Search the web
Sign In
New User? Sign Up
L3DT_users_group · L3DT users' group mailing-list
? Already a member? Sign in to Yahoo!

Yahoo! Groups Tips

Did you know...
Real people. Real stories. See how Yahoo! Groups impacts members worldwide.

Best of Y! Groups

   Check them out and nominate your group.
Having problems with message search? Fill out this form to ensure your group is one of the first to be migrated to the new message search system.

Messages

  Messages Help
Advanced
SDK progress report   Message List  
Reply | Forward Message #162 of 177 |
Re: [L3DT users' group] Image overlay

Hello, I've had a chance to look at the image overlay and basically
it seems fine in terms of visual use. It's a great help. ;)
There are a couple of minor artifacts in the image- don't know
what's causing that, but they haven't affected what I was doing in
any way. I don't know for sure but the image overlay may be slowing
down the paint tool process. I think the redraw is slower with the
overlay.
I've had a crack at reproducing a tile of the ME-DEM topo. This
would be typical of the minimum tile size a user would be working on,
that is a 1000 px square. (this would exist a 100 px square in the
database before being upsized by the user to 1K before editing and
resubmitting- in the old proposed setup anyway). I used 16 x 16 DM
map with 20m res- ie 20 KM on a side. L3D uses a 1024 tile but we
would be presumably importing into L3D a 1K square tile from the base
DEM (?). The image overlay was a 1K square jpeg overlay covering 20Km
(again this would exist as the corresponding 100 px aquare topo tile
before users would upsize it as I did).
It was a basic exercise in transcribing what was in the topo. I'm
pleased to say that I got some pretty good results- even at that low
DM resolution- though it did take me some time to get there.:).
The topo contained simple mountains with ridges and is probably
representative of the simple end of the topo's visual complexity
scale. It *might* be necessary to go to 32 x 32 but I don't think any
higher than that (out of necessity anyway). Having said that, more
control is always better.
The method I used was to build up the DM pixels using the altitude,
almost exclusively in 200m increments- painterly fashion. The max
heights of the terrain I did were reasonable 'real world' and fairly
accurate to the topo. I used the peak, fractal roughness and
terracing settings from the Fjord tutorial (great by the way). Ditto
for the approach with global and targetted erosion (same numbers). It
was noob stuff.
I've satisfied myself that it is totally feasable to use L3D as of
now for ME-DEM, now that the image overlay is in place. :)


I've got a lot of ideas/ feature suggestions:

(Please accept any apologies that might be due for a total noob
making such outlandish feature requests :D)

It would be EXTREMELY useful to be able to have the last generated
contour map (perhaps even just the contour lines if possible) from
the heightfield operation as an additional layer to the DM overlay.
Using this painterly method I found myself going back and forth
between the DM and HM countless times. This was a direct consequence
of the painterly method and the approach we are using- we have less
free rein as we are more or less transcribing from a topo, thus a
lot of control over the terrain is needed.
Once I performed the height operation I was having to visually
eyeball and commit to memory areas from the contoured heightmap map
that I needed to alter for when I got back into the DM map view. If
this was implemented it would be useful to have an on/off toggle and
independent transparency for any image layers. The coord readout in
the heightmap view was of no use here because the DM map coord
readout is in DM pixels. It would help if the two could coincide in
some way but the overlay would be VASTLY better. ;)

Because I was going constantly back and forth between the design map
and heightmap operation would it be possible to have a selective
update (or do you already do that ?)- ie only those DM px that were
changed would be used in the next heightmap operation ? just a
thought- I don't know what code implications that has. This would in
theory speed up the heightmap operation.

A hotkey for the heightmap operation would be another help.

An (optional) altitude display on the pencil cursor itself might be
handy as in Wilbur. So that all the info was at the pencil point (ie
pixel to be set and its current height). I didn't find the colour
scheme I was using that helpful. I didn't really have a look into the
options (I was using classic) but I think this would probably be
unchanged for the others because of the precise(r) nature of what we
are doing.

