Photo Manipulation

Inland paddling
User avatar
Mark R
Posts: 24087
Joined: Thu Jan 10, 2002 6:17 pm
Location: Dorset
Contact:

Photo Manipulation

Post by Mark R » Tue May 09, 2006 11:26 pm

I've been reading through some of Galen Rowell's stuff, interesting if you excuse the extremely frequent use of the word 'I'.

Here he is on the subject of phony/ manipulated/ colour filtered pics (writing pre-digital and Photoshop)...

"The long-term dignity and viability of nature photography as art and a source of information about our planet is becoming as endangered as a rare species".

In other words, he says that manipulation is killing the credibility of photography and devaluing it.

Thoughts?
Mark Rainsley
FACEBOOK

User avatar
Tom_Laws
Posts: 8122
Joined: Mon Aug 09, 2004 8:37 am
Location: North Wales
Contact:

Post by Tom_Laws » Tue May 09, 2006 11:50 pm

Agree and Disagree.

Ill quote James F if I may...
When Turner painted his masterpieces of the Thames, did he bother to paint the dead dog floating about in the current? No, he did not. We can learn from Turner here. If there is something unwanted in the scene that, try as you might, you can't avoid including in the image, you may be able to erase it with Photoshop. Hey, sue me! I don't care.
Also some sequences look pretty cool.

I try and avoid having to "tidy up" or brighten/darken/tweak/fettle images if I can avoid it. The odd bit of cropping and straightening doesn't go amis mind you.

Tom

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 7:53 am

I've just bought Galen's book and still disagree with him on that particular issue, the old curmudgeon.

David McCraw

Post by David McCraw » Wed May 10, 2006 8:20 am

I'll go even further and say that I disagree completely.

All photography is the result of a vast chain of artistic decisions - 35mm or medium format? Selecting film (or white balance) to achieve the desired colour. Aperture and shutter speed, choice of lens... portrait or landscape (panoramic)?

All of these things can have a profound effect on the end result. I like James's quote about Turner - Galen would probably have manually removed a traffic cone in an otherwise awesome scene, similarly affecting its versimilitude. It that any different to removing it in Photoshop?

Is the use of filters and different film types to produce a desired colour / contrast any different from doing this in Photoshop? What about high ISO film for that grainy effect, or adding grain in Photoshop?

The only convincing argument against it comes from people who photograph using a fixed lens / aperture / shutter speed approximating the human field of view, and would leave the litter on the beach because they really are trying to portray reality.

I suspect most of us are going for the same end result as those who painted portaits to a 'romantic ideal' in centuries past.

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 9:22 am

I see Photoshop as the digital darkroom. With Film you would have sent your roll off to a processing house who'd process the film and send you the pictures back. For a professional photographer there can be a lot of tweaking at this point to get the photo how you want it.
With Photoshop you get to do all the tweaking and I personally see it as half the fun of taking the photos. From simply straightening out horizons to improving the contrast, shadows & highlights and more. My camera is not very contrasty in strong light so I have to tweak that to get the most out of my sunny pics.

I don’t think it’s becoming in endangered. I just think digital gives the average photographer the flexibility they could have in a darkroom without the hassle of playing with chemicals in the dark.

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 9:41 am

I took my darling out to the aquarium
Sea creatures stared at us and we stared back at them
My baby freaked when she peeked at that eight legged blob with a beak
She was too scared to speak
And I said, don’t be upset, it is only an octopus
Don’t bother it, and I am sure it won’t bother us
Please don’t be upset my darling, please don’t get upset


Is that a story about a trip to an aquarium, or about taking your girlfriend/wife/partner to a river and scaring the bejabbers out of her/him? Where's the puzzle in the literal?

Steve B
Posts: 5699
Joined: Wed Jan 16, 2002 2:36 pm
Location: Taunton, Somerset

Post by Steve B » Wed May 10, 2006 1:33 pm

I haven't read anything of Rowell's, but I know his name from the photography forums and I've just taken a look at his gallery at www.mountainlight.com to refresh my memory.

Is he saying his images are not manipulated? It depends on your definition I suppose. His particular combination of film/exposure/printing produces oversaturated colours which are not true representations of the world. He used tricks like extreme wide angle to give a distorted representation for dramatic effect. It's not that he didn't approve of manipulation, it's that he had decided how much / what kind of manipulation was ok for him and wished to impose that standard on everyone else.

