On 1/15/2025 at 9:27 AM, Adventurer said:

Don‘t get me wrong (Camera/RAW)Profiles (Portrait,Landscape,Neutral,Standard, Adobe Landscape etc) + ColorProfiles (AdobeRGB, sRGB) are to a large amount available in both, camera and raw converter. So with your phrase above I now struggle to understand if you refer to ColorProfiles or (Camera/RAW)Profiles.

 

The fact that (Camera/RAW)Profiles also heavily influence color balance does not make it easier for everyone to communicate about this 🤭🤣.

 

I have never used Canon, but with my Nikons I can change Picture Control settings (Standard, Landscape, Portrait etc) and the color profile (RGB, sRGB). Both settings will only affect the jpg files, but not the RAW files. I leave my Picture Control to "Standard".

 

In Lightroom you can choose a camera profile (also Landscape, Portrait etc) which I also leave to "Standard".

 

For export jpgs for web browsers, I export in sRGB, which will affect the colors and can alter the "blue". That's why I stated that in RGB and viewed on a good monitor (wide color space) the blues look usually better in my pictures than in sRGB.

On 1/15/2025 at 9:27 AM, Adventurer said:

@ChrisH let’s come back to warm strobes or gels topic. Mr. Ross pointed out that sometimes he might not be close enough with his Z240s. This is an important factor. I see with you that you are mainly mastering the fisheye lense which usually gets you ultra close! In this case I would subscribe to your algorithm. The warm color temp of the strobes will definitely hit the foreground and also use the warm spectrum.

 

But now let’s imagine you have to be more far away and will use a more tele end lens, such as 17-55 or 14-35. You want to shoot it on the zoomed in end of wide angle underwater photography. This is not uncommon when shooting portraits of mid sized pelagics such as sharks or dolphins or shy reef sharks with strobes. There the warm color in the foreground will be quickly absorbed and have no use, because you are simply not close enough.

 

Same applies when you do not have sufficient stuff to fill the frame in the foreground.

 

In that scenario I would opt in on the cooler strobes and try to avoid diffusers, as I want the maximum energy of the strobes to penetrate the water column as far as possible. Cool color temp of the strobes spectrum will travel more far than the reddish parts emitted or filtered.

 

There is a second thread going on right now where that concept is discussed and exaggerated by even adding cooling filters to video lights or strobes. This is also the reason why Backscatter offers blue Diffusors for there HF-1 strobe.

 

However if you want maximum energy penetration through the water column then cool native color temp strobe, without any diffusors or filters is the way to go.

 

Yes, the effect of warm strobes will be not as pronounced when shooting from a greater distance. But for underwater you will usually want to avoid shooting from a greater distance and instead try to get as close as possible.

 

I have a 16-35mm lens, but I used it only a couple of times and with not much success. So I don't really get it, why I should not use a Fisheye, unless I need a lens without the fisheye distortion. If you need more flexibility, a fisheye zoom used with a TC will get you there.

 

Also, Retra makes a diffusor that they call "Shark diffusor", so it is for pelagics. On their website it says: 

"The Shark diffuser shifts the color temperature by -1000K and reduces the overall light output by -0.8 F-stop. 
The Shark diffuser is designed to prevent overexposing shark bellies and to give an especially warm light output for a more vivid blue background. The Shark diffuser is suitable also to other large animals and schools of fish."

So Retra also thinks it is a good idea to use warm color light, even if you can't get as close as with reef scenes.

 

I have attached some pictures, that show how warm color temperature strobes render colors with sharks and that it is absolutely possible to get close enough for fisheye shots.

As I wanted to show the "out of cam" colors, I have attached pictures that I didn't process because they are nothing special. Only some adjustments to highlights, shadows and some clarity were made now. Colors are as you get them with auto white balance, warm strobes, fisheye. Nothing fancy, simple as that. Note the colors on the 3rd picture, shot with the same equipment on the same dive, nothing done to colors, not even saturation added. Just as it is.

 

So that brings me back to the question: why would I not want warm color strobes? Maybe for shooting in greenish water, yes. But in warm ocean water? Why use cold light temperature? For video I get it. But for pictures? Why not go for the strobe with warm color temperature?

