Idea: Terragen Visual Reference Sheets

Started by PCook, August 29, 2012, 12:14:21 PM

Previous topic - Next topic

PCook

Just casual reading of the Planetside forum will reveal a lot of requests for Terragen documentation. In those discussions over the last several years it's clear that the staff of Planetside is very aware of the need for documentation and has been addressing that need with a wiki-based node reference. While a node reference is a good place to start, I'm finding it difficult to translate text descriptions of settings into practical application when working in Terragen. I'm sure I'm not alone in this particular challenge.

The demands of a business require that priorities be established which, in the case of Planetside, is clearly to create powerful software for terrains, which it is doing remarkably well. But that priority, and limited staff, leaves documentation (and I'm sure many other projects Planetside would like to do) on a lower priority, but nonetheless important. While we (myself included) have been pressuring Planetside to give us more documentation, there may be something the Terragen community can do to help.

I've constructed a settings reference example to illustrate a way to publish visually oriented Terragen reference material quickly. For discussion purposes, I'll call it "Terragen Visual Reference Sheets" (TVRS). I've simply taken screenshots of low resolution renders along with the applicable settings window, and show them side-by-side to allow a quick visual comparison. I chose lower, middle and higher settings to illustrate the effect of the setting. Note that I didn't take a minimum, mid or maximum setting – rather just enough to show the effect of the setting. This provides a visual of how a setting affects the image. Such TVRS's could be placed in a wiki to be indexed by the upper left title and the descriptive text. My example TVRS image is below, which is not intended to be technically accurate, just as example of a TVRS.

A couple benefits come to mind of this type of visual reference:

a) Persons from the community can easily make these, even with the non-commercial version of Terragen 2 because the render size is low. In this case I used Image Width=450, Image Height=300, Detail=.85, AA=3. These are well within the TG2 non-commercial version's rendering capabilities (and they render more quickly). I don't feel the preview image would work for TVRS's.

b) Other than Terragen itself, no special software is needed. I used Photoshop, but there are many alternatives including such low power apps as Window's Paint. In most cases, a TVRS author could have a zero monetary investment.

c) Provides a very fast visual, good for when an author just wants to know the affect of a settings with minimum technical description and much less experimentation in a real Terragen project.

d) This approach isolates the setting from relationships with other settings to aid understanding what the specific setting does.

e) Since the intention is not to create a specific landscape image, TVRSs need only focus on a specific setting and do not need to be tutorials. In fact, the actual TVRS image can be anything that illustrates the target setting's effect.

f) When a future Terragen release affects published TVRS's, only those TVRS's need be revised/republished.

This is clearly a project for the Terragen community on the whole, since there are potentially hundreds of such TVRS's. One person would take months or years to produce a usable TVRS collection. What I'm thinking is that a wiki be constructed (or Planetside makes its wiki available) where persons could publish their own visual reference sheets into the wiki assuming that a TVRS didn't already exist for that setting. Anyone could tackle any VRS they chose, whenever they like. The *original* author would get attribution on the TVRS page.

A couple key success factors of this project are; a) isolate as much as possible - avoid setting relationships as much as possible, b) keep technical text to a minimum to not slow down publishing new TVRS's – detailed text can come later. Keep the TVRS's highly visual to help them make it to the wiki quickly with minimum author pain. A consistent structure will also be important so that the reader doesn't encounter variations of the TVRS layouts. In summary; keep TVRS's very specific, highly visual, consistent in layout, easy to make, fast to publish and even fun to create.

Anyway, that's the idea for all to consider and comment on.
-Pat

[attachimg=1]

RArcher

Sort of like the user editable wiki already available that mostly gathers dust... http://planetside.co.uk/wiki

The problem is the same as it always is each time these threads are created.  Coming up with a plan is great but pretty much everyone would rather be building their own scenes instead of building documentation.  Maybe if you started it off and did up 50-100 or so others may jump in?

PCook

Ryan, you referenced the official Planetside wiki. Is that wiki gathering dust?

As I'm sure you yourself have discovered, creating content yourself is the slow way to success. The exception would be when you have the time, ability and motivation to create large enough amount of content to attract a good number of people to the website. Else what happens is that people stop returning to the site because they lost confidence that there will be new content for them to review and consider. Point is, with this project, a single person can't get there fast enough to get regular traffic that uses the content. When traffic is not there, contributions slow or cease because it's not worth the effort for a small traffic volume.

A single-person project suffers from a couple of handicaps. First, the project becomes "owned" making it seem off limits to others. It becomes proprietary. Second, a single person is not likely to get enough content in place fast enough to get the project out of first gear, and indeed may actually slow the project down considerably. Contributors are more likely to wait until the single person gets his work done, which chews up valuable time to establish the resource. In short, a commitment from those 50-100 people you mentioned need to come before the content is created, else the result will likely be a Terragen resource with sparse content not seemingly worth spending the time to contribute to.

Projects die in their idea stage because the support is not there, not because the idea doesn't have merit. But putting the burden of proof on the person with the idea doesn't make the project any more viable. If the Terragen community wants this, the community will make it happen. Else it dies along with the other ideas and hundreds or thousands of Terragen users get to keep experimenting with Terragen settings until a nice image appears.

-Pat

RArcher

Gathering dust only in the User Contributed sections.  Jo and Oshyan update the official parts as often as they are able.  You were asking for a community driven wiki, well feel free to log in to the wiki and edit to your hearts content.  One clarification on the previous message - I did not mean 50-100 people would jump in and help, but perhaps if you, personally, created 50-100 "Terragen Visual Reference Sheets" then perhaps others might step up eventually as well.

I'm not knocking your ideas at all, they would be great.  However the reality of the situation is that these sorts of threads are created quite regularly and I've yet to see one move any further.

PCook

Quotethen perhaps others might step up eventually as well

"perhaps", "might" and "eventually" are some fairly uncertain words. But as I explained, this needs to be a community project, not my project.

Quotethe reality of the situation is that these sorts of threads are created quite regularly and I've yet to see one move any further

And this one may also suffer that same fate. But wouldn't it be better if we spent our time discussing how it could be done rather than predicting it won't be done? If the idea is any good, let's not kill it with history.

-Pat

masonspappy

I think Pat makes a lot of valid points - they are well thought out and he has the gift of explaining them clearly and concisely.  My viewpoint differs in this regard: if the documentation is to be created, then one person will need to own that mission to ensure it gets done properly.  The reality is that the Planetside folks should be doing this and, as far as it is possible, I think they've tried. But let's face it: documenting a process is boring and painstaking. Who wants to do that when they could be satisfying their creative yearnings? Most of the folks on these boards are artists, and artists follow their passions. Unfortunatly, artists also have a reputation for being, well ... let's call it "a little bit undisciplined."    They will spend countless hours trying to make a beautifull image, but will always have trouble finding the time and focus to document how they went about making that image.  I honestly belive that the only way this documentation effort will ever come to fruition is if one person takes it upon themselves to own this effort and treat this as a project - with themselves as the project manager - and actively strive to pull other in people in to contribute their knowledge and ideas.  To be clear about this: the project manager doesn't do the technical work. The project manager solicits input from other people (on these boards) for their creative input, and then puts that input into a document that can be reference and expanded upon.   But - and this is a big 'but' - the Planetside folks should really be approaching someone in these boards to take on that tasks.  Ultimatly, PLanetside should be overseeing this effort.
But that's just my opinion...
- Cam

mhall

If I recall correctly, someone created a