What Viewing Light for Prints?

Hi everyone,

So we all know that prints can look very different from one lighting

to the next. What is a good light to use for viewing prints? I'm

not going to make some box in my appartment for viewing-- something

inexpensive would be best.

The ideal scenario is viewing the print in the light it will be displayed in. The standard is to view it in a D50 lightbox, which just means a box with good, 5000-degree-kelvin light that keeps other light sources from influencing your perception of the print. You could probably build one using good lights and a box of your own devising. Beyond that, real experts would have to chime in...

Hmmm... apartment... I would get the Sol-Source desk lamp by GretagMacbeth with the D50 lamb option. It is the size of your basic adjustable desk lamp but will provide a 5000k viewing light. I bought mine from http://www.chromix.com For what its worth, the viewing boxes are not all that big. Mine is 25" wide, 20" tall and 15" deep. I use the GTI PDV3D D50 viewer, also available at Chromix.

 

Ron

Most specialty lighting stores will sell 5000k compact florescent bulbs for about $20-25. I have some 55watters that are super bright, near perfect 5000k, and can be screwed into any standard lamp socket.

 

Too accurate in fact since using this type of light to evaluate prints will often yield prints that require equally good lighting.

<a href="http://www.keithlaban.co.uk">Keith Laban Photography</a><p>The one thing that is certain here is that your prints are going to end up in all kind of situations and will be viewed in all kinds of different light.<p>Take a tip from artists throughout the ages, use north light daylight to assess your prints. If artists aren't tearing their hair out over this why should photographers?

That would be south light daylight if you are from Oz :-)

I think there is some confusion above. Artists have used north light because a) there is no reference painting that they are trying to match and b) because north light stays relatively consistent in colour temperature for the whole day giving the painter more time to work with the same colour of light (ie, the painting doesn't "change colour" over the course of the day).

 

But that is far different from a photographer who took a photo, brings it into the computer and tags it with a colour model, views it on a monitor, makes changes to colour/contrast/tonality on the computer and (hopefully) coming up with something that they like, wants to print it out as they see it on the monitor. When the print is produced it can only be evaluated in terms of its consistency with what one sees on the monitor when viewed under 5000k light. The reason is that 5000k is an underlying assumption to the model upon which colour management is based.

 

It is true that the print will be viewed in non-5000k light. It's also true that the monitor can and often is set to 6500k (north light colour btw). It's possible to profile for another white point temperature for specific situations, for example gallery viewing. But the point of viewing the print under controlled lighting is to evaluate it for colour accuracy, colour cast, etc. in relation to what one saw on the monitor. Or to put it another way, if you want to know that what you see on your monitor is what you get on your print, you have to view the print under controlled 5000k light.

 

Noah, I'm in the same boat as you. I want to find a nice cheap compact 5000k viewing solution. Unfortunately, I haven't found one yet and don't think it exists. Its a game of compromises. Flourescents nominally at 5000k apparently are very spiky through the colour spectrum and don't just emit 5000k light. Ott lamps may fill the bill and are apparently available with 5000k bulbs but not always so you have to make sure which bulb you get with your lamp. Andrew Rodney, a well known colour management maven also claims that matching intensity of the light between viewing box and monitor is critical. He uses the GTI box with a dimmer, even more expensive than the GTI viewing box alone and far too expensive for the typical photographer's budget.

 

There have been several good discussions on this topic on robgalbraith.com colour management forums.

Just yesterday, I bought some 6500 degree Kelvin fluorescent "daylight" bulbs, 48" for about five bucks apiece. There were others at 5000 deg. K and 6200 deg. K. - all in various sizes. These might be useful?

 

I have an "artist's lamp" - double-arm table clamp lamp - that combines a ring of fluorescent around a regular 60 w incadescent. About a hundred bucks at an art supply store, and both handy and useful.

