Questions about World Machine v1.2

Questions about 1.2?

Post em in this Thread!

These are not really questions about v1.2, more to be understood as feature requests for the next version. :slight_smile:
[list][*]For the terrains there is still used 15 bit instead of the full 16 bit accuracy – only 0 … 32767 (instead of -32768 … 32767). Stephen, could this be expected for v1.3 or v1.2.1?

[*]When there is a complex Node network that does not fit to the screen, would it be possible to add scroll bars? Yesterday I saved a file which worked nice with 1400x1050 resolution. I wanted to continue with the file on another computer with 1024x768 resolution, but could not see all the nodes.[/list:u]- Karsten

For the second point, you can do this :

  • Keep the right mouse button pressed, and move the mouse, it will scroll the network
  • Or use the mouse wheel to zoom in/out the network…

OK, I downloaded v1.2, but is there a new pdf help file somewhere for it? Or do we just have to rely on the change descriptions?
–Dale–

Rhalph: Thanks. :slight_smile:

Hmm Karsten, interesting point to consider with the accuracy issue however: WM does not have negative values at all in it. It is always 0-1. To ever use the negative values, the terrain would have to be remapped with some negative values, perhaps all below .5 for example (there would have to be some kind of rule, anyway). But this would end up strange in Terragen I think, with potentially very large negative values. Or am I misunderstanding and do the negative values not correspond to actual negative heights? Is there an offset specified in the file?

  • Oshyan
Or am I misunderstanding and do the negative values not correspond to actual negative heights?
There are two ALTW values (each 2 byte integer), for Height Scale and Base Height. After these values follows a "row" of 2 byte (16 bit) height values. Terrain file specifications [url=http://www.planetside.co.uk/terragen/dev/tgterrain.html]here[/url].

You can download TerrainScout, it shows quite good all terrain chunks.

Or an example. Terragen saves a terrain with height range -40 … 20 as:

Base Height = -10
Height Scale = 60
(with height values -32768 … 32767)

  • Karsten

/edit: For any replies should be opened a new thread.

OK, I downloaded v1.2, but is there a new pdf help file somewhere for it? Or do we just have to rely on the change descriptions?
Hi Dale,

I’m sure docs will be made available very soon…
I believe WM 1.2 came out primarily to fix some issues regarding large memory uses and saving files, and other things. It was released quite fast, in fact, so there was not much time to finish documentation on the new features.
For that reason, I believe, Stephen created this thread so that it would serve as a temporary replacement for docs…
I also know Stephen for giving us presents on spetial occasions, so having WM1.2 for Thanksgiving was probbly intentional as well.

So right now, the place to read documentation is here… I know there isn’t much yet… but questions posted here will be answered!

Thanksgiving is around the corner, and that may delay some replies. So I might find some time tomorrow (now + 10 hours) to write up a small “review” of the new features I liked most. Hopefully that will help a bit.
I will also be out for the weekend, so I’ll write it before then.

Cheers,

@Osyhan: While WM might calculate values internally within the range of 0…1 (floating point), WM sure has to convert the heightfield data when exporting to Terragen terrain files, for example. And TG requires the data in signed integer format (16 Bit) which ranges from -32768 to 32767.
I know you know that but let’s have a short look at two histogram screenshots for the same terrain (WM and re-saved in TG):


As you can see there’s a lot of accuracy “wasted” because WM currently doesn’t use the full range of 16 bit signed integer (i.e. the whole negative range of it). That’s what Karsten was referring to…

It’s exactly the same “problem” that 3DEM once had. :wink: Depending on the terrain this might not be visible at all if you have only low hills etc. (where even 8 bit would be sufficient) … But with deep valleys and very high mountains, it would certainly better to take advantage of the full 16 bit range…

Ralf.

All the explanation wasn’t necessary. All I needed was to know there was an offset (base height), that says it all. :wink:

  • Oshyan

I agree that it was a bit lengthy :wink: but although I adressed it to you in the first place, most of it wasn’t targeted towards you.

However, BaseHeight doesn’t “say it all”, seriously…

Ralf.

hi guys,

as Fil said, I really wanted to release 1.2 before thanksgiving… a users’s manual addendum should be upcoming next week, including hopefully some other documentation stuff that has been put off for a while now (such as the device reference), although that will take longer than a week to get done.

:stuck_out_tongue: OK thick old wrinkly here. Can I overwrite 1.0 with 1.2 or do I old 1.0?

The most recommended install is to install into a new folder. Installing 1.2 atop 1.0 may work fine, but your best bet is to not.

:stuck_out_tongue: Thanks Stephen :stuck_out_tongue:

I think i found a bug. Using an inverter on a file input is a bit wacky. My input file is black and white. The inverter inverts it but the lowest elevation clamps up to midland. Doesnt make sense? Try it. :stuck_out_tongue:

Ignore that, it seems like the real problem is that BMP is being input at half height. TGA works fine though.

-Ben

Yup! That is already known, and Stephen already corrected it for next version :wink:
As a workaround you can place a Clamp after the File Input with Normalize Input on. This will give the correct behaviour in the case the BMP file only has grey colors.

There are a few other glitches that were noticed that are also corrected…

Bugs are a frustrating species of living beings that hide inside programs and get discovered almost everyday… I am very happy that there is no (known) crashing bug around (yet)… But given Murphy’s cumplicity in this matter, someday there will be one of those :stuck_out_tongue: (to be corrected shortly after) :wink:

Cheers,