If my banner is 32.25x85 how do I get stockphoto that will fit it?

It will get pixelated, because you are stretching the maximum size of the largest image.

You can enlarge a photo to about twice its native size in photoshop without it getting pixelated depending of course on the ISO it was shot at and its dpi. Make sure to use “image size” in photoshop to do your enlargement. First enlarge on “bicubic smoother” to a size a bit bigger than you want it to be (so maybe 100 across), then reduce its size to the size you want it to be (85 across) with the image size set to “bicubic sharper.” This is the best way to avoid that typical “pixelated” look when you make an image bigger.

@fireinthepriory

What has ISO to do with enlargement? It has to do with film speed/light sensitivity. The DPI of the image also has little to do with it, the whole issue is total image size and sharpness. There are many stock photos capable of being printed at those dimensions without difficulty – but they will cost more. You also have the Photoshop technique wrong – with enlarging you use bicubic smoother, not sharper.

@DarkScribe
The DPI has a lot to do with it. In fact all that really matters is the image size and DPI. I think fireinthepriory meant if it has a quick ISO, then there is low light, meaning maybe some artifacts that would be exagerated in expanding.

If its still not large enough, PS has some very good enlarging techniques. I would google for them.

@DarkScribe you and @fireinthepriory are in agreement regarding the photoshop technique:

First enlarge on “bicubic smoother” to a size a bit bigger than you want it to be…
…then reduce its size to the size you want it to be… with the image size set to “bicubic sharper.”

@jackm In fact all that really matters is the image size and DPI.

DPI is an indicator of resolution, often incorrectly used to refer to both screen display and printed image resolution. The image can be displayed or printed at ANY resolution. If the image is large enough then that resolution will suffice for an acceptable output at a large print size. That means that only the image size and CLARITY is important. No matter what else you have, if the source image is not sharp enough you cannot make and acceptable large scale print. As for ISO, it is to do with taking the photograph – nothing to do with printing an image. I have no idea what you mean by a “quick” ISO. (And I have been a Photographer for more than thirty-five years.)

PS has some very good enlarging techniques. I would google for them.

It doesn’t. End of story. It cannot create detail that doesn’t exist.

I have been using Photoshop for nearly twenty years – since before it was bought and renamed by Adobe.

@Beta_Orionis _First enlarge on “bicubic smoother” to a size a bit bigger than you want it to be…
…then reduce its size to the size you want it to be…_

Why? You enlarge to the final print size (or at least the shortest edge ) then crop if necessary. If you need to smooth you can run filters to reach your desired result. There is absolutely no point in degrading an image more than is necessary.

@DarkScribe
From wikipedia: “Highly sensitive films are correspondingly termed fast films. A closely related ISO system is used to measure the sensitivity of digital imaging systems. In both digital and film photography, the reduction of exposure corresponding to use of higher sensitivities generally leads to reduced image quality (via coarser film grain or higher image noise of other types).”

It don’t know how you could be a photographer for 35 years and not know this. Fast films gave a lower image quality.

And DPI does matter. Let me give you an example
I have one 10in x 10in image at 1000 DPI and one at 50 DPI. Typically print should be 150 DPI (100 if your stretching it), so the 1000 DPI image can be enlarged to about 100in x 100in. The 50 DPI would have to shrink to look good in print.

Photoshop CAN create detail that doesn’t exist. There are very complex filters and algorithms that can do this to some point. Not infinite, but enough to matter.

I have worked with photoshop for a long time. I worked in a print shop for over 2 years, and did graphic design for much more.

@DarkScribe I didn’t suggest it! I agree that it’s pointless.
I’m just saying you misread the comment to which you first responded!

@jackm

The reference was to quick ISO, not high ASA film speed.

Photoshop CAN create detail that doesn’t exist. There are very complex filters and algorithms that can do this to some point. Not infinite, but enough to matter.

Nonsense.

@DarkScribe
Thats what I was talking about.

Nonsense.
I am currently an electrical engineering student and I work in signal processing and imaging all the time. Trust me when I say this: Algorithms exists that do this. They range from the very basic, to the math intense. Since photoshop is the top of the line, they employ some very smart engineers to come up with good enlarging techniques.

Look at this website if you do not understand. Some of the more basic ones are described in it.

I don’t know the ISO as these are stock photos; from my understanding that is something done when taking the picture. Please correct me if I’m wrong.
If I take an image that is 9.5×14.3 @ 300 dpi how big can I get that image. Is there a way to know before I purchase the photo if it’s going to work on the banner?

@jackm _Trust me when I say this: Algorithms exists that do this. _

Maybe you had better let Adobe know – they warn students that this is one thing that Photoshop can’t do.

As for your credentials, I am not an Electrical Engineering student, I am an Electrical Engineer, ex Navy who now works in print media. I am a journalist/editor who specialises in IT, technology and Lifestyle. I work on a daily basis with Photoshop and InDesign plus the rest of CS4 on a less regular basis. Photoshop CANNOT create missing detail, it simply isn’t possible, it can create an illusion of detail, but that will not stand up to large format printing.