 

 

 

 

7 hours ago, Adventurer said:

When a photo is underexposed, it alters how colors are rendered in the image. The effect of greenish water appearing blue in an underexposed photo can be explained by several factors:

 

1. Camera White Balance

 

Cameras adjust white balance automatically or manually to ensure colors appear correct. In underexposed conditions, the camera’s white balance often shifts toward cooler tones (blues), as green and yellow tones reflect less light in dark conditions, making blue tones more dominant.

 

2. Spectral Light Absorption in Water

 

Water absorbs red light more strongly than blue light. This natural property makes water appear bluish, especially in low light (e.g., during underexposure). When the image is darkened, the green component of the water reflects less light, allowing the blue hues to stand out more.

 

3. Perception and Color Saturation

 

Underexposed photos tend to increase the perception of color saturation and contrast. This can enhance any slight bluish tint in the water while muting green or yellow tones.

 

4. Camera Color Profiles

 

Digital cameras have specific sensitivities in their sensors. These sensors often respond more strongly to blue light under low-light conditions, which can make greenish water appear more blue when underexposed.

 

In summary: Due to the physical properties of water, the camera’s processing (e.g., white balance), and sensor behavior, greenish water often appears bluer in underexposed images.

While I don't doubt you see this effect when taking your photos, I don't see it myself and this is likely down to perception.  On these points :

 

1.  yes the WB may vary my experience is that is varies only minimally on autoWB with my cameras.  Adjusting the WB in my RAW converter to compare images taken at various shutter speeds within the range that it varies WB, the adjustments are so small that I can see no change in the WB.  The WB may also generally be influenced by the subject and the light received from the flash.  If the subject is large in the frame this may well dominate and be the larger impact.

Different camera brands may respond to this differently of course but my OM-1 seems quiet consistent with similar subjects.

 

2. While the spectral absorption is correct, the camera knows nothing of this and only records the photons it receives.  If you shorten the exposure the blue light is reduced by the same extent that the green and indeed the red light is reduced.  All that happens is that the is the total luminosity is reduced.  The only way for the colour tone to change is to change the WB and We've established that doesn't change much with exposure.  There's no room for argument on this the RAW file has the same red/green/blue ratios as you reduce the exposure, just less total photons.

3.  Perception of colour can indeed change - I don't see any change in colour tone with exposure.  You might see something and I won't argue with that.  It's purely subjective.

4.Using camera profiles with RAW processing only provides a preset or a starting point for processing.  If you find one you like it certainly saves processing time.  These are proprietary to each camera brand and anything discerned from them applies purely to this camera brand.  They could well be non linear and do weird stuff that is not expected, but again if this is advantage you see it only applies to your camera model and possibly others of the same vintage using that same algorithm to work out how to adjust the preset for your image.  The underlying RAW image remains the same.

 

 

4 hours ago, ChrisH said:

 

 

As I wanted to show the "out of cam" colors, I have attached pictures that I didn't process because they are nothing special. Only some adjustments to highlights, shadows and some clarity were made now. Colors are as you get them with auto white balance, warm strobes, fisheye. Nothing fancy, simple as that. Note the colors on the 3rd picture, shot with the same equipment on the same dive, nothing done to colors, not even saturation added. Just as it is.

 

So that brings me back to the question: why would I not want warm color strobes? Maybe for shooting in greenish water, yes. But in warm ocean water? Why use cold light temperature? For video I get it. But for pictures? Why not go for the strobe with warm color temperature?

 

 

Nice pics Chris, I agree with your argument - Retra understands what is going on with lighting UW.  Yes as distance increases the impact of strobe colour decreases somewhat- the solution ----  Shift the light further to the red and that way you still get the enhancement from using a warm strobe.

 

While it is true that red light doesn't travel as far as blue light , there is still a lot of blue light in the strobe light and overall the loss of light is not that great and certainly not great enough to need to change to using a strobe with a cooler colour balance.

 

Of course this only applies to clean blue water, temperate waters you generally don't see the benefits.

 

The use of cooling filters for video is a different issue all together - they are trying avoid the issue where the WB varies dramatically with distance with video lighting, the effect is where are subject swims close and looks very red.

 

3 hours ago, Chris Ross said:

If you shorten the exposure the blue light is reduced by the same extent that the green and indeed the red light is reduced.

