Inches vs units/pixels

LadieStorm · October 2016

I'm starting a new map for someones, and I've been given dimensions as to how big they want the map to be. Problem is: they gave me the dimensions in inches, and I need to translate it, so I know how big of a map to make. Math is NOT my strong suit.

So.... If someone wants a map.roughly 16 inches by 12 inches.... What would that translate to in cc3+ units? What would be the pixel ratio?

Monsen · October 2016

Is it a battle-map type of map? If so, a common grid size on physical maps is 1 inch squares, so if that is the case, the map is 16 by 12 squares. Under many popular systems, one square is 5' on the map, so that means a 80' by 60' map. Now, I made a lot of assumptions here, so you need to correct for those, but that is basically the formula.

(I am assuming that they don't want a map with an in-game size of 16 by 12 inches, if so, it would be a 4' by 3' map, which is hilariously small.)

If I were you, I would ask them what the in-game (or in-world) size of the map is supposed to be however, not the output size on paper (Although that also helps to know, because it affects the detail level you can use in the map)

LadieStorm · October 2016

This isn't a battle map, Remy...sorry I should have been more specific, but I also don't know what I can say, and what I can't.

It's not a Battle map, More of a regional...overland type of map....that can be put on a large.screen display, and possibly small poster size? It also has to be printable....

Monsen · October 2016

Then there is no simple formula, because you don't have enough information

You must ask them for either the in-world size of the map, OR the scale of the map, without that information, 16 by 12 inches means nothing. You can draw a continental map on that, or the immediate outskirts of a city. Just pick up an old-fashioned school atlas from your bookshelf, and look at all the size variants that all fit on a standard 210x297mm page.
If they give you the in-world size of the map, just use that as units in CC3, if they give you the scale (miles per inch, hopefully), just multiply the scale with the size and use that for the CC3 size.

Ralf · October 2016

One thing to keep in mind, is that you need a good dpi (dots per inch) resolution for the printing of the map. This does not matter for the CC3 map units, but will matter for the export. 300dpi is a good resolution to aim for and for a 12" by 16" map that translates to a 3600x4800 pixel export. Normally I like to export from CC3 somewhat larger than needed (say 4000x6000) and then scale it down to the desired size in Photoshop.

LadieStorm · October 2016

That's what I was needing Ralf... Sorry Remy, I was asking the wrong question. Thanks for trying to help:)

Loopysue · October 2016

Are you ok exporting larger and reducing the size in a paint app?

It seriously improves the finish, and it means you don't have to use antialiasing during export, which takes quite a lot of time.

JimP · October 2016

I think Irfanview is very good for reducing pngs, bmps, jopgs. andf its free.

http://www.irfanview.com/main_download_engl.htm and this page will show you which is the fastest site for you.

Dogtag · October 2016

I'm confused. Why would you export to a larger size and then reduce the size when you can export to the correct size from the start?

Monsen · October 2016

Because how the resampling happens during a resize. It will take multiple pixels into account when producing each finished pixel, generally producing less harsh edges. This is the same CC3 itself does when you check the anti-aliasing option. It will export an image of a higher size than specified, then will use an image editor (Image Magic) to reduce it to the requested size, thereby creating the anti-aliasing. There is much science behind good resampling routines.

Dogtag · October 2016

I guess my question/surprise is that there's any resampling at all when exporting to a given size. But I suppose I'm still thinking of CC2, where the maps were drawn with vector art rather than raster sources.

Loopysue · October 2016

Reducing graphics by 50% for final print run has been standard practice in the graphics industry since I was at art school.... ummm.... 29 years ago. It means that any tiny flaws in the original artwork are even less noticeable in the finished item.

Monsen · October 2016

Posted By: DogtagI guess my question/surprise is that there's any resampling at all when exporting to a given size.

It's not (Unless AA is on). And this causes some rather sharp edges in the export, which are then improved upon by then subsequently reducing it (Which causes resampling to be employed, softening the edges, and make things generally more natural, assuming you chose a proper algorithm)

Shessar · October 2016

Which is better when reducing, bilinear or bucubic resampling? Does it matter?

Monsen · October 2016

As far as I know, bicubic is better because it creates a smoother result, but slower. On a modern computer, the speed is probably only an issue if you are resampling huge amount of images or a server running batch jobs.

Loopysue · October 2016

Bicubic, because its a lot more advanced than bilinear

jslayton · October 2016

When upsampling photos (making them bigger), you definitely want bicubic. When downsampling, it doesn't matter as much. Bicubic typically involves using a neighborhood of 16 pixels around the sample point, while bilinear typically just uses the nearest 4 pixels.

One way to get really pretty downsampling is to start by doing a small (1-pixel blur if you're shrinking by half), doing a bilinear downsample, and then apply just a little bit of sharpening after the shrink. That way you spread the high-frequency information around a little bit by the blur which reduces the potential loss of high-frequency information when you reduce the size. The sharpening then punches up the edges a little bit. I have been told that Photoshop's "bicubic" is actually a polynomial that does this sort of thing (including a little bit of sharpening) when downsampling, but I like to have a bit more control over things.

For the most exact results, you would want to use a wider reconstruction kernel like lanczos. https://cartographersguild.com/showthread.php?t=2596 is a lengthy discussion on the topics of bitmap filtering and resampling if you want more information.

Shessar · October 2016

I've wondered this for a long time and never knew who or where to ask. I never thought to ask the question here. Duh!

Thanks for the answers and the added detail! This will give me something to experiment with while I'm laid up for a few weeks.

LadieStorm · October 2016

This is going to be a HUGE map!!!

ProFantasy Community Forum

Inches vs units/pixels

Comments