Would it be possible to implement a 3D preview, or realtime 3D
editing, but using only the Design map pixels down the line. It
wouldn't be necessary to have a realtime interaction with the
heightmap post heightmap operation. I imagine that this would be easy
on the system and a more palatable code pill to swallow. :) The 3D
view would consist of the DM pixels but their faces interpolated to
smooth them. Any image overlays would function as above (draped over
it as a texture say) with individual layer toggles/transparencies.

I found that one had to individually click the DM pixels because the
refresh (or whatever) was too slow to keep up even with moderate
painting speed. This is the slowing down I mentioned, possibly
associated with the image overlay. For a 16 sq map this wasn't too
much of a problem, but is for higher res DMs. A shift click between
two points would be handy. Vectors, polylines and polygons would be
heaven :) [I have a P4, 2 GB RAM and 256 RAM vid card]

I didn't use any of the special overlays so some of the stuff above
may be addressed with that. It would great to be able to have a
combination of stamp-like tools and the painterly option. I only
briefly looked into the mountains overlay yesterday. From that
admittedly cursory look I have found so far that I prefer the
painterly approach but can see massive advantages of cookie cutter
like tools.


I'll post the results on ME-DEM and send you two fellas (Aaron and
Oshyan) a couple of pics if you like ;)

Cheers,
monks


--- In L3DT_users_group@yahoogroups.com, "monkschain" <lingardc@g...>
wrote:
>
> Awesome, thanks for that Aaron. I've downloaded it and will give
it
> a whirl first chance I get.
>
> monks
>
> --- In L3DT_users_group@yahoogroups.com, Aaron Torpy
> <aaron_torpy@y...> wrote:
> >
> > Hello Monkschain,
> >
> > > The most useful short term addition would be the
> > image overlay
> > > (with transparency) so that terrain could be
> > 'defined' from a
> > > topo reference. Like tracing over it. I see that
> > that is
> > > on the dev roadmap so that's great.
> >
> > Okay, I've uploaded an update of L3DT Pro that
> > includes a rough first-attempt at the image overlay
> > feature. I'll send you a direct e-mail with an
> > activation key for the beta-release and instructions
> > on how to download the Pro version.
> >
> > Anyhow, assuming everything else works OK, you can
> > load the overlay image (bmp, jpg or png only, at the
> > moment) using the 'view->overlay image...' menu
> > option. This image will then be blended/stretched over
> > the top of any map that you create/view/edit. You can
> > also set transparency in the control dialog.
> >
> > This is really quite rough, but I thought it might be
> > good to get your feedback before I try to smooth the
> > edges.
> >
> > Cheers,
> > Aaron.
> >
> > PS: I should hopefully have time reply to these posts
> > more thoroughly on the weekend.
> >
> > --- monkschain <lingardc@g...> wrote:
> >
> > > Hello, monks here from ME-DEM. It's great to see
> > > that our lil' baby
> > > elephant still has a caring family- (trumpets
> > > appreciatively) hehe.
> > > You're right about the l...o...n...g term time
> > > frame- we'll all be
> > > long in the tooth ere the task is done !
> > > I just wanted to pipe in with a couple of details
> > > if that's OK Aaron.
> > > The most useful short term addition would be the
> > > image overlay (with
> > > transparency) so that terrain could be 'defined'
> > > from a topo
> > > reference. Like tracing over it. I see that that is
> > > on the dev
> > > roadmap so that's great. With this we might be able
> > > to go ahead with
> > > using L3D straight away (without all the other cool
> > > stuff mentioned
> > > of course).
> > > The abilty to load existing terrain and create in
> > > this manner in
> > > *replace mode*. So one would load up a low res dem
> > > with the
> > > corresponding image overlay (topo map) and then use
> > > the elevations of
> > > that low res dem for reference (one would need some
> > > kind of cursor
> > > feedback for height or something) along with the
> > > topo overlay. So one
> > > would know, here is a mountain and this is it's
> > > approx height. Define
> > > it with the L3D tools and bingo.
> > > The ability to load up tile sets that are multiples
> > > of x Km (on the
> > > first pass each tile will be 1K px on a side at 1
> > > px= 20m....20 Km on
> > > a side, on the second pass (down the line) each tile
> > > with a factor of
> > > 10 larger scale- that is, each tile being 1K px
> > > still, but at 2m res,
> > > so 2 Km on a side). This would be really useful as
> > > users could work
> > > on larger areas (which is probably the preferred
> > > approach).
> > > ps Joe Slayton of Wilbur (he's helped us out
> > > already). ;)
> > > ps I'm still convinced that L3D could fundamentally
> > > change the way
> > > people build terrain models.
> > >
> > > monks
> > >
> > >
> > >
> > >
> >
>