To answer the question in more general terms, I agree with the others. The camera is a tool, to be used creatively. The things you do before taking the shot (composing the scene, tidying up that litter, applying make-up, asking someone to look at the camera, or not) are valid parts of the creative process. The shot itself obviously is. And the work you do on the recorded image, whether in the darkroom, on the bench in an art class, or in Photoshop, is all equally valid.

Obviously there is a difference between creativity and misrepresentation. But this is not new. Whether it's fake fairies at the bottom of the garden (there are some famous pictures from the early 20th century which were purported to be real) or Isambard Kingdom Brunel with his cigar removed for schoolbooks, photography has always crossed that line from time to time. And like creativity, misrepresentation can happen before or after the shutter is pressed - what is the difference?

I'll use my last paragraph to steer us somewhere near on topic. If you were to use Photoshop to create a picture of yourself running a rapid that you didn't actually run, that would be telling a lie, just as surely as if you told it in words. That doesn't devalue photography, it devalues the person telling the lie.
Steve Balcombe

User avatar
Mark R
Posts: 24087
Joined: Thu Jan 10, 2002 6:17 pm
Location: Dorset
Contact:

Post by Mark R » Wed May 10, 2006 1:39 pm

Steve B wrote:Is he saying his images are not manipulated? It depends on your definition I suppose. His particular combination of film/exposure/printing produces oversaturated colours which are not true representations of the world. He used tricks like extreme wide angle to give a distorted representation for dramatic effect. It's not that he didn't approve of manipulation, it's that he had decided how much / what kind of manipulation was ok for him and wished to impose that standard on everyone else.
He says that using polarising lenses and neutral graduated filters is acceptable, as it adjusts the lighting/ contrast of the picture in the way that your eyes would. Anything done subsequently to the pic is heresy.


I am happy to admit that some of the photos of my own that I like best, have involved me pressing the 'auto levels' button in Photoshop. I've always felt very uneasy about this.
Mark Rainsley
FACEBOOK

Steve B
Posts: 5699
Joined: Wed Jan 16, 2002 2:36 pm
Location: Taunton, Somerset

Post by Steve B » Wed May 10, 2006 1:45 pm

MarkR wrote:
Steve B wrote:Is he saying his images are not manipulated?
He says that using polarising lenses and neutral graduated filters is acceptable, as it adjusts the lighting/ contrast of the picture in the way that your eyes would. Anything done subsequently to the pic is heresy.
Well, the bit about polarisers is just plain incorrect, the eye/brain can't do what a polariser does. The neutral grad does, in a very primitive way, emulate the way the brain handles very high contrast scenes.

No, he has his version of the truth and I have mine.
Steve Balcombe

Dave Thomas
Posts: 1734
Joined: Sat Oct 11, 2003 10:36 pm

Post by Dave Thomas » Wed May 10, 2006 1:48 pm

I think Steve B has hit the nail on the head. Photography has two uses, to record or to make an artistic statement. OK, there are shades in between these black and white (sorry!) extremes, but lets keep the argument simple.

If the purpose of a shot is to provide a factual record, then any manipulation risks accusation of dishonesty. If it is to provide an attractive pictorial statement - art, if you like - manipulation is OK either in taking the shot or in post-processing it. After all, haven't we all changed a viewpoint to get overhead cables out of a shot, or to hide a parked car behind a bush?

Going back to Steve's axample of running a rapid leads to an interesting point - the fact that different people may see the purpose of a shot in different ways. If I want a photo of me running a rapid, it is probably as a record of me doing just that. Other people, assuming they don't want a record of my appalling choice of line, would only be interested in the shot as 'canoeing art' (and possibly not even as that, with my ugly mug in the frame!).

Discuss!

Dave Thomas

Discuss

User avatar
pwilkinson
Posts: 422
Joined: Thu Mar 24, 2005 7:29 pm
Contact:

Post by pwilkinson » Wed May 10, 2006 1:48 pm

MarkR wrote: He says that using polarising lenses and neutral graduated filters is acceptable, as it adjusts the lighting/ contrast of the picture in the way that your eyes would.
If I'd know that i wouldn't have spent so much on fancy sunglasses!

[edit] beaten, twice!

