Thanks for the feedback guys - so glad you like the site. PelleTm, the idea of a glossary is great, I've started work on one, and I'm also planning some additional diagrams. I'll post an alert here as I add them.
If any of you come across anything that seems confusing on the site, or needing more explanation, or just plain wrong, PLEASE feel free to post your questions here.
There is something I observed today and you may found this completely obvious but I got intrigued for a while. It was late afternoon, really dark and wet from the rain. There were some red lights from the cars that were reflected by the sidewalk. However the edges of these reflections were affected by dark blue light and went into purple while the core of the red light didn't "sink" into the sidewalk. I found photo which shows very similar thing.
So to me the centre light rays that straight into the eye are more dense, but on the edge there is more diffused light and lack of red is complimented with blue. Those two colours then mix in the eye and as a result I see purple . Is it good thinking?
Last edited by Farvus; February 27th, 2008 at 03:18 PM.
My guess is that the longer wavelengths are not absorbed as much at these angles from your eye and are therefore the ones you see 'haloing' the edges. But this is just a guess and probably just as wrong. The fact we see it in the photo means it is not a subjective physical effect produced by the retina at least.
From Gegarin's point of view
I have a question about Photoshop & light, but first- thanks a million for putting the site together. A truly huge undertaking both in the research for and the creation of the site. It's an amazing resource that I'm already referencing constantly to answer questions.
My question is maybe slightly off-topic, but I thought I'd post it anyway. I noticed on the page "Effects of Colored Illumination" that the "Multiply" blend mode in Photoshop suggests the effect of colored illumination. Do any of the other blend modes have similar suggestions of the physics of light?
Specifically- for me a sort of Photoshop holy grail would be a way to have a base layer with the local colors for a given object, with a layer on top with a single color for shadows, and another layer with a single color for diffuse light, such that the local colors on the base layer could be changed and the "shading" (light & dark) would remain consistent visually. In other words, you could change the local colors on the base layer, and the relative shift in dark & light produced by the two upper layers would remain the same regardless of the chroma and value of the colors on the base level. I understand this is somewhat dubious in terms of the physics, and might be impossible given what Photoshop has to offer, but it would be very useful for me in a lot of situations.
I've done a fair bit of experimenting to no avail. "Multiply" works acceptably well for the shadow layer as long as the shadow color is almost totally neutral- something I find undesirable as in some situations I like to shift the color of the shadows slightly (towards blue or brown usually). I've yet to find anything that works well for the light layer across a wide range of base local colors. "Screen" seems to come closest, but the relative shift in lightness is much greater for darker base colors, so that the base color can't be changed freely.
In general though, I'm also just curious how blending modes relate to the physics of light...
Without being there I can really only guess too, but if you had an uneven wet surface, then each reflection would be centred on the point where a smooth surface would be at just the right angle to bounce light from the light source to your eye. At that point, most microfacets on your uneven surface might reflect light from some point within the area of light source, but moving out from that point, fewer and fewer microfacets would do so, and more and more would reflect the dimmer light from the sky around the light source.
Wherever these two components were not visibly distinguishable, the resulting stimulus would be the result of additive mixing. I'm not quite sure from your description whether the sky was actually bluish, or was a dull grey that appeared bluish by contrast with the red lights. Mixing with bluish light could easily make a purple, but mixing with dim white light could also account for a more subtle shift towards pink.
The term "mixing in the eye" is widely used for situations like this, but may give the impression that the process is more subjective than it really is. What we are really talking about is mixing of physical stimuli that for one reason or another can not be distinguished separately. Many if not all of these effects can be photographed.
Your question came up as I was about to post this answer. Could you post an example of your best attempt so far, and maybe email me the psd file (djcbriggs at the dreaded gmail.com)? Glad you like the site!
Tim, I'm not totally sure if this is what you were after, but you can use Adjustments on the colour layers to vary both the colour of the object and the colour of the ambient illumination in any way you want.
