A snag I could not have anticipated

So after my tests proved that I could get just as good results from gathering images digitally as I had been able to on film, I committed to buying all the kit necessary.

I made my first digital shoot underwater in the UK and then set off for Mozambique for my first real digital expedition.

While I was away, the UK photoshoot went to layout and repro. When I got back I was shocked to see a horrible set of proofs. Why?

Well, when you supply an image on a piece of original film, the printer is duty-bound to match it. The art director may use a low-res scan for layout but the film is then scanned in high res by a skilled technician. However, supplying an image as a digital file (TIFF) allows the art director to take it into Photoshop and make a few changes and send the changed file to the printer. (Art directors always seem to like to make a few improvements!)

However, messing about with a TIF file in Photoshop does not necessarily do the image any favours. The art director sees a low-resolution (72dpi) RGB image on his screen but the printer uses a high res. (350dpi) CMYK image. There is a big difference.

What a mess! We are now in a situation at the magazine where I have insisted that files are only changed when in RAW format. The art director insists he knows what he is doing of course. The results so far are very poor.

Come back film. All is forgiven!

Those art directors will die out and be replaced with knowledgeable people. It may take some time...

Hi John,

It's probably easier to replace the art director than to go back to film !

Just a quick question. Why are you supplying TIFFs and not JPG 12s or RAW images ? If you don't save the JPG file over and over, there should be very little loss.

JP

I supplied RAW files and converted uncompressed TIFFs. The problem is that people prefer to work certain ways. Obviously the a/d prefers to use conventional Photoshop controls rather than adjusting RAW data. The CRT screen is very seductive but not CMYK. I am interested on how it looks on the page. JPEGS tend to get copied and lose quality - so that is asking for trouble.

Hi John

Welcome to the world of variable colour reproduction! That said, if you look at the reproduction of the fish pix I supplied and which were used in the June issue of Diver, you will be happy to know that they are pretty accurate - and I supplied them as scans on CD!

Today, you SHOULD have no problems with files (scans or digitally originated) providing your viewing system is callibrated to a 'reasonable' extent - if you do then its down to either a set-up which is very much at odds with your own, or someone messing around with your files. If you are supplying an 8-bit rgb tiff file, then it simply shouldn't be messed around with to any significant degree (as doing so can badly affect reproduction) and perhaps this is the message that you need to get across. If there is an issue with the way your scans reproduce it might also be in the cmyk conversion or your/the printers callibration.

The issue of digital versus film should by now be over. An inability to handle digital files is something that I now find baffling but all too common. I know designers/printers who are well aware of the problems and who adjust very carefully and produce excellent results from any media. And I have dealt with others who operate on a very hit or miss basis.

Given that this is a problem which has got to be dealt with (digital is NOT going away) you need to sort it out sooner rather than later.

Given that this is a problem which has got to be dealt with (digital is NOT going away) you need to sort it out sooner rather than later.

<{POST_SNAPBACK}>

I dunno. Apparently type-writers and carbon paper are making a comeback in 2007. So magazines will be digital for pictures and analogue for words again.

Hi Paul.

The problem is that the a/d CAN fiddle with the file in Photoshop. I have been supplying TIF files of product shots since I first spoke to you about going digital about two years ago. It has proved very successful (eg. Just Surfaced and Diver Tests).

However, in this case the a/d thought he would get rid of the green water and make it blue (like he thought it ought to be!) and once he got started...

No-one else gets to see what I supplied (unlike a piece of film which stands permanent testimony). The shots looked excellent on my own calibrated CRT and printed beautifully sharp on my own lazer printer. So once the magazine comes out it will be time to discuss things. I am sure the guys at DYK Mag in Scandinavia (where my stuff also goes) will prove the difference when they print it.

It is a political problem rather than a technical one! But a problem nevertheless.

Hi John

An additional problem is the lack of tonality in large areas of blue and green. In my experience ANY adjustment here can rapidly lead to gaps in the tonality and blocking of colour - posterisation. And as soon as this occurs there is a temptation to adjust it to correct - making matters worse! Mind you I've found this from scans too! So you can't win.

