Saturday, December 29, 2007

Soft Proofing

The top image is the original, the bottom a duplicate with soft proofing applied. The differences aren't huge and you probably want to click on the image to bring up the large version to really see the changes.

I was reading a discussion about the value of soft proofing on Luminous Landscape and while the discussion degenerated into abuse along the lines of 'my expert is better than your expert' it did make me think more about the process of soft proofing. I'd watched 'From Camera To Print and Jeff Schewe in particular espoused the value of soft proofing. Time to look into this a bit further. I took the Athabasca Falls image that I showed the editing on for the previous blog entry and made a print on Enhanced Matte, knowing that soft proofing shows the differences with matte paper especially.

Know what, Jeff was spot on. First reaction to turning on soft proofing - 'yuck, what happened to my lovely image'. But then I did as instructed and held up the print made 90 degrees to the monitor and in the same lighting and flipped back and forth between monitor and print and guess what, the soft proof explained all the differences between my carefully edited image on monitor and the resultant print. Very impressive.

Now the trick is to learn how to change the image so that the soft proof matches as closely as possible the original. Again I did as Jeff suggested, I duplicated the image so I could compare the soft proof side by side with the original.

I used a combination of five adjustment layers to to match as closely as possible the original. I increased overall saturation by +7 (Jeff mentions that this is usually needed when printing to matte paper). I created two curves layers which could prob. have been combined, to increase contrast slightly and give a bit of punch to the dark areas - and I noted that the white water took on a bluish tinge which didn't match the rest of the image so I used a selective colour adjustment layer, working on whites to yellow them up a tinge.

The end result on screen looked darn close to the original. Of course the blacks weren't as black (because the soft proof is emulating the lower dynamic range of matte paper) but overall it looked good. I made a print. The print looks like I went just a bit too far. I'm going to try again with all the added adjustment layers set to 66% opacity, or perhaps 50%. Ah, 66% works perfectly.

I can see that with a bit of practice this is going to be very helpful in getting images as close a possible to what I want - also to making the same image on two types of paper, matte and glossy, and having them match as close as possible.

It's intersting, the difference means that rock that looked wet in the on screen image, now looks wet again in the print - it didn't before I soft proofed and adjusted.

If you are not soft proofing, you might want to start doing so. Clearly an experiment of one isn't the whole answer, but so far it's looking useful and I will update down the road.

I need to try the technique with black and white images too.


orcasmac said...

I've found soft-proofing to be quite valuable, and I'm still learning good ways to adjust a soft-proofed image so that the print more closely matches my expectations the first time around. I have found that soft-proofs for matte papers seem to require more adjustments than those for glossy papers.

George Barr said...

Agreed, of course there is a bigger gap between the original and the soft proof with matte papers because of the reduced dynamic range and scattered light coming off the colour and affecting both hue and saturation of the colours. I note that with the Harman FBAL gloss paper, there is very little (but some) difference when you turn on soft proofing. Getting back to the original has got to be simpler and easier.

alw said...

Is it not possible instead to use Photoshop's CONVERT TO PROFILE (i.e. convert to the desired printer-paper profile), work the resulting screen image and finally print that image with NO COLOR MANAGED?

George Barr said...

That's a really interesting question. To be honest, I don't know the working of colour enough to give you a technical answer of what's going on, but a simple experiment gave me the practical answer - no. When I converted an image to my enhanced matte profile, the image got brighter but shadows didn't get weaker as one would expect when representing a matte paper and it looked nothing like the soft proof image.

alw said...

Are you soft proofing with Simulate PAPER WHITE turned on? It could be that you are only seeing the effect of Photoshop turning off simulate PAPER WHITE when CONVERT TO PROFILE is used. This feature has nothing to do with the eventual print quality.

David said...

I always thought I knew a lot about printing, but when I started looking into soft proofing and such I found that the process works differently with different printers. I print at a lab where the printer can not read the ICC profile so the steps I have to do is

-Edit in sRGB and get the image looking good.

-Soft Proof using the labs printer profile. 9times out of 10 the print looks really good and there are no gamma warnings. If there is then I just adjust the problem colours saturation and luminosity time I have tamed it.

-Convert the image to my labs printer profile.

-send it in

I have a good batting average this way and the only colour I have a problem with is yellow and I have heard that is because the Fuji paper my lab uses has problems with yellow.
But when printed I get a very close print compared to my monitor

George Barr said...

It's not too surprising that you have a reasonable match since sRGB is such a small colourspace that there are virtually no printers not capable of reproducing it. Problem is, you are significantly limiting yourself by doing so - basically saying I will never work with really saturated colours so I can't miss what I never saw. Most cameras are capable of significantly better and assuming you are shooting raw, are capable of outputing in Adobe RGB which is significantly bigger space or as I and a lot of others do, into Prophoto RGB which takes absolute maximum advantage of the really good printers and inks of the last year or two.

David said...

You are very right George and that is why I capture in RAW, that gives me the option on later using larger Colour Spaces. But one thing you are not taking to account is that the printing industry is behind on the colour space thing.
Yes printer today are able to produce a colour gamut equal to Adobe RGB and some as high as ProPhoto, but that doesn't mean the labs that use them take advantage of that fact. In fact most labs in North America and all in my area calibrate their machines for the sRGB colour space....yes even the so called "Pro" labs. They may advertise that they except aRGB, but as soon as they get them the first thing they do is convert them to their printers profile which has a gamut no bigger then sRGB. The main reason they do this is for technical reasons, they just don't want to deal with the headaches a large gamut would produce.

If you are lucky enough to afford a wide format printer at home you are in a better situation the us "Lab Rats" because most mid to high range printers from companies like Canon, Epson and HP can handle Adobe RGB and both Epson and Canon have come out with wide format printers that can produce 16-bit Prophoto prints...but those printers start at around $8000 range.

So till the Labs catch up with the colour space there is no real advantage to even dealing with aRGB or ProPhoto...the Labs jsut can produce the colour gamut that these spaces use. Just like images for the web, if the image is only gonna be use on the web, what is the use of even using a large colour space?

George Barr said...

Checked with my local professional printer, ABL, here in Calgary. They are printing to an Epson 9800 and use Adobe 1998 as their standard colour space. they do not supply profiles for soft proofing though with Epson I suspect that if you used a canned profile for soft proofing, you'd do very well.

David said...

Must be nice! lol I'm in Ontario and here there is a huge selection of labs but most just use the high output machine like Noritsu or the Fuji Frontier. These machines don't even read embedded ICC profiles, to get acqurite results you really need to convert you image to the printers profile.

But here is a funny situation. A lot of "Pro" labs in Toronto use these does Costco lol.

Maybe I should be looking around a little more in my area. For my wedding and portrait work Costco is very worth while, but it would be real nice to produce some high quality fine art prints.

George Barr said...


aren't the printers you mention outputing to chemical processed photographic paper rather than printing inkjet? I know that ABL offers both lightjet and inkjet output.


David said...

George, Yes they do, so they are limited to pretty much sRGB.

But inspired by your lab I called around in Toronto and found a lab that no only offers the Epson 9800 but also the new Epson 11000 series printer (will arrive today).

But talking with the tech there he was saying that depending on the paper you are still limited to aRGB.

Carlo said...

Good Job! :)

Chandresh said...

a printer which employs a printing process that uses heat to transfer dye to a medium such as a plastic card, paper or canvas. The process is usually to Portable GPS lay one colour at a time using a ribbon that has colour panels. Dye-sub printers are intended primarily for high-quality colour applications, including colour photography; and are less well-suited for text