wrong color temperature Godox ? Do I do something wrong

[xpatUSA]

No, the exif says wb manual. Probably he used a value of 5500 for the manual wb.

There is no "probably" about it! 5500 is clearly stated in his JPEG's 'Maker Notes' which are part of the EXIF specification:

"MakerNotes: A tag for manufacturers of Exif writers to record any desired information. The contents are up to the manufacturer, but this tag should not be used for any other than its intended purpose."

Some people seem to think that "EXIF" is the few bits of data that a Viewer might list but there are literally hundreds of defined tags which may or may not be used by Camera Manufacturer.

The main point is that when I shoot a gray card under normal exposure, I expect a histogram with the three channels being equal and somewhere in the middle. That would be my start point.

As was already stated as being expected by Ruud, so why is that expectation being repeated?!

A further investigation might deal with the color temp of the used flash.

How would that be done? The manufacturer's spec. has already been investigated and stated!

As you can see in DxO the color temp is 5154 and tint -7.

In the raw or the JPEG?

If I change the wb setting to flash, I get a gray image, the three channels are covering each other nearly. If I go higher the covering of the channels is getting better.

That's just DxO ... other converters/editors will vary.

His manual wb setting is wrong.

LOL

His manual WB setting is absolutely correct for his Godox flash and, furthermore, the physics of Xenon gas hasn't suddenly changed, as far as I know.

[rlast]

That could bring us to the question of flash power, about which I know very little.

Your flash has a guide number of 60 meaning 60 meters at 100 ISO:

http://www.godox.com/EN/Products_Cam...sh_V860II.html

What I don't know is which power (out of the 22 steps from 1/128 to 1/1) should be set or occur automatically when shooting at 0.73m. Others here should know.

I was shooting in TTL mode, which has proven itself to work before. Next time I can go to manual mode - and write down the power. I guess it was quite low, as my flash was only 1.5 meter away from the subject

I can assure you that neither the raw nor the JPEG contains an embedded ICC color profile of any kind, which means that ...

I see what you mean! C1 does not show the (intended) color profile I selected in the camera.
I've just checked my color space on my camera again, and it is sRGB.
When I inspect the RAW file attached with RAWdigger, it also reads sRGB, so I guess that is oke.
Capture one shows sRGB when I select the JPG, that I shared before. That is different from what you see.
If needed I can attach a screenshot.

... I don't know what is meant by that, sorry.

That is a color profile I can select in C1 one. It enhances the color and contrast. I've disabled it, for this flash - analysis

My gut feeling is for an error in how the flash itself was used, rather than there being a problem with the flash.

As to the profile, if no profile is found, viewers assume that the file image content is sRGB which I assume is what your Sony is set to ...

yes
[QUOTE]


cheers, Rudolf

[Stagecoach]

Rudolf,

Do you get the same result when using the cameras 'pop up flash' as when using the Godox?

You may have tried this already, I'm just thinking 'elimination'.

Edit:

Just another thought to throw in, I notice the higher shutter speeds (and earlier explanation) and wonder if there is a possibility that the Godox is somehow working in an HSS mode due to this. Short bursts of lower powered flash could this affect colour temp from the flash?

[xpatUSA]

Pure speculation but I have just read "They [xenon lamps] are an excellent source of short wavelength ultraviolet radiation".

So if the subject gray card's surface can be excited by that UV then maybe it emits some blue as a result. Quite unlikely, but we do have "whiter than white" paper with OBPs in it for example ...

[rlast]

Rudolf,

Do you get the same result when using the cameras 'pop up flash' as when using the Godox?

You may have tried this already, I'm just thinking 'elimination'.

Stagecoach, I haven't tried this, so I will check this tomorrow!

Edit:

Just another thought to throw in, I notice the higher shutter speeds (and earlier explanation) and wonder if there is a possibility that the Godox is somehow working in an HSS mode due to this. Short bursts of lower powered flash could this affect colour temp from the flash?

