Month: October 2021

Remove A Colour Cast from Your Photos

Remove A Colour Cast from Your Photos

One of the most annoying things associated with photography is getting the white balance (WB) wrong, especially at night or under artificial lighting. While auto WB should help to overcome most lighting issues, it does pay to make sure that you use the right WB option available to you in camera for the situation you are filming in. But what about if you don’t make the adjustment there and then? Is it game over? Can you recover a scene that has a strong colour cast for example? Well, yes, in general you can and it’s all thanks to the power of the Curve tool together with a little help from our old friend WB. These two tools are found in practically every decent photo editor available today from Snapseed on your mobile phone through to Photoshop, Affinity Photo and Gimp.

What’s a Colour Cast

A colour cast is basically where the underlying colour tends to shift towards a particular tint. It’s most often seen in night shots where photos can adopt an orange tint or in the studio where the WB is not corrected for the types of light being used. While in some night shots a golden orange tint or glow can enhance a photo, it certainly doesn’t look good in studio photography or where age has spoiled a family heirloom. This article presents some ways for you to improve your photos where a colour cast is spoiling things.

EXAMPLE 1 FIXING THE COLOUR IN AN OLD IMAGE

Let’s start with an example on how to remove the colour cast in an old photo.

Copyright unknown. Image used purely for educational purposes

As can be clearly seen, this image has a very strong orange colour cast. It’s also not helped by the fact the predominent colours in the image are earthy tending towards browns, red and orange. Even so, bringing out the blues and dropping the reds and browns has to be a good start in making this image look more natural.

To recover this image I have used two techniques. The first is to identify an object in the photo that should be white and then to use this information to find other areas within the image that should also be moved towards white. Here the most obvious white colour is in the boys tee-shirt. In the first photo above this has a pinkish tinge especially around the neck. In this technique we create a rectanglular box, colour pick the pink tint from the teeshirt and apply it to the box. We then pull out the handles until the box completely covers the image. Setting the blend mode for that layer to Divide results in other areas of the image that should be white or close to white becoming white lifting the whole image towards more natural tones. Here’s the result of undertaking that process. Notice that the white tee-shirt now looks white while other tones within the image have improved. However, to my eye the photo still exhibits a strong bias towards earthy colours which might not be desirable deping on the intended use.

Copyright unknown. Image used purely for educational purposes

To improve this image still further, especially with regards the peol in the photo, we can additionally apply a curves adjustment to the image to remove much of the red and to better control the greens and blues. To do this we simply shift the RGB values in the image towards what they were more likely to have been. For example, jeans are typically blue so act as a guide here. The resulting image below shows just how powerful the curves feature can be in everyday editing.

Copyright unknown. Image used purely for educational purposes

And below is the final image with pretty much all of the colour cast removed. Applying a white balance correction could improve this image still further but as a demonstration of technique, it pretty much serves it’s purpose.

Final image with curves applied. Copyright unknown. Image used purely for educational purposes

EXAMPLE 2 PROBABLY THE QUICKEST FIX THERE IS

In this example we are going to use the very simple technique discussed above of applying a colour picked rectangle over the image and setting the blend mode to Divide. This only works if you can identify an area within your original image that should be white. Here, the jacket does have some white so I have carefully picked on this shade (see image 2) in order to colour the rectangle that covers the image. As you can see in image 3, this is a distinctive shade of cream rather then the white it should be. In the final image I have set the Blend Mode to Divide in order to neutralise all of the colours in the image bringing them quickly back to what they were on the day of shooting.

As the final picture of the series shows, this is a remarkably quick way to compensate for a colour cast in your images where you have a known area of what should be white. To work though your editor must include the Divide function within its Blend Mode options. On1 for example does not have this option so this is not a teqhnique that will work if you have On1. In such cases I recommend using curves and white balance modifications which I describe in more detail below.

EXAMPLE 3 RECOVERING FROM WHITE BALANCE ISSUES

Here’s another example. In the first of these photos the image again has a very strong orange colour cast. Here the colour cast is caused by an incorrect setting of the white balance in camera during a photo shoot. The second image shows how easily the image can be recovered using the simplest of work flows.

