Help with color issues

[DennyN]

Newbie question.....
What accounts for the amazing vivid colors I see with photos taken by folks that know what they are doing. My guess is there are many variables but how much of it is post processing and in what way ? My stuff always seems a little bland.
My stuff - Denny Noll

Any input would be appreciated.

 

[Jeff1547 47]

First, you have to make sure that your computer screen is actually giving you the correct colours, I use a Spyder 5 for this. Then in your photo editing software you adjust to your taste.

 

[weepete]

can you give us an example Denny?

 

[TCampbell]

I'll make a few comments...

You might think... just grab the saturation slider. That would be a mistake. There is a place for using the saturation adjustment but it's usually the last control I reach for.

First... you'll want to make sure you understand how "white balance" works so that you can capture accurate color when you're out shooting. White balance simply means you are compensating for the fact that the light illuminating the scene may have a "color cast" to it... and that will alter the colors of the subjects being illuminated by that light. The compensating control is take something that you know is color-neutral (such as a photographic "gray card") and shoot a photo of that subject the SAME light you will be using to shoot everything else. The computer knows the RGB values for each "gray" pixels should basically be identical. So if the "green" and "red" values (which, when combined produce orange) are stronger than the "blue" values, the computer can compensate and this neutralizes the color cast.

Second... make sure you can trust your monitor. A color calibration tool is really helpful. Jeff mentioned the Spyder (that's made by DataColor). The other popular product is the ColorMunki (made by X-Rite). Both will get you great results. BTW, if you do your own printing (many people send out prints to labs), there are versions of these which can also color calibrate the printer (really the printer/paper/ink combination).

Third... look at the histogram of the photo in your processing software. What does the light distribution look like? You can often dramatically improve the look of an image by properly setting the white point & black point and/or using the "levels" or "curves" adjustments. What this REALLY does is boosts the "contrast" of the image. You might think to just use the contrast adjustment, but the contrast just makes bright stuff brighter and dark stuff darker without actually knowing anything about the image. When you set the white & black points and/or use the "levels" adjustment, you are in more control of the contrast (suppose the blacks should only be slightly blacker... but the whites need to be much whiter... you can control this with the levels adjustment whereas a contrast adjustment would just apply equal amounts ... and that wouldn't get the correct result.) There's also something called "Clarity". Think of this like "contrast" except it's really mid-contrast. Whereas "contrast" stretches the tonality of the entire image... "clarity" ignores the very light and very dark and only adjusts the middle. (btw, there's some debate on exactly how it does this).

This controlled contrast adjustment (via white-point/black-point/gray-point, levels, and curves) WILL result in the appearance that you've boosted color saturation.. even though you never touched the saturation yet.

Fourth... if shooting on an overcast day, sometimes just cheating up the white-balance to warm up the image will create the illusion that it was shot on a sunnier day and possibly during the golden hour... even though it was really shot on an overcast gray day. When you do this, it ALSO creates the illusion that you applied color saturation.

Sixth... Vibrance & Saturation adjustments.

I'm short-cutting this to avoid it being too long of a post because there are a few controls I MIGHT grab depending on the photo. Most editing software has a "colors" adjustment where you can selectively boost either the saturation or luminance (and even adjust the hue) of SOME colors in your image, without doing all them. Also be aware that many of the better editors allow you to "brush on" an adjustments rather than having to apply the adjustment globally to an entire image.

So ignoring the ability to selectively edit just a few areas or selectively tweak just a few colors... we'll get to vibrance & saturation.

I've said elsewhere... saturation is strong juju. You can easily over-use this tool and make things worse. I don't even trust my own judgement. When I apply these adjustments, I often adjust to my liking and then... WAIT A DAY... then come back and take a 2nd look (it causes me to be more objective). I often think "What the @#$@# was I thinking?" and then I tone things down a bit. So just be careful not to get carried away. If the image LOOKS like you adjusted the saturation... chances are you over-did it.