Sorry Chris, but this is wrong and I need to correct you for physical facts.

 

Water absorbs light in a wavelength-specific manner:

• Blue (~450–495 nm): Absorbed the least and penetrates the deepest.

• Green (~495–570 nm): Absorbed more strongly than blue but less than red.

• Red (~620–750 nm): Absorbed almost completely within a few meters.

 

Absorption Coefficients

• For blue light (~475 nm): Absorption coefficient ≈ 0.015 m⁻¹.

• For green light (~525 nm): Absorption coefficient ≈ 0.04 m⁻¹.

 

This means that green light is absorbed at approximately 2.67 times the rate of blue light in pure water.
 

While the speed of light in water is nearly identical for both wavelengths, the absorption rates dictate how much light remains visible, making green “fall out” of underwater photos more quickly than blue.
 

You can utilize this effect by underexposing. The green part of the histogram will overproportionally be underexposed if you lower your exposure by one stop.

 

Sidenote: underexposing slightly for decent blues underwater is a technique coming from the film days where underwater photographers had no control over white balance in a non-digital camera. It‘s a proven relict.

 

However if you do not meter and expose in a proper way you may be unable master this technique with your camera. Unfortunately it blows the scope of an Internet forum to teach this here and the thread is on it‘s way to go off-topic.

Edited January 17Jan 17 by Adventurer

Double needs to be deleted

Edited January 17Jan 17 by Adventurer
Duplicate reply

On 1/9/2025 at 6:57 PM, Adventurer said:

In most cases this comes with the price of sacrificing strobe power as the warming gels and diffusers will absorb photon energy.

Hi @ChrisH I totally see your point and in your specific configuration and setup it all makes perfect sense. You are very strong fisheye user, hence your warm strobes (ex-factory warm) play out perfectly. Next influencing factor: the Nikon blues in AWB got you covered. Also a (minor) last factor: Your SiGMA Fisheye, instead of a Nikon or Canon FE. I previously owned this and remember that when even shooting this topside this lens would render the shots a little more golden than a Canon or Tokina FE.

 

Anyway, just change your perspective to the average Joe who does it different and bought one of the cooler strobes out there, internationally sold in multiple quantities. Even Nikons, Sony and Canons System strobes are manufactured with quite cool color temperatures.

 

So let’s look at my initial statement quoted again above. Does it always make sense to „rape“ your strobes energy output with warming diffusors ?

 

I would like to hear more different opinions from others. You and @Chris Ross have made your point.

Hi Adventurer,

 

When you are referring to different wavelengths of light being absorbed at different rates it is obviously with distance travelled through water, nothing significantly to do with time.... as you say "speed of light in water is nearly identical for both wavelengths". So I can't understand how exposure time (shutter speed) could possibly make any difference to the relative proportion of different wavelengths of light reaching the sensor (i.e. colour).

 

I use shutter speed to vary the deepness of background colour from pale blue to dark blue. I use Inon strobes and a Sony camera on auto white balance with white balance adjustment in Ligfhtroom. If I didn't use the warming diffusers it would be the same change in deepness of background colour but with slightly more green aspect.

uses

4 hours ago, Adventurer said:

I would like to hear more different opinions from others. You and @Chris Ross have made your point.

 

Since you ask I can tell you that it is not my opinion, but a physical fact, that it is impossible to influence the spectral properties of ambient light by adjusting the shutter speed of a camera that is used to take a photo. By changing the shutter speed the amount of ambient light that is collected varies, but its spectral properties remain what they are...

 

I believe that confusion in this discussion arose, because the effect of automatically occuring WB, color profiles and other processing tools are mixed with the effect that shutter speed has on spectral properties of the data collected on the sensor (=raw file).

 

#1.: The crucial point is the discussion of the effect of shutter speed on the original raw data collected by the camera. Regarding the raw file, this file does not even contain any color information, just intensity is encoded, from black to white in different shades of grey (either 12 bit (=4096 different shades of grey), 14 bit (=16384) or even 16 bit (=65536 different shades)). The only color information that is contained in the raw file is the type of Bayer array of the sensor (https://en.wikipedia.org/wiki/Bayer_filter). Color information is introduced at a later stage, when the file is opened and the type of Bayer array is taken into account to process the raw data and produce a colorful picture. As Chris Ross rightfully states, shutter speed influences intensity values on all the different pixels on the sensor the same way, no "if and but" is possible at this stage ...