The psd file is here:
briggsy@ashtons - Thanks for reply. I just wasn't entirely sure it it was something about light or that it didn't loose it's hue but mixed with surrounding colors in the eye. You explained it more thoroughly than me. The sky was rather cool gray but everything around appeared a blue like in late afternoon. I couldn't say right now if it was contrast to red lights or more general phenomenon.
Today I realised that color brilliance mentioned in Loomis book is something a bit different than saturation and it all suddenly makes sense with all these colors in the shadow. Going back to basics definately helps .
Blending modes in Photoshop
To answer dose's main question (as best I can at the moment), the three blending modes that I've found to more or less emulate effects of light are multiply, screen and normal.
In multiply mode the relative brightnesses of the R,G and B components of the two layers are multiplied together, e.g. if the relative brightnesses of the two components are both 0.5 (normalized to 1.0), the resulting brightness will be 0.25. Multiply mode accurately emulates the effects of subtractive mixing, including the effect of coloured illumination. The only proviso here is that if you employ fully saturated colours, you will be emulating light of greater colour purity than comes from most light sources, or is reflected/transmitted from ANY actual coloured surfaces or filters respectively, so you may get some unrealistic results.
In screen mode, the differences between each component and one are multiplied together and then subtracted from one, e.g. if the relative brightnesses of the two components are both 0.5, the resulting brightness will be 0.75. The results of mixing in screen mode often resemble those of additive mixing qualitatively, though generally not quantitatively. Because of the nonlinear response of our visual system to light energy, a light that has twice the energy of another light will look brighter by a factor of, not two, but between the cube root and the square root of two, i.e. around 1.37. However, adding two identical lights together in screen mode results in a light that is brighter by a factor of nearly two for dim lights, reducing progressively to one (i.e. no increase) for bright lights. I assume that some sort of flattening of response of this sort is inevitable, given that brightness has a finite range in RGB space, unlike its open-ended range in the real world. In any case, mixing of bright lights in screen mode can give quite different results to additive mixing. For example, yellow and magenta at maximum brightness mix to give pure white, not the reddish white that would result from additive mixing (and which you in fact get if you mix them at 50% brightness).
I used layers in screen mode to emulate additive mixing in many of the diagrams and interactive animations on the Dimensions of Colour site. For this animation demonstrating additive mixing, you can see the individual layers that I used to make it (against a black background) if you slide any two of the three sliders to the left:
Finally, layers in normal mode (at low opacity) can be used to emulate the effects of atmospheric fog or turbid water.
I should add that I make no claims to be any sort of Photoshop expert - the little I know I've picked up by having to make the illustrations for the site. If anyone with some real experience has anything to add, please jump in and help out.
When I was first trying to find out about the various blending modes I found that a lot of what I read was not very specific on exactly (i.e. mathematically) what each blending mode does. I found this link to be helpful:
I grabbed a bit of time over the Easter break to reorganize a few pages (although the re- part could be a euphemism), fix some typos (and add others), and add a few more diagrams (on CMY and CMYK, and demonstrating advancing and receding colours, below).
By the way, dose says the stuff I just posted was a big help but he's a bit too busy to drop by and say thank you.
Here's another new diagram explaining an efficient way of painting a complex multicoloured surface turning out of a light source. Explanation here.
Hey Dose, glad you made it back! I'm not looking for a prize, but it would be great to get a bit more in the way of specific feedback, especially on my stuff about digital painting/Photoshop. Otherwise, since I'm entirely self-trained and relatively inexperienced at the latter, it's hard for me to know if what I'm posting on those subjects is actually helpful, or just common knowledge.
David solved some issues of chroma for me. The chart of chroma reduction is an excellent concept. Thank you.
O...M...G... I'm guessing there aren't too many people walking around who can claim to have solved some issues of chroma for Graydon Parrish. Thank you! (Looks like I won my prize after all...)
Did you get my PM in response to yours? I wrote a bit of feedback in there. I can post it here if you think it would be valuable to the discussion.
I've been giving some thought lately to the placement of the specular highlight. I was taught correctly that the highlight is not at the point of the form closest to the light (which, as you say, is often taught incorrectly). However, I was also taught to largely ignore specular highlights, as for the most part they undermine the illusion of form on a 2D surface. There are, however, times when it useful to put them in when it enhances the feeling of form or conveys critical info about a surface's material. Any thoughts on this?
