One of the most common mistakes I see on workshops is students making lots of local changes, but ignoring the global changes. When they ask me how I would approach their photo, I typically get it right where they want it with basic global changes and a few local changes.
I’m convinced that making global changes first will have a huge impact on how you print your photographs. I’ve emphasized this in my latest video EXAMPLES – The Making of a Photograph: Mono Lake Sunrise.
In this photograph, it’s the global changes that do 95% of the work, with local changes that refine the photograph. It still needs local changes…but they are easy to make when the heavy lifting has been done by the global changes.
You’ll need to view it in HD to see the numbers in the info palette. Also, the color change I make at the end is very subtle, and it didn’t carry through in the video format, but it’s there in the real file.
Use the blog comments section to ask questions about this video and further the dialog on how and why I made the adjustments I did.
A very good video Rich. Its nice to see how others “see” and then apply the tools. Well done.
Just discovered your blog Rich, a very helpful place. Very nice to see others work flow and be able to learn from it. Nicely done, Thanks
I had a question as to how you determined just how far off the color imbalance, referring to the magenta correction, was? Did you just guestimate or do you have a process for determining the proper balance of color? I could see where you could over-or under-color balance, some color shifts are not easily seen. Also, nothing on the sharpening process? To really output a fine art print, what kind of sharpening processes are you using? I know that typically there is a capture sharpening and output sharpening process, but what about creative sharpening? How many passes do you usually find yourself doing within the capture sharpen and output sharpening stages, what is your approach? Thanks.
A. Larsen. Good Questions!
The correct color balance was first determined visually on a color accurate monitor. Then a proof is made that can be used for further refinement if necessary. It’s all by eye and experience since what I’m trying to “match” is an experience in my head and heart.
Seeing color shift and imbalance is a matter of practice. It’s a lot like learning to tune a instrument. The more you do it, the better you get, and the better sound you are able to make.
I haven’t covered sharpening yet. This video is the guts of how to make a Master File. Since it is a scan of film there is no initial sharpening. If it were a RAW digital camera file, there would be some sharpening in the RAW converter, and we’ll look at that when I do a video working with a digital camera file. Sharpening for print takes place at the “Targeting” stage in my workflow, where the Master File is “Targeted” to a specific use. A “Targeting” video is somewhere on my to-do list of videos.
Fantastic tutorial Rich. Thanks for taking the time to put that together.
Nicely done again Rich! Sorry it took me so long to look at this.
One question: you mentioned the values in the yellow band of sky being around 200 to 211 I think, and said that you wouldn’t want these to be up around 230 or 240 in the raw scan. This puzzled me a bit, because at 230 or 240, as long as none of the individual channels were blown out it seems like you’d have plenty of detail to work with, and could easily darken this area if desired. Maybe I’m just too used to working with Raw files now, where I’d rather get the exposure as light as possible without anything being washed out, knowing that trying to lighten the image overall – which is what you did here – would tend to bring out noise. I know that noise isn’t an issue with the Tango scanner, so as long as the shadows aren’t at 0, 0, 0 you’ve got detail if you need it, but I’m still wondering what the problem would be with highlight values in the 230 or 240 range.
Good Question Michael!
I’ll give you the 5 minute version on why I do it this way, and over dinner sometime we can dig into it deeper…
In the broadest strokes, you are seeing the difference between how film and digital sensors record highlights, and how you handle those differences in the processing software, and also a difference in preference on how to do things.
With the chrome in this example, the highlight values were not recorded hot, because that is not what gets the subtleties I want in highlights. So I don’t need to try and recover a hot value on scanner, I just want to record the film values in a way that gives me the flexibility I want in Photoshop. There is a range of what the highlight value can be for my desired results, and these values at 200-211 were about in middle of the range, but you don’t have to be perfectly precise. As long as you are in the range, you can move it around it Photoshop with good effect. (But too hot, as I mentioned in the video will not get the look I want.)