 

#2.: The other point is how processing of these raw data leads to different colors (=adjusting WB). Usually people adjust temperature and tint of their UW photos in post via software, e.g. LR. To some extend, such processing is already done automatically by the camera/processing software (Auto WB/ color profile) and this is how the photo appears on the screen, as a first suggestion for further adjustments by the photographer. The way of automatic adjustments varies from camera model to model and, beyond that, can be finetuned in the camera menue - hence an almost infinite number of possibilities exists and the outcome can be different from model to model and from photographer to photographer. I would say this outcome is, more or less, unpredictable. Here comes ChisH into the play, when he says rightfully that in his hands camera X and flash Y give best results already as they work by themselves and little adjustment of colors is needed in postprocessing (under these conditions shutter speed may even influence temperature/tint of the "original" image on the screen, depending on camera model and also how the photographer tunes the automatic features of the camera in the menue)...

 

=> As the bottomline I would say everybody is right in his own way, but please nobody confuses spectral properties of water with, in practice unpredictable, automatic processing of raw data (spectral properties of raw data are constant and invariable by shutter speed)...

 

 

Wolfgang

Edited January 17Jan 17 by Architeuthis

19 hours ago, Adventurer said:

Sorry Chris, but this is wrong and I need to correct you for physical facts.

 

Water absorbs light in a wavelength-specific manner:

• Blue (~450–495 nm): Absorbed the least and penetrates the deepest.

• Green (~495–570 nm): Absorbed more strongly than blue but less than red.

• Red (~620–750 nm): Absorbed almost completely within a few meters.

 

Absorption Coefficients

• For blue light (~475 nm): Absorption coefficient ≈ 0.015 m⁻¹.

• For green light (~525 nm): Absorption coefficient ≈ 0.04 m⁻¹.

 

This means that green light is absorbed at approximately 2.67 times the rate of blue light in pure water.
 

While the speed of light in water is nearly identical for both wavelengths, the absorption rates dictate how much light remains visible, making green “fall out” of underwater photos more quickly than blue.
 

You can utilize this effect by underexposing. The green part of the histogram will overproportionally be underexposed if you lower your exposure by one stop.

 

Sidenote: underexposing slightly for decent blues underwater is a technique coming from the film days where underwater photographers had no control over white balance in a non-digital camera. It‘s a proven relict.

 

However if you do not meter and expose in a proper way you may be unable master this technique with your camera. Unfortunately it blows the scope of an Internet forum to teach this here and the thread is on it‘s way to go off-topic.

The first part is correct light is absorbed by wavelength unevenly, that is well established.  Secondly speed of light in water is identical for all wavelengths, however this has no impact on the current subject.  

 

In any photography above or below water you have a continuous flux of photons from a continuous light source such as the sun, by this I mean photons/square meter/second as long as the light source remains constant - which the sun does as long as a cloud doesn't move across it.  To get the number of photons multipy flux by time: photons/sqm/sec x sec = photons/sqm.  The absorption doesn't come into it except to set the ratio of red to green to blue photons that exists due to the distance the light has travelled from the surface to the subject, through the lens into your camera and falling on the sensor.

 

Changing the shutter speed doesn't change the the ratio of red/green/blue photons, just the total number of photons that arrive.  This is not opinion, it is indisputable fact.  It is important to get this right to avoid confusion when discussing these topics.

 

What the camera data does with this set of photons it has recorded is a different issue and is buried in the proprietary algorithms used to take the sensor data recorded and calculate a RGB value for each and every pixel.  Any such changes are global, though WB adjustments could potentially be non linear treating shadows WB slightly different to highlights WB.  Any thing that is due to the camera processing is specific to that brand and potentially even the hardware/firmware of that particular model.

23 hours ago, Architeuthis said:

Regarding the raw file, this file does not even contain any color information, just intensity is encoded, from black to white

Yes, but for three channels! 🤦‍♂️ R G B

 

RED GREEN BLUE

 

Your long text is highly misleading for the sake of making an argument.