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 1:52 pm

My 'Workflow' (I shoot JPEG):

*I always press auto levels and auto colour, to see what happens. More than half the time the photo improves, in that it looks more like a believable (non-distracting) representation of the scene. I can always undo if the auto buttons don't work and go into curves/channel mixer (if I can't find another image that doesn't need so much work).

*I usually fiddle with the brightness and less often with the sharpness.

*At some point along the way I will probably have cropped the image.

* If there is a distracting area of brightness in an otherwise dark area of the scene I may clone over the top of it.

That's about it. Anything else takes far too much time.

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 2:03 pm

My Work flow (I shoot RAW)

I will set the white balance on the RAW file then convert it to TIFF

Open this in Photoshop, and bring the bottom and top of the levels in normally from between 15 - 25. I do this on the majority of my sunny photos as my camera can't seem to cope with the sun.
Alternative - spend more money on a decent camera

If there are a lot of shadows then I will use the Shadow/Highlight to bring some detail out
Alternative – could use a graduation filter

Adjust the colour balance slightly if required
Alternative – This could be done on the enlarger in a darkroom

Crop and sort out the horizon if it’s wonky
Alternative – Again a professional could do this in the darkroom

Occasionally use the unsharp-mask to sharpen the picture.

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 2:14 pm

Does RAW have a wider contrast range? That is, with my D70 on JPEG, if there is any contrast in the scene, all the highlights will be burned out unless I spot metered on them (in which case the darker area are too dark).

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 2:18 pm

Don't know, I will have to do some tests shooting in both RAW and JPEG and see what happens.

Dave Thomas
Posts: 1734
Joined: Sat Oct 11, 2003 10:36 pm

Post by Dave Thomas » Wed May 10, 2006 2:40 pm

James F wrote:Does RAW have a wider contrast range?
In theory, no - just finer gradation (typically 12b-14 bit per colour rather than 8bit). In practice - don't know!

I have just started shooting in RAW, using Pixmantec 'RawShooter Essentials' for initial processing. This does have very handy one-touch controls for bringing up shadow brightness ('fill lighting') as well as varying shadow and highlight contrast separately. These operations are much less simple to apply in Photoshop Elements - though maybe full Photoshop can do something similar to JPEGs etc?

Dave Thomas

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 2:43 pm

Shooting in RAW also means you can change the White Balance at a later stage. With JPEG it's set once the photo's taken.

David McCraw

Post by David McCraw » Wed May 10, 2006 3:17 pm

James - I also have a D70, and do most of my prints at 16x12" (approx. A3). Even at that size I, personally, cannot tell the difference between a meticulously processed RAW and using the D70's Normal (not Fine) JPEG setting.

I shoot a lot so using JPEG delivers a significant money (CF cards) and time (processing) advantage. It also doesn't fill up the buffer as quickly, meaning I can shoot longer sequences.

The solution to blowing out highlights, I have found, is to underexpose by a stop and then lighten in Photoshop. This saves time-consuming and error-prone spot metering while someone is on the run in to a drop!

The only real killer is leaving it on ISO 1600, flourescent WB from the previous evening's pub antics - now that *is* difficult to process out of a JPEG.

Sophie
Posts: 108
Joined: Fri Mar 17, 2006 3:28 pm
Location: Leeds

Post by Sophie » Wed May 10, 2006 3:21 pm

Just to throw a spanner in the works...

Does the ability to manipulate so much in photoshop detract from really knowing how to take a proper photograph. I mean is this type of software mostly useful to those who don't understand the concepts, in the same way as auto settings allow for those who don't understand apeture and shutter speed?

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 3:28 pm

I spend a lot of time trying to make the best use of contrast. White water, especially when thrown in the air, needs to be a few stops less exposed than the surroundings as for whatever reason it is very 'bright'. I can either chose to have properly exposed water and everything else black as night, or properly exposed background/kayaker and the white water burnt out.

Photoshop, without spending vast amounts of time selecting and adjusting different areas, can not overcome this fundamental problem.

The solutions I have found are:

a. compositional (leave out as much white water - and sky - as possible)

b. aim to take the photo when the subject and highlights coincide and spot meter for the highlights - everything that is not the subject is effectively junked.

I have been thinking that what would be neat-o would be for the camera to have a wider latency for correct exposure. Am I dreaming?