On the question of whether or not to show the highlight, I think it all depends on whether we're talking about drawing or painting, and if the former, what kind of drawing. A specular highlight won't confuse the form as long as it reads clearly as a highlight. The thing is, though, if you are drawing in pencil or pen on white paper, it takes a lot of rendering to satisfy this condition. This effort may be superfluous if the drawing is what we call a form drawing - one in which the focus is more on showing the 3D form of the subject than its appearance in a particular lighting and atmosphere. If on the other hand we are talking about either a tonal painting or a highly rendered tonal drawing, in which the focus is on the visual appearance of our subject, then I think we would normally want highlights. Not only are they a major component of the visual appearance but, as you say, they convey information about the different surface characteristics of the materials represented (By the way, I think that this distinction also applies in general to the question of whether to show or to suppress other complications like variations in local colour, multiple shadows, or multiple secondary light directions within the shadow zone).
In your PM you talked about how you had been establishing local colour with an underlying layer and shading this with an overlying dark layer in multiply mode, whereas I had represented the pattern of light generated by each light source by a layer in screen mode, and then applied the local colour using an overlying layer in multiply mode. I think the latter method makes it at least easier to represent the additive combination of multiple light sources (in fact I would really have to sit down and think to see if that is even possible the other way). But I think that the real trouble you were having, if I understand you correctly, is that you were trying to change the colour of the darkness, instead of the colour of the light, in the shadow zone.
re: specular highlights
Ah, good. That's kind of what I was leaning towards- that there's sort of two "modes" of drawing- one to build/convey form and one to convey surface information, and that in the latter the highlight needs enough context to make sense. I hadn't had the chance to test it out for myself- especially because these days I rarely have time for any detailed rendering. I'll still have to do that to get the ins and outs of it, but your response leads me to believe I'm on the right track.
This also helps me better explain a big difference between the French impressionist technique and the Russian impressionist technique I was taught. For the most part, French impressionists seemed to be mostly concerned with the tonal effects of the light, while the Russians borrowed some of the French impressionist techniques but retained more of an emphasis on form. I had understood this difference before, but it gives me better insight as to why it may have happened.
re: Photoshop layers
Yes, I had been going about it backwards. The method you presented is much more flexible- and honestly one that I had heard & used before and is more common. It's closer to the idea of a grisaille plus glazing. Somehow I'd gotten turned around...
All of your color work
is very impressive
Would be great
to pop down
for a visit and
do a workshop
at your digs.
David, thanks for the site. A wealth of knowledge.
I would just like to give some feedback if I may, after reading through the entire site over a few days. I am a beginner to colour theory, so even terms like 'hue' are new to me.
I found, particularly at the beginning, some of the language to be inaccessible. I tried to unravel what you were saying a few times but eventually learned to just read on hoping it would become clearer, and it did. For example, despite the excellent Hue-Chroma-Lightness diagram fig 1.5, I did not fully understand what hue was until the Hue-Chroma-Lightness section and this sentence: 'A colour wheel represents chroma on a radial axis from the centre, and hue by position around the wheel...'. For my own understanding I translated the hue definition to 'On Fig 1.5 a single hue is the sliver of colour under the chroma line, and the level of chroma in that hue is a point on that line. The single line hue circle as depicted in fig 1.5 represents a range of hues having the same chroma.'
I can't say I understood much of the scientific stuff in the trichromacy and opponency section but more than likely I just need to read over it again.
I think you did a good job adding as many visual aids as possible, but I still would liked to have seen less text and more painting or photographic examples (more than one wouldn't hurt) to describe the phenomenon you were discussing, you are dealing with predominantly visual thinkers afterall. In particular I was disappointed by your explanation of your principles. Principles 1 and 2 I understood, and even did some experiments in photoshop to confirm, but I thought 5 to 8 were too briefly explained.
