(I’m going to put up a series of posts about the artistic aspects of how I take photos. This is not meant to be a definitive work and certainly not the ‘last word’ on creating images, just my point of view. As usual.)
I was going to title this “Framing and Composition”, but even though the two are intertwined they are somewhat separate things. Composition is the more all-encompassing, so Framing could be considered a subset of it. But as the song says “you can’t have one without the other”.
I think of the two like this: Framing is what is in the picture, Composition is how it is in the picture.
So like a bad stereotype of a movie director, we first have to consider the framing. And framing includes not only the initial image aspect but any changes made to it post-shoot with cropping. We are fortunate that digital photography easily frees us from the trap of fixed aspect ratios and standardized formats first inflicted on photography by the wet plate process. If you’re still stuck in the “8×10” mindset, you need to get away from that.
(Apologies that the photos are rather dull and I did not put much effort into the technical aspect as that isn’t important to demonstrate the principals discussed. They are in monochrome to avoid the distraction that colour sometimes causes.)
Very strange picture of a toy truck and some chess pieces.
That images doesn’t make much sense, does it? Or if it does it’s on some level of consciousness (or perhaps unconsciousness) I have yet to achieve. No matter. Through the power of cropping we can change what it looks like.
Voilà! It’s a picture of a toy truck.
This makes more sense, or at least is less incongruous than the original. The unnecessary and distracting chess pieces are gone and we can concentrate on the intended subject. We know it is the intended subject now because it’s the only recognizable item in the frame. Framing!
If we go the other way …
Unless this part of the original was the intended subject, in which case different framing achieves the desired result. (Notice how the knight and rook are a bit fuzzier for being just a little bit further forward? Not much depth of field in extreme close-up shots!)
If you really want to concentrate on the truck …
Squared format, fill the frame (even though the subject is basically rectangular).
This all seems ridiculously basic, doesn’t it? Well along the way here I’ve given some hints of things to come. One final clue for next time:
There are more things in Heaven and Earth, Horatio, than are dreamt of in your photography!
I see a lot of posts on various platforms discussing which digital cameras are best for “that film look”. Immediately Fujifilm’s offerings come to mind, as some have film simulation recipes built-in with the additional ability to enter and save your own. To some degree you can make similar adjustments to most digital cameras, at least the ‘higher-end’ ones. But there is a question which no one seems to be asking, and that is this:
Just what do we mean by “that film look”?
It’s about the final image, not how you achieved it.
You see if you have a Fujifilm X-series and you want a recipe for Kodachrome 25 you’ll probably find not the recipe for it, but a half-dozen different ones. That’s because ‘what film looks like’, even (or especially) a specific film, is subjective. When there was no choice but film, you could use two rolls of the same type and get different results from each. This is because in addition to the variables of your picture-taking there were inevitably further variations due to different emulsion batches and processing. So even though you shot the ‘same’ film with the same camera at the same time and had it developed by the same lab – the results might be noticeably not the same. Now add to that the vagaries of personal optical interpretation (and sometimes memory), not to mention differences in digital displays, and you have a recipe not so much for perfect film-like pictures every time but rather something that falls just short of being photographic chaos. Frankly it’s a wonder we get a recognizable image at all. (Honestly it’s really easy to mess a shot up so that you don’t get anything usable from it.)
Don’t believe me? Do a search for colour discerning tests. Those are the ones you see promoted with titles like “Can You See The Numbers?” or “What Colour Is The Dress?” and such. Even discounting the fairly common blue-green colour blindness in its many degrees, we don’t all see alike. There really is no objective standard for what a film looks like; it’s more art than science. We are all about art, but … it’s time to dive into the science. Just a little bit.
In order to determine what is meant by “film look” we have to examine the similarities and differences between film and digital imaging.
Digital images depend on pixels being ‘on’ or ‘off’. Arguably film is also digital in that it depends on grains of silver being turned black or not depending on how much light hits it. Okay there’s shades of grey in between, but the higher the contrast the fewer intermediate tones. By the time you hit lithography it’s 1’s and 0’s just like digital. Even before that, much of the grey scale is implied rather than absolute as it is based on how many black or white dots are seen in a tiny area which our eyes observe as contiguous. This is the basis for half-tone printing, btw, and is not unrelated to resolution (which varies with eyesight as well as with camera lens quality. TRUST ME ON THIS! *LOL*)
When we add colour in to the mix we get further similarities between the two media. Film uses dye layers to separate red, green, and blue tones (and then re-blends them in viewing) and digital uses a filter (usually Bayer) on the sensor to separate the pixels into individual colour receptors. The number and combination of them in any given area determines what colour is there, as per the grey scale analogy above. In viewing the reverse happens; a screen is red, green, and blue LEDs near each other and turned on or off to make an area appear some shade that is a composite of the base colours (you can examine this with a magnifying glass). One of the flaws of this system is that the colours are separated side-by-side with digital (except for with Foveon sensors), but layered on top of one another with film. This affects digital resolution, and is the reason why I have a camera that takes terrible colour images but does okay in black & white (try it on your own camera and see if you notice a difference).