@cbazelewick
As I said before you can get to about 100 DPI if you want to look good maybe even 72 DPI.

@cbazelewick

What stock agency are you using? Most will indicate the suitability for various size prints.

istock photo

@cbazelewick

Istockphoto is not the best, but it is inexpensive compared to some. What is the image subject or stock number? (Are you casual or do you have an account?)

there are a few options. I’m looking for a photo of St Mark’s Square\
http://www.istockphoto.com/stock-photo-2329298-piazza-san-marco.php

@DarkScribe
Did you read the link I posted? Lets do a thought experiment. Say we have a 10 by 1 pixel image. Its a white to black gradient. This is only 10 different shades. Now, lets enlarge this gradient in PS to 1000×1. We will get much, much more shades. This is called interpolation, its photoshop guessing at the data that should be there. The algorithm you first suggested useing, bicubic smoothing, is a form of interpolation. That is photoshop creating data that did not exists before.

There is no such thing as the illusion of detail. Its there, or its not.

I am sorry for doubting your credentials, but you seemed not to understand ISO, DPI or enlarging techniques, so I figured you were just guessing at things.

@cbazelewick
did you try http://www.sxc.hu
Its a good free stock photo site

@jackm Lets do a thought experiment

Let’s not. You are clearly not a professional – you are waffling about thing that don’t work in practice and I have little interest in your theories. I work in the real world of print media.

@DarkScribe
As do I. Simply because I am younger than you doesn’t mean I cant understand things. In the quickly changing world of digital media, it probably means I understand things better than you.

The fact that you ignored my contradiction to your argument doesn’t make it less true.

Try admitting when your wrong.

@jackm This is called interpolation, its photoshop guessing at the data that should be there. The algorithm you first suggested useing, bicubic smoothing, is a form of interpolation. That is photoshop creating data that did not exists before.

This could be almost amusing. Guessing as you describe it is not creating. What it the original was quite different?

@DarkScribe
I never said it would be the exact correct data. I said it could create data. Look through my posts

@jackm

Say I take a photograph of something really quite simply. A flyscreen with a fly sitting on it. If I use a nice sharp lens, get a really sharp shot, one that shows the fly in sufficient detail to identify its species, then another with a softer lens and select a lower resolution, one so that the fly is just an indistinct blur. Are you suggesting that software can make the second image sharp enough to see that the blur is actually a fly?

@cbazelewick
Another technique I use for finding large images is google search. Search for the image, and then define the size you want on the left side. If you see a good one, contact the owner of the image and ask for permission.

@DarkScribe
No, I am not suggesting that. Until we have artificial intelligence, that can’t be done.

@cbazelewick

Sorry I am being distracted. That image is not of a high enough standard to produce a poster at the dmensions that you want. There are many other such shots though. Have you tried the “Flickr” commons?

I know that there are several shot similar there, I saw them a few weeks a go.

@jackm No, I am not suggesting that. Until we have artificial intelligence, that can’t be done.

Right, but although filters can interpolate and make the screen appear much sharper, they cannot add the small amount of missing details that constitutes the fly. They can analyse the large area of screen, find edges, and sharpen that way, but once the pattern is broken they have nowhere to go.

@jackm I never said it would be the exact correct data. I said it could create data. Look through my posts

… and I said “missing” detail. Look through my posts.

do I just contact the owner of the photo and ask to purchase on Flickr commons?

@DarkScribe
Yes, But all I argued was that Photoshop can create data.
The example I used was just a basic interpolation that could be calculated by hand using a matrix. There are much, much more sophisticated algorithms, such as edge detection, bayesian analysis(slightly different, comes from many low res images, not just one), and many others that can predict and create this “missing detail” that you mentioned.

Computers are smart. Your brain can guess at what detail is missing, and so can a computer. It won’t be able to reveal what’s behind a tree in a JPEG, but it could enhance a face, or fly, supplying the missing detail.

@cbazelewick do I just contact the owner of the photo and ask to purchase on Flickr commons?

The commons are the royalty free areas. if you find one that is copyrighted, talk to the owner and ask for the image. Many people there are quite flattered when someone want to publish one of their images and will often agree as long as they are credited with taking the photo.

@jackm
My comment:

It doesn’t. End of story. It cannot create detail that doesn’t exist.

Your response:

Photoshop CAN create detail that doesn’t exist.
The discussion was about photographs and printing at large print sizes, requiring high resolution, not about adding non relevant detail.

that’s great. I found a photo; I hope it will be big enough. As for the flattery, it’s for a pretty recognized brand, so i’m sure they will be happy.
Thanks for all your help.

@DarkScribe
No matter how much you argue, photoshop CAN create detail that doesn’t exist. An interpolation creates detail that doesn’t exists, I showed this to you with my thought experiment.

