K-5II P-TTL overexposure with tilted flash - test and analysis

krebsy75 · 01-05-2013, 12:11 AM

Bounce exposures are wild with our K-01 as well.

**bjdavey** · 01-05-2013, 12:56 AM

Indeed - my workarounds whenever I need bounce flash are to either dial in -2 stops underexposure, or to use the above mentioned optical wireless mode (which works fine, but has you awkardly holding the camera with one hand and the flash in the other). This confirms that the camera/flash combination is capable of the correct P-TTL metering actions with bounced light. Out of interest I have just held the flash in wireless slave mode immediately adjacent to the hot shoe - so that it is in almost the same position as the tests above where the flash was mounted. I repeated the sequence of direct flash, slight tilt, and full tilt - all shots are properly exposed, so it is definitely not a problem related simply to light direction.

As reported by the OP, downward tilt does not cause the problem, even though the rear flash LCD shows the tilt icon activated. However sideways rotation in both directions (which also activates the LCD screen tilt icon), does cause the overexposure problem in the same way as I reported previously (normal exposure when the rotation is only to the point where it has not yet activated the tilt icon, and overexposure when rotated just a tiny bit more, activating the icon). The overexposure occurs in green (auto), P, and Av modes.

It seems pretty clear that that the fault is somehow linked to the "signal" which the flash generates to indicate to the camera metering system that the flash is tilted upwards, or rotated sideways (all of the above with the AF540FGZ). That said, I'm not sure why this "signal" is necessary in the first place, as the TTL metering should work regardless of the pathway light travelled before entering the lens.

Noam · 01-05-2013, 01:30 AM

Originally posted by bjdavey

I'm not sure why this "signal" is necessary in the first place, as the TTL metering should work regardless of the pathway light travelled before entering the lens.

Other flash systems (I think Nikon and Sony, maybe Canon) do incorporate a method where the focusing distance (reported by the lens) is used to determine flash power when the flash is pointed straight forward. Compared to pre-flash TTL metering, It can yield a more accurate and consistent exposure that is not affected by non-gray subjects or reflections of shiny objects (which is a problem especially with flash pointed forward).

I am guessing that Pentax has a similar method that works in tandem with the PTTL metering. A simple way of testing this hypothesis would be to take several controlled shots on a tripod with the flash pointed forward, and the lens focused at different distances (either manually, or fooling the AF with e.g. your hand and locking AF).

**bjdavey** · 01-05-2013, 01:37 AM

Originally posted by Noam

EDIT: More precisely, it depends on the appearance of a distance scale on the flash LCD. When my flash is tilted down (slightly, for macro shots), the tilt icon appears yet the distance scale is still there, and the exposure is the same as straight (i.e., not overexposed).

I just read Noam's edit above. Indeed I can confirm using the AF540FGZ that the presence or not of the distance scale on the rear LCD is consistently related to the overexposure, in all tilt or rotation directions. The distance scale disappears when the tilt icon appears, with the exception of downward tilt when the distance scale remains visible. If the distance scale is showing, exposure is correct. When it is not showing, overexposure occurs.

**bjdavey** · 01-05-2013, 01:45 AM

Originally posted by Noam

A simple way of testing this hypothesis would be to take several controlled shots on a tripod with the flash pointed forward, and the lens focused at different distances (either manually, or fooling the AF with e.g. your hand and locking AF).

Nice suggestion - I just tried it. Using manual focus and direct (non-tilted) flash, there is progressive overexposure as focus distance increases, confirming a link between the camera's calculated focus distance and flash light output.

**bjdavey** · 01-05-2013, 01:57 AM

Originally posted by bjdavey

Using manual focus and direct (non-tilted) flash, there is progressive overexposure as focus distance increases, confirming a link between the camera's calculated focus distance and flash light output.

... and further to the above, that only happens if the distance scale is showing on the flash LCD. If the flash is tilted or rotated to the point where the distance scale disappears, overexposure is consistent throughout the focus range.

tzeentchiangod · 01-09-2013, 08:59 AM

Forgive me if this was mentioned up-thread... but has anybody had a conversation with someone at Pentax about this?

**Canada_Rockies** · 01-09-2013, 11:35 AM

Originally posted by dosdan

My pet hate when using bounce flash at an angle of 60°, is when the top half of a person ends up brighter than the bottom. Whick is OK for bust shots, but a problem with full-length shots. What do you guys do to reduce this? Set the flash to straight up? Flash towards the rear ceiling if the rear wall is close enough and light coloured?