Now back to the subject at hand.
Okay, both media have an ISO rating. Film is fixed, digital is variable from shot to shot. Both have contrast, which tends to go up with ISO. Film has grain, which also tends to go up with ISO, whereas digital arguably has both grain and ‘noise’. The difference? Noise is random colour dots which do not fit with the over-all colour tone of the picture area in question appearing as, say, a bright red speck in the midst of an otherwise greenish field. In monochrome it’s not so noticeable, showing merely as additional grain (which in digital is separate ‘dots’ of the area in a like tone).
With colour we have more factors: Both types of images will have colour temperature. What is that? In digital it’s called “white balance”. In film it’s a misleading statement like “daylight” which means some degree of Kelvin that the manufacturer of the film thinks is daylight wherein white appears as, well, white and not yellow or cyan. For that reason some “daylight” film is a little warmer, and some a little cooler. There is also colour saturation, which is how intense the colour is, and colour bias; the tendency to favour one base shade over another. An example of this would be Kodachrome 64 vs. Ektachrome 64: the first is warmer, more saturated, and favours red whereas the second is cooler, less saturated, and favours blue (cyan). Or at least that’s how it seems to me; your interpretation may be different. Digital sensor will have similar characteristics based on the the filter, sensor, and not least of all processing engine.
When it comes to digital film simulation recipes there are, if your camera has them, settings that adjust most of these factors and perhaps even the dynamic range (how well the camera handles shadows and highlights); the fine-tuning of these characteristics can vary from one model to another. What changes you can make, and how effective they seem to be to you, are individualized. For example you might have sharpness, grain, and clarity all playing a part in simulating the ISO/contrast/grain aspect of film. I have noticed some of these ‘recipes’ cheat and let ISO and/or white balance go automatic. To me fixing those is the first step to getting ‘the film experience’, if not the actual results. Your opinion may vary. One thing I would suggest is to start out by changing nothing, then take the same shot over and over making one adjustment at a time (and be sure it’s enough of a change to be noticeable; +/- 1 when there’s a potential variance of 100 won’t show much) and see if it gives you an image you like more – or less. One of the best parts of digital is that you can shoot a lot of pictures and delete them if they’re not up to par; you’re out only your time. In the meantime you’ll get a better understanding of your camera and how to get the results you want out of it.
To my way of thinking the cameras that do the best film-like rendering ‘right out of the box’ so to speak are lower resolution, CCD sensor ones. These are invariably older as well, as that type of sensor is no longer used as far as I know, and the industry has become obsessed with high pixel numbers and outrageous ISO values. I have four working cameras that meet the criteria; the Fujifilm Finepix F80 EXR, the Canon Powershot G11, the Pentax K100D Super, and the Kodak Easyshare P850. In addition to those, the Canon 1Ds and Olympus E-410 both do a fairly good job of producing pictures that look like film, despite having CMOS sensors (both are ‘lower’ resolution by today’s standards). On the whole, I wouldn’t say any of the other CMOS cameras I have do it well at this, regardless of how you change the camera settings. I think the reason may be that the CCD sensors have a smoother tonal gradation than the CMOS ones, allowing for more subtle differences in colour across the spectrum. I have no way of testing or demonstrating this though, and so I could be completely wrong.
Fuji F80Canon G11Kodak P850
“Film-like” images from three different cameras, all unprocessed except for sizing.
I came across one photographer who described Fujifilm’s in-camera film simulations as “cute” but unnecessary, as he could make any such adjustments post-shooting as he saw fit. I understand what he means, although selecting a particular film style before using can force a certain way of looking at a scene on you and thus can be useful at least for learning artistic application. In any case it’s clear that when we talk about digital “film simulation” we are discussing a characteristic that is subjective and mainly a matter of personal taste.
Disclaimer: This is an opinion piece, not the final word on film simulation for digital photography! It took me a long time to write, and as such I kept losing the thread of what I was trying to say. As such it may still be highly inaccurate even as a matter of opinion. One of the failings of ageing, I’m afraid.
I don’t even know if this will get published, but it’s a rant (a little tongue-in-cheek too) anyway so maybe it’s just as well.