 

23 hours ago, Architeuthis said:

Chris Ross rightfully states, shutter speed influences intensity values on all the different pixels on the sensor the same way, no

Wrong. Because of cited absorption coefficients in water and distance of blue/green background water column.

 

Just try it , and talk to a knowledgeable underwater photographer who still shot chemical film, if you know one, for confirmation.

23 hours ago, Adventurer said:

Yes, but for three channels! 🤦‍♂️ R G B

 

RED GREEN BLUE

 

Your long text is highly misleading for the sake of making an argument.

 

Wrong. Because of cited absorption coefficients in water and distance of blue/green background water column.

 

Just try it , and talk to a knowledgeable underwater photographer who still shot chemical film, if you know one, for confirmation.

No, Wolfgang is in fact correct, all colours are reduced proportionally when you use a faster shutter speed.  Other effects like picture style might change things but the RAW files are the same colour shade just less bright in all channels for the shorter exposure.

 

I dug through my files to find some examples I've shot, first is a basically a snapshot taken on a safety stop one at 1/80 and the next at 1/200.  First the uncorrected file, only change is to reduce highlights to stop clipping and the auto WB equalised for both frames - it was a very small adjustment.  Next the images are equalised in exposure . 

The images are in Adobe RGB to avoid having to try to recover the blues, they need to be viewed in a colour managed browser or other colour managed application.  If the blues look flat, try downloading the file and opening in the browser directly of PS or LR.

 

unprocessed 1/80 and 1/200 exposures

 

Exposures equalised

 

Next is a WWII Zero fighter quite close to shore and water is a bit greener overall.  again side by side comparison with exposures at 1/160 and 1/100, unprocessed and exposures equalised.

 

Unprocessed 1/160 and 1/100 shutter speed

 

Exposures equalised

 

To my eye the blues are very close in shade and I haven't changed the green noticeably by under exposing.  The blues are of course not the finished product and I can clean up the bright water in the first set some.  Equalising the exposures is done by using an exposure layer and is equivalent to changing exposure in camera.  and takes away inclination for the eye to see shade differences that aren't real. 

On 1/19/2025 at 7:55 AM, Chris Ross said:

No, Wolfgang is in fact correct, all colours are reduced proportionally when you use a faster shutter speed. 

 

When I corrected Wolfgang @Architeuthis I especially refered to his wrong general statement about RAW files:

 

On 1/17/2025 at 11:42 AM, Architeuthis said:

this (RAW) file does not even contain any color information,

 

RAW files contain color information, as the the bayer array has sorted light into all three channels.
This is usually encoded in the manufacturers unprocessed file written to the buffer and card.

Wolfgang is right though when he talks about the initial sensor before photons go through the bayer array, but at that stage no RAW file in the camera has been stored. This is from Wikipedia, for the non-believers: 

 

Quote

"Raw files thus contain the full dynamic range (typically 12- or 14-bit) data as read out from each of the camera's image sensor pixels. The camera's sensor is almost invariably overlaid with a color filter array (CFA), usually a Bayer filter, consisting of a mosaic of a 2x2 matrix of red, green, blue and (second) green filters. (...) This RGB raw data still needs to be processed to make an image file, because the raw RGB values correspond to the responses of the sensors (..)"

 

Source: https://en.wikipedia.org/wiki/Raw_image_format

 

So now let's discuss the following statement which some of the previous authors cling to here.

 

Quote

all colours are reduced proportionally (in a camera underwater)

 

…when we talk about the blue or green water column, not illuminated by a strobe, in the background of our wide-angle underwater photos, let’s consider the following to think this through one more time. I’m referring to the early stage of photon collection, without any white balance, profile adjustments, or final color space corrections, etc. This process is very similar to the days of chemical film, where light hit your unprocessed slide, and underwater photographers had only 36 shots per dive.

 

Imagine you have a light source emitting the following simplified RED, GREEN, and BLUE values of light (photons):

 

Light travels at the same speed underwater for all wavelengths, but RGB values don’t decrease proportionally when underexposing because of wavelength-dependent absorption. In tropical seawater, the absorption coefficient for green (~525 nm) is ~0.04 m⁻¹, while for blue (~475 nm) it’s ~0.015 m⁻¹. Green is absorbed ~2.67 times faster than blue, which corresponds to a logarithmic difference of 1.42 stops.