I have tried different speeds from 1/200 (flash specified to 1/300sec and shorter) to 1/1000, and haven't noticed any shift in color.

Thanks!, I get back with the results from the internal flash.

[george013]

There is no "probably" about it! 5500 is clearly stated in his JPEG's 'Maker Notes' which are part of the EXIF specification:

"MakerNotes: A tag for manufacturers of Exif writers to record any desired information. The contents are up to the manufacturer, but this tag should not be used for any other than its intended purpose."

Some people seem to think that "EXIF" is the few bits of data that a Viewer might list but there are literally hundreds of defined tags which may or may not be used by Camera Manufacturer.



As was already stated as being expected by Ruud, so why is that expectation being repeated?!



How would that be done? The manufacturer's spec. has already been investigated and stated!



In the raw or the JPEG?



That's just DxO ... other converters/editors will vary.



LOL

His manual WB setting is absolutely correct for his Godox flash and, furthermore, the physics of Xenon gas hasn't suddenly changed, as far as I know.

I thought this was past.

The color temp is either given or calculated. If the exif says manual, than it's given. So if Ruud fills in the value 5500 then it's 5500. But if I change that setting in the converter to flash, another predefined value, than I get a rather good histogram. So did Ruud by using the gray point method.
And so you can do it with RT.

I've been thinking what you said about the distance. Might be an issue.

Ruud can also try to measure the color temp of his flash. Check double check.
A trick I once read was to cover the lens with a white tissue, aim the flash right in the lens and make a preset of the wb. It was said to have the same accuracy as metering from your hand or grass. One can work with it.

George

[rlast]

Pure speculation but I have just read "They [xenon lamps] are an excellent source of short wavelength ultraviolet radiation".

So if the subject gray card's surface can be excited by that UV then maybe it emits some blue as a result. Quite unlikely, but we do have "whiter than white" paper with OBPs in it for example ...

hi Ted, not sure if I understand you, so just to check:
Every portable flash (at least the popular / reasonabe price) uses a Xenon lamp? So my camera's popup flash and Godox are likely of the same type and radiate the same spectrum (so need same WB setting?)

The UV / greycard combination - then everybody should have this problem, isn't it?

Thanks for thinking with me …. :-)

[xpatUSA]

hi Ted, not sure if I understand you, so just to check:
Every portable flash (at least the popular / reasonable price) uses a Xenon lamp? So my camera's popup flash and Godox are likely of the same type and radiate the same spectrum (so need same WB setting?)

Agreed, in theory. We need more input from flash experts, though.

The UV / greycard combination - then everybody should have this problem, isn't it?

Pure speculation on my part. Of course your logic is impeccable, so I withdraw the remark.

Found your card's technical data including spectral reflectance curves - they look normal enough for your purpose ...

http://www.mennon-usa.com/documents/mennongreycard.pdf

In there, it recommends lighting the card at an angle, not straight on.

Remember too that a flash unit puts out a lot of near infra-red. Some sensor CFAs have some IR response, for what that is worth.

Hope this helps ...

[Manfred M]

Everyone seems to be assuming that flash output is a constant colour temperature. That is not necessarily a good assumption.

[xpatUSA]

Everyone seems to be assuming that flash output is a constant colour temperature. That is not necessarily a good assumption.

Good find, Manfred. I was looking for the same or similar information, but with less success. Thanks for posting.

[Stagecoach]

Everyone seems to be assuming that flash output is a constant colour temperature. That is not necessarily a good assumption.

From some practical experience and experiments with my macro work some time ago I believe I came across this.

I often shoot my subjects on what to my eyes is a perfectly grey rough round large pebble. If when using flash I leave the camera WB set to 'Daylight' as it's left on for 98% of my shooting there is a very distinct change in colour of that stone depending on flash power used. It comes out very bluish when flash power is very low.

If on the other hand I set the camera WB to 'Flash' the difference towards 'blue' is far less significant with the low powers.