So this is the opposite of what you described, because with film I’m not facing a noise problem. I can pick the chrome with the highlight detail I want, and scan it so that it is of a value that lets me push it around to taste in Photoshop. As long as I have good highlight detail on the chrome, highlight recovery is not necessary.
If I did want to recover highlights on a scan from film, I would do it on scanner first as that is the better tool for it, so in that case you wouldn’t see me doing it in Photoshop in the example video. The scanning software is analogous to RAW converters and what you can do at that step.
Furthermore, If the highlight value is too hot, I don’t have as much flexibility as to how I shape the roundness of the highlights, or the local contrast and separation which happens through the s-curve. Digital tends to have a much sharper cut-off of highlight values, which makes it harder to achieve the the same result in digital. So I’m explaining in the video how I play to the strength of film, even though this particular scan doesn’t need the s-curve to achieve the look I want. Highlight flavor is really important to me, so I’m always working to achieve a previsualization I have in my mind of what they should be. Even with digital camera files, I like to do it this way so that I can get the highlight look I like. (I confirmed this with Terrance Reimer and Michael Jones, my Master Printmakers at West Coast Imaging, and they like to start with darker highlights on both film and digital capture as well.)
I find it interesting that Cinematographers are facing this same issue, and many prefer to shoot film as a result. Kodak even has a whole micro-site http://motion.kodak.com/US/en/motion/Products/Customer_Testimonials/index.htm devoted to this. The reason for most cinematographers to shoot digital is cost and convenience, but a large contingent still prefer film is for aesthetics. Not trying to say that film is better, but to say that even the folks that spend far more money on “photography” that us still photographers do see the difference in the capture methods.
Lastly, at West Coast Imaging we’ve tested this workflow on tens of thousands of images and we’ve found it to give us results we like best when working from film. I created this approach because doing it the inverse way (inverse to what you describe for digital capture) wasn’t producing the results I wanted.
And the great part is you don’t have to understand why it works to get it to work. If you scan like I describe and print like I show in the videos, these highlight qualities are a natural consequence of the workflow. Really my workflow is based on using tools that through trial and error I discovered gave me the results I like, and with highlights and curves, I’m just using basic curves as defined in sensitometery to craft the shoulder to the shape I want to produce the exact look I want the highlights.
Thanks for your detailed answer Rich. I guess I’d have to actually see the difference to know what you’re talking about. Since I never got both lighter and darker Tango scans of the same image, it’s hard for me to compare. Certainly I’d rather not have to “recover” blown-out highlights in my digital camera files, at least not if there’s color involved. If the red channel is pegged at 255 in a photo of yellow aspens, for example, then even after using the Recovery tool in Lightroom or Camera Raw I’d end up with areas that look splotchy, where the leaves are all the same color and lack the natural separation in tones and colors that you’d expect them to have. But I’d have no problem with seeing the brightest values in that red channel at 230, as that would give me plenty of room to work with, and natural-looking variations in tone and color. However I’ll do some experimenting and see if I can actually see the difference when the highlights are at, say 210 instead of 230.
So here’s another question, when you have time – no hurry here. Would you rather work with a scan where the brightest highlights are at 210, but shadows at 0, 0, 0, or with highlights at 230, and the darkest shadows at 10?
Happy Thanksgiving!
Michael,
Looking over your questions, I don’t really know how to answer them in a way that will make sense to you because I think we are cooking in different ways, and that makes us view the recipe differently. While I like the results from the way you cook, it’s not the way I cook and it does not produce strictly the same result. To be clear, I’m not saying either result is better or worse, or right or wrong, just different, in the same way two experienced chefs handle their ingredients differently. I think prints cooked with many different recipes can be enjoyable, but it’s only natural that I prefer my own cooking most of the time!
Highlight values are an important part of achieving the look I like, so I want to try to expand on the video in an attempt to give a better look into my decision-making process.
