Hey guys,

I hope everyone is doing well. I feel obligated to drop by every once in awhile and ask a stupid question just so you guys know I'm still using the program in spite of all my other options these days.

Lately I've been rendering a lot of Jpegs and I notice when I pull them into Photoshop they are always 72 DPI. Is there any way to increase the resolution of the image sot say something more print ready like 300 DPI or something like that?

I've sniffed around and I don't seen any easy way to do it myself. Which means it's probably staring me right in the face.

And yes I know you can change the width and height of the camera view and that the program calls that "resolution" but that's not what I mean by resolution. I'm referring to DPI.

Anyway, thanks for being one of the best on line communities I've ever run into.

Cheers.

One of the harder once to understand .

Before we start: DPI stands for Dots per Inch.
DPI is only a way to say the printer what he should do with the pixel-data given.
That means: The effective resolution will not be increased if you have a JPG with 300 dpi in it. Only the printer will print 300 of the pixels in the image in an inch instead of 72.

The effective, digital resolution of a 300 dpi-image is not larger than one with 72 dpi. It is all just about the pixels here.
One pixel in this circumstances can be seen as one dot. (-> Dots per Inch)

You can see that here:
1.) Open Photoshop and load a JPG with 72 dpi.
2.) Go to "Image > Imagesize" (or whatever it is called in english... it is the menu where you change the image size with).
3.) Below the inputfields you will find a checkbox "Interpolation Method" (or something like that). Deactivate it.

Now you will see that the pixels above will not be altered anymore, no matter what you put into the DPI or the documentsize-fields.
That is because the pixel-amount will not be interpolated anymore. (interpolation -> the try of the computer to create more (or less) pixels on base of the existing once. This will often result in blurry images, etc. (depending on the algorithm used) and will not increase the quality of the image)

So for any digital purpose, DPI is just not necessary. And if you need your image for printing, you can just increase the pixel-size of the rendering and tell the JPG in Photoshop to increase the DPI-values.
That will decrease the documentsize / printsize, but not the pixel-amount / size which stays the same no matter what if you are not interpolating anything.

Hope that was understandable...

See you
*Fuchur*

I am so glad you beat me to answering this Fuchur... because I was gonna mess it up.

About every year or so this topic comes up and a heated discussion soon ensues.
Fuchur has stated this in just about the best way I've heard it said before.
What's been fairly entertaining in the past was when folks tried to argue this out out with Martin and.. um... he's one of the the guys who wrote the earliest software used to operate printers.

Here's how Vern said it a few years ago:



As Fuchur and Vern relate, DPI is all about printer resolution.

Just to muddy the waters a bit, if you want a "hi-res" render, just render it bigger. e.g, an 8x10 inch image at 72 dpi would be rendered in AM at 576 x 720, but an 8x10 hi-res (300dpi) image would be rendered at 2400 x 3000. They're both essentially 72dpi images, but for print you would want the second formula.

The Sun still hasn't set on British Measure System. DPI forever!

I'll just add that you can reverse calculate your res if you know the size the image needs to print at and the DPI that is desired.

If your image needs to fill a 5x3 inch space on paper and the printer can do 300 dpi, render at 1500x900 (5x300 by 3x300) to fully exploit the capability of the printer.

In actual practice you could render half that res and use it as 150 dpi and no one would likely notice the difference.

Pixels are pixels. Resolution just tells you how big those pixels are.

Resolution really matters with straight lines. Type and line art, for instance, at low resolutions will appear to have jagged edges or appear "soft" because the pixels are big enough to be noticeable. Renders from A:M are like photos and can cheat the resolution more, since we don't expect photos to be particularly crisp. We're used to photos having softer focus. Of course, if the point of the image is to be sharp (ie an image of a product, for instance), resolution will be really important.

Generally, for small items (like the 5x3 size Robert mentioned), you want as much resolution as you can, because the image is small and likely to be held relatively close to the viewer's eyes. A poster or billboard, on the other hand is going to viewed from a distance and so, the relative size for the viewer will be much smaller than its actual size. 150 is often used for outside signage and large billboards drop even further.

That said, it's always better to have more resolution than you need.

Let's muddy it even more, and mention the line screen of your method of printing!

To this day i still don't understand how, in the pre-computer age, they converted continuous-tone photographs to black dots of varying size to give the impression of shades of gray.

I did it in the 80s when I was in college for the school newspaper. I don't remember the particulars, but you had screens that you would load into a large machine kind of like a sandwhich of lightboxes along with the photo and it would expose them onto photo paper that would be developed and have the dot screen. My only real memory was that it was a chore and I didn't like doing it. :-)

I'll guess it was an extrapolation of the age old methods engravers used in printing and artists used in stippling, crosshatching and such in sketching. To an extent this equates to the splotches of ink and paint that form patterns and shapes that we percieve as familiar... recognizable... an interpreted facsimile of the original in all artwork.

In this sense a higher DPI is just as advantageous on a painter's canvas as it is on paper as a painting with only 10SPI* is not going to have the detail and definition a similar painting would have with 300SPI. Of course, it'd be a whole lot quicker to paint that painting with 10SPI. For all practical purposes the image is the same but the impression on the real world surface makes the difference in what we perceive.

Added: Silk screen printing is another good example of this in that it allows the paint to go through the very fine grid in the screen.
(Makes me wonder what the SPI of a typical silkscreen is)

*SPI=Splotches Per Inch

I understand how dots create the impression of grays. But how did the dots get made at all?

I'd expect that putting a screen between a photo and some sort of receiving plate would just produce a screen pattern of same-sized dots filled with different shades of gray. But that's not what happens. Dots of differing sizes, all black, get made.

