My wife wants to submit some pics of her paintings to a magazine. Their requirement is 300 dpi. Shouldn’t that be “pixels per inch”?
Anyway, how the heck does one determine if an image is that high definition?
I look at the size of some pics I have taken and they range from about 600 KB to 2 MB, taken with the same camera. I right click on one of these and go to Properties and it gives the sizes in bytes.
How do I determine what dpi these images are?
Unless you know the physical dimensions of the image, it’s meaningless.
I guess you could assume an 8/5" x 11" image, and get (8.5 x 11) x 300 x 300 = 8.4 Mpixels.
Digital images don’t have DPI in any meaningful way. They have dimensions; say 1200x1600 pixels. How big you intend on printing it determines your DPI.
File size means nothing, since it doesn’t represent image dimensions directly due to compression algorithms. Ask the magazine how big the image will be printed and work backwards from there.
The size varies depending on “quality” as well as most cameras have a setting for size. You can save the same dpi photo using less memory, but the result is a bit fuzzier in detail.
2MB at typical high quality is probably about 6Mpixel. (What does you camera claim it is? You can probably google it.) In windows 7 my computer shows the details if I just hover over a JPG in the Pictures folder listing. Otherise, right click, properties, and I have a Details tab that lists the size.
6Mp in a 3:4 aspect ratio is around 2800x2100; at 300dpi (dots per inch, pretty much same as ppi) that’s about up to a 9x7inch photo.
Fancier photo programs can save a purely meaningless dpi attribute with photos as well. But really, photo processing is just “how big do you want this picture?” If the DPI is too low, say less than about 100dpi, the picture will be noticeably fuzzy. The magazine just wants to say we want a fairly good resolution photo for the article, not some fuzzy phone camera stuff.
You set DPI by specifying print size.
Photoshop resize does this. Tell it the dpi you want and print size. It will resize the pixel resolution to whats needed. Usually that means dropping unneeded pixels. Image isn’t degraded. But, be sure to use save As, and rename. You want the original image as a backup.
You never want to increase pixels. That will degrade image quality significantly.
Modern cameras produce huge pixel resolution. Getting a 4x6 300 DPI image is not a problem. Even 8x10 at 300DPI is possible with high end cameras.
300 dpi is not that high/big a resolution. Any picture taken with a camera bought in the last four years will easily have that resolution (and usually way, way more), as long as you haven’t zoomed on a detail or altered the settings on your camera, or had the picture e-mailed to you (mailing programms often make the dpi smaller so the picture is less KB.
4x6, 300dpi => 1200x1800 or about 2.2Mp
5x7, 300 dpi => 1500x2100 or about 3.2Mp
8x10, 300dpi => 2400x3000 or about 7.2Mp
You’ll have a hard time nowadays finding a camera that does less than 7Mp.
That’s not to say the picture is automatically good. It still has to be in focus, well lit and the camera held steady for a good shot.
Some people have a bad habit of taking picttures at 2Mp or worse (You can set some cameras down lower), because it still looks great on a computer monitor or in facebook. But printed or in a magazine, a picture of 100dpi will look noticeably less “crisp”.
300 dpi, on its own, doesn’t mean anything useful. A one megapixel file can be 300 dpi, as can a 36 megapixel file. You need to know dpi and dimensions to get an idea of what resolution is requested.
And there’s nothing wrong with upsampling. If you have a good file, you can get away with enlarging it quite a bit before the file starts to break down, especially using algorithms like Genuine Fractals (or whatever the current incarnation of it is called.) I have 2.7MP files printed up to 13"x19" at 300 dpi that look good, even though that resolution is about 6.5"x4.5" at 300 dpi without interpolating.
yes, I have a 2.2Mp file from a Fuji camera; the sensor was a diamond pattern, and the camera itself interpolates to 4Mp. I hav a print that is blown up to 11x14. it is somewhat grainy on close inspection, but otherwise quite good. it’s slightly better than 100dpi.
