What's involved with 'processing' an image for the magazine?

[Andy Y]

Or what happens after photos have been taken.........

 

From this:

 

 

To this:

 

 

Periodically I get asked questions about images and how to do such and such so I've recorded the processes (somewhat abridged!) with screen capture software and a narrative full of umms and aaahs (it really is difficult when you're thinking what you're doing rather than what you're talking about!).

 

I chose a scene photographed on CK's excellent 'Engine Wood' which features in the next issue of BRM, it's not an image that's in the article but from a similar position of one which will be.

 

If you do want to follow the tutorials and see exactly what actions are carried out I'd recommend clicking on the full screen option and select HD resolution when each video is playing.

 

 

Part 1 - Image stacking for greater depth of field

 

 

If you don't have Photoshop free open source software, CombineZP,  can be downloaded here - http://www.hadleyweb.pwp.blueyonder.co.uk/

 

 

 

Part 2 - Removing unwanted background

 

 

 

Part 3 - Tidying up and improving the image quality

 

 

 

Part 4 - Adding steam & smoke

 

 

 

Feel free to ask away with any questions as part of the exercise is to help readers get better images from their model railway.

 

[Ian Abel]

I have to say, this is bloody brilliant - a great "quick" lesson and I'm more than grateful for the explanations as I have often wondered how it's done, and even though I have Photoshop, only ever use 1/100th of its capabilities, I'll have to have a go now

[Grafarman]

Finally!  The 'secret' told in an understandable format; lots to play with there...

 

Thanks for taking the time to put it up

 

 

 

David

[Enterprisingwestern]

I preferred the first picture.

 

Mike.

Coat on, hat nearly on.

[Andy Y]

I preferred the first picture.

 

Mike.

Coat on, hat nearly on.

We ran a sweepstake; your name wasn't in the hat but the time interval was about right.

[bgman]

Now where's that roll of film and my Box Brownie ? Or should I get my wax crayons out again!

All joking apart, it was a useful tutorial thank you Andy

 

Regards

Grahame

[jonny777]

I hope I have not missed this somewhere in the text, but how long would that complete process take, approximately?

[Andy Y]

I hope I have not missed this somewhere in the text, but how long would that complete process take, approximately?

Removing the background is the most time consuming element so it will depend on the complexity of the background and any foreground detail which needs to be worked around. Typically a feature image for the mag will take a couple of hours but once all the hard work's done it doesn't take long to have a further play if you wanted to do something different.

 

[gwrrob]

Is this something that could be shewn/taught at RMWeb live.Great to see 'Engine Wood' in print again.

[jonny777]

Removing the background is the most time consuming element so it will depend on the complexity of the background and any foreground detail which needs to be worked around. Typically a feature image for the mag will take a couple of hours but once all the hard work's done it doesn't take long to have a further play if you wanted to do something different.

 

Dusk.jpg

 

 

Thanks.

 

That dusk photo looks rather good. I had never thought of being able to vary the time of day to that extent.

[gwrrob]

Thanks.

 

That dusk photo looks rather good.

 

Isn't it just.I wonder if he could do something similar with one of my/his creations. Even the signal lamp is modelled.

[Miss Prism]

Great tutorial, but I got lost completely in Part 4: 

 

- I don't understand why you duplicated the top layer (why not diffuse the original top layer??)

 

- I understand why a new smoke source layer was introduced, but why did you 'render clouds'?

 

- I found the sequence from 3:30 to 4:00 very hurried and difficult to understand, were you painting smoke on or removing areas from a layer to let the smoke source layer appear?

[Andy Y]

Great tutorial, but I got lost completely in Part 4: 

 

- I don't understand why you duplicated the top layer (why not diffuse the original top layer??)

 

- I understand why a new smoke source layer was introduced, but why did you 'render clouds'?

 

- I found the sequence from 3:30 to 4:00 very hurried and difficult to understand, were you painting smoke on or removing areas from a layer to let the smoke source layer appear?

1. I duplicate the main layer to create the 'Haze' layer so I can backtrack if necessary for a future image without effects.

2. It's just a way of 'randomising' the creation of toned clouds.

3. I clone from the 'smoke source' layer onto the 'smoke layer' (with thin opacity).

 

Hope that makes sense.

[Miss Prism]

Ah, so the 'render clouds' is a new creation in the smoke source layer, and not a rendering of the existing clouds in the picture!

 

I'm not too familiar with cloning, but presumably you just take an area of it from the 'smoke source' layer and in effect spray it on the 'smoke layer'?

[Andy Y]

Ah, so the 'render clouds' is a new creation in the smoke source layer, and not a rendering of the existing clouds in the picture!

 

I'm not too familiar with cloning, but presumably you just take an area of it from the 'smoke source' layer and in effect spray it on the 'smoke layer'?

1. Spot on - I can see where the confusion may have arisen.

 

2. That's it; it sprays a thin duplication onto the 'smoke' layer.

 

I'm not saying my approach is the definitive way as it's self-taught rather than books or tutorials but it's just a method that works for me. I just wish there was a quick and easy way of masking around OHLE, trees and lattice-post signals!

[Steve Williams]

 I just wish there was a quick and easy way of masking around OHLE, trees and lattice-post signals!

Would the Magic Wand tool be any use for that Andy?