It's been too many years, but the machine was called a PMT. I did some photos making halftone screens, but usually I used it for my cartoons, which used no screen. I vaguely recall that you would take the exposed sheet and another sheet and feed them both into a developer. In the end, you would peel them apart and one had the image.

Just googled and the actual image was a PMT. I guess we just called it the PMT machine. It stood for Photo Mechanical Transfer. You didn't need a negative in this process, because the activator sheet would transfer the image over to the sheet, resulting in a print.

I can't seem to find any pictures of the contraption online, though.

Somehow they've been doing half-tone dots since about the 1870's.

If some one said to me "we project the image through a screen of holes that are all the same size onto a petri dish and the brighter the light shining through each hole, the larger a dot of mold grows on the petri dish" I'd say that at least somehow connects different light intensities to different size dots.

But i dont' think there's mold involved.

Is there a paper like that mold?

The science of Half tone images
From that page:
"The size of the dot behind each grid square is proportional to the intensity of the light falling on it. Therefore the tones of the original photograph are converted to dots of varying size on the high contrast film. "

Thanks Nancy. Here's the part that explains it for me...



The high contrast film is converting fuzzy circles into distinct dots.

Thanks for the replies and lively discussion. I was in school and finishing work for a client (in A:M I might add) or I would have been back sooner.

Believe it or not my education in Graphic Design has allowed me to understand ALL of what you guys said. However... allow me to muddy the waters even more.



DPI is used as sort of a "catch all" for resolution and actually doesn't serve print or digital very well because of the confusion. Truth be told most digital resolutions would best be described as PPI or Pixels Per Inch rather than DPI. And printing resolution is best described as Lines Per Inch, LPI or sometimes Line Screen (especially in reference to a printer's capabilities as most printers don't have a set DPI but a range) in some cases (offset printing using halftone printing process) and DPI in digital printers.

All digital presentation is at 72 PPI (but if you wanted to you could set your monitor at 92 PPI and everything would appear smaller... weird huh?) but most printing is done at 300 DPI or 150 LPI. The calculation for determining proper LPI is to take the PPI and divide by two so a PPI of 300 gives you a proper resolution for printing of 150 LPI (there are some things like magazines and newspapers that print at lower LPI... I think my professor said that magazines are typically 250 DPI or 125 LPI and Newspapers even less.. wikipedia says newspapers are about 85 LPI).

And since 72 pixels = 1 inch of digital presentation you can do the math and try and figure out resolution adjustments by adjusting the size of the camera frame and moving the model closer to the camera. So an image you render at 10 inches by 10 inches at 72 PPI will convert to just under 2.5 inches at 300 DPI. That's assuming you don't use resampling which I (and most designers) don't like to do because despite our advanced graphics technology resampling still isn't... great. And you can get blurring and pixelation out of your final image using it.

So yes. I get it I can adjust the frame size and model position and make (theoretically) any resolution I want. If I'm willing to do the math. What I was hoping was that A:M like almost every other graphics related program out there, had a selectable menu somewhere in the render functions that would do the math for me so I didn't have to jump through that hoop. The fact that it doesn't is kind of a shortcoming (in my opinion) in an otherwise amazing piece of software.

So thanks for the answer guys. I didn't even go into dye sublimation or stochastic printing, or SPI for scanners (hee hee) to further muddy things up. Print may have gotten a lot simpler over the years but it's still a pretty complex business.

And lastly one more stupid question... is there a way to render a model with a transparent background? Another feature I'd love to see but suspect doesn't exist. Just thought I would ask.

Thanks again everyone you are AWESOME!

p.s. If anyone cares this is probably the best and most succinct article I've found on the differences related to the various "whatevers" per inch measurements. http://www.designtalkboard.com/tips/dtp/dpi.php

We have obviously failed you here. <insert head hanging very low here>
That you suspect A:M doesn't have this... ouch. You've hurt me personally... deeply.
We must do something about this immediately.

A computer graphics program with out RBGA (Alpha Channels) isn't worth having IMO.
Alpha Channels also just happen to be one of my favorite topics.

Look for the Alpha Channel option in the Render Options Panel under 'Buffers' when using formats like EXR, TGA and PNG.
Some image formats do not support transparency (they are RBG only) and for those you'd have to use key colors and a program to target that specific color for transparency masking. The GIF format is an image format that you can often find with key'd colors for tranparency.

LOL. Rodney you are the man. Thanks so much. Got nothing but love for you guys.

I hope with all the details that were laid out your question was answered, I know I could not have given as much information without a lot of research on my part.

However, it was not a stupid question. Asking what color donkey poop would be if it ate red cabbage instead of green - would be a stupid question just about anywhere but in the Agriculture Industry. Asking for help with a serious question - no matter how mundane or trivial the subject may be (as long as it is germain to the profession, industry, person or community to which the question is posed) is never stupid because you are seeking to grow through the question.

Humble is one thing but asking questions and feeling stupid doing so is another. Give yourself more credit for trying to learn and grow. Next time perhaps you might say, "I have a question that may have been answered previously but I am unable to find the answer myself, can someone please help".

Remember this always: everyone on this forum started out knowing NOTHING about Animation Master. Don't feel bad when you need to ask for help, thats what we are all here to do - help each other.

Mark

What exactly do you mean by "adjust the frame-size / model position"?
You only have to do that, if you are using another ratio for the image. In any other case you can just increase the camera-/rendering-settings and everything will be the same.
It really doesnt matter if you choose 1280 x 720, 1920 x 1080 or 3500 x 1969px concerning the model-position or the frame-size. It will render it at that resolution then and that's it.

If you go for another ratio (between width and height) it will show another part of the scene...

See you
*Fuchur*