My simple solution has been to mount two bulbs of the same wattage, one

household incandescent & one 'daylight,' in reflectors equidistant from a print-

viewing table. The result is the same as buying an expensive viewing box

that switches between the two temperatures. A 'good' print is one that looks

OK under both (a) incandescent light & (b) both bulbs. I hope nobody is

hanging my prints in broad daylight. The bias should be in the other direction,

because many galleries use ordinary track light bulbs.

Judging a transparency on a 5000K lightbox makes perfect sense because I wouldn�t view a transparency any other way. Judging a monitor image set to 5000K makes perfect sense because I wouldn�t view a monitor image at any other setting. But why judge a print this way when it is unlikely to ever be displayed in 5000K light?<p><a href="http://www.keithlaban.co.uk">Keith Laban Photography</a>

Actually judging a monitor image set to 5000K doesn't make any sense at all, most monitors are set to 6500K.<p> A painting made by north light is going to look right in most daylight display situations. Same goes for a print.

The reason you have to judge a print in 5000k light has nothing to do with its applicability to the real world. It has to do with correctly (accurately) using your computers colour management controls and the theoretical system which underlies it. The system is predicated on 5000k as the foundation for all its computations. So if you want to know that what you see on your monitor is what you're going to print, the only way to make that judgement and have it be reliable and consistent is to use 5000k viewing light.

 

Of course you want to judge your print in all kinds of light, esp. the light it will be displayed in. The problem is that when you work on your monitor and then print out, the question arises "under what conditions can you say... yes, the print looks like it does on the monitor." Answer... under 5000k light.

 

There are analgous issues with sharpening. You don't just "sharpen an image" but sharpen it for a given viewing distance. And this isn't just relating to computer darkroom. In the OCT-NOV LensWork Bill Jay writes in End Notes how he went to visit Lisette Model to look through her prints. He was led into a "dim room where she set up a hard, straight-backed chair and then measured (yes with tape) the exact viewing distance from an easel. Floodlight was then angled just so, towards the easel..."

 

So here's the dilemma we face. If we are going to ignore "Colour Management Systems" in our computers' operating systems and image editing software, we'll get prints which are either ok or not on a random basis and we won't be able to predict what will happen based on the image we're viewing on screen. We'll spend a lot of time tweaking and printing and tweaking and printing each image at a great overall cost of time and ink and frustration. If we use the CMS systems in Photoshop and in the APIs of our operating system then we have to follow through to a very high, almost anal degree of fussiness in order to be sure that things are working under the controlled system inherent in CMS. Please don't take my word for it, read up from such colour mavens as Bruce Fraser, Andy Rodney, Jeff Schewe, Norman Koren, etc. I have links to all these people's stuff on my web site at: http://www.stonequay.co.uk/learn

 

And if you think this is anal, try reading what the guys over at the film and processing forum write about how to choose a developer and how many milliseconds at tenths of a degree to develop film x or read Ansel Adams explanations in The Print and The Negative

 

PS

Your better off setting your monitor to 6500k. Read Real World Color Management by Bruce Fraser to learn why

<i>"The reason you have to judge a print in 5000k light has nothing to do with its applicability to the real world"</i><p>Well, you said it Eric.<p><i>"And if you think this is anal"</i><p>Oh yes.<p>By the way Eric, I just visited your website and community page here on photo.net hoping to see how you apply all this to your images :-)<p><a href="http://www.keithlaban.co.uk">Keith Laban Photography</a>

Keith

 

I'm not sure what your agenda here is. In answer to the poster's question, I wrote a professional answer. It's something I know about. Are you suggesting that I'm wrong or are you just unhappy with the answer to the question.

 

If you're looking for my photographic images, you can find them at my personal web site www.kiiquu.com. If you read my bio, you would have seen that.

 

If you want to play my dog is bigger than your dog, I'm happy to say that I've looked at your photography before and I think its stunning. You obviously know what you're doing with a camera. But I don't think you're much of an expert on computer colour management. That's what the original poster asked about. That's what I wrote about.

 

Lastly, when it comes to making friends and influencing others you seem to get a kind of kick in sinisterly suggestive innuendo. Not flattering for a guy who takes such pretty pictures.