On the right hand side of the above image you can see the various layers applied to the photo. Firstly I duplicted the image, just is case I do something stupid which affects the integrity of the base image, then I used a Curve to modify the colours in the image. After this I used an Unsharp Mask to affect a little increased sharpness in the eys, hair and pearls and finally I took the opportunity to remove a few tiny flaws such as flyaway hairs etc in the overall portrait. An alternative strategy could have been to complete the basic work up to and including the Curves adjustments and then to take the image into say, On1 Effects Ai and or On1 Portrait Ai to apply some adjustments. To be honest, Affinity is excellent for portraiture so I’m not convinced there’s a huge lot of benefit of going into On1 to achieve a better resullt. I also have the option though of using Anthropics PortraitPro but again, here I didn’t feel the need to do this.

So, let’s now look at what the tone curve would need to look like in order to achieve the above result.

This is a pretty simple curve adjustment but of course, for each RGB channel ie Red, Green and Blue as well as the Master channel you can apply much more complex curves in order to achieve your required goal. As it is, these simple modifications work pretty well although as mentioned, I did also reduce the temperature in the white balance to add a little more blue.

So, that’s pretty much it really. In some cases, particulalry where the lighting is used to effect a particular artistic look it is much more difficult to return the image to a neutral colour pallet because this actually doens’t exist. It can also be difficult where orange, green or blue lighting is used to create an effect, for example on buildings at night. However, where your camera misunderstands the colours due to processing confusion as in the above example, it is really quick and easy to return it the neutral colours which actually existed.

This final image of Sarah, from the same set as those above, has been further processed to illustrate additional ways to enhance the basic image. As above, the original image was also colour tinted orange and has therefore undergone a similar transformation.

Model: Sarah Dowrick. Image copyright Dave Collerton

I very much hope this short introduction to colour correction using very basic tools available within any quality editing suite has been helpful.

Dunked your camera? It’s not all over, you’ve got a fighting chance to save it!

First off let me say that the experience of falling in water with your camera around your neck is not the best fun you can have. My personal attempt did result in good scores from all judges especially as I bounced on some fairly large boulders before hitting water. Sadly, my Fujifilm x-T1 also saw action and while I have now dried off, and indeed so has the camera, the 18-55mm f2.8-4 lens that was attached to it is proving to be a little stubbon. Here’s what I have been doing to recover from this semi-disasterous situation.

Your Next Actions Make a Difference

Firstly, let’s rewind a little because the actions you take when this sort of thing happens really matter. So, while still lying in the water I had the presence of mind to pass the camera to a passer-by who came across to help. That move was practically instantaneous which meant that while the camera was under water for a few seconds, it was not in water for minutes. It was also turned off, another plus point. The other big plus point here, if there can be one in this situation, is that the water was practically pure as well as being fast moving as it was generated by a cascading waterfall on Dartmoor. Had the camera fallen into seawater, or a murky algae laden pond I would not be as confident that I can recover the camera or the lens. So, what can you do to improve your chances of a successfull outcome?

  • Turn the camera off or keep it off
  • Remove the battery and memory cards
  • Wipe down the camera with some form of absorbent cloth as soon as you can – inside and out (not the sensor though)
  • Wipe down and dry the lens as much as possible
  • Open all of the comparment doors to enable them to circulate air into the camera
  • Protect the sensor with a body cap or use another another lens
  • Get the camera and lens into a dry, warm environment as soon as possible
  • Do not be tempted to turn on the camera for several days. Put it aside and leave it alone!

The Camera Body

Firstly, let’s talk about the camera body. Although emersed for a few seconds, all the compartment doors were at least closed and the camera was off. Once I was back on my feet, and although drenched, I immediately removed the lens, the battery, the memory card and I opened up all of the compartment doors. I was fortunate in that I was carrying a microfibre cloth and although my canvas camera bag did get a little wet, it didn’t get soaked. That was lucky because it also had two spare batteries and a Viltrox 23mm lens inside. First job, remove every drop of water off every part of the camera body I could access. Amazingly, no water had got into the sensor chamber, it was bone dry thanks to the 18-55mm that was attached. There was a little water in the various compartments but overall it cryed off externally very quickly due in no small part that it was a nice dry and warm day. Normally I would have hightailed it home at this point but as it was a family day out and we had our grandchildren who are only three and five, I decided to trudge around for the next hour or so. In fairness, other than being drenched and a little sore I came off pretty lightly and my youngest grandson George was a delight helping me up and down the routh terrain at every opportunity. As I walked, I allowed the camera to sit in the breeze with all the compartment doors open. I had fitted the Viltrox 23mm by this point as I didn’t want to add fuel to the fire by allowing dust and debris into the camera body via the open sensor compartment. Ideally I would have used a body cap but of course, who carries one of these around just in case you fall in a river, stream, pond or the sea!!

Over the next couple of hours the light breeze dried off the camera a treat and I eventually made it home about three hours later. When I did, the first thing I did was to remove the Viltrox 23mm and to put some cling film over the lens apperture. It was still bone dry so I think I got really lucky here. I then opened the window and hung the camera on the window pole to gently continue to dry out in the breeze. By this point in time there was no water apparent anywhere so I thought this was perhaps a good first step. I left this overnight and then in the morning, I put it on a radiator (this was on very low setting) to allow a little more heat to access the internals. I didn’t however put the camera directly on the radiator, I lifted it up about an inch to allow the warmth to envelop it. All the doors remained open, including the LCD panel which I had pulled out earlier to dry off and had kept it there. After a day and a night more of doing this, I decided to test the camera. Interestingly, the on/off switch, which on the x-T1 is part of the shutter button, was really stiff which worried me a little. However, with a little persuasion it moved and turning it on and off a few times (with the battery out) resolved the problem. I also checked all of the other dials and buttons for similar problems but these were all good, no stickiness as far as I can see.

Did the camera fire up? Amazingly it did, without issue other than it looked like the damp had forced a factory refresh as I needed to reset the language, the date and the time. All of my eight Presets had also been wiped but no big deal, I only really use two slots and I know the settings off by heart. Having shot a few bland test shots I turned off the camera and set it to one side. It’s being sitting in the corner since then and I’ve just tried it again and it’s working fine.

The Lens

As I mentioned, the Fujifilm 18-55mm really didn’t do well in it’s encounter with water. Unlike the camera it was sopping wet. The worry here then is that even if I manage to dry it out, it is very likely that despite the water being really pure, there are going to be some water marks on the internals of the lens etc. It is also a toss up as to whether the electrics will ever work again. Now I’m prepared to sit it out for as long as it takes to dry it out totally as I still have the gorgous Viltrox 23mm working perfectly on the x-T1 plus I have the Samyang 12mm manual to play with which is huge fun. All in all, and while I might miss the 18-55mm, I can still shoot perfectly well with what I’ve got. Anyway, let’s talk a little now about what I have done to try to recover this lens.

The first thing I did was on the way home we stopped off and bought a huge bag of rice. Rice is brilliant for sucking up moisture so it’s a good first step, after drying the lens externally, in sucking up the moisture in the lens. Now, given the amount of water drops I could see on the internal lenses, this was never going to be a quick solution. It could take days if not weeks to remove all of the water, even in a warm home. Anyway, I popped the lens into a plastic container completelysurrounded by the rice. I used a fair amount of rice for this as it’s cheap and plentiful. I then put a cleanfilm over the top and let it sit for 5 days before sneaking a peak. Interestingly, the top lenses were dry, or at least appeared so, while the lens closest to the sensor, which had ben bone dry whan I started drying out the lens, was covered in water drops. This tells me that water is moving freely around the lens and that the process is working, albeit very slowly. In order to try to speed up the process, although why I don’t know, I have now put it on the radiator, which is only slightly warm, to hopefully help with the process. I also felt it wouldnt hurt to move the rice around. I’ll add some photos later to illustrate the process of camera and lens drying.

Initial Conclusions

So, as it stands I have at least achieved some success. My body is repearing itself nicely and the x-T1 appears to be working OK, no apparent problems although time will of course tell how it fairs in the longer term. The 18-55mm although initially not looking particularly healthy has now, and this is after 9 day sitting in rice in a warm environment, dryed out significantly. It might actually be bone dry but you know what, I don’t need to push it and while another few days won’t affect me it might be good news for the lens. Examination today shows that there are some marks on the lens surface, but these aren’t substantial. Whether it will ever work again, I still don’t know. The chances 50/50 I’d say. If the electronics work I have found an interesting article on dismantling and accessing the top two elements to clean them which I will try if necessary. I suspect that this will be some time yet though. Whatever happens I will update this blog post and let you all know how things turned out.

I hope that this article is helpful. Having a good plan of attack in such circumstances is key to success. You also need patience. Everything has to be 110% dry before you even consider turning the camera on and that probably means waiting the 4 or 5 days as I did.

What’s the best lens for street photography?

Right off the bat I just want to say that there is no right or wrong answer to this question. In truth, the lens you choose is probably going to reflect your confidence levels as a photographer. If you like mixing it up on the street then a 28mm is probably perfect as it gets personal at that focal distance. If you are a little shy or worried about photographing strangers then a longer prime or even a short zoom is probably going to favour your style of photography. The essence of this article then is that, there is no perfect lens for the job, it’s all about you and your confidence levels as a photographer!

All this being said, it’s hard to imagine anyone doing street with a long zoom. Candid / street photography requires you as much as possible to blend into a crowd and nothing shouts photographer like a DSLR armed with a long zoom. Equally, a fisheye or an ultrawide lens is pretty much useless in my opinion unless you like standing toe to toe with your subject. With that lens you are going to see more that the whites of their eyes and that means getting very personal. Great for urban landscape photography for sure, but candid people biased photography, in my opinion, leave it in the bag or at home. Now while some photographers also spout long and hard about the benefits of vintage glass, which is often manual focus of course, forget it. It’s not 1961 when you had no choice, it’s 2021 when you certainly do. Rely on a manual focus lens / camera combo and I guarantee you that you’ll miss that pulitzer price winning shot when the opportunity arises. In fact I’ll go as far as to say that you’ll miss a lot of really good shots. In street, great images don’t come along everyday so when they do, you need to be nimble and nimble means more often than not, nailing it on auto.

For me, the sweet spot for street and candid photography is going to fall somewhere between 24mm – 75mm in full frame terms although my favourite squeeze for candid photography today is the pocket sized Canon G7X, purely because when I started to get really interested in street and candid photography, that is what I bought. I use it a lot for street because it’s small and unobtrusive and easily fits in a trouser pocket so is highly portable. The 8.8 to 32mm lens combined with the 1″ sensor seems to capture great shots even in low light. In full frame terms that’s 24mm to 96mm so pretty handy for street. In truth though I’d be equally happy with any good quality 1″ sensor compact nowadays, such as Sony or Lumix. Another good option of course is the Fujifilm x100V (or an earlier variant) and the x-Pro series ie version 1 through 3. All great cameras for street and candid photography.

The diminutive Panasonic Lumix GX80 with 12-32mm kit lens.

Another favourite of mine is my Panasonic Lumix GX80 which has an M43 sensor. Again it’s small, descrete and very capable. Armed with the Lumix 12-60mm f2.8 (or even just the f3.5 if cash is tight) this is a great street combo because again, it’s small and compact. In fact it might equally be as good with the kit lens, the diminutive 12-32mm. What I certainly wouldn’t take out with me is my Nikon D600, even with a tiny pancake lens. Way too big for serious street work. And although I have used my Fuji x-T1 a lot for street photography, especially with the 18-55mm f2.8-4 and the Viltrox 23mm, I still think that this size is too big for serious street. This is because with street photography you really just want to blend into the background, not make a statement along the lines of “hey, look at me, I’m a photographer and I am photographing you!”.

France, 2018. Image taken with Lumix GX80 with Lumix 12-60mm lens

Now I am a member of a couple of Fuji facebook groups and the buzz on these groups is about SOOC photography ie straight out of camera shooting. Like many, I do like the jpegs that Fuji cameras can create, especially with a filmic simulation applied in camera but I maintain that for good street photography it pays to blend in and you’ll do that best with a small, innocuous camera like the Canon G7X or something similar. Because I am as happy playing with RAW as I am with JPEG’s, I don’t worry too much about SOOC although as I mentioned, with the right light, Fuji cameras can produce stunning yester-year film quality results. As mentioned above, for the serious Fuji street photographer I would suggest that you look at the X100V or perhaps the X-Pro3 if you want lens interchangeability. These two cameras for me, along with many modern compacts are about as perfect as it gets for street and both sit alongside my Lumix GX80 and Canon G7X for this very reason.

Recolouring B&W photos using Affinity Photo

Affinity Photo is not for the faint-hearted but put your soul into learning and prefecting it and you’ve got the perfect editing partner for life. That’s not to say that there aren’t some amazing photo editors out there, many for free, just that it is a well supported, hugely capable product from a major author. Serif have left nothing to chance with Affinity Photo and with each release, it just gets better and better. Add to that a healthy, growing userbase with many users adding quality tutorials to YouTube and elsewhere and you’ve got the perfect, practically zero cost alternative to Adobe Photoshop.

