Amazed at Mis-information

Not so long ago, when first converting to digital image gathering and aware of my ignorance, I asked a lot of people about it. I was later amazed at the number of people who had their credibility invested in mis-information. Paul Kay was the exception. I'm still not an expert but I know how to produce a file that reproduces in web-offset. (It's a pity that computers allow so many people who should know better to tamper in areas they don't understand, so often before it gets to the printer.)

It still goes on.

My wife recently took a couple of files to be printed up to mural size. The guy at the (retail) store told her that her files were CMYK and therefore could not be printed. I had to convert them to RGB so that he would accept them. I guess it was a question of giving him what he was used to getting. I wonder what colour inks their printer uses! No doubt someone else down the line will convert them back to CMYK, just as they do for the RGB jpgs supplied for iPhoto books.

At Diver Mag our repro is not as good as the exact same files reproduced in our companion Scandinavian mags DYK. That's because the art director of DYK used to share an office with a repro expert and he obviously learned a lot about the parameters needed for their printer. Diver Mag's art director is only just resigning himself to the fact that my files are best left untouched - although we have had some disappointments in the past because he likes to add black to make up for the fact his monitor faces a window. (Mine is in a darkened room.)

I provide him with a RAW file along with my adjusted CMYK tif but he is unwilling to use the RAW converter even though I gave him my spare copy of CS4 RAW. He prefers to stick with a way of working that he knows and does everything in Photoshop. People prefer to retreat to what they know rather than learn new tricks, but then we'd all be shooting film still.

Film scanned by me and supplied as a Nikon file was originally not accepted by our production man on the basis of not being good enough until I got a digital camera and converted my scans to Photoshop tifs. Then, because he thought they were from the digital camera, they suddenly became OK. I remember Kurt Amsler telling me of a similar experience.

Similarly Diver Mag receives lots of pictures converted to RGB jpgs that will not reproduce because although they might look very nice on a PC monitor, there is lots of data missing. Often they are over-sharpened too. It seems people do not understand how Photoshop works and nor will they if so many people dish out incorrect info.

I'm not an expert. Apart from a bit of 'mustardisation', I do all my picture adjustments in RAW, converting to suitable CMYK parameters at the last moment and with all sharpening switched off. I'm sure there are plenty of ways to skin a cat so I do not profess to know what's best but take my tip and be aware that many experts are not that expert!

Of course, if your pictures are only ever going to be projected/viewed by computer, you wasted your time reading this!

[EDIT by MOD - add subtitle]

I know what you mean. The guy at one of our local photo labs outright REFUSES to believe that negatives can be scanned at a higher DPI than 72 from his or any other scanner. It's not possible and it's useless if so he says, pointing out how "large" the 3MB JPEG produced is .

In a similar theme, I ran into a photographer on a dive boat the claimed that film had an effective resolution of 150MPixel and had much wider dynamic range than digital. I tried to explain to him that this wasn't correct, but wouldn't hear of it. He considered himself an "expert"

As far as ink jet/dye sub, etc prints... They actually do use RGB files, not CMYK, even though they print in CMYK+. The do a conversion and do do well when supplied with CMYK.

Jack

RGB can represent brighter colors than CMYK this is why you often get that annoying “out of gamut” warning when converting to CMYK. Unless a particular version of an image is going to be only used in magazines and books I would save it in RGB. JMHO

There's RGB and RGB and RGB, of course, each with a different gamut...

.. and then the hardware: printers that make an internal CMYK conversion, but need an RGB input (and that may or may not accept profiles). I'm not sure what Jack means, as many office printers will try to make a CMYK conversion of the CMYK offered, and get their ******** in a twist!

So, a commercial operation may be happy to take CMYK files, or may want a profiled RGB file to make their own CMYK conversion.

Two principles: deliver what is asked, don't second-guess the print professionals; and RTFM!

Tim

>>> Two principles: deliver what is asked, don't second-guess the print professionals

A "pro" lab recently informed me that I needed to submit my images to them in sRGB as "that's the color space of all digital printers" ... it took me a little while, but I finally managed to convinced them that was not the case.

We do all our own CMYK conversions for Wetpixel Quarterly. Some images take 30 seconds, and some take upwards of 15 minutes. It is often not a simple thing to do, even if you haven't done anything to your images. For example, if you take a camera into the blue waters of the Bahamas and snap a shot of the blue and do a default RAW conversion, saving your image as 16-bit AdobeRGB 1998 file, it will be 100% out of gamut when you try to convert it to the CMYK profile we use for the magazine.

I've been thinking of writing an article on it, but it is so much easier presented as a workshop than described using screen grabs and text...

... and finally, some photographers become known for submitting images that are consistently out of gamut. These images all look great on the web, but the photographers really do ruin the image to make that happen (contrast, saturation, and sharpening to 11 = FAIL).

We do all our own CMYK conversions for Wetpixel Quarterly. Some images take 30 seconds, and some take upwards of 15 minutes. It is often not a simple thing to do, even if you haven't done anything to your images. For example, if you take a camera into the blue waters of the Bahamas and snap a shot of the blue and do a default RAW conversion, saving your image as 16-bit AdobeRGB 1998 file, it will be 100% out of gamut when you try to convert it to the CMYK profile we use for the magazine.

 

I've been thinking of writing an article on it, but it is so much easier presented as a workshop than described using screen grabs and text...

 

... and finally, some photographers become known for submitting images that are consistently out of gamut. These images all look great on the web, but the photographers really do ruin the image to make that happen (contrast, saturation, and sharpening to 11 = FAIL).

So if you're not planning to write an article, and I don't have any workshops of yours on my calendar, any good reading resources you would recommend regarding preparing images for print?