Three blogs I read this week caused me to grumble. Let’s take the mildest one first.
It’s about film recipes. There’s this person that does them and does them well, except that they all kind of look alike when you get right down to it. Proudly promoted as “Kodacolor” or whatever, they’re all low-saturated, cool-toned, and bluish. I guess no one but me remembers Kodak’s standard of rich, warm colours meant to please the average consumer. Most of these digital recipes look like they should have the “Ekta-” prefix, not the “Koda-” one. If you don’t know the difference, that’s part of the problem. In later years Kodak literally toned down their colour experience because too many people couldn’t remember Uncle Bob spending so much time in the sun that he looked like he was part lobster. Whatever. Save it for post processing is my advice. The same with the cyan prints, which can easily be done that way. Oops! Did I just give something away?
Second complaint: someone said forget about getting a Pentax Spotmatic for a first film SLR. SERIOUSLY? Must be the STUPIDEST photographer on Earth! You want fantastic film photo results, get that Spotmatic with its incredible Takumar lenses – if you can. The idiot was recommending only newer, electronic film cameras. Yeah, right. Good luck finding one that still works or doesn’t break down two frames after you start snapping away. They weren’t dependable when new. You’ll learn more about photography with the manual camera, people. A lot more.
Complaint number three is rooted in someone once again declaring all cameras save pro DSLRs and smartphones are dead. Kind of misses the fact smartphones suck six ways to Sunday. Go ahead; change my mind. I’m waiting. I know lots of people who have and use them and think they’re fine. I should video their performances as they constantly swear at the things and repush touch-spots on the tiny screen trying to get the damn device to do what they want. Yeah I had that with the Lumix and it’s my #2 complaint about the camera. Frankly anyone managing to get a smartphone to do anything right is just plane lucky. The things are pure techno-trash. Undependable, unreliable, and unimpressive.
You can see I’m in a good mood. No, really I am. I’m actually having fun with my good cameras and have a few quite remarkable shots coming up to share. In the meantime here’s a “bleach-bypass” version of a picture taken with the cursed ZS60*:
Better times ahead!
*Footnote: the Panasonic Lumix ZS60 actually seems to work better in 3:2 ratio rather than 4:3. I don’t know why, and it’s not a major improvement.
Sometimes you take a picture that just isn’t very good. This is an example of such an image:
It’s hard to track a flying bird at full zoom in a cloudy sky. The focus is off, the exposure is off, and the composition off. This is not a good picture. But can anything be done to save it?
Maybe.
I cranked up the exposure, increased the contrast, desaturated to B&W, inverted the colours, and recropped the framing. Voilà! A “MOMA quality” image that looks like it might have been made in a completely different medium such as paper or even cloth. Dramatic lithographic tones and severely asymmetrical composition turns a dull, blurry image of an unrecognizable bird against and indistinct gray sky into a signature piece. You see less artistic offerings on the market every day.
Limited edition prints available in the gift shop. *LOL*
There are a number of people creating film-simulation ‘recipes’ for digital cameras, with good reasons and results. Especially note the fine efforts of Ritchie Roesch of Fuji X Weekly. Fujifilm cameras are particularly set-up for this sort of thing, but other ‘better’ cameras have similar settings adjustments in them. My Lumix ZS60 in fact has some specific film simulation settings, albeit buried deep in the touch-screen menu system. I’ve done some such setting changes with my Canon as well; it is in fact defaulted to a Kodacolor simulation with rich, warm tones and slightly elevated contrast – because that’s how I like to shoot.
Film has four basic characteristics: speed (ISO 100-800 as not many cameras can truly go above that level), colour type (including white balance [temperature] and saturation), contrast (one of the simplest variables), and grain. Only grain presents difficulty for digital simulations unless you have a camera (like certain Fujifilm models) that has a specific setting for that. Otherwise you’ll just be varying the resolution, which is the “digital equivalent” (“grain” introduces a fine noise effect).
Now here’s the thing. One of the advantages of digital over film is that you’re not stuck shooting 12/24/36 exposure on all the same ‘film’ type. You can easily change from shot to shot, as you see fit. The film recipes can allow you some preset choices to go with, providing your particular camera makes is easy to keep the simulations easily to hand. My old Kodak P850 has three “user defined” selections right on the main function dial, and all I had to do was remember what I had them set for. It’s somewhat more difficult with the other cameras, as the custom settings are notoriously menu-accessed and therefor not the easiest to get at – or remember.
