actually does resulting in
overexposure.

Yeah a dillema isn't it?

A weaker pre-flash will make it even more difficult for the camera
flash meter to measure the pre-flash and set the proper flash
exposure. There already are enough problems with flash consistency
that this would be the wrong direction to go. I only rarely
encounter the lazy eye effect and am more concerned with consistent
flash exposures than the lazy eye effect. Since those complaining
of the lazy eye effect, or at least some, seem to think this is due
to the time lag, it would seem to have nothing (or little) to do
with the power of the pre-flash.

Well the fact that using omnibounce reduce the chance of lazy eyes means that lazy eyes can also be fix by reducing pre-flash intensity. Sadly it may worsen other effect like inconsistent flash (I OTOH seldom has this, seems that only using omnibounce/difuser and bounce give me this problem)

I think using 1 pre-flash may even be less powerfull than using
multiple pre-flash because even each pre-flash in multiple
pre-flash scheme still have to lit entire scene (making a 14 times
pre-flash more powerfull than 1 time pre-flash).

You recognize that in HSS mode, with several bursts during the
exposure, the effective power of the flash is reduced to give the
same amount of light during each burst while the shutter is open.
It would need to be the same with a multiple-burst pre-flash; each
burst has to be of the same power, and multiple bursts likely means
that each burst is weaker than if the system used only one
pre-flash burst.

See above. I think the difference between HSS and pre-flash is that in HSS the reduced intensity is a must because of the limitation of number of burst and shutter speed combination. Pre-flash intensity OTOH does not bound by that limitation.

The minimum intensity required by 1 honeycomb sensor is in fact also enough for the other 13 (because the light illuminate entire scene that include all 14 of the honeycombs). So if we make a honeycomb system that can read/meter the reflected light at the same time then we don't need the other 13 pulse and the intensity is much reduced and a fix for lazy eyes as well.

I think until Sony make Off-The-Sensor flash metering then we can't really solves this problem. It all interconnect and if you try to solve one then you can make other problem worst. OTS is the only answer in this. Unless Sony find a way to isolate mirror movement (faster) with AS!

Reply   Reply with quote   Complain
Gil Knutson
Veteran MemberPosts: 4,239Gear list
Like?
Re: Flash System Operation (Long)
In reply to hanugro, Jul 2, 2006

hanugro wrote:

We all hope Sony will find a way to do Off-the-Sensor(OTS) TTL
flash metering as it will solves ALL problems (red eye because of
pre-flash and slow mirror flap, and also inconsistent flash
exposure in bouncing/diffuser/etc). With OTS flash metering there
will be no debate on slow mirror, flash intensity, lazy eyes, etc.
It will be the perfect Minolta's flash film exposure that is the
best at it times!

It HAS to be doable!!! Pentax DOES it now... with their digital
cameras.

-- hide signature --

Gil
Sardis, BC
Canada

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads