Welcome to my second tutorial, looks like we’ve made it for another one! Today we’ll talk about enhancing colours in a photograph. In particular, we’ll talk about increasing the saturation or vibrancy in an image. Not so much the changing of, or adding of color casts, but enhancing the look of the colours that are already in the photograph to begin with.
This is one of those subjects that if you ask 10 photographers, you’ll probably get 10 different answers on how they do it. For me, I have two methods I love to use, to enhance the colours in an existing photo. They are both Photoshop techniques.
This method is quick and convenient, the feel of the image isn’t as good as the next “LAB method”, but for most purposes it works very well. This method is also often called “Velvia” action, mainly due to the look it produces which has a similar appearance to the highly saturated look of the famous Fuji Velvia transparency film.
This is a non-destructive method, meaning that by making this adjustment (with my method), you will not destroy or alter any pixels in your image. It uses an adjustment layer. What we do is create a new channel mixer adjustment layer.
For reference, this is what our test image looks like, before any adjustments:
What we see now, is for each of the output channels red, blue, and green, you can adjust the source contributions from the three channels. By default the red channel is made up of 100% red, blue channel is 100% blue etc. Now what we do is for each of the output channels we make an adjustment as follows:
- red output channel: R=150%, G=-25%, B=-25%
- green output channel: R=-25%, G=150%, B=-25%
- blue output channel: R=-25%, G=-25%, B=150%
Note that for each of the output channels adjustments, the sum of the three source channels is always 100%.
Now that we have added the adjustment layers, we see the colours in the photo are much more saturated and it does a lot better job than simply using the saturation tool.
To control how strong the effect is, we can adjust the opacity of the adjustment layer.
LAB mode a and b channel curves adjustment
This method to me creates nicer looking results. It is not as convenient as the channel mixer method however. It involves adjusting a and b curves while in LAB colour (most people work in RGB colour).
The disadvantages if this method is that it is a destructive edit, you can not apply this as an adjustment layer. To convert to LAB mode you really need to flatten your layers also (in certain cases it may not be necessary). But we can get around this by using backup layers and doing the adjustment as a separate image copy. We must keep in mind also that to convert from RGB to LAB, then from LAB to RGB again, is a lossy operation, i.e you will lose detail. But working 16-bit mode can help minimise losses.
I find this method very pleasing, it is in fact my favourite method as it adjusts the true colour “saturation” without affecting the overall luminosity.
To start, we have image we would like to enhance or bring out the colours a bit. We need to convert it to LAB mode, but since we have multiple layers already, we do not want to just convert it straight away as it would require us to flatten the layers. We need to bring the image into a new window and work from there.
So we go to Image -> Duplicate… to duplicate the image
We now have a new image with all layers intact, we work from here.
If you have multiple layers in your image, you will be asked to flatten them, yes you can since you are working on a duplicate copy.
Now you are in LAB mode, you’ll see under Channels there are no more red, green and blue channels anymore. Instead you now have L, a and b channels.
We now add an curves adjustment layer, and apply a change to a and b channels. The a and b channels contain information about colour, it is given a value between -128 and +128. The further away from zero it is, the more saturated a colour is. We then apply a S curve with the following control points (input, output):
- 0,0 (this should already be there)
- 128,128 (this should already be there)
The curves adjustment dialog should like this (screenshot shows adjustment curve already set up):
It is important the curve is symmetrical, otherwise you will get a colour shift (more accurately, a shift in hue). The numbers I have given above are just guidelines only and is what I use, but any symmetrical S curve can be used. The stronger the S, the stronger the saturation effect.
Now go to OK and you will see you should have a lot more colour in the photo.
By adjusting the opacity of the curves adjustment layer you just made, you can adjust the strength of this effect on your image. When you are finished, you can flatten the layers and copy the layer back into the original file, as a new layer.
Automating with actions
If you like those two methods, why not record an action. There’s no reason you need to be setting those parameters over and over again, just record an action and add it to your workflow.
A word about Lightroom
Adobe Photoshop Lightroom has a adjustment called Vibrancy which works well as a quick edit method. For non-critical images, this works pretty well so we mustn’t forget this method!
Oversaturation and clipping
Its easy once we start enhancing or boosting colours in a an image to go a bit overboard and over-do it. We must keep this in the back of our minds. There are a few things to consider:
- it is now easy to produce colours in your image, that is not physically possible to be printed on paper
- it is now possible to have colours in an image that is not reproducible on a monitor, and vice versa i.e you can see colour on the monitor that doesn’t exist in the image
These concerns and the management of colour is another article in itself. But for those who already know a bit about colour management, here is what you can do:
- When performing these adjustments, use the Gamut Warning option (in Photoshop) to see when colours are being clipped. You will of course need to select the correct ICC profile for your paper/ink/printer combination.
- Use soft-proofing to approximate how your print will look.
- Keep in mind, what looks great on a computer monitor may not look so good in print, so that’s why its important to make your adjustments, with the output device and profile in mind.
In screenshot above, you can see the grey areas in the middle of the image. This is an image which has had the channel mixer method colour boost applied. The grey areas shows colours which are now clipped and are outside the gamut of a Fuji Frontier and lustre paper combination. By using layers masks, we can reduce or mask out the adjustment in that area by careful use of the paintbrush, or by pulling back the opacity of the adjustment layer (which would affect the entire image).
The image used in this example probably isn’t the best to show the effects of gamut clipping, but take a shot like this, full of deep blues. Before adjustment:
Now after a little curves adjustment and a channel mixer adjustment layer for some colour boost:
Now you see why it is important to keep an eye on your colours blowing out. This is what it looks like without the gamut warning on. Okay, I admit it, it does look overdone and a bit too blue! This is an extreme example. But I think we start to get the picture here.
If you are using a CMYK output (for offset printing, magazines etc), the available colour range is even smaller, so you are even more likely to blow out.
No one ever said you must apply these adjustments to your entire image. If you only desire the adjustments to be made to certain areas of your image, then take advantage of layer masks, go wild with your paint brush or use some gradients and only apply the changes to those areas that need it.
There are a few other techniques that other people use. I don’t personally use these methods, but I will list some just in case you might like to investigate them:
- increase saturation under hue/saturation
- using vibrancy in Lightroom (I do this if I’m in a hurry and don’t plan on using Photoshop)
Any comments or questions, leave a reply down below and catch you in the next one!