Sun Nov 20, 2005 11:28 pm

monkschain
Offline Offline
Send Email Send Email

Forward
Message #162 of 177 |
Expand Messages Author Sort by Date

Awesome, thanks for that Aaron. I've downloaded it and will give it a whirl first chance I get. monks...
monkschain
Offline Send Email
Nov 17, 2005
9:02 pm

Hello, I've had a chance to look at the image overlay and basically it seems fine in terms of visual use. It's a great help. ;) There are a couple of minor...
monkschain
Offline Send Email
Nov 20, 2005
11:28 pm

Hello, ... slowing ... slower with > the overlay. Yup, it should slow things down. The overlay code is mixing the image-buffers each time the screen is drawn,...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 21, 2005
6:27 am

-yeah I realise the overlay was put in pretty speedily. OK, can't say fairer than that :) Cheers. monks...
monkschain
Offline Send Email
Nov 21, 2005
1:11 pm

Hello, ... ;) Hardware is the problem here. L3DTVi2 can load big mosaic maps (I’ve looked at 32k x 32k maps before), but you still have to keep your...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 18, 2005
7:23 am

... Yeah, only time will make this more feasible. 64 bit OS and hardware, multi-core CPU's, etc. ... Sounds good. I agree that by the time we get enough data...
Oshyan Greene
ticketseller...
Offline Send Email
Nov 18, 2005
7:46 am

Hi Oshyan, ... Yup, and I'm hoping and praying for faster hard-disks too. The disk-fetching delay can be a real pain when you’re dealing with so much...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 18, 2005
2:17 pm

... Yes, this is a very good point. It was fairly long ago that we passed the point where the data transfer capability of the interface surpassed the HD's ...
Oshyan Greene
ticketseller...
Offline Send Email
Nov 18, 2005
9:31 pm

My head is spinning trying to follow all this discussion regarding Middle Earth DEM. It's quite interesting and a lot of good discussion going on. I have...
Rofar
rofar_ds
Offline Send Email
Nov 18, 2005
11:14 pm

Hi Rofar, I don't know if you know of the ME-DEM project. Here's a link: http://me-dem.ashundar.com/index.php You're welcome to register and get involved. :) I...
monkschain
Offline Send Email
Nov 21, 2005
1:01 pm

_____ From: L3DT_users_group@yahoogroups.com [mailto:L3DT_users_group@yahoogroups.com] On Behalf Of Oshyan Greene Sent: Friday, November 18, 2005 4:32 PM To:...
Michael & Kristina Ho...
crxb16a91
Offline Send Email
Nov 19, 2005
3:02 am

Hello, ... Yes, but the tile-loading is automatic when you read/write to a pixel in a tile that's not in memory. This means that the program itself (and the...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 19, 2005
6:12 am

... Ah, sounds delicious. The only issue would be if the tile was not local it would have to be downloaded, causing an additional delay in the overall process....
Oshyan Greene
ticketseller...
Offline Send Email
Nov 19, 2005
6:16 am