[Neil]

I watched the 'remove unwanted background' tutorial as it's that aspect which I thought might be of most use to me. Fascinating, the image program used is unfamiliar to me, seems far more complex than anything I've ever used. Would it be a stupid suggestion to use a toll of bright pink paper as a photo backdrop as it would be so unlike any of the foreground colours that the program could identify it even round the fiddly bits?

 

Also I think I see where Entreprisingwestern is coming from with his preference. I'm not in total agreement, but I'm not sure the image stacking adds to the realism, though it might add to the clarity. I've read somewhere that at any given moment only a surprisingly small percentage of our field of view is in sharp focus. I think I prefer a photo with soft areas; it may be less perfect, but in a like for like comparison I think it adds to the myth of realism.

[Andy Y]

I've read somewhere that at any given moment only a surprisingly small percentage of our field of view is in sharp focus. I think I prefer a photo with soft areas; it may be less perfect, but in a like for like comparison I think it adds to the myth of realism.

You're absolutely correct Neil and I nearly covered this aspect in one of the videos but thought it would ramble on too much; it's certainly worth addressing though. At any insatant what we actually see if we were looking at the real thing would be muted and indistinct outside the focus area so the representation below is closer to what we see if we focus on a fixed point, very limited depth of field with reducing detail and colour perception as we move towards the periphery. Our brains and optic nerve compensate for this with the speed at which they can move focus points and assemble a complete picture in a short space of time which we memorise to a degree.

 

 

What would be more correct to say is that "attempting to reproduce, in a model's image, the depth of field captured in a similar photograph of the real thing* increases the perception of realism".

 

* Dependent on camera, lens, exposure, aperture and environmental conditions.

[Andy Y]

Would the Magic Wand tool be any use for that Andy?

About as much use as a real magic wand would be Steve; it just cannot differentiate with sufficient consistency in areas of similar colour and tone without a lot of additional work after a best stab.

[robmcg]

Great stuff Andy, I very much like pic no.2 in the first post.

 

As you know I put a lot of pics up in RMweb and I use my own self-taught methods, no layering or cloning but lots of bending, re-sizing, painting, and copy-pasting. As you say there are many ways to do things, I never quite got the hang of layering or cloning, but then I don't have to photograph lots of subjects for magazines. I suppose you could call my method 'floating selections', selections in new images are messed-about then copy-pasted, with a lot of dithering and opacity and paintbrush effects. 

 

Here is Hornby's latest with about 5,000 little tweaks. Enjoyable stuff doing this, it's not at all stressful, sometimes a bit like knitting, and whatever I don't like I can usually un-do. I enjoy making pics mostly because my modelling skills are limited, and I've always had a great interest in British steam railways. Thanks again and cheers,

 

Hornby Duchess of Abercorn LMS 6234 in forward gear, coming off shed for a hard WCML turn..... I hope it's not too irrelevant.

 

Edited April 3, 2014 by robmcg

[Steve Williams]

About as much use as a real magic wand would be Steve;

.......ah, now it just so happens. Expelliarmus!

[westerhamstation]

A really useful tutorial with lots of tips, and explained in an interesting way, thanks all the best Adrian.

[Kenton]

 

Great tutorial, but I got lost completely in Part 4:

Have to agree. loved all of it up to Part 4 - the purpose of which I can't agree with and consider as destroying all the positive work preceding it.

 

Background removal can only enhance, image stacking (so that's how it is done!) and tricks to put right failings in the photography - all good and worthwhile. But adding things that are never going to be there in the original scene - well travesty why not make up everything and delude the viewer completely? Just because it can be done does not make it something that should be done.

 

Still great Parts 1-3, Thanks

[eastwestdivide]

 Have to agree. loved all of it up to Part 4 - the purpose of which I can't agree with and consider as destroying all the positive work preceding it.

 

Background removal can only enhance, image stacking (so that's how it is done!) and tricks to put right failings in the photography - all good and worthwhile. But adding things that are never going to be there in the original scene - well travesty why not make up everything and delude the viewer completely? Just because it can be done does not make it something that should be done.

 

Still great Parts 1-3, Thanks

Perhaps our take on part 4 (steam and smoke) depends on the perceived purpose of the photo - is it trying to

a. represent the model accurately, or

b. do essentially what all models are doing, and pretend to be the real thing?

Is part 4 just one more degree of pretending? In that case it is any different to adding a sky?

[Kenton]

 

Perhaps our take on part 4 (steam and smoke) depends on the perceived purpose of the photo - is it trying to

a. represent the model accurately, or

b. do essentially what all models are doing, and pretend to be the real thing?

Is part 4 just one more degree of pretending? In that case it is any different to adding a sky?

I'm afraid I see photography as taking an image of what is actually there. I do not see it as an art form inventing and misleading the viewer.

 

So prettying up a fault in the photography or its surroundings by retouching (electronically) is fine. These are all techniques that could be at the disposal of the photographer by using a better studio/backdrop, or in the focus by using a different lend/camera/setting/lighting, or by removing blemishes as in retouching/masking.

 

But adding smoke/clouds/atmosphere or making other additions that are essentially lies within the image just are not photography and are misleading.

 

I also believe that they do nothing to add to the image and only detract from the quality of the model.