In your first question you said:
Here I think we have a different approach, so let me explain mine:
When I talked about highlight values in the scan, I was talking about a general approach to scanning.
Over a large sample of photographs, both my own and others that I’ve worked on at West Coast Imaging, the low 220s stick out in my head as numbers that consistently make it easy for me to achieve the results I like.
At these numbers, highlights just naturally fall into place and make a really good looking print, with the flavor I like.
It also has to do with this idea of making printing simple that I keep talking about. If my highlights start falling out of this range, I have to start fighting the file to create the look I want in the highlights. At the low 220s, my experience is that I’m not fighting the highlights, but working with them. In general, once they start getting brighter than that, the more I have to fight them. So when I say 220s, I’m not setting a hard ceiling. Higher numbers may work for me on an individual photograph, but on average, they make me have to fight it harder. At 240, in most cases, I wouldn’t have the headroom to work on the highlights with the pots and pans I like to cook with, so I don’t go there with a scan or with a file out of camera raw.
To be clear, it’s not about making compromises to make printing easier. These values help me achieve the look I like.
Now, this is only talking about values in the uncorrected scan or file that comes out of camera raw. Once I’m in Photoshop, I don’t care what the values end up at, because I’m looking at it visually and responding to what my eyes see. But I need enough information in the uncorrected scan or camera file to let me play in Photoshop. It’s like getting a block of wood that is the rough shape I need, but leaves me enough room to carve out my expression.
This is what works for me, and the way I like my highlights to look. If you prefer a different look, you’re probably using a different approach, just like Bobby Flay will season a Thanksgiving turkey differently than Paula Dean will, but both will still be enjoyable meals.
All I can teach is they way I do it, which produces the results you see in my prints, and is also the same look we seek to produce for our clients at West Coast imaging. I say very close because, while the printmakers and I approach highlights in a very similar fashion, we still each have our own way of cooking, as does every chef.
So in the video I was talking about my approach, in general. Now let’s apply that to the Mono Lake photo in the video.
In this case, the original piece of film has strong density, and the scan records this. When the scan was made, the values were not being depressed to achieve their final value. Instead it’s just the value that was rendered by the setting used, which I suspect was making no correction to the overall density, or maybe lightening it a small amount. Regardless, it gave the values as shown in the video. Those values followed my approach of keeping enough headroom in the highlights so I can apply my working style to the image. When I actually worked on it, it turns out that I didn’t really need that headroom at all for this specific photograph. But that is of no consequence, because scanning it brighter would not have given me a better result. The scan, as it was, let me achieve the result I wanted even though this particular photo called for a different curve adjustment than my typical photos do. This is what I want from a scan…I want it to let me be free to create and not be limited.
I think it might also be helpful to explain what I mean by highlights, as that is a qualitative term.
So I’m not misunderstood, let me briefly state what I DON’T mean. I don’t mean a white-point. I know most Photoshop books teach that the first thing you should do to a file is to set a white-point, usually by looking at the histogram and using the levels tool. I don’t use that workflow because I don’t like the look it produces, which is typically highlights with no detail and no texture. I used to use it long ago, and gave it up once I realized it was preventing me from getting the look I wanted. I don’t worry about white-points, because according to the zone system scale, white or Zone X are paper-based whites, and are typically only specular glare or light sources within photos. Those do a fine job of being white all by themselves! Furthermore, I so rarely use levels that I can almost say I never use it, and I don’t look at histograms while printing. I also don’t color balance by white-points or neutral values.
When I talk about highlights, I’m talking about Zone VII through Zone IX as defined in Ansel Adam’s classic “The Print” in the 1989 printing. And these 220 RGB values are what I’m trying to render Zone IX at when scanning. I need to do that so when I work in Photoshop, I can use a curve shape that not only prints the Zone IX values at the proper print density, but also allows the texture and feel to fall into place correctly for Zone VII and VIII. Those zones are heavily influenced by the shape of the shoulder section of the density curve, and I want to be able to have a lot of control over how they look.