[dje]

Hi Ruud

I had a look at your raw file in all the raw converters I have that would open the file (ie support your camera ). The As Shot temps were all around 5200-5500 but these are clearly wrong as the image is nowhere near white balanced with these settings. Using the white balance picker, which assumes your card is true gray and determines multipliers that equalise the RGB values, I found the following (average values)

Capture One 7250
Affinity Photo 6700
ACDSee 6700
Darktable 6700

The color temps displayed by the software are estimated from the WB multipliers mentioned above. It seems that C!'s estimate differs considerably from that of other software I tried. Personally I don't think you can rely on the Godox spec for your flash col temp either.

Out of interest, I shot a gray card with my Sony a6000 and a Godox TT350 flash. For this, using the wb picker, I found actual col temps of about 6000 but once again, C1 was reading higher at about 6500. The As Shot temps weren't that much different to those determined with the wb picker (about 200 higher). Also, putting the flash in HSS mode caused the temps to drop by about 600-800.

Incidentally most camera color profiles are designed for white balanced data so if your wb is out, the profile may not work optimally.

Dave

[george013]

Hi Ruud

I had a look at your raw file in all the raw converters I have that would open the file (ie support your camera ). The As Shot temps were all around 5200-5500 but these are clearly wrong as the image is nowhere near white balanced with these settings. Using the white balance picker, which assumes your card is true gray and determines multipliers that equalise the RGB values, I found the following (average values)

Capture One 7250
Affinity Photo 6700
ACDSee 6700
Darktable 6700

The color temps displayed by the software are estimated from the WB multipliers mentioned above. It seems that C!'s estimate differs considerably from that of other software I tried. Personally I don't think you can rely on the Godox spec for your flash col temp.

Out of interest, I shot a gray card with my Sony a6000 and a Godox TT350 flash. For this, using the wb picker, I found actual col temps of about 6000 but once again, C1 was reading higher at about 6500. The As Shot temps weren't that much different to those determined with the wb picker (about 200 higher). Also, putting the flash in HSS mode caused the temps to drop by about 600-800.

Incidentally most camera color profiles are designed for white balanced data so if your wb is out, the profile may not work optimally.

Dave

DxO 7355 with the gray picker somewhere in the middle. But with a difference of a little more as 7000 to as high as 7700 depending on where you click.
I don't think he used HSS mode for this camera, post 5.

George

[Stagecoach]

Also, putting the flash in HSS mode caused the temps to drop by about 600-800.Dave

That's good to hear Dave as it confirms the colour/power theory.

[rlast]

Everyone seems to be assuming that flash output is a constant colour temperature. That is not necessarily a good assumption.

Thanks Manfred for bringing this up. Very interesting.
Based on this and other suggestions, I did a series, with the flash in manual mode and varying the power (1/1 to 1/16), compensated with f/stop (f16 to f4) to keep the exposure constant.
Furthermore I put the greycard in an angle (as suggested), and put the WB on 5500K.

A picture of the 5 exposures:


When I use the whitebalance picker in C1, the temperature decreases with power going up, going from 7360 for 1/16 to 7090 for 1/1.
SO THE EFFECT IS THERE BUT NOT AS STRONG AS FROM THE TABLE

I did a comparison with my pop up flash, don't know at which power but I think it is the max available (shooting at ISO100, with f4).
Shooting the greycard with the popflash, also results in the same blue-ish picture, which needs a WB setting of 7360
to become grey. (result from the WB picker / histogram)

SO POPUP FLASH and GODOX give similar results!