Dan.

Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post.

I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading.

tabl10s · 01-09-2013, 10:17 PM

Originally posted by Canada_Rockies

Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post.

I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading.

I'm gonna try that.

**fs999** · 01-10-2013, 09:02 AM

It would be interresting to see if the new AF360 FGZ II has the same problem...

arnie0674 · 01-10-2013, 12:05 PM

Originally posted by Canada_Rockies

Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post.

I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading.

The K10d does not have this bug, so I would expect it to expose correctly

dosdan · 01-10-2013, 03:01 PM

Interestingly, I've just tried a Sigma 530 Super on the K-01. It works in direct and bounce. However, in low light, only the focus assist lamp on the K-01 is used, not the one on the flash. With the K-5, it does use the focus assist lamp on the Sigma flash gun in preference to the one in the camera.

Dan.

Pheo · 01-10-2013, 04:50 PM

Just done some slightly scientific testing here, looks like its an issue on my K-30. However I can cope as long as I figure out what is consistent about it (i.e. if its always a stop, this is fine, I just flash exposure comp to -1.0)

Also, presumably metering on the K-01 is likely on the sensor is it not? As normally the metering is part of the mirror path, but there is no mirror? I may be wrong. Anyway the results from my K-01 seem more consistent.

Andreas_L · 01-11-2013, 07:59 AM

Hello all

I bought a Metz 58 af-2 (V3 firmware) for my new K5II and it behaves no different. My workarround is as some already stated above:

P-ttl at 0 degree and A-mode when tilted. Works perfect!

As a side note:

I dont know if anyone already mentioned it but when I use the 2nd fill flash available on the Metz, in P-ttl when flash is tilted it (maby as expexted due to the already existing problem?) overexposes even more. I use the fill flash at 1/4 of its full power indoors when the flash is tilted and it works great in the flash A-mode (as I understand i very good in the Metz 58AF-2).

Take in consideration that Im NOT a pro but for my demands that solution to use A-mode as a workarround works perfect.

Pheo · 01-11-2013, 08:32 AM

Originally posted by Andreas_L

Hello all

I bought a Metz 58 af-2 (V3 firmware) for my new K5II and it behaves no different. My workarround is as some already stated above:

P-ttl at 0 degree and A-mode when tilted. Works perfect!

As a side note:

I dont know if anyone already mentioned it but when I use the 2nd fill flash available on the Metz, in P-ttl when flash is tilted it (maby as expexted due to the already existing problem?) overexposes even more. I use the fill flash at 1/4 of its full power indoors when the flash is tilted and it works great in the flash A-mode (as I understand i very good in the Metz 58AF-2).

Take in consideration that Im NOT a pro but for my demands that solution to use A-mode as a workarround works perfect.

I was trying out my Metz, and did find that A mode was similar to P-TTL with a -1.0 FEV compensation. Although A mode is not perfect too. I guess until you go manual, its all alogrtihms with the computer working it out, and sometimes it might not get it right.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
K-5: P-TTL External FLASH overexposure	gflauti	Pentax K-5 & K-5 II	32	10-20-2013 03:58 AM
K-x popup flash AND wireless P-TTL with AF-360 FGZ?	runswithsizzers	Flashes, Lighting, and Studio	5	07-29-2011 11:21 PM
For Sale - Sold: Bower SFD926P - Digital Autofocus Flash - Power Zoom P-TTL with swivel and bou	mackloon	Sold Items	2	07-26-2011 08:08 PM
Will Pentax AF-540 FGZ P-TTL Shoe Mount Flash work with the K-x?	justtakingpics	Ask B&H Photo!	6	04-04-2011 02:11 PM
P-TTL overexposure, no more negative flash compensation	Eruditass	Flashes, Lighting, and Studio	11	10-09-2010 05:35 PM

01-05-2013, 12:11 AM	#61
krebsy75 Veteran Member Join Date: Jun 2010 Location: Chester County, Pa. Posts: 804	Bounce exposures are wild with our K-01 as well.