But if you think that film will make a comeback, great. I have a nice collection of Subal and Aquatica paperweights which might regain some value!

When I made TV commercials, we we shot on film. By and large the agency accepted what the director gave them. Then video-editing was introduced and the film was transferred to G-spool. That meant that everyone from the client and the agency could be a director too. Thats when I left!

Does it make a difference whether you supply a tranny or a digital file ? I imagine the the art guys would be just as likely to want to "enhance" the scanned image than they would be to retouch the file you provide on a CD.

Alex, have you dusted your Hasselblad, yet ?

The difference is that the art director never gets to touch the high res scan of the trannie. He works with a low res scan for layout and then this goes off to repro with the original trannie. Obviously if you supply scans (on CD) there is a risk they can be adjusted (for the worse) but scanning is a skill in itself and those that operate drum-scanners tend to know how to get th best out of a film original 0or at least they know better than the designer.

John

Are you supplying rgb or cmyk tiff files?

In my experience printers do as little adjustment as possible (time is money after all) and I am surprised that an art director actually adjusts the hi-res files. Actually I find it can be a problem that printers and their designers don't adjust files as, as you say, they generally know how to get the best out of the files (whether produced by their scanner or a supplied file) and their printing set-up.

This said, many don't seem to understand that files need to be 'balanced' to their set-up (or more likely, can't be bothered) and fail to operate any unsharp masking at all.

On a personal note, I cannot even imaging going back to film - digital is, for me, simply better as I am producing better images than I ever did on film.

John

Are you supplying rgb or cmyk tiff files?

In my experience printers do as little adjustment as possible (time is money after all) and I am surprised that an art director actually adjusts the hi-res files. Actually I find it can be a problem that printers and their designers don't adjust files as, as you say, they generally know how to get the best out of the files (whether produced by their scanner or a supplied file) and their printing set-up.

This said, many don't seem to understand that files need to be 'balanced' to their set-up (or more likely, can't be bothered) and fail to operate any unsharp masking at all.

On a personal note, I cannot even imaging going back to film - digital is, for me, simply better as I am producing better images than I ever did on film.

I supplied both the original adjusted 16-bit RAW files and a TIF file of each simply converted to 8-bit CMYK. The problem stems from the a/d's reluctance to work from the RAW and preference to use the TIF. When I supply product shots (CMYK TIF) he just sends my file to the printer - result success - but when he got an interesting subject he obviously started playing around to make it look better - but that was only on his screen!

The political problem stems from the fact that the production man never gets to see the original so does not know what it could have been like. You will know what I mean when you see the mag in a couple of weeks.

I encountered a similar problem when I tried supplying scans from trannies with a Omnicon or Nikon icon. The production man in that case proved to be the problem. He would not accept them. He knew that I had original film available. So I took them into Photoshop (untouched) and said I shot them on digital. Then everything was then OK because he recognised the icon.

He thinks your fish shots were all shot on a digital camera (thats OK) but I bet you scanned some from film. He takes the view that if it was shot on film he must have the original for scanning. Its a political stance not a technical one. However, in the real world we have to work with other people.

Hi John

First, I simply won't supply cmyk tiff files. In my opinion the cmyk conversion should be the printers job as should final use of the unsharp mask, since the first demands knowledge of the specific printing set-up for optimium results, and the second is only possible when precise output size is known. This follows various experiences as well as advice from a friend who runs a graphic design company amongst others. Some printers do apparently try to charge for the conversion though (cheeky!). Adjusting an 8-bit tiff file without knowing how much 'optimisation' has already taken place is asking for trouble as tonality can easily become compromised. I supply 8-bit rgb files but I process in 16-bit and check the final histograms to ensure that tonality is still reasonably smooth.

Have a look at the various forums on www.robgalbraith.com and you will no doubt find a lot of discussion about such issues. I doubt you are alone with having such irritations!

For your info ALL the fish shots were from 35mm via a Coolscan 8000!