Hello, ... causing an ... least it would ... There would also be the option to pre-load the tiles you're going to use. The L3DT subsystems have this capacity,...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 19, 2005
6:34 am

... From: "M O" <rhoneranger2001@...> ... Maybe I've misunderstood what you want, but surely the point of the sdk is to allow you to write exporters....
Noisecrime_PIPEX
noisecrime
Offline Send Email
Nov 7, 2005
9:02 am

Hello, Okay, one more e-mail before I go to bed. ... True enough. L3DT doesn't use vertex data, but it will be supported indirectly in the SDK, since plugins...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 7, 2005
3:05 pm

Great Work. For my personal use, i love to have only c++ language, and a simple dll (not com object). I don't know if is usefull to have unicode in the sdk, i...
Admin Kiraya
pmkiraya
Offline Send Email
Nov 7, 2005
12:03 pm

Hi Andrea, ... Excellent! So far this is 4/4 votes for C/C++, and 4/4 votes for non-unicode (I'm counting myself as the 4th voter in each case, of course)....
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 7, 2005
2:42 pm

... From: "Aaron Torpy" <aaron_torpy@...> ... I think i misunderstood what you meant, I thought you meant including plugins in general, not plugin access...
Noisecrime_PIPEX
noisecrime
Offline Send Email
Nov 7, 2005
3:45 pm

Wow, I missed so much stuff in one day. Well Aaron for your answers: 1. I think everything I want/need has already been included or mentioned. 2. For standard...
mark troutt
marktroutt@...
Send Email
Nov 7, 2005
8:44 pm

Hi Mark, ... I'm terribly ignorant of Java, but from what little I have read, it seems that it can be a bit of a pain in importing 'non-managed' functions from...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 8, 2005
1:14 am

Well, when I say "besides maybe java" I don't mean to imply that I've ever used it. All I know is that it can be more portable to other operating systems and...
mark troutt
marktroutt@...
Send Email
Nov 8, 2005
12:37 pm

sorry but java is not faster than c++. it's only clearer,easiest and portable (because have virtual machine). ... From: mark troutt To:...
Admin Kiraya
pmkiraya
Offline Send Email
Nov 8, 2005
12:49 pm

Oh, then my source was wrong. I was told that it was faster, my mistake. -Mark p.s. Thanks for correcting me. ... -- Gravity. It's not just a good idea, it's...
mark troutt
marktroutt@...
Send Email
Nov 8, 2005
3:19 pm

Hello, I agree with the C++ is faster than Java bit, but if the Java zealots are to be believed the difference isn't that great these days. It probably depends...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 9, 2005
6:58 am

Hey! ... Me too! :) I just wonder if it would be too hard to provide a Delphi "version" as well. You know, I'm "learning" C++ for a while now but I'm wondering...
yahoo at ralf-lehmann...
loslalfos
Offline Send Email
Nov 9, 2005
12:08 am

Hi Ralf, Yup, I think I could write wrappers in Delphi (Lord knows I need the practice). I’ll probably do the C++ release first, get some feedback and make...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 9, 2005
1:15 am

Hi Simeon, ... I want the SDK to be as powerful as possible, so that other developers can make programs like L3DT without having to write their own...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 9, 2005
6:30 am

Aaron, What I meant by no Header.... Is in the top of every other file you export now, you have a header l3dt version 1.x ....... etc etc In the SDK you...
M O
rhoneranger2001
Offline Send Email
Nov 7, 2005
2:25 pm

Hi Matt, The file headers are there to identify the file type, because file extensions are ambiguous. This is a pretty standard ploy in both binary files (eg...
Aaron Torpy
aaron_torpy
Offline Send Email
Nov 7, 2005
2:59 pm
 First  |  |  Last 
Advanced

Copyright © 2009 Yahoo! Inc. All rights reserved.
Privacy Policy - Terms of Service - Guidelines - Help