Depending on viewing distance, about 75dpi to 100dpi is as bad as you want to get. (However, if you look at photo posters, even commercial ones, you will see the quality can be quite fuzzy - we’re just meant to look at them from farther away. )
300dpi is an arbitrary but doable number because the magazine has to set a minimum quality. You don’t want fuzzy pictures in a quality magazine. part of National Geographic’s appeal is that the photos are not only striking, but usually highly detailed - which comes from the resolution of the photos themseves. it also depends on detail. Fuzziness in a flat expanse like a face is less noticeable than in a detail like hiar or clothing weave.
I know that even with the 2.2Mp Fuji, when I first had photos done at 4x6 the phenomenon I decribed was “painfully sharp”. being used to crappier point-and-shoot prefocused film cameras, the resolution quailty of digital was overwhelming. I did a photo book in My Publisher at 11x15 of these pictures for a trip I did to Italy, and some at full page size still look good.
Isn’t DPI an archaic term now with color lasers and inkjets?
DPI was a dot matrix printer term. Literally dots per inch that it printed. DPI was a mechanical limitation of the printer.
I’ve read the specs on our HP Color laser at work. From what I understood, it doesn’t print dots like a dot matrix did.
The scanner on my PC has a dpi selection range of 50 to 600, and the difference in quality is obvious. The higher the dpi the sharper and clearer the scan. The magazine is simply asking for quality scans that can be printed on the page. A lower quality would be useless as they probably don’t want to print blurry unclear pictures.
No, it still has meaning.
Even though lasers and inkjets might be able to print variable-sized dots, they still have a maximum physical resolution, which is determined by the mechanics (and electronics and optics) of the device. This sets and absolute upper limit on how many dots per inch can be printed on a page. Often, this number is higher in one axis than the other.
DPI originally was used to describe the halftone screenused to print B&W photos. The usage become a little more complicated when color halftones were introduced, and then even more complicated when stochastic or diffusion dithered printing techniques were implemented.
Most images don’t go through a scanner, they are either straight from a camera or an image processing program. The question isn’t high quality over low quality, but how large dimensions are needed to achieve 300 DPI for the physical print.
I agree that most images from modern cameras are close enough to satisfy the number of pixels needed. But not all pixels are equal. Compact P&S cameras may produce large images, but they’re rarely crisp and clear unless shot with good light.
There’s more to it than that, particularly if it’s going to be printed in a magazine. There is a lot more to image quality than MP–sensors can vary in quality with the data they can capture, noise levels, etc.
They use a more sophisticated technique to get better image quality but they are still printing digital pixels and there is still the concept of DPI to express printer resolution capability and printed images based on source image resolution (also scanner resolution). Laser printer spec sheets include the DPI capability, like this HP LaserJet 500.
So, when I’m working w/ an image in P’shop and I resize it from 22 x 16 inches and 180 pixels/inch (straight from the camera) to 300 pixels/inch (unchecking “resample image”) it now becomes 13.3 x 10 inches. Have I altered the “quality” of the image?
No, DPI does relate to the number of dots per inch printed onto a sheet of paper to form an image, but it isn’t a dot matrix term and doesn’t relate to the printer type. Magazines are normally printed either by sheet fed offset litho (where the image transfers from an inked plate onto sheets of paper in a stack) or more likely by offset web printing (similar to litho, but the paper is on a big roll - cheaper than sheet fed but lower quality). Some shorter run magazines are now printed digitally, but the ‘dots on paper’ method of forming an image remains the same.
If you can look at a magazine under a magnifying glass, you will see the dots on the paper.
OP, 300 dpi is standard high resolution required for litho and digital printing, but it’s meaningless unless you know the dimensions of the final printed image as it will appear in the magazine. ie, a small 300DPI image won’t work if the magazine want to use it as a double-page spread, as blowing the image up reduces its dpi.
Do you have a photo editing programme, like Photoshop? This would allow you to see the dimensions and DPI of the image, and adjust them.
No.
You’ve just made the pixels themselves smaller.
No - the image dimensions go down as the dpi goes up, but the same amount of data exists in the image. You will note that the file size (in MBs) will not have changed.
What you can’t do is increase the dpi whilst retaining the same image dimensions, as you would be effectively trying to add information to the image that doesn’t exist, and ending up with a fuzzy image.
All he’s done is change the meta data for the image file. Nothing else has changed. The program responsible for printing may interpret the data differently but no information has been gained or lost.
I believe that’s what I said…