Eric, I�ve calibrated everything but the cat�s arse and got profiles for things I could get arrested for. My monitor and prints are as perfectly matched as I would ever want them to be. I�ve had profiles supplied to me for testing by someone who like you sells profiles over the Internet, but come Judgement Day I use my eyes and God�s own light, Hallelujah.<p>Seriously, what really winds me up is when people who should know better imply that a perfectly calibrated and profiled system will automatically result in the perfect print. Printmaking is not an exact science; in fact it�s not a science at all. If I have offended you in any way then I apologise, but this is something that I tend to get a little heated about :-)

<i>Seriously, what really winds me up is when people who should know better imply that a perfectly calibrated and profiled system will automatically result in the perfect print.</i><p>

 

Yes, not only will a perfectly profiled system not automatically result in a perfect print, its a physical impossiblity to have a perfect match between monitor and proof, a point I've emphatically made many times at photo.net if you care to check up on my recent posts. I also carefully avoid the implication at my business and all my customers know that I caution them against this kind of thinking.

<p>

And the reason I started Stone Quay Studio profiles is because I wanted to have an inexpensive alternative here in the UK, a kind of profiling for the rest of us like Cathy's Profiles in the US precisely because I feel that colour management is an overhyped and over priced but valuable tool. So I'm totally with you on this.

<p>

The point I've made now 3 times on this page and the only point I'm trying to make is that if you ask yourself ...does my print accurately look like what it looked like on the monitor (notice the absence of the word exactly but accurate is what you will get) you need to view the print under (pure) 5000k light.<p>

 

And I tried to make this point in answer to several comments about other colour temperature lamps/light or even flourescents which seem to be a good 5000k source on the surface of the issue but turn out not to be so satisfactory in terms of answering the question. I pointed out that not only is colour temperature an important factor but also equal luminosity is critical. And I pointed out that unless one becomes anal about things, you wind up making compromises which at some point everyone has to make. At which point I at least, and you also apparently, get back to the thing we love which is creating visions.

<p>

Given the above, I still don't see why you're all wound up with emotion at me. By implication, in your last post you're putting words in my mouth by talking about unspecified profilers who promise the world and should know better. Well since I'm the only profiler around in this thread and since you engaged me earlier, it does seem you're implying this about me. I don't argue anywhere that you/anyone should not judge the quality of the print by one's artistic eye which in your case is really excellent in my opinion. And even without an art school education, I do the same although I have lots more to learn.

 

They're simply two separate issues. Issue one, when is a print good? You have that down pat. You have my respect. But its not the issue which was asked here. Issue two which is the issue of this discussion thread, what is a good light for viewing prints in the context of a digital darkroom and that is the question I've tried to answer. And your answer about your eye being the final judge has nothing to do with the question.

<p>

Further, I do argue that if you want to use a highly complex tool like a computer colour management system, you need to understand it and do it right or you add complications which quickly lead to the same frustration you were trying to avoid by taking the system on in the first place.

Wow, guys! Thanks for all the input. Even the fighting was quite amuzing, but has left me a little confused. Frow what I've heard from everyone (on this site and off) I gather the cheapest option is to get some 5000k bulb. Some have recommended a flourescent bulb, which is very inexpensive, but I've heard hear they are not good. What about the Ott-lights (making sure I get a 5000k bulb)? Should that be my route? I'm really not going to spend a lot. And this question of "something with the right brightness"- how do I figure that out. Is this just a matter of the correct watt # on a 5000k bulb. What watt? I've calibrated my monitor with a Spyder (Photocal), which adjusts my monitor to the correct brightness.