SO... there's also something called "vibrance". Vibrance is similar to saturation... except it tries to protect colors that match skin tones. E.g. if you take a portrait of someone and you think... their outfit looks a little dull... you're thinking maybe you should boost the saturation to give it more pop. Then you notice they have oompa-loompa tangerine skin because the saturation sider boosts everything. So imagine you boost the saturation, but selective grab the skin-tone colors and bring them back. That's sort of what vibrance does. Colors that are not similar to skin tones will get a much stronger saturation boost than colors that are similar to skin tones.

Notice how I put the actual saturation & vibrance controls at the end of the list... they should be the last things you use.

 

[DennyN]

can you give us an example Denny?

OK, here is a Niagara Falls picture of mine - Niagara Falls & Area Nice and realistic but what if a want a more dramatic effect.

Here is one from someone one Flckr (hope it's ok to post this link, if not let me know and I will edit it off) - Rainbow (Bridge) Rainbow aside, big difference in richness of colors.

 

[Derrel]

Excellent advice on the "wait a day " bit above! A fine post by TCampbell above.

 

[smoke665]

OK, here is a Niagara Falls picture of mine - Niagara Falls & Area Nice and realistic but what if a want a more dramatic effect.

Hopefully will be there in late September, can I reserve giving advice till then? LOL. I didn't see it mentioned anywhere but what are you using for software? In addition to Tim Campbell's excellent points I've found the Dehaze Slider in LR very helpful, especially on landscapes. It seems like they always have air contaminants that cause the image to look flat. The Dehaze Slider gives you brighter colors and sharper details.

 

[DennyN]

OK, here is a Niagara Falls picture of mine - Niagara Falls & Area Nice and realistic but what if a want a more dramatic effect.

Hopefully will be there in late September, can I reserve giving advice till then? LOL. I didn't see it mentioned anywhere but what are you using for software? In addition to Tim Campbell's excellent points I've found the Dehaze Slider in LR very helpful, especially on landscapes. It seems like they always have air contaminants that cause the image to look flat. The Dehaze Slider gives you brighter colors and sharper details.

I use PhotoScape for minor quick adjustments. The only software I have for more in depth processing is GIMP.

 

[Ysarex84 78]

It's all post processing if you're talking about digital photos. In other words a digital photo is only realized via heavy post processing. Digital photos start with an electronic sensor in the camera which generates a charge relative to exposure to light. The sensor is B&W only and so to record color an array of tiny red, green and blue filters is laid over the sensor. To create an final photo that filter array has to be interpolated off into the colors of the scene and the processing begins. You can actually see a digital image before that stage but it's pretty useless. It'll look kinda like this:

and I show you that to make the point that processing is certainly required. Many digital cameras at the stage in which you see the image above have the option to save that version of the photo for people who want to do all the processing themselves. The camera's software will then proceed to process the image as well. Much has to be done. Bare minimum:
1. The red, green, blue filter array has to be interpolated off the image and color values assigned to each pixel. To do this a color space must be assigned.
2. The white balance of the image has to be set.
3. A color profile must be applied.
4. A brightness level must be set.
5. A tone curve must be applied to balance shadows and highlights.

This is all post processing and it's all done by the camera software immediately as the photo is taken. The photographer has input control over many of those steps just noted. Typically the photographer can use the camera controls to moderate at least brightness, the color profile used, white balance, the color space assigned, and parameters that effect contrast, shadows and highlights. This is all post processing.

That's my favorite test subject above: Cheapy Smokes. It's a block from my house and every time I get a new camera I head over to Cheapy Smokes on a sunny day.

That's the photo created by my camera set to it's processing defaults. I have two processing routes I can go:

1. Save that raw unprocessed version of the image above and process that myself or
2. Work with the camera controls to vary the output from the camera.

For example:

This version of the photo was also produced by the camera. Notice the difference in the sky color -- it's turning green compared to the first version. That's because I changed the camera's color profile. I also raised the contrast.

Here's another:

And now the sky is turning purple, yet a different camera profile, and I lowered the contrast. This is all processing control built into the camera. It can be a little demanding to have to make those decisions on the fly as you're taking photos, but many people get quite familiar with their cameras and skillful at doing that.