Hi John

First, I simply won't supply cmyk tiff files. In my opinion the cmyk conversion should be the printers job as should final use of the unsharp mask, since the first demands knowledge of the specific printing set-up for optimium results, and the second is only possible when precise output size is known. This follows various experiences as well as advice from a friend who runs a graphic design company amongst others. Some printers do apparently try to charge for the conversion though (cheeky!). Adjusting an 8-bit tiff file without knowing how much 'optimisation' has already taken place is asking for trouble as tonality can easily become compromised. I supply 8-bit rgb files but I process in 16-bit and check the final histograms to ensure that tonality is still reasonably smooth.

Have a look at the various forums on www.robgalbraith.com and you will no doubt find a lot of discussion about such issues. I doubt you are alone with having such irritations!

For your info ALL the fish shots were from 35mm via a Coolscan 8000!

Just spoken to production department and as I thought the repro (including any scans from film done by the repro company) is sent to the printer as a PDF so that nothing is changed. If you remember, when I first started shooting on the S2 Pro, I had a conversation with you about settings for repro and what I subsequently did was come up with a formula that gave acceptable results. If you look at the Diver Tests and Just Surfaced pages you can see the results.

What this posting was about is the fact that there IS the opportunity to adjust files in the stage between the photographer and the PDF that is sent to the printer. This gives the opportunity to get some unfortunate results.

Talking about tonality I checked the levels on the files sent to the printer from these shots in question and the levels had a huge number of spikes!

The problem for me is that the Editor will never get to see what I originally supplied. When I supplied film it was not my problem if they screwed up the repro!

PS. By the way I have been told on good authority that you CANNOT do good scans on a Nikon Coolscan!!!!! So keep that secret!

Hi again John

It is a tricky one. The printer needs to be able to adjust some things (unsharp mask at least) but should not adjust others (anything affecting tonality for example). Add to this the problem of translating what you see on an rgb screen to what is feasible on the printed page, along with a lack of any form of hard proof and we are back to a fundamental problem of digital photography.

I've read, and quite happily believe, that we as photographers are at the cutting edge of this technology - others, who involved in the utilisation stage of digital images, may still in a state of catch-up. Add to this the unfamilar constraints of low contrast and easy blocking of seawater colour by any significant adjustment and we have a recipe for the inevitable problems such as yours to arise.

All this said, I now find that the results that I am achieving from my digital set-up underwater are better than those I ever achieved from 35mm film. Whilst I accept that not everything is sorted for underwater yet, the ability to achieve better fill flash balance and the incredible amount of detail delivered by the hi-res cameras means that I would not even consider shooting on fim again (a personal opinion).

There was a wonderful response on another web forum to the question "What do you miss about film?" - ans. "Those handy little plastic containers" - I'm in this camp now too.

Hi again John

It is a tricky one. The printer needs to be able to adjust some things (unsharp mask at least) but should not adjust others (anything affecting tonality for example). Add to this the problem of translating what you see on an rgb screen to what is feasible on the printed page, along with a lack of any form of hard proof and we are back to a fundamental problem of digital photography.

I've read, and quite happily believe, that we as photographers are at the cutting edge of this technology - others, who involved in the utilisation stage of digital images, may still in a state of catch-up. Add to this the unfamilar constraints of low contrast and easy blocking of seawater colour by any significant adjustment and we have a recipe for the inevitable problems such as yours to arise.

All this said, I now find that the results that I am achieving from my digital set-up underwater are better than those I ever achieved from 35mm film. Whilst I accept that not everything is sorted for underwater yet, the ability to achieve better fill flash balance and the incredible amount of detail delivered by the hi-res cameras means that I would not even consider shooting on fim again (a personal opinion).

There was a wonderful response on another web forum to the question "What do you miss about film?" - ans. "Those handy little plastic containers" - I'm in this camp now too.

Update.

The magazine is printed and the digital shots do not look too bad, if a little woolly, despite being manipulated at 8-bit CMYK stage.

Alas, not digital photography, a different feature with a DPS that was from an original 35mm trannie was accidently printed from the low res. layout file rather than from the drum-scanned original. It looks very different from what was expected but I wonder if any reader will notice?

That is the problem. It does not matter how good a job you do if someone further along in the process decides to screw it up!