01-05-2013, 12:56 AM	#62
bjdavey Site Supporter Join Date: Mar 2011 Location: Alexandria, VA Photos: Gallery \| Albums Posts: 162	Indeed - my workarounds whenever I need bounce flash are to either dial in -2 stops underexposure, or to use the above mentioned optical wireless mode (which works fine, but has you awkardly holding the camera with one hand and the flash in the other). This confirms that the camera/flash combination is capable of the correct P-TTL metering actions with bounced light. Out of interest I have just held the flash in wireless slave mode immediately adjacent to the hot shoe - so that it is in almost the same position as the tests above where the flash was mounted. I repeated the sequence of direct flash, slight tilt, and full tilt - all shots are properly exposed, so it is definitely not a problem related simply to light direction. As reported by the OP, downward tilt does not cause the problem, even though the rear flash LCD shows the tilt icon activated. However sideways rotation in both directions (which also activates the LCD screen tilt icon), does cause the overexposure problem in the same way as I reported previously (normal exposure when the rotation is only to the point where it has not yet activated the tilt icon, and overexposure when rotated just a tiny bit more, activating the icon). The overexposure occurs in green (auto), P, and Av modes. It seems pretty clear that that the fault is somehow linked to the "signal" which the flash generates to indicate to the camera metering system that the flash is tilted upwards, or rotated sideways (all of the above with the AF540FGZ). That said, I'm not sure why this "signal" is necessary in the first place, as the TTL metering should work regardless of the pathway light travelled before entering the lens.

01-05-2013, 01:30 AM	#63
Noam Forum Member Join Date: Apr 2011 Location: Haifa, Israel Photos: Gallery Posts: 74 Original Poster	Originally posted by bjdavey I'm not sure why this "signal" is necessary in the first place, as the TTL metering should work regardless of the pathway light travelled before entering the lens. Other flash systems (I think Nikon and Sony, maybe Canon) do incorporate a method where the focusing distance (reported by the lens) is used to determine flash power when the flash is pointed straight forward. Compared to pre-flash TTL metering, It can yield a more accurate and consistent exposure that is not affected by non-gray subjects or reflections of shiny objects (which is a problem especially with flash pointed forward). I am guessing that Pentax has a similar method that works in tandem with the PTTL metering. A simple way of testing this hypothesis would be to take several controlled shots on a tripod with the flash pointed forward, and the lens focused at different distances (either manually, or fooling the AF with e.g. your hand and locking AF).

01-05-2013, 01:37 AM	#64
bjdavey Site Supporter Join Date: Mar 2011 Location: Alexandria, VA Photos: Gallery \| Albums Posts: 162	Originally posted by Noam EDIT: More precisely, it depends on the appearance of a distance scale on the flash LCD. When my flash is tilted down (slightly, for macro shots), the tilt icon appears yet the distance scale is still there, and the exposure is the same as straight (i.e., not overexposed). I just read Noam's edit above. Indeed I can confirm using the AF540FGZ that the presence or not of the distance scale on the rear LCD is consistently related to the overexposure, in all tilt or rotation directions. The distance scale disappears when the tilt icon appears, with the exception of downward tilt when the distance scale remains visible. If the distance scale is showing, exposure is correct. When it is not showing, overexposure occurs.

01-05-2013, 01:45 AM	#65
bjdavey Site Supporter Join Date: Mar 2011 Location: Alexandria, VA Photos: Gallery \| Albums Posts: 162	Originally posted by Noam A simple way of testing this hypothesis would be to take several controlled shots on a tripod with the flash pointed forward, and the lens focused at different distances (either manually, or fooling the AF with e.g. your hand and locking AF). Nice suggestion - I just tried it. Using manual focus and direct (non-tilted) flash, there is progressive overexposure as focus distance increases, confirming a link between the camera's calculated focus distance and flash light output.

01-05-2013, 01:57 AM	#66
bjdavey Site Supporter Join Date: Mar 2011 Location: Alexandria, VA Photos: Gallery \| Albums Posts: 162	Originally posted by bjdavey Using manual focus and direct (non-tilted) flash, there is progressive overexposure as focus distance increases, confirming a link between the camera's calculated focus distance and flash light output. ... and further to the above, that only happens if the distance scale is showing on the flash LCD. If the flash is tilted or rotated to the point where the distance scale disappears, overexposure is consistent throughout the focus range.

01-09-2013, 08:59 AM	#67
tzeentchiangod Junior Member Join Date: Feb 2012 Photos: Gallery Posts: 34	Forgive me if this was mentioned up-thread... but has anybody had a conversation with someone at Pentax about this?

01-09-2013, 11:35 AM	#68
Canada_Rockies Loyal Site Supporter Join Date: Nov 2007 Location: Sparwood, BC, Canada Photos: Gallery \| Albums Posts: 12,385	Originally posted by dosdan My pet hate when using bounce flash at an angle of 60°, is when the top half of a person ends up brighter than the bottom. Whick is OK for bust shots, but a problem with full-length shots. What do you guys do to reduce this? Set the flash to straight up? Flash towards the rear ceiling if the rear wall is close enough and light coloured? Dan. Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post. I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading.