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 3:39 pm

Oh dear, poor Marcus... he has become a wraith

Image

This is just an illustration as it was an extremely high contrast scene (sun almost directly behind) so I certainly would need a 'dream' camera to have Marcus come out with any detail. I spot metered on the foam pile - which as you can see turned out acceptably well as did the water behind.

To correct this scene in photoshop I would need to 'select' Marcus and adjust the levels for him - there will be enough detail recorded for some buffing up. However, this takes ages and never really looks believable

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 3:48 pm

Sorry to hijack this thread...

Here is a perfect example of what happens if matrix metering (average of the whole scene) is used for sunny white water shots. Look at the water where I am going to land...

Image

The whole thing is SO difficult.

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 3:48 pm

James F wrote:To correct this scene in photoshop I would need to 'select' Marcus and adjust the levels for him - there will be enough detail recorded for some buffing up. However, this takes ages and never really looks believable
James try using Adjustments > Shadow/Highlight then adjust the Shadows Amount. It's an easy way to bring out a bit of detail in the shadows.

User avatar
James F
Posts: 1669
Joined: Sat Jan 12, 2002 3:04 pm
Location: Weymouth, Dorset
Contact:

Post by James F » Wed May 10, 2006 3:53 pm

You can do something similar with Levels and Curves - but you can only take it so far before it starts looking weird.

User avatar
jon a
Posts: 172
Joined: Wed Aug 17, 2005 8:42 am
Location: Yorkshire
Contact:

Post by jon a » Wed May 10, 2006 3:57 pm

I've always had problems with Shadows/Higlights making things go grainy, but to honest that is probably a combination of uderexposing eveything (Not intentionally!) and my poor Photoshop skills!

J
You learn something new every day, most times it is that what you learnt yesterday was in fact wrong!

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 4:02 pm

Obviously there is a limit to how far you can go, in James's photo Marcus is shadowed so much it would be difficult to get it perfect.
It is also dependent on what ISO you were using, if you find the Shadows/Highlights makes things very grainy you may be using a high ISO (and/or trying to push it too much).

Steve B
Posts: 5699
Joined: Wed Jan 16, 2002 2:36 pm
Location: Taunton, Somerset

Post by Steve B » Wed May 10, 2006 6:25 pm

Dave Thomas wrote:
James F wrote:Does RAW have a wider contrast range?
In theory, no - just finer gradation (typically 12b-14 bit per colour rather than 8bit). In practice - don't know!
Actually the answer is yes, in theory and in practice. I'll have a go at explaining (sorry, long answer)...

The very first stage in recording a digital image is that each individual pixel on the sensor stores up an electric charge, in proportion to the amount of light that hits it. This is an analogue process - more light simply puts more energy into the sensor. And it is linear - twice as much light results in twice the charge. If enough light hits the sensor it reaches a point where it is saturated and can store no more (more on that later).

When the exposure is complete, the analogue charge in the pixel has to be converted into a digital value for processing and saving in a file. Supposing we wanted to distinguish 100 different levels of brightness. If the pixel is 23% charged, we store "23" - in binary form of course - if it is 95% charged we store "95" and so on. In practice, though, we need to distinguish many more levels, and a typical value is 4096 which is the number that can be stored in 12 binary digits ("bits") - 12 bit analogue to digital conversion.

If your camera can save RAW files, that is the end of the story. For each pixel, a value from 0 to 4095 is saved in the RAW file for future processing.

The next stage is another conversion, from 12 bit RAW to (usually) 8 bit JPEG. The values stored in a JPEG file are designed to reflect the way we perceive brightness, and unlike the camera sensor that is not linear. We are *much* less sensitive to changes in bright light than we are to even quite subtle changes in shadows. So the conversion uses a kind of graph - a response curve - to convert the 12-bit linear RAW data into the 8-bit brightness data that we need. That is most often saved in JPEG format.

To pick up Dave Thomas's point, the conversion doesn't take the entire range of 0-4095 values and squish them down into the 0-255 of JPEG. To do so would result in very flat, low contrast images because too much of the scene would fall in the middle, grey part of the range. In practice most well-exposed shots have very little, if anything, close to zero in the RAW data, and equally little at the extreme upper end of the range. So a default conversion discards the lowest and highest values and makes a nice, punchy JPEG with good mid-tone contrast from the middle part of the RAW data. It's not the middle 8 bits - remember there is also a linear to nonlinear conversion going on to complicate matters, but it is a bit like that in its effect.