I've been thinking of doing some books, and this topic has got me thinking whether I should approach preparing my images differently.

What hasn't been said clearly enough in this thread is that there is a massive difference between preparing files for Photographic/Home Printing and preparing images for Magazine/Book Printing.

Most photographic/home printing is designed to work from RGB files. The main problem here is over processing - but the process is quite forgiving on people ramping up all the levels. Banding and noise from pushing colous around are the main problems - but this type of printing does not really have the out of gamut problems associated with over processed magazine images. The printers do a good job most of the time these days.

The CMYK printing of books and magazines is quite different. I think unless you really know what you are doing, then submitting in CMYK is not desirable (particularly as different print houses use different CMYK profiles), the exception is if you are a staff photographer for a single magazine (I have a friend who shoots for a sports magazine and he has the "company CMYK profile" on his computer and submits entirely in CMYK).

For almost everyone else, it is much better to produce an Adobe 98 RGB file - that you have not pushed about too much in processing. Shoot it right and submit the file with only minor adjustments and most of the time the magazines will be able to adjust it and print it right. You can Gamut check in and adjust in Photoshop if you want (CMD+Shift+Y) - but second guessing the CMYK profile and process of the magazine is not always reliable.

Most magazines and books will tell you to give them good quality RGB files that have not been pushed about too much (the missing data that John refers to) - so there is lots of data there for them to make the CMYK conversion and then adjust the colours to fit their CMYK printing gamut.

Eric makes a fair point about rich blues being hard to print, even straight out of the camera, but most of the time it is photographers ramping up their levels in their files that causes the problems. Plus if the photographer has made adjustments to their RGB file - then the file is not in as good condition for the magazine guys to adjust later.

Alex

I think it's interesting that so much blame is being directed toward overprocessing when the failure is the printing tools. An image produced in a color-managed environment that displays well on a monitor should be able to be "displayed" well on a printed page. Otherwise, color spaces aren't the universal "language" they are designed to be.

I realize there are limitations in the tools. For example, water that turns purple during print is a tool failure, not a photographer mistake. It's disappointing that such tool limitations exist, but even more disappointing that there isn't greater outrage that they aren't remedied. I appreciate the skill required to produce a good printed image but wish such skill weren't necessary.

I think it's interesting that so much blame is being directed toward overprocessing when the failure is the printing tools. An image produced in a color-managed environment that displays well on a monitor should be able to be "displayed" well on a printed page. Otherwise, color spaces aren't the universal "language" they are designed to be.

 

I realize there are limitations in the tools. For example, water that turns purple during print is a tool failure, not a photographer mistake. It's disappointing that such tool limitations exist, but even more disappointing that there isn't greater outrage that they aren't remedied. I appreciate the skill required to produce a good printed image but wish such skill weren't necessary.

I think this is a good point; and in some ways is likely to become an even bigger problem as electronic capture devices and displays increase their gamut and dynamic range capabilities. Paper, and inks have certain physical limitations and the gap is only likely to get wider. As I understand it, one has to think about compressing the gamut of the electronic images to the more restrictive gamut of physical media when preparing for print.

Yes, that is true. I consider that job to be part of the printing process but, judging by comments here and by the opinions of many experienced photographers, many think it's a a shared burden with the initial image processing.

One thing that has always bothered me is that images are either "scene referred" or "output referred" and the transition occurs during the raw conversion. There is no intermediate form. Some print professionals I'm familiar with perform raw conversions with a particular output device in mind because the raw converter itself. I understand why that is but it's counter to how many people work with their images. It makes me wonder why these people bother working with intermediate color spaces.

A philosophy I believe in is that things should be as simple as possible and no simpler ("Everything should be made as simple as possible, but not simpler." - Einstein). What we need, and what we have, is an intermediate color language for representing images divorced from either the input method or the output device and what is happening when people smash all the processing together is that things are being made "too simple". Doing so allows us to assign blame more easily, because there's only one processing step to point to, but I think it retards progress. One thing's for sure; there's a lot of room for improvement.

We have a real set of dilemmas here. In an ideal world we would have systems in place which would operate on a WYSIWYG basis throughout but unfortunately we don't. So we use gamuts, colour profiles, colour conversions and a variety of viewing devices ranging from LCD monitors to looking at a book in the sun or even viewing it indoors under low energy bulbs. Is it any wonder that we have problems?

It seems to me that there are only two scenarios which we can operate under - the fully controlled loop where the original image is taken knowing the final output requirements (a very low percentage of images) and the rest where we shoot not knowing how the image will be used and produce a 'generic' file - which is a compromise - and I'm very doubtful that this situation will change much in the near future.

We also have another problem which has been touched on - over processing of the image file - to ensure that it looks good on screen. Problem here is that if a specific file isn'tover processed, it may not look good on screen, and if it doesn't look good on screen it probably won't look good anywhere else. So we have a catch 22 situation. I think the answer is probably optimal capture but this isn't always possible - for example in extreme low contrast conditions (try photographing mud dwelling creatures in Welsh waters!).

Lastly, personally I always try to bear in mind that theory aside, photography is fundamentally a very practical application of technology - for example I send my files to my local printer after having adjusted them carefully but my final step is to remove the colour profile. Why? Because I've found that this is the best way to achieve the most accurate and consistent print quality. For big, expensive prints, I produce a file and small print (on the same paper and print system) and request a match at larger size from the printer. This works very well because the printer can make a direct comparison by eye and adjust as required.

PS, thanks for the comment John - not sure its at all warranted as I still have an disconcertingly enormous amount to learn about photography .... its never ending I should say.