Washed out highlights in prints is one of the most common mistakes I see when viewing prints. This is true of both digital and analog prints. I find that the taste of the highlights sets the mood of the whole photograph for me, so how I choose to place them is very important to me.
Now that is a slightly more expanded explanation of what I was talking about in the video, but the good thing is that you don’t have to understand a word of this explanation to put it into practice. It’s exactly like how you can describe how a baseball is thrown with physics, but you don’t need to know the physics to learn to throw a great pitch. If you just follow what I show in the videos, things should work out with little effort.
So with that, I’ll try and answer your latest post.
That may or may not work. It’s not something that can be measured for the same reasons a group of people can go to the same restaurant and all have a different opinion of how it tastes. It’s opinion. You may well try it and decide you don’t like the result, but that is just your preference.
As an experienced photographer, you have your own preferences and biases you’ll be bringing to the tasting, and I wouldn’t expect it to be any different. I’m not writing this blog to change the preferences of photographers who have settled on a workflow that produces results they like. I think there is room for all of our preferences in printmaking.
Instead, I’m writing it for the photographers who I constantly encounter who are still learning the process, and who I know can quickly gain a much higher degree of control and expression over the process with tools that are really simple to use.
Moving on, you ask:
I’d follow the approach outlined in this question. In my experience, 10, 10, 10 shadows are just as detail-less as a 0,0,0 shadow, so there is no difference between the two for me, and no reason to prefer one over the other.
And to put this in context for everyone, you should know that I’ve known Michael Frye since 1995 when he was my neighbor in Yosemite Valley, and I worked for his wife at the Ansel Adams Gallery. I have the highest respect for him and his photography, and I consider him to be a good friend. I think he makes beautiful photographs, and has a unique vision and signature that is all his own, and I very much enjoy viewing his work. I recommend you check out his website at http://www.MichaelFrye.com, as well as his books.
What you are reading with our questions and answers back-and-forth is part of a conversation between two friends…me trying to answer his questions, and also me trying to put it all in context for the rest of you, so I can achieve my purpose in this blog: Helping others unlock their potential to express themselves through photography.
So that’s another layer peeled off the onion (so to speak) of how I approach printmaking. If it intrigues you, great! Add me to your RSS feed and keep following along!
Wow – I really didn’t intend to make you work so much! Thanks for the detailed reply. I have to digest this a bit. I understand the part about having more leeway to shape the upper part of the curve if the highlight values are lower. As I said, I’ll have to work with some different images with an eye on this question. I always have to see a like this difference for myself – as I know you do too! I completely agree with you about white points. I rarely want anything to be pure white in my photographs, especially color photographs. And thanks for the kind words about my work. I really like your work too – I’d just like to see more of it!
No worries about making work. Answering made me articulate things I’ve never had to commit to paper. This is still the 5 minute version BTW
Interesting video and thanks for blogging about it. I was wondering how did you save the “Red 2″ as a separate selection in the Hue/Saturation layer?
Great tutorial, thank you so much for sharing your information, I for one really appreciate it. Gorgeous image too.
Rich,
Great information and tutorials!! I really appreciate you sharing the behind the scenes of your workflow and the decisions that you made each step of the way. That has given me some ideas to incorporate!
Thanks again for sharing.
John
(found your posting via Michael Frye’s blog post)
Seth, go to hue/saturation, use the pulldown to choose “blue” (or any color other than red), then go click on a red in a photograph. You’ll see that the name changes to “red 2”, and that is how your selection is saved. You can use this for any color too, not just red.
Steven, check out my Color Range tutorial Part 1 and Part 2. You can select any color using the same method shown in those videos and in the EXAMPLES video.
Thanks everyone for all the kind words, and I’m glad this has been useful to you.
I’d love to read more