large scale printing
Posted 2 years ago
I have previously printed large scale images from film at 20 x 30 inch but I have been asked how big I can print one of my digital images. So my question to you guys is how big comfortably can I print an image shot with a nikon D80. I know I could interpolate it to any size but what do you feel is the maximum with out loose too much quality is 20 x30 achievable. The last thing I want to do is say a size then not be able to achieve it.

Hope you can help

Ian
 
King 
Posted 2 years ago
I say that anything over 10 mpx is capable of printing well up and above 20x30"

From a 4 jpeg mb file I printed a 20x30" out of a point and shoot camera and the quality is stunning. Saved it originally at 300 dpi. I can probably go a lot larger than that. With some of my aurora shots done with a 21.1 mpx camera I find I have to go large...otherwise the details won't come out. 8x10" won't show all the details in the print, it won't nearly look as cool as a 24x36". The minimum size I do now is 13x19" to ensure that at least some of the detail will be visible. I don't know how or why it works, it just does.

The only thing I'd never do is physically increase the size in photoshop (increase image size) as that will only smear the pixels and smudge details. With large prints they are often not viewed from close up, anyway, so looking at how an image looks from arm-distance isn't practical if it's meant to go on a wall and be viewed at least 3-5 feet back.

Bill boards are another example of this. In newsprint too. A 60kb file at 72 dpi can look very similar to a 3mb file at 300 dpi in newsprint. Quality of the print is much more dependent on the output process, though, than anything.

In all practicality DPI and PPI are absolutely meaningless when you get down to it.

 
King 
Posted 2 years ago
That said I recommend trying out a size at least once to see how you like it. Imagine a client purchases a $1200 30x40" canvas stretched at 0.75. You see it on a website so you figure yeah why not, in theory I could do that. I think it would be worth trying out the printing house to see how well they can do it. Many labs give out free samples.

Sometimes even if it is feasible often a client doesn't know how large 24x36" really is.

30x40 and 40x60 are the other two "ultra large sizes" i have available.
40x60 being nearly double a 24x36, which on paper doesn't look like much but in print -WOW-

 
Posted 2 years ago
It depends a lot on wahat kind of image you print. I've made 110 x 200cm prints from 10 Mp images (approximately 43 x 80 inches). This was a landscape photo with quite a lot of detail. And it came out very nice. I used photoshop for the resizing, but increased it in smaller steps (recommended).
I believe that an architecture shot with sharp lines might be more difficult to scale up that much.
Karl Johnston wrote
The only thing I'd never do is physically increase the size in photoshop (increase image size) as that will only smear the pixels and smudge details.


I think photoshop is a bit smarter than that. But Genuine Fractals is supposed to be better (top of my shopping list now).
I make my prints on a Epson 9900

 
King 
Posted 2 years ago
I use lambda printers exclusively, though I remember reading that it was a bad idea to enlarge via photoshop even when printing giclee. Here's a quick article I found about it on google:

Why you should not enlarge

I use a epson 3800 for general giclee prints and portrait though it does create some amazing prints. Especially black and white. The BW mode is intense.
 
Posted 2 years ago
Well, at some point it has to be enlarged. If you don't do it in photoshop the printer software will enlarge it.
The link seems to be directed on people who don't usually handle images in software
 
King 
Posted 2 years ago
I suppose, though at the example posted I would leave it up to the pro lab to handle that stuff and just send them my TIFFs :p let them figure it out
 
Posted 2 years ago
Karl Johnston wrote
I suppose, though at the example posted I would leave it up to the pro lab to handle that stuff and just send them my TIFFs


Ok then. The 9900 is in house, my house..
 
Posted 2 years ago
Thanks guys for your feedback.

Lars what kind of pixel step do you use when you enlarge an image? a description of your work flow would be cool.
 
King 
Posted 2 years ago
Lars Grepstad wrote
Karl Johnston wrote
I suppose, though at the example posted I would leave it up to the pro lab to handle that stuff and just send them my TIFFs

Ok then. The 9900 is in house, my house..

I'm not sure I understand, I don't doubt the location of your 9900. My epson 3800 is in my house too..cool eh? We're practically brothers now
 
Posted 2 years ago
Ian James wrote
Lars what kind of pixel step do you use when you enlarge an image? a description of your work flow would be cool.


I try to step it up by about 10% until I have the size I need (at 240dpi). This method has been recommended by several people with experience in the field. But, I don't know how much difference it makes in the latest versions of ps. It seems like a direct upscaling works well too. Not really tested it, but tried a few times. Then you have Genuine Fractals.. they claim you can go up something like 3 times as large as you can in ps with the same quality. I have a large format project I will print soon, 3,2meters wide from one frame. Will try out Genuine Fractals for it.

Karl Johnston wrote
My epson 3800 is in my house too..cool eh? We're practically brothers now


Hello little brother...:-))
 
Posted 2 years ago
I think Lars is saying that as he prints at home he does all the scaling himself rather than sending out for someone else to do it at the same time as printing.
 
King 
Posted 2 years ago
Oh! I get it now. Was tired at that time in the morning :p woops.
 
Posted 2 years ago
I believe the 10% step-up myth is to do with older algorithms that were not as good as todays versions. Some time ago, as a test (in cs3) I enlarged a shot both in one go and one using 10 x 10% steps method. Image matching the result showed no difference between the two methods at all.
 
Posted 2 years ago
Peter Davidson wrote
I believe the 10% step-up myth is to do with older algorithms that were not as good as todays versions. Some time ago, as a test (in cs3) I enlarged a shot both in one go and one using 10 x 10% steps method. Image matching the result showed no difference between the two methods at all.


I thought so too, and I found the same. But never really tested it. Still Genuine Fractals is on my list. Anyone with experience?
 
 
Compose a reply
You must sign in if you want to post a reply.
Fine Art Prints  -  Our books  -  Work with us  -  FAQ  -  About 1X
© 1X Innovations AB 2007-2011. All rights reserved.
 
 Stumble 1X