 

When you underexpose by minus 1 stop (halving the light), green weakens further compared to blue due to this absorption difference. For example, starting with (R, G, B) = (0, 120, 255), underexposing by minus 1 stop results in (R, G, B) = (0, 22, 127), as green fades significantly. With minus 2 stops (quartering the light), the values become (R, G, B) = (0, 11, 64).

 

This demonstrates that blue light dominates as green diminishes more rapidly with underexposure, disproving the idea that RGB values decrease proportionally. The key factor here is the wavelength-dependent absorption of light in water.

I am afraid you are describing the effect of distance rather than exposure time.

Sorry, hit it too fast. Lambert-Beer’s law describes the remaining light as a function of extinction coefficient AND distance (in the lab, that is the thickness of the cuvette).

11 hours ago, Adventurer said:

 

When I corrected Wolfgang @Architeuthis I especially refered to his wrong general statement about RAW files:

 

 

RAW files contain color information, as the the bayer array has sorted light into all three channels.
This is usually encoded in the manufacturers unprocessed file written to the buffer and card.

Wolfgang is right though when he talks about the initial sensor before photons go through the bayer array, but at that stage no RAW file in the camera has been stored. This is from Wikipedia, for the non-believers: 

 

 

Source: https://en.wikipedia.org/wiki/Raw_image_format

 

So now let's discuss the following statement which some of the previous authors cling to here.

 

 

…when we talk about the blue or green water column, not illuminated by a strobe, in the background of our wide-angle underwater photos, let’s consider the following to think this through one more time. I’m referring to the early stage of photon collection, without any white balance, profile adjustments, or final color space corrections, etc. This process is very similar to the days of chemical film, where light hit your unprocessed slide, and underwater photographers had only 36 shots per dive.

 

Imagine you have a light source emitting the following simplified RED, GREEN, and BLUE values of light (photons):

 

Light travels at the same speed underwater for all wavelengths, but RGB values don’t decrease proportionally when underexposing because of wavelength-dependent absorption. In tropical seawater, the absorption coefficient for green (~525 nm) is ~0.04 m⁻¹, while for blue (~475 nm) it’s ~0.015 m⁻¹. Green is absorbed ~2.67 times faster than blue, which corresponds to a logarithmic difference of 1.42 stops.

 

When you underexpose by minus 1 stop (halving the light), green weakens further compared to blue due to this absorption difference. For example, starting with (R, G, B) = (0, 120, 255), underexposing by minus 1 stop results in (R, G, B) = (0, 22, 127), as green fades significantly. With minus 2 stops (quartering the light), the values become (R, G, B) = (0, 11, 64).

 

This demonstrates that blue light dominates as green diminishes more rapidly with underexposure, disproving the idea that RGB values decrease proportionally. The key factor here is the wavelength-dependent absorption of light in water.

This is not what happens.  The colours you see are the result of sunlight shining through the water and reflecting off your subject and reaching your cameras sensor.  The absorption results in the loss of longer wavelengths and is the result of distance travelled only. 

 

The popular analogy is a bucket.  You have a constant stream of light entering your bucket and if you double the the amount of time the tap is on, you get twice as much water.  Shutter speed and aperture are interchangeable in the bucket analogy, with aperture you use a bigger hose to get more light as you open up. 

 

To express this mathematically you use the term flux, this is the number of photons falling on a surface, in this case our sensor  per square metre per second.  The surface area of the sensor is fixed so to get the number of photons arriving you multiply flux by time to get photons/ sq metre.  If you reduce the time the shutter is open you get less photons.  Now this is the important part the colour of the light is already set by the distance it has travelled through the water and you have a constant number of photons of every colour arriving at every sq metre every second - the tap is right right at the sensor - the shutter curtain.   The sensor records the colour of the light that arrives at its surface.    Absorption doesn't change with time it changes with distance travelled and the camera records the light after it has travelled that distance through water at the time of exposure.  There is no room for argument here, this is how light behaves.

 

Now enough of the science behind it, I provided some examples above where I can't see the effect - If you want to argue further please provide some examples to prove your point. 

Thanks Chris, I was writing something very similar but you beat me to it.