Now, as I mentioned, Affinity Photo is not for the fainhearted but if you know Photoshop, you’re already more than half-way there. I bought my version of Affinity Photo back in 2019 and I used the pandemic lockdown to throw myself into it. That doesn’t make me an expert, far from it, but I have yet to find something that Affinity can’t do. Where it differs from Photoshop, other than in price and the lack of a yearly subscription, is in the number of people developing presets for it. Whatever you need to do, Photoshop probably has a preset to help you automate the task. That’s not to say that Affinity Photo doesn’t have similar features, just that currently at least there are fewer people building add-ons (Serif calls these Macros) for Affinity Photo. Now doubt as time goes on, that situation will change.

Before I go futher, this article is not meant to be a HOW TO for Affiity Photo. These are best created using video as you really need to see the process rather than read about it. All I am trying to do here is to introduce an interesting creative process which is to convert an old 1940’s post card sized photo into a colourful, period acurate image. Lets take a closer look at this particular image.

Home Guard, West London, 1940’s

The primary reason for doing it was because it contains an image of my grandfather Frank, who served in the home guard during WW2. Frank, for those interested, is bang in the centre, top row. This then is a labour of love and is still very much work in progress as far as I am concerned. So far I have completed perhaps 80% of the editing work with the concrete road and finishing refinements still to do. There’s no hurry on my part as there’s no client to wory about but had there of been, a job like this would take perhaps 4 – 6 hours depending on the level of detail required, historical accuracy and overall refinement needed.

Now, although I have said that I don’t intend to talk about the exact process of re-creating this image in colour, I do need to talk though the strategy and process as this is important in understanding how we can create a colour accurate image of wartime 1940’s. The only real way to do this then is to find relevent colour images from that period or failing this, find appropriate colour images from things such as renactments which utilise, for example, authentic uniforms from that same period. Same goes fo rthe buildings, roof and road colours. Given the lack of colour images of this type from that period, the latter approach then is much more feasible. For a start, in the 1940’s B&W was the predominant film stock used and while colour film existed, google searches don’t always deliver the necessary results. The other thing to remember, since we are going to gradient map the colours we find in more modern photos onto the B&W image above, the highlits and shadows in that image play an equal part in contributing to the reality of the final result. For this reason we not only need to find a colour accurate scene, we also need to find one with similar lighting and tonality conditions ie light and shade.

Having found relevent photos to use as a guide, we can now create swatches that will help us to recreate the colours, for example of the uniforms, on the B&W image. Here, you can see that the sampled colours from the above photos have been applied to the B&W image as reference colours. On Affinity Photo these are on a seperate top layer. From left to right we have uniform lows, mids and highs, in the centre facial colours again from lows, though mids to highs and on the right, the colours sampled from the door of the barracks, again lows, mids and highs. Amazingly, this is all we need to begin the fascinating process of recolouring the B&W image.

As previously mentioned, the process of applying the colours from the swatches is done using a Gradient Map. For example, in the case of the skin colours we simply sample the colours above setting the low on the gradient map (which is in red) to the darkest shade, the mid colour (green) to the mids and the lightest colour to the highligh (blue) colour. This is perhaps better explained by the reference image below which shows the default colours applied to the image. Here red depicts shadows, green depicts the mids and blue depicts the highlights in the original B&W image. Of course we aren’t just talking about 3 tones, you can add as many points on the gradient map as you like by simply sampling colours between low and high and adding them to the map.

This all being said, let’s take a look at where the image is right now as I write this article so I can talk more about the process and the refinements that I still need to make. In this image, I have applied four gradiant maps. One is applied to the uniforms, another to the skin tones, a third to the buildings and finally, a fourth to the roofs. Of these I am least happy with the ashphelt roofs and will revisit this shortly to lower the harshness of the blacks and to add some texture.

I chose to create this article to inspire others to have a go. We all have family photo treasures, many in B&W, and in my view, simply adding colour to a face immediately brings it to life. The last time I saw my grandfather was 40 years ago but looking at his face here, irrespective of the fact that the resolution of the image is terrible, it immediatly bought back wonderful memories of spending precious time with him when I was a boy. Master builder, mandolin player and a great father and grandfather, as is so often the case we don’t know how good things are until we lose them.

Affinity Photo is an amazing image editing tool. Serif offer a short trail period if you wish to give it a go. With special offers, Affinity Photo can be purchased for about £25 which is basically stealing. If you do decide to try it though, my suggestion is to persevere as if it was your only editing tool. You can’t pick up and put down Affinity as you can something like Luminar Ai or even Lightroom, it requires practice and patience. Put the effort though in and you’ll be rewarded by some some amazing editing tools on which to build real editing ability.