01-09-2013, 10:17 PM	#69
tabl10s Veteran Member Join Date: Jun 2011 Location: Sacramento(formerly from B'Ham, England). Posts: 1,424	Originally posted by Canada_Rockies Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post. I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading. I'm gonna try that.

01-10-2013, 09:02 AM	#70
fs999 Loyal Site Supporter Join Date: Jul 2008 Location: Luxembourg Photos: Gallery Posts: 8,638	It would be interresting to see if the new AF360 FGZ II has the same problem...

01-10-2013, 12:05 PM	#71
arnie0674 Senior Member Join Date: Jan 2012 Photos: Albums Posts: 173	Originally posted by Canada_Rockies Whenever possible, I bounce the flash into a wall/ceiling three way corner behind me. It's like using a large umbrella and evens out the exposure a lot. It can cause "interesting" problems with white balance, depending on the colours of the walls and ceiling. It also assumes your flash can be tilted and rotated to flash behind you and up. Different brightness of one wall vs the other vs the ceiling can also cause the shadows to come out different, and that is hard to fix in post. I use a K10 and AF540 and have had a flash exposure that I could not handle in post processing. It isn't perfect by any means, but it is within a stop. I've never found flash to be a perfect exposure medium except when using my manual studio flashes in umbrellas and setting the camera based on a flash meter reading. The K10d does not have this bug, so I would expect it to expose correctly

01-10-2013, 03:01 PM	#72
dosdan Pentaxian Join Date: Dec 2007 Location: Brisbane, Australia Photos: Gallery \| Albums Posts: 1,741	Interestingly, I've just tried a Sigma 530 Super on the K-01. It works in direct and bounce. However, in low light, only the focus assist lamp on the K-01 is used, not the one on the flash. With the K-5, it does use the focus assist lamp on the Sigma flash gun in preference to the one in the camera. Dan.

01-10-2013, 04:50 PM	#73
Pheo Veteran Member Join Date: Nov 2006 Location: Brighton, United Kingdom Photos: Gallery Posts: 311	Just done some slightly scientific testing here, looks like its an issue on my K-30. However I can cope as long as I figure out what is consistent about it (i.e. if its always a stop, this is fine, I just flash exposure comp to -1.0) Also, presumably metering on the K-01 is likely on the sensor is it not? As normally the metering is part of the mirror path, but there is no mirror? I may be wrong. Anyway the results from my K-01 seem more consistent.

01-11-2013, 07:59 AM	#74
Andreas_L Inactive Account Join Date: Jan 2013 Posts: 1	Hello all I bought a Metz 58 af-2 (V3 firmware) for my new K5II and it behaves no different. My workarround is as some already stated above: P-ttl at 0 degree and A-mode when tilted. Works perfect! As a side note: I dont know if anyone already mentioned it but when I use the 2nd fill flash available on the Metz, in P-ttl when flash is tilted it (maby as expexted due to the already existing problem?) overexposes even more. I use the fill flash at 1/4 of its full power indoors when the flash is tilted and it works great in the flash A-mode (as I understand i very good in the Metz 58AF-2). Take in consideration that Im NOT a pro but for my demands that solution to use A-mode as a workarround works perfect.

01-11-2013, 08:32 AM	#75
Pheo Veteran Member Join Date: Nov 2006 Location: Brighton, United Kingdom Photos: Gallery Posts: 311	Originally posted by Andreas_L Hello all I bought a Metz 58 af-2 (V3 firmware) for my new K5II and it behaves no different. My workarround is as some already stated above: P-ttl at 0 degree and A-mode when tilted. Works perfect! As a side note: I dont know if anyone already mentioned it but when I use the 2nd fill flash available on the Metz, in P-ttl when flash is tilted it (maby as expexted due to the already existing problem?) overexposes even more. I use the fill flash at 1/4 of its full power indoors when the flash is tilted and it works great in the flash A-mode (as I understand i very good in the Metz 58AF-2). Take in consideration that Im NOT a pro but for my demands that solution to use A-mode as a workarround works perfect. I was trying out my Metz, and did find that A mode was similar to P-TTL with a -1.0 FEV compensation. Although A mode is not perfect too. I guess until you go manual, its all alogrtihms with the computer working it out, and sometimes it might not get it right.