If your camera can only save in JPEG format, the above process still has to happen, but it is entirely inside the camera. The sensor still records RAW data, but it is immediately converted to JPEG by the camera's electronics.

If you have a more sophisticated camera that saves RAW data, you have the option of doing the conversion on a computer. This is potentially very important. You will have noticed, if you have been paying attention, that the RAW file holds more information that the JPEG. Much more. If you do your own conversion you will still eventually have to discard some data, but you get to chose which parts to keep. In particular, you can choose in favour of best highlight detail, or best shadow detail. For portraits, which often have no very bright highlights or very deep shadows, you might favour mid-tone contrast. Whatever, the choice is yours instead of just an average value picked by the camera.

In photographic terms, a RAW file might contain around two stops more data than a JPEG. That is, you can recover the shadow detail of a one stop longer exposure, or the highlight detail of a one stop shorter exposure. ***Or you can convert the same RAW file twice, once to get a JPEG with best shadow detail, once to get a JPEG with best highlight detail, and combine the two in Photoshop.***

Photoshop CS2 (which I don't yet have, unfortunately), has a utility which takes two images produced as per the previous paragraph and combines them automatically to make one "high dynamic range" image. That is possibly the ultimate solution to the problem of high contrast white water shots. One day perhaps I'll be able to try it.

In writing the above I have completely omitted all reference to colour. It's a pretty big omission, but we are far enough off-topic as it is. If you're interested: http://en.wikipedia.org/wiki/Bayer_filter

---------------------------------

Just a quick word about the Highlights and Shadows adjustment in Photoshop. It is much cleverer than anything you could achieve by using Brightness, Contrast, Levels or Curves adjustments. In a nutshell, it analyses the image and if it finds an area which is dark compared with its surroundings (but not necessarily dark in an absolute sense), it lightens it. So it is defining a 'shadow' not as something which is simply dark. It is something which is dark compared with everything around it. Similarly for highlights. This is how it can appear to fill the shadows with light, without damaging the rest of the image.

Highlights are usually more problematic than shadows. Think back to the beginning of all this - individual pixels filling up with charge. Eventually, in very bright areas, they become full and can no longer record any further detail. These areas have "burned out". The linear response characteristic means that this happens very suddenly - unlike film. If you darken them in an attempt to recover lost detail, they all become the same shade of grey - the detail simply isn't there. This is what leads to David McCraw's suggestion of deliberately underexposing. That way none of the pixels reaches its maximum value so you won't lose highlight detail, and midtone/shadow detail can be recovered. Well, there's no such thing as a free lunch. For reasons to do with the linear response of the sensor and the non-linear response of our eyes, you will get an image which has much better all-important midtone contrast (without which images look dull) and also lower noise in the shadow areas, if you expose correctly - right up to the top of the range. If you do as Dave says your image quality will suffer. Sorry. But in a sense Dave's advice is good - you may get a slightly less than ideal image, but at least you won't get a ruined image.
Steve Balcombe

TomS
Posts: 16
Joined: Wed May 10, 2006 9:27 am

Post by TomS » Wed May 10, 2006 6:30 pm

Image

Steve B
Posts: 5699
Joined: Wed Jan 16, 2002 2:36 pm
Location: Taunton, Somerset

Post by Steve B » Wed May 10, 2006 6:31 pm

jon a wrote:I've always had problems with Shadows/Higlights making things go grainy, but to honest that is probably a combination of uderexposing eveything (Not intentionally!) and my poor Photoshop skills!
Yes, underexposure does make matters worse. Also the default settings in Photoshop are to high - try 20% or less instead of the default 50%. Shadows are supposed to be dark! - don't overcorrect them.
Steve Balcombe

User avatar
Ian Beith
Posts: 112
Joined: Thu Jun 16, 2005 3:32 pm
Location: Dorset
Contact:

Post by Ian Beith » Wed May 10, 2006 6:34 pm

Steve B wrote:Photoshop CS2 (which I don't yet have, unfortunately), has a utility which takes two images produced as per the previous paragraph and combines them automatically to make one "high dynamic range" image.
WOW, haven't found that tool yet. Will have to give it ago (name?).

I always go for, if in doubt under expose by a couple of stops and keep the ISO low, that way you've got room for tweaking in Photoshop.

Post Reply