Bill

I read this discussion for some time and started to wonder about the role treshold values of the camerasensor.

 

@Chris Ross and @Architeuthis are is right when they argue that "Changing the shutter speed doesn't change the the ratio of red/green/blue photons, just the total number of photons that arrive."  True, but after arriving there must be enough photons to surpass the treshold value of sensor elements of the camera. As the ratio of red light is the lowest,  underexposing  may cause a too low amount of red photons to activate the 'red light sensor elements'.  Not all in 'red sensor elements'' but relatively..

 

I am not an expert, but it sounds convincing to me 🙂

4 hours ago, Floris Bennema said:

I read this discussion for some time and started to wonder about the role treshold values of the camerasensor.

 

@Chris Ross and @Architeuthis are is right when they argue that "Changing the shutter speed doesn't change the the ratio of red/green/blue photons, just the total number of photons that arrive."  True, but after arriving there must be enough photons to surpass the treshold value of sensor elements of the camera. As the ratio of red light is the lowest,  underexposing  may cause a too low amount of red photons to activate the 'red light sensor elements'.  Not all in 'red sensor elements'' but relatively..

 

I am not an expert, but it sounds convincing to me 🙂

There really is no threshold with camera sensors, if photons arrive they are recorded, you may not get enough to click the reading from 0 to 1 if you are that deep.  But for exposing the water this is somewhat irrelevant - the red value being 0 or 1 or even 3 won't change the shade of blue a great deal at all.  The main thing is the green/blue ratio, but even at 30m depth dark blue water might read R:G:B of 0:25:80 (from checking one of my shots) And that's plenty to get away from any threshold activation issue. Even if you were very deep and G:B was 0:50 for example, changing that to 1:50 wouldn't notably shift things it would still look very blue.  You would also need to be underexposing at normal depths for this to be a possibility.

 

Even though the readout in PS is restricted to 256 shades of each colour internally in 16 bit there is still a reading stored even if it's not enough to click from 0 to 1 in intensity.  even with film it will still record, though that indeed has a threshold, but when it starts to record, the initial response is very flat plus we are talking about 1/250 sec exposures or thereabouts so reciprocity won't be an issue and there is some overlap between R:G:B channels as well.  Camera sensors though are much closer to linear and in the shadows seeing a change in shade is more difficult.

 

So the the fact that R:G:B photon ratios stay the same still stands.

Noise in underexposed areas of my pictures is not « colored » unless the white balance is (purposefully) off. So I don’t believe that thresholding plays a role here.
I think the quantum yield is in fact lowest for the GREEN pixels in the Bayer array, but did not look up any reference for that. I vaguely remember that’s why the square array has a 1xB /1x R /2xG geometry, but frankly one should look this up again to be sure.

But in the end, why bother? As long as we’re satisfied with our 50 shades of blue, does it matter how we achieved it?

4 hours ago, Klaus said:

Noise in underexposed areas of my pictures is not « colored » unless the white balance is (purposefully) off. So I don’t believe that thresholding plays a role here.
I think the quantum yield is in fact lowest for the GREEN pixels in the Bayer array, but did not look up any reference for that. I vaguely remember that’s why the square array has a 1xB /1x R /2xG geometry, but frankly one should look this up again to be sure.

But in the end, why bother? As long as we’re satisfied with our 50 shades of blue, does it matter how we achieved it?

The point of all this discussion is that changing shutter speed doesn't change white balance, the way it is does done for the water alone is to use a warm strobe and you can bring the WB to cooler temperature as a global adjustment.  For the image overall you can make the water less green by adjusting the tint but the subject increases in magenta.  It is important to understand how things are achieved so that you can take advantage of the phenomena to get your pics looking the way you want.

Initially I was looking for some smart inputs to ditch cto gels or warming diffusers on strobes with cooler color temperatures and harvest the maximum photon energy of these cooler illuminators while still getting decent blues in tropical waters.

 

I can adjust the hue and tint in my cameras auto white balance by several mired. In fact more mireds then some of the suggested gels or diffusers mentioned in this  forum.
 

I have the assumption that the additional color dynamic range when firing warm light into the foreground may not be needed (anymore) with recent full frame mirrorless cameras.

Edited February 2Feb 2 by Adventurer