So failing memory and increasing laziness affects my own choices, and I resort to some pretty simple solutions. Namely not making the changes in the camera. I’m notorious for warning against shooting in B&W, and that’s not just because the shot might look better in colour. My experience with digital B&W is the out-of-camera results tend to be rubbish. Perhaps something like the Leica Monochrom can produce fantastic results (it had better for the price), but the average camera trying to assemble a black-and-white image from the RGB detection under a Bayer filter tends to come up short. More often than not I find the contrast lacking, and if I crank that up the dynamic range goes ‘poof!’ (or other humorous sound effect of your choice). Thus I shoot colour and desaturate if I think it will look better in B&W.
I have adopted a similar stance for other versions of film recipes. Basically, shoot what looks best to you as full colour coming out of the camera, and then adjust the individual picture on the computer if you think it will look better rendered a different way. Sounds weird coming from someone who also boasts about not doing a lot of post-processing, but for end results – which are all that matter – it’s a good way to avoid missing a shot that would look better under different settings. The other option is to shoot it over and over in all kinds of ways and then pick the best from a dozen images. Ergh. Who wants to do that? If you post-process you can also have pre-set film recipes and not fiddle around endlessly adjusting settings by tiny amounts and wondering if that’s ‘perfect’ or not. Film has latitude; let your digital images have it as well. Perfection is not required for art.
Examples time. Here is the infamous red shed shot with the Canon T100 set with my Kodacolor recipe:
“Kodacolor”
One picture, incorrectly exposed I admit, and now we will process it to look like different film types. First, how it would look with the settings “normal” for the camera. This is not a different picture, just different processing to look like the standard output from the camera.
“Normal”
This one is simply desaturated to black-and-white:
Monochrome
Now let’s look at some colour variations. Starting with the silvery-blue, low saturation appearance that some people like.
“Bleach bypass”
Or you can crank up the warm tones (red, yellow, and magenta):
“Warmed”
Or crank up the cool tones (blue, cyan, and green):
“Cooled”
Which changes you make and how extreme you make them is up to you. What the subject is will also determine what effect you use, of course.
I’m not saying this is the way everyone should do it or even the way you should do it all the time; it’s just presenting an option to trying to get the film simulation right in the camera. You may find it easier to do this way than to adjust camera settings, or you may not have the camera settings to adjust. Likewise what software you have can limit your post-shoot processing. I just use the very simple GIMP program, and my files are JPEGs not RAW. As such there are certain effects not available to me on the computer which are available in the camera (or with the camera in the case of using colour filters). But it works for my “professional snapshot” style of photographic art.
By the way, one of the most fun things I’ve done with my Canon T100 was simulate a Kodak Brownie 127 camera from the 1960s: Shooting with the Canon Brownie. That’s not just film simulation, but camera simulation and even photographer simulation!
Once again we are having not-good-for-picture-taking weather, with temperatures so cold the cameras try to reset or shut down completely. Really the best of them are only meant to operate down to -10°C and yesterday it was -18°C. So I’ve been inside thinking about shots and trying some experimental things which may or may not lead to more involved photo shoots. All taken with the Nikon P610.
First we have a picture of the sky where you get to guess which is the original and which were processed:
Next we have the antique inkwell. You get to pick which you like best. They all ‘work’, but I think the original colour version is the best.
Otherwise, I’ve been reading interesting blogs again including one from someone who actually has a new Fujifilm X-Pro3 – and has confirmed my opinion of it: So this happened. Great camera. And a clever bit from Eric L. Woods about being fiscally responsible – and still getting the camera you want (essentially).
Only a couple more posts for the year. Another silly one with silly pictures and a silly one with less silly pictures.
Always have a sense of humour, especially about yourself.
Some random words on things photographic because I haven’t got any ‘projects’ at the moment.
People talking about “chimping” – the practice of previewing damn near every shot you take – often say it’s something “old film photographers frown on”. Well, I am that guy.
It is, I suppose, another aspect of the modern digital photographer not learning photography the right way. *Old man voice – ’cause it’s the only one I’ve got* “In my day we couldn’t preview any shot! We had to wait ’til the film was developed!” Now in the digital age we can, and frankly it’s a godsend. You see, it’s not that you should never look to see if you just got the shot, it’s learning when to.
Some personal, anecdotal examples:
The Canon is my experimental camera. So far it’s done over 2000 shots and I don’t think there’s even one I haven’t previewed. Some of them I couldn’t see (night shots), but I had to look and see I couldn’t see them. Er, yes.
The Nikon has the wonderful ability to keep its LCD to itself, turned flat against the back of the camera protected and off. Yet it’s best for self-portraits because you can flip it out and turn it all the way around and see yourself staring back at yourself. This doesn’t count as ‘chimping’, really. The point is I hardly ever preview with the Nikon.