Additionally I found myself many times wishing there were some photoshop exercises I could follow along with to confirm each section of the site. Perhaps a simple sphere painting exercise in Photoshop or Painter for a lot of the sections could be devised where the reader had to translate what is learnt into a correct colour selection for any given situation?
There is certainly more to praise than criticize however. I did get a lot out of the site. As a beginner I don't think I could have a better start in my understanding of colour theory. I feel I now I have a rock solid foundation of understanding that will serve me well for ever more, so thanks again.
With regard to warm and cool colours, interior decorators use warm colours to make a large room seem smaller (make it more cosy), and they use cool colours to make small rooms seem bigger. In effect putting to use in a practical way the advancing tendency of warm colours and the receding tendency of cool colours. I find thinking of colours in this context helps me to classify whether a colour is warm or cool. For example, I once mistakenly thought a dark brown coloured rock with a blueish tinge that was sitting by a flowing river, all slicked over with water, was a cool colour, but it is not despite its context. All I had to do was imagine this colour painted on the wall of a room to realize it would make the room seem more 'cosy'. The same thinking can be used to clarify purple as a warm colour. Yellow is the colour that doesn't quite work when thinking like this however - paint a wall yellow in a room does it make the room more cosy? Perhaps its lightness and closeness to white makes it difficult to judge? There is no doubt warm and cool associations can be ambiguous.
Lastly, I submit the typos I noticed:
051 - In programs such as Painter that cleverly [simulate] the appearance and physical behaviour...
063 - Figure 6.11. Effect of adding white to opaque pigments. Unlike transparent pigments, addition of white to an opaque pigment like Cadmium Red Deep does not increase the chroma, but steadily [reduces] it.
074 - In digital painting programmes, the usual conceptual fr[a]mework for hue is the...
075 - Quinacridone Magenta, Pthalocyanine Blue (Green Shade), and any bright Lemon Yellow mix [is] about the greatest gamut...
077 - Since "warm" and "cool" are psychological associations, it is not [surprising] there is a great deal of inconsistency in usage.
082 - Figure 8.4. Chroma and colorfulness. A surface of a given chroma is more "colorful" in higher illumination (B,D) than in low illumination (A,C). [Tonal] painters would observe this difference in...
083 - Screen (RGB[)] colours have a more complex...
083 - Screen (RGB colours have a more complex geometry for a largely different reason. Among the full-chroma RGB (screen) colours... [You interchange 'screen (RGB)' and 'RGB (screen) - not consistent.]
083 - These differences [can] be accounted for by the fact that in the secondary colours, pixels of both the adjacent primaries are glowing.
083 - These differences be accounted for [because] in the secondary colours, pixels of both the adjacent primaries are glowing. [Pet hate, the use of 'the fact that' - up to you, just thought I'd highlight. Refer Strunk&White The Elements of Style pg24 ]
101 - My advice is that anyone who doesn't understand colour should not use black, but now that you've got this far you're certainly ready [to] join the black-using party...
From nothing everyday, and teach
The morning stars to sing'
-William Butler Yeats
That's a huge compliment, bone doctor!!!! It sure would be great to cross paths with you sometime soon.
Thank you very much, that's exactly the sort of specific feedback I've been hoping for.
I completely agree that the section on principles needs more content - I've already said on a few occasions elsewhere that that section is still very much a work in progress. I've recently been making various additions and corrections in the earlier sections, and I hope to get onto that section over the next few months when I have time.
I hadn't really thought of the site as an introduction to colour theory for complete beginners, so it's really good to know that with some improvements it might serve well as such. Some accompanying exercises in Photoshop is a great idea! The site in its present state is very much a reaction to the predominant confusion and disinformation about colour circulating in print and on the internet. I guess I had in mind a target audience of somewhat more advanced students, practitioners, and teachers who were already familiar with these myths and legends. I structured things to try to get everything straight for that audience from the ground up, but it looks as if this layout is giving the impression that the site is pitched at a more introductory level than it really is at present. You've done extremely well to go from learning the meaning of the word "hue" to finishing reading the site in a few days!
Thanks very much also for the help with the typos, I'll deal with them in the morning. (Glad you didn't see the site a month ago, though!)