Our newer cameras can have a pretty extensive range of options to control the final output but with one big CAVEAT: Anything done is done to the whole photo. Our camera's built-in processing controls are all global in application. For example they can't change only a local area in the photo.

WB: In all of the above photos I let the camera set the White Balance automatically. On a sunny day with a subject that contains a good collection of white and netural areas auto WB can work fairly well. In other circumstances it can crash and burn. That can be a major processing concern. It's possible to set a custom WB on the camera and it can be very worthwhile depending on what you photograph.

This last version of the photo is hand processed by me. I've done things that the camera software can't do because I was able to make local changes. The local changes I made are subtle except for squaring up the building, but they have a cumulative effect. I got the color right by setting a custom WB and by adjusting the camera profile I applied. I darkened the sidewalk a tad compared to the rest of the photo and put a light vignette into the corners. I made some local contrast changes as well.

Software: You have multiple choices that require different levels of involvement and learning and spending.

+You can focus on the processing controls in your camera and make sure you're using them to best advantage.
+You can then apply additional processing to the images produced by the camera. This is limited in what's possible but can be quick and easy and rewarding and often accomplished with inexpensive software. You noted Photoscape and GIMP. GIMP is quite capable and finally out in version 2.10 (Yeah 16 bit!!), but in both cases you're using software that adjusts the already processed JPEG image from the camera. That has limitations: here's some reading: classnotes
+You can use a camera that saves raw captures and do it all yourself. Higher learning curve here but ultimately more capable of doing whatever you want. Software is typically more involved but there are some inexpensive options (and plenty of $$$$ options).

Joe

 

[smoke665]

The only software I have for more in depth processing is GIMP

Then mine and earlier comments may or may not be applicable. I haven't used GIMP so maybe some who have will chime in. I believe the dehaze function can be simulated via levels and unsharp, in GIMP but again I haven't tried it.

 

[weepete]

Well there's a couple of things that jump to mind. First is time of day, the one from flickr was shot during the golden hour at dusk (you can see the lighting difference if you have a look at the buildings) and that will have an effect on the colours in the shot. There's also a bit more contrast in the shot which wil deepen the darker colours and to me it looks as if the blue channel has been boosted (I think in this shot both the luminocity has been turned down and some saturation added as the sky is begining to get that unnatural look).

 

[KmH]

Unfortunately, you get what you pay for with GIMP.
But, what release of GIMP do you use? The latest - 2.10.0?

 

[DennyN]

Newbie question.....
What accounts for the amazing vivid colors I see with photos taken by folks that know what they are doing. My guess is there are many variables but how much of it is post processing and in what way ? My stuff always seems a little bland.
My stuff - Denny Noll

Any input would be appreciated.

Unfortunately, you get what you pay for with GIMP.
But, what release of GIMP do you use? The latest - 2.10.0?

Yes newest one 2.10

 

[DennyN]

Thanks for all the suggestions and info, very helpful..............

 

[Tim Tucker 2]

Newbie question.....
What accounts for the amazing vivid colors I see with photos taken by folks that know what they are doing. My guess is there are many variables but how much of it is post processing and in what way ? My stuff always seems a little bland.
My stuff - Denny Noll

Any input would be appreciated.

Reading through this thread I'm puzzled why nobody has suggested the obvious:

If you want to know how to edit colour then learn about colour.

Colour is not a property of light, colour is how the human eye differentiates between wavelengths. It is also worth noting that the human eye has not evolved to see in absolutes but to give us the best chance of survival. We see by recognising slight differences, in fact your eye is tuned to exaggerate these.

The upshot of this is that colour is perceptive, not absolute.

We live in a world of subtractive colour, we see a world of subtractive colour. We do not see in RGB and RGB does not explain colour or have anything to do with how editing programs handle colour. All RGB explains is how an additive system such as a computer screen displays colour and tricks the eye into seeing (almost) the full spectrum by using only red, green and blue.

It is important to realise the above, very important if you're to understand how to lift your images.

As we live and see in a world of subtractive colour, and colour is perceptive all editing programs are based on perceptive colour theory which is the standard subtractive colour model.