The Kodak P850 has its preview function turned off so it doesn’t waste battery. I did that when I was using it with the old battery to judge if it was worth buying a new one for. Haven’t turned it back on yet. Don’t miss it a bit.
The Kodak V1003 … ah, well. No point, really. Can hardly make out the screen to see what I’m shooting, never mind review it.
The displays are fine for showing you camera setting, although I’d prefer that info indicated by a line pointing at a number on the respective, dedicated control for the function. In all honesty, how much can you see of the image on your screen? They tend to be small and cramped and lacking in resolution. Can you really judge if you got the shot by ‘chimping’? Only in the most minimal of ways. You can see if the framing was good enough or there’s approximately the right exposure, but it isn’t going to look the same when you bring it up on a computer screen.
A cell phone photo is another matter. Since that likely is the final medium of display, you will see what you will be looking at. It’s kind of amazing that many of them have far better screen resolution than most computers, albeit considerably smaller in overall size.
Before you go previewing everything I would suggest you ask yourself one question: what am I looking for? If you think it’s “the final result” don’t bother pushing that button ’cause you’re not going to see it. If you honestly need to check that everything is in the frame or check focus and/or exposure (especially for manual shots) then do it. But don’t get obsessed. In fact I recommend you train yourself to take pictures without checking any of the views. Go through the equivalent of a full roll of film (24 exposures) without looking at any until you’re done. It will require some willpower, but it will make you a better, more confident photographer.
And now here’s my cat Hannibal (aka “Puff-Puff”) being warm and fuzzy:
Today I have a couple of “test shots” to make involving some different lens configurations on the Canon. No doubt I will preview them all. Also I want to test a couple of objet d’image shots for a future project. Those I won’t look at until they’re on the computer because that’s the only place where it will matter what they come out like.
Also, a certain company has put a camera I do not need but kind of like on sale at a significant discount and I might buy it. Trouble is, it’s not the really nice one I want which is much more expensive (6X the price). I probably wouldn’t use either enough to justify the expense, with or without ‘chimping’. I’m just bored and looking for some inspiration.
Everyone makes mistakes now and then. Usually if you flub a picture it’s no big deal. But sometimes you really wanted that shot and … you didn’t get it. With film photography you were pretty much stuck; there was only so much exposure latitude available for any given film. Slides were the worst, as a couple of stops either way meant it was unusable. Oh maybe you could copy it and take up some of the exposure error, but that meant you added a generation of imaging to the final and lost some resolution as a result. There was no way around it.
Digital is somewhat more forgiving. It doesn’t rely on Farmer’s Reducer or chromium intensifier to alter the image. It relies on computer technology, which in this case is somewhat more reliable and predictable. Plus, if it goes wrong you haven’t ruined anything; you just start over.
So let’s start with a severely underexposed image. This should have been shot at 1/125 @ f5.6. Instead it was purposefully shot at 1/1000 @ f16. I make that 6 stops under.
You can barely make out that it’s an image. Now let’s crank the brightness up 125 and the contrast up 125:
Not a perfect image, but it has gone from ‘unviewable’ to ‘minimally acceptable’. Here’s another way of doing it, by applying GIMP’s “auto white balance”:
Either of these could be worked on more to get better results. The question is, how long do you want to spend reworking one image? The answer is: how important is that one image?
And this is what it “should” look like:
Now what happens if your exposure mistake is in the other direction? Chances are you lose the picture. That’s because overexposed means the image information is not there; it’s just blared out to white. I tried 6 stops over for this experiment and got a field of white from which no image could be extracted. 4 stops over yielded much the same result. Here’s the worst I could get away with:
After much post-processing (Equalization, adjustment of hue, saturation, and lightness, plus white balance changes) this is what it yielded:
You can make it out, but it’s never going to hang in the Louvre. With more complex software and more hours of work it might be made better, but really you’re relying on the computer to fill in colour info that isn’t there based on its ability to extrapolate from what little is there. I doubt you’d ever manage to get it to this:
The moral of the story is: if you’re going to flub the exposure try to err on the side of under rather than over because that stands a better chance of recovery.
In case you were wondering (and you probably weren’t) this post grew out of my recent adventures with old manual lenses used on the Canon. I found it easier to ‘set and forget’ exposure and let the latitude handle variations in light. As demonstrated, erring on the side of under proved better than over. In general two stops either way would still produce usable results. In fact I have actually changed the settings for auto exposure to go under by ⅓ stop because I find the camera consistently overexposes when left to itself.
Wandering Words 