Last edited by briggsy@ashtons; April 16th, 2008 at 10:37 PM.
I've been reading through your site and there's so much information. Most of it is out of my league and I have no idea what you are talking about mainly because I am a beginner at color theory, english isn't my first language and there are so many words which are very particular to color theory(case sensitive) and are new to me. So I have more reading to do and going through the vocabulary.
But I think what would help like some people have said is adding more demonstrative pictures. The more the better. It's easier to understand hard case sensitive words if I can see it in a picture with the context if you know what I mean.
Also one feedback about the site design. I would very much encourage you to change the brown background texture behind the text into a single color non pattern background. The current background really entires ones eyes and especially since there is so much to read. This would be my biggest concern.
I can't express how thankful I am that you've seen the effort and time used to create the site and share it with everyone. Hope my feedback helps and keep at it.
Havent read it all yet but i just wanted to stop by and say: Woow!
Big massive Thank you!!
Just downloaded the Colorspace program after seeing you mention it on an old thread. Seems like a really useful program and I look forward to playing around with it.
I was wondering if you could explain why you use the YCbCr space to plot the color space diagrams here and on your color site...
First off, you've put together a really nice resource. I'm sure a lot of hard work went into it and it shows. The presentation/color scheme of the site is simple and effective.
You stated earlier that the site wasn't necessarily meant to be a resource for beginners. As a beginner, I think I understand why. A lot of my problem is even if I understand the concepts, I don't have anything to apply them to. I have so little knowledge of basic color theory. Which brings me to my question: What are some good resources for beginners to color?
I've searched around some on the internet and CA for anything, and it looks like there are one hundred and one different ways to approach color. The sheer size of the subject and opinions on it have in many ways baffled me. I have no idea where to start, so I decided to bring the question to you, someone who's opinion and knowledge of color I respect.
Any advice would be appreciated greatly. Thanks for making such a valuable resource available to everyone. (hey, I may not be able to apply it to my art yet, but I know value when I see it. )
Thanks a million...Gr8 help for beginners like me.
I dont understand but noooothing at all... thought I saved it on the computer knowing it GOT potential not yet discovered by my totally confused brain, I think a lil theory (written) could accompany this graphic since I'm asking so much questions about color theory!
Thanks for the comments everyone and sorry I haven't been back to this thread for a while!
Dose, that's a very good question and one that I'm going to have to explain better on the site. For now the short answer is that YCbCr is quite good as a representation of hue-value-chroma: distances outwards from the Y axis correspond very closely with Munsell chroma for a given hue, while Y corresponds fairly closely though not exactly with lightness. I'll add some diagrams on the site to show this more clearly.
To everyone who is finding it hard to know where to start, the really essential thing is that you get used to thinking of colours in terms of three-dimensional spaces. That means hue, value and chroma when thinking of colours of surfaces, but also hue, brightness and saturation when thinking in terms of colours of light. The former dimensions and their importance are relatively widely known among painters, but currently even the best available resources generally miss the significance and even the existence of the different set of dimensions for colours of light. The site is really an attempt to clear this and a few other matters up for people who, like many here, are already familiar with these available resources. I'm still planning additions to make some things easier to understand, but a resouce written for complete beginners would be a quite different site.
In my courses I do take colour from absolute basics for complete beginners, and one day I might put this stuff together, perhaps as a book. But in the meantime, to get a grasp of hue, value and chroma I very highly recommend the exercises and text in the The New Munsell Student Color Set. Once you've worked your way through that, I doubt if my site will still hold any terrors for you.
Having said all that though, I'd be happy to try setting and commenting on some exercises in thinking about colour on this thread if anyone is interested.
I figured it was something like that. I'll keep an eye out for those diagrams on the site- would love to see more specifics. In the meantime I'll read up on YCbCr.For now the short answer is that YCbCr is quite good as a representation of hue-value-chroma
This is something I've sort of intuitively started doing over the years as I've learned about color, but reading that sentence really crystallized something for me.the really essential thing is that you get used to thinking of colours in terms of three-dimensional spaces.