Last test I conducted is to use a different WB settings for the f/4, 1/16th power combination:
WB FLASH
WB AUTO
WB Manual 7400K,

See picture below:



Gaining some understanding, but not there yet …. :-)


Rudolf

[rlast]

Hi Ruud

I had a look at your raw file in all the raw converters I have that would open the file (ie support your camera ). The As Shot temps were all around 5200-5500 but these are clearly wrong as the image is nowhere near white balanced with these settings. Using the white balance picker, which assumes your card is true gray and determines multipliers that equalise the RGB values, I found the following (average values)

Capture One 7250
Affinity Photo 6700
ACDSee 6700
Darktable 6700

The color temps displayed by the software are estimated from the WB multipliers mentioned above. It seems that C!'s estimate differs considerably from that of other software I tried. Personally I don't think you can rely on the Godox spec for your flash col temp either.

Out of interest, I shot a gray card with my Sony a6000 and a Godox TT350 flash. For this, using the wb picker, I found actual col temps of about 6000 but once again, C1 was reading higher at about 6500. The As Shot temps weren't that much different to those determined with the wb picker (about 200 higher). Also, putting the flash in HSS mode caused the temps to drop by about 600-800.

Incidentally most camera color profiles are designed for white balanced data so if your wb is out, the profile may not work optimally.

Dave

Dave, thanks a lot for conducting your experiment which tells me C1 is doing something different than the others, and that a greycard - flash combination does not result in 5500K either (as indicated by Manfred before, that it could be far off)

To be sure we use the same settings: to what setting did you put your curve under base Characteristics?
I put mine on linear Response for this experiment, to minimize the processing C1 is doing!

Thanks ! Rudolf

[rlast]

From some practical experience and experiments with my macro work some time ago I believe I came across this.

I often shoot my subjects on what to my eyes is a perfectly grey rough round large pebble. If when using flash I leave the camera WB set to 'Daylight' as it's left on for 98% of my shooting there is a very distinct change in colour of that stone depending on flash power used. It comes out very bluish when flash power is very low.

If on the other hand I set the camera WB to 'Flash' the difference towards 'blue' is far less significant with the low powers.

Stagecoach, thanks for sharing, and in my latest experiment (today) confirmes your method:
I also got better results putting the flash @ WB -FLASH mode.

Rudolf

[rlast]

Agreed, in theory. We need more input from flash experts, though.



Pure speculation on my part. Of course your logic is impeccable, so I withdraw the remark.

Found your card's technical data including spectral reflectance curves - they look normal enough for your purpose ...

http://www.mennon-usa.com/documents/mennongreycard.pdf

In there, it recommends lighting the card at an angle, not straight on.

Remember too that a flash unit puts out a lot of near infra-red. Some sensor CFAs have some IR response, for what that is worth.

Hope this helps ...

Thanks… if you have time and interest, please see my latest results from today, shot with the greycard in a 30% angle …

cheers, Rudolf

[rlast]

I thought this was past.

The color temp is either given or calculated. If the exif says manual, than it's given. So if Ruud fills in the value 5500 then it's 5500. But if I change that setting in the converter to flash, another predefined value, than I get a rather good histogram. So did Ruud by using the gray point method.
And so you can do it with RT.

I've been thinking what you said about the distance. Might be an issue.

Ruud can also try to measure the color temp of his flash. Check double check.
A trick I once read was to cover the lens with a white tissue, aim the flash right in the lens and make a preset of the wb. It was said to have the same accuracy as metering from your hand or grass. One can work with it.

George

George, I'm interested to do my experiment with a surface, other than my greycard. Though it has to be a reliable / defined setting for me to make sense.
I once red, that (white) teflon tape is a neutral white. Have you or anyone experience with that?

Regards, Rudolf

[dje]

Dave, thanks a lot for conducting your experiment which tells me C1 is doing something different than the others, and that a greycard - flash combination does not result in 5500K either (as indicated by Manfred before, that it could be far off)

To be sure we use the same settings: to what setting did you put your curve under base Characteristics?
I put mine on linear Response for this experiment, to minimize the processing C1 is doing!

Thanks ! Rudolf

Rudolf the curve was set to Auto but fiddling with the options just now I found it made no significant difference which curve was selected (nor should it).

Dave