noah

Worst case scenario: during the day the prints get indirect natural light ("northern window") that has lots of blue, then in the late hours the print gets incandescent light that has very little blue and lots of yellow. In this case, there is no way to make a print that looks good in both kinds of light, so a comprimise somewhere in between is needed (good luck on that one) The real issue here is that many artificial lights tend towards the "warm" look: incandescent, halogen, and warm white fluorescents. I also found the "daylight" compact fluorescents that Scott mentioned, but only online... all the ones at the local hardware stores are warm white (to look more like incandescents they are replacing) Similar issues with large fluorescents... cool white being the common culprit. Using "daylight" 5000K is a good start, but the CRI is just as important (less spikes, more even distribution of color sprectrum) I'm hoping the daylight bulbs will become more popular (and available) They have come down in price, and also recommended for better mental health, so I don't see why anyone would use anything else (previously, they were too expensive, but that's changing) As long as people still use incadescents, warm white, ect. there will always be this issue for prints (unless you supply a light with the print)

Noah, as I mentioned, I'm also looking for an inexpensive solution to this lighting issue. I did some reading at imagereview.com. The Ott lamp comes in several flavours (5000k, 6500k and others) and various people had negative statements to make about it. The general consensus from what I read was that the best/cheapish solution was Solux lamps. You can do google search for suppliers of solux, there are many. The Solux are available as bulbs and came at different temps and wattages and the one recommended was the 4700K bulb which is slightly warmer than 5000k but has a very high CRI which is its purity rating. You can buy this as a bulb or as one of several styles of lamp ranging from clip on to floor standing. There is a dimmer but the problem with it is that it dims and also changes the colour temp of the bulb and it wasn't recommended. But the bottom line was that its cheap and close enough and this from the likes of Andy Rodney and Bruce Fraser who are among the real central chaps in the field.

 

The brightness issue you mentioned is simply this. The intensity of the light of either the monitor or the 5000k bulb affects the apparent intensity of a colour either on your monitor or the print. So if your monitor is brighter than your 5000k bulb the colours on your monitor will look brighter than your prints, and vice versa. So the idea here is to get them even. The cheap compromise is to move the lamp further from the print to equalise the "brightness". Is this a good solution if you're doing colour critical professional work? No. Are you? No.

 

So what I'm leaning towards right now as a result of my research is to buy a solux bulb and cheap lamp and see how things go from there. As I said, you are always going to have to make a compromise somewhere.

 

Lastly, don't forget that all of these points aren't about looking at the print in the environment in which it will be seen and whether or not it will look good in that environment. These endless colour management points are about trying to create the right environment for analysing the closeness of the print to what you saw on your monitor.

Eric, that sounds like a great light source. Color rendering (CRI) is very important... maybe more than the rated kelvin temps. Actually, most of what I know about lighting is from previous experience with greenhouses and indoor gardens, and I'm used to looking at spectral response... the color temp alone doesn't say much about the range of colors or the peaks (spikes)

It's ironic Jeffrey that I keep finding daylight bulbs advertised in the UK under embroidary and sewing.

 

Anyway I found this link for solux and you can see the CRI is 98 or 99!

 

http://www.solux.net/

 

Also, GTI has an excellent pdf which covers the technical specification of why/how all of this works. There is an ISO standard: ISO 3664:2000 which describes how and why one needs to use 5000k light and control lamp intensity/monitor intensity. Unfortunately I downloaded it from imagingreview.com and that requires (I think) membership before you can see it. But its a GTI document (they make viewing booths).

Great! Now I'm getting somewhere. The Solux site looks like what I'm going to use. Just for others interested, it seems that at this point they do not just have bulbs that work with regular lamps you might have, and you need to buy one of their lamps (the clip on one is only $60 something, though). So I'm thinking maybe the narrow-flood would be a good angle, but what wattage? They have 35 and 50.

Thanks,

ns

Easy to find daylight flourescent bulbs. I have a light viewing test srip that indicates the above bulbs from WalMart that are installed in my digital studio are accurate 5000k bulbs. You need no box or expensive lights. These bulbs last a couple of years.

Noah, I'd recommend the Solux 50w, 4700k, 36 degree. If the light will be very close to your prints, maybe a 35w if they make it. Any standard 12 volt halogen fixture should work.

Daylight balanced screw in flourescent. I bought mine at Walmart a few dollars each. I use two of the highest wattage they sell, in my office. I have a GATF/RHEM light indicator that tells me my office lighting is 5000K.