I am done arguing here, there is nothing more to prove. Maybe you are stubborn, maybe you are ignorant, either way, I am not going to change your mind.

@jackm

You are missing the point between adding irrelevant detail and re-creating missing detail. I use it to add copyright information – that is adding detail, but is has NOTHING to do with improving the quality of the image by adding detail that ineffective photography, low resolution, poor focus, limited file size etc., has created.

@DarkScribe
How is correctly interpolating a gradient irrelevant? But beyond that, photoshop can add relevant data, using the algorithms I described, and linked to. It can enhance edges, It can find patterns and use that info.

This has become an argument of whether the algorithms do a good job or not. I say the data they add is relevant. It fills in what I need.

You come from the days of when we were all taught that computers can do nothing for themselves. That we have to tell them everything. Those days are gone. Computers are smart, bordering on intelligent.

@jackm Computers are smart, bordering on intelligent.

No they aren’t. Some running AI software might appear that way, but the most powerful computer is as dumb as a stump. They will be smart one day – when we have artificial sentience. That day is not close.

@DarkScribe
Thats a matter of opinion. What’s the difference between appearing smart, and being smart? Computer intelligence already outsmarts us in certain niche fields. They are quickly taking over more and more areas. I believe that the singularity will happen in our lifetimes.

Either way this is now way OT, we are into philosophy and futurology.

@jackm Computer intelligence already outsmarts us in certain niche fields.

There are no intelligent computers. A computer is no more intelligent that a book, it can contain knowledge, but not act on it. It can only follow a program. A program means a planned series of events. The closest you can get aside from AI are things like chess programs – but even then the computer is not “thinking” it is just reacting to input as per its program.

@DarkScribe
That is the old view of computers. Computers can learn, can modify their code, can create music. They can deduce, solve problems, interact with humans

Computers can tell if we are lying, can land a plane, can analyze martian soil.

Computers have done things they were not programmed to do. They have surprised their own programmers. It really depends on your definition of intelligence. I say they have intelligence (and sentience soon)

@jackm They can deduce,_I say they have intelligence (and sentience soon)

Sorry – you are really naive and gullible. (That was the politest thing that I could think of.) Computers are boxes of electronics that can be programmed to respond to input. Just tools, nothing more.

Conversation over.

@DarkScribe

Thanks for being so kind. I am sorry you don’t understand all of this. Some things just go over our head.

@jackm Some things just go over our head.

Some of us are taller than others in that case.

Wow. Quite the convo I missed! I readily admit that I know less than either of you – as an undergrad I was essentially a photography minor, so all I can tell you is that I was taught to use the procedure I described in photoshop to enlarge an image. The reason you enlarge to a bit more than you want and then downsize again is that it will look sharper than if you simply enlarge to the size you want – I know this because I’ve done it and compared. Yes I was talking about ISO in terms of digital, but @DarkScribe it makes sense in terms of film, too – as I’m sure you know, 35mm 100 ISO film will show MUCH less grain than 35mm 1600 ISO film printed at the same size. Similarly, if you try to enlarge a digital image shot at 1600 ISO to twice its size using the above method your photo will look like shit, if it was shot at 100 ISO you’re fine.

Anyway, from what I read (ok, occasionally skimmed) above you’re BOTH right. I don’t think there can be any argument that, yes, algorithms used in photoshop can create new information allowing an image to be the same dpi and larger but not “pixelated.” The argument was about whether this information “existed” or not. @DarkScribe is right, what the algorithms are creating isn’t “recreating” the exact detail that existed in real life but wasn’t there in the original photo – however @jackm is right that it’s using the available data to make a pretty good guess that will often approximate the real thing, thus creating new and meaningful information.

Hope that made sense… I’m very sleepy and off to bed now!!

@fireinthepriory . Similarly, if you try to enlarge a digital image shot at 1600 ISO to twice its size using the above method your photo will look like shit, if it was shot at 100 ISO you’re fine.

This is entirely camera/sensor related – some cameras can perform as well (noise wise) at 1600 ISO as others at 100. I will often shoot at 1600 with barely discernible noise where a couple of years ago 400 was about the limit. Nowadays with some sensors and using Ninja and Neat Image you can work with up to 6400 ISO in a pinch.

@DarkScribe
You can’t just let someone make a statement, can you? You have to prove your intelligence.

@jackm You can’t just let someone make a statement, can you? You have to prove your intelligence.

Certainly I can – as long as it is accurate. (I don’t need to prove anything – if I did here would not be the place to do it.) This is a question and answer forum, the presumption is that people want accurate answers.

BTW, experience is not intelligence.

@DarkScribe I know it’s camera-dependant, but sadly I have a Canon Rebel so for me 400 is still the limit!

look for one that is a good crop, or make a montage, or shoot your own.
this blog changes its banner daily—look at some of these
Perfect Duluth Day