Sign in to follow this  
Followers 0
david.atwell

DPI and YOU - Understanding Resolution for Print and Web

76 posts in this topic

Almost two years, actually.  :-)  I was on a bit of a forum sabbatical between 2009-ish and last month.

 

Glad you liked it!

0

Share this post


Link to post
Share on other sites

Glad you liked it!  Thank you!

0

Share this post


Link to post
Share on other sites

David, I was going to leave a short comment:  This was great!!!

But then decided this tut calls for a longer comment:  This was grrrreeeeeaaaaaattttttttt!!!! :P

0

Share this post


Link to post
Share on other sites

Thaaaaanks for your appreciaaaaaaaaaaaaation :-)

0

Share this post


Link to post
Share on other sites

^^ What did vowels ever do to deserve this?

0

Share this post


Link to post
Share on other sites

You're right.  Vowels don't kill people.  Ppl kll ppl.

0

Share this post


Link to post
Share on other sites

This isn't Wheel of Fortune. It's not like you're charged by the vowel here on the forum.

 

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

2

Share this post


Link to post
Share on other sites

Short answer: You can't.

Long answer: Yyyooouuu cccaaannn'''ttt...

Have you ever seen this scenario played out on TV? The CSI shows do it as often as

. Even
(and literally everything that comes out of his mouth is gibberish in that clip), though I'm willing to forgive him since it was the 80's and NO one knew what they were talking about. (But this is probably the funniest parody of Hollywood's computer illiteracy.)

Anyway, they're lying to you. There's no computer system in the world that can do what the computers in CSI are depicted as doing. While an image can easily be resized smaller, it can never satisfactorily be made larger. Why is that?

Because the image is made up of pixels, remember? [...]

Asking the computer to "enhance" it, like on CSI, is like giving someone a book with three out of every four pages missing and asking them to reconstruct the story. It can't be done very accurately, because the information isn't there, and there will never be any technology that will let us do that.**^

So, what can you do if you need the picture to be bigger?

Well, the easiest way is to just take a bigger picture. You'll never make an enlargement look as good a picture that was taken at the size you want.

But there are tricks you can employ. First, leave the drop-down menu in Paint.NET on "Best Quality." It'll automatically choose the best way of "guessing" at the missing information. Second, try running Gaussian Blur, followed by Sharpen. Repeat several times, and you'll get something passable. Keep in mind, though, this really only works at images that have been doubled in size at best. More than 200%, and you're getting into impossible territory.

And if you come up with a technique that works better, share it below! I am open to suggestions, most definitely.

[...]

**^ At least, not with our current understanding of technology.

 

This tutorial is still helpful to me, though it is almost six years old. If I may, I'd like to point out some advancements that have been made in more recent times.

 

There is a certain photo editing software (that isn't Paint.NET... yet) that has intelligent upsampling with its Preserve Details feature. Before this feature's release, the software had the ability to upsample and fill in the missing pixels without scaling down the image size. However it relied on the bicubic smoother, which actually made the edges and the image on a whole too soft (rather than being super sharp and pixel-y). Now the Preserve Details feature does a really good job of filling in the missing pixel information while still remaining close to the original image quality. It's a thing of beauty.

 

I will keep using this tutorial until this feature appears in Paint.NET. Perhaps I'll even be the one to write the plugin for this... in twenty years. (I'm a newbie at coding.) Thank you for all the help you've inadvertently provided me all of these years!

0

Share this post


Link to post
Share on other sites

Paint.net has bicubic resampling. Check out the options under Image > Resize > Resampling. These algorithms are used to 'fill in the blanks' when you resize an image to be larger than the original.

It is no surprise that the resultant details are softer. You're using (very clever) methods to smear the existing pixel data over a larger area.

There is still no way to "enhance" an image like CSI would have you believe. The data is just not there.

0

Share this post


Link to post
Share on other sites

This was an outstanding explanation. It helped me understand it better than any other place I have read this information at. I've saved it onto a .pdf

 

I sincerely thank you.

0

Share this post


Link to post
Share on other sites

Glad to help! :)

0

Share this post


Link to post
Share on other sites

Can I translate this important article in my native language, for the usage in my future Paint.NET tutorial blog ? Thanks !

0

Share this post


Link to post
Share on other sites

Feel free!  Please just link back to the first page of the tutorial.  I'll even be happy to help if you have any questions about it!

2

Share this post


Link to post
Share on other sites

Feel free!  Please just link back to the first page of the tutorial.  I'll even be happy to help if you have any questions about it!

Thank you so much !

0

Share this post


Link to post
Share on other sites

nitenurse79 - I'm glad you found this useful!

 

So, I'm going back through and reexamining my old tutorials to make sure nothing needs to be updated.  Are there any burning questions anyone happens to have about DPI (either in general, or as it relates to PdN in particular)?  I might do an updated version in the not-so-distant future.

Is there a image format that is above others in keeping resolution etc etc ?

0

Share this post


Link to post
Share on other sites

Always save a copy in paint.net's own format '.pdn', if you intend to re-edit later (it preserves layers and is lossless).

For single layer images .png (portable network graphic) is good as it saves transparency and is 'lossless'  and widely used.

1

Share this post


Link to post
Share on other sites

Dots per inch

I think he knows. He built the program. ;)

Is there a image format that is above others in keeping resolution etc etc ?

Red Ochre's reply is correct. PDN if there are layers, PNG if not.
0

Share this post


Link to post
Share on other sites

hallo David, what does the DPI-value cause here in the application report? (see file attached)

It is set to 96 DPI, and I am afraid this is why when I resize Images having a high resolution the result is not as it should be.

Example: I have a picture with 300 DPI and over 1200px width. I resize it (I shrink it) without changing the resolution and the effect on screen is a bit blurrier as expected... 

Maybe this is a pre-setting I can adjust?

 

app_report.txt

0

Share this post


Link to post
Share on other sites

-The writers of CSI are dirty rotten liars. (Well, they don't know computer graphics as well as they claim to. Same thing.)

 

 

Hmmm... I doubt that. It is simply called TV.  :D  :D  :D

0

Share this post


Link to post
Share on other sites

Excellent, I'll try to remember the content as much as my old brain will allow me to.

 

Thank You. Vulpec UK

0

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0