I will try to demonstrate.

There is no separation between colour and luminosity, this is just the construct of what I like to call the "narrative of absolutes", a logical framework many apply because they find it easier to understand. Colour has three properties in the standard colour theory, hue, saturation and luminosity. Contrast exists between all of them, it is not a property of luminosity alone. Colour (hue) affects luminosity and luminosity affects colour (hue).

Here is a yellow against blue, the yellow is slightly desaturated and at 100% luminosity and the blue is more saturated and at 68% luminosity. Note the apparent contrast:

Here is the same yellow against green. The green is exactly the same saturation and luminosity as the blue. Do you see more or less contrast between the two? Remember all I've changed is the hue:

So I can change the contrast just by changing the hue, leaving saturation and luminosity the same. It proves that there is a visual contrast between hues, or colours.

Here is the first image again but this time I've reversed the luminosity, the yellow is 68% luminosity and the blue is at 100%:

So where did the contrast go? Different colours have different visual luminosities and if you try to reverse this contrast and colour will disappear. As you see above compared to the first image, exactly the same hue and saturation but I've simply reversed the luminosity and look what happens to both the impression of colour and contrast. (NB. This is a similar effect to excessive tone mapping where the natural luminosity of colour is reversed. The impression of both colour and contrast is greatly reduced. It is also important to understand that all cameras and editing programs work on a perceptual model, that is to say that they convert or change the luminosity and saturation of colours differently in line with human perception. it is not an absolute model but one based on perceptive colour theory).

My choices of colours here are no accident, I deliberately chose colours that show the greatest variations in combinations and when reversed because I understand colour theory. Look at many tone-mapped images and you will see the colour more similar to the bottom image, look at images where the colour appears like the first one and you will see more vivid colour.

It is important to note that the saturation of the first and third images is exactly the same for the colours, it is not that which makes colours pop, it is quite simply:

Understanding colour.

Contrast changes saturation, it's as simple as that. In line with subtractive colour theory, adding white, (all the colours of the rainbow), de-saturates colour. This both changes the colour and your perception of it because you've altered the luminosity. What the editing programs do is to increase saturation in line with perception as you increase contrast. The idea is to make the colours appear as stable as they can. It is not that you perceive the colours to be more saturated because of the increase in contrast but the exact reverse, you see the colours as less saturated, (in the editing program. In line with perception and the real world increased contrast, i.e. a sunny day, produces a greater impression of colour. It is just difficult for a photo to show because the reduced contrast of a screen or print either pushes the colours towards white and black or reduces their contrast and therefore your reduces the impression of colour. Subdued colour can look better in print simply because it suits the contrast range of the paper better.), so the editing program makes a very real increase in saturation to compensate. This increase is in line with perception and is not linear but varies dependant upon hue, different colours behave differently perceptually as you see in the third image. Light yellow against deep blue is visual far different from a deep yellow against a bright blue. If you want to see a much lighter blue against a darker yellow here it is:

Now you would think, with a conventional "narrative of absolutes" that increasing the difference in luminosity between the two colours should increase the contrast. But the visual effect is the complete opposite. There is no getting away from the fact that we perceive yellow as a light colour and blue as a dark one, and no denying that in terms of colour and contrast the first image is far more vivid even though it has less difference in luminosity than the fourth image, (52%/100%).

If you want proof that editing programs edit colour based on the subtractive colour model the see what happens when I overlay the blue at 50% opacity over the bright yellow. If it works in line with an additive colour theory then the combination of the light blue and light yellow should produce a colour that is the combination of the light combined. Subtractive theory though, dictates that only the common wavelengths are reflected and all others as absorbed, so the result will be a darker grey. See for yourselves:

The editing program has not combined the light of the two colours emitting from the screen, instead it has done a rather complex calculation based on a perceptive model and worked out what colour you would expect to see with a real human eye in a subtractive model such as the real world and displayed that.

Vivid colour is not saturation, vivid colour is colour edited by someone who understands colour.

Though it is not essential to fully understand why, I am hoping that I've shown enough to allow you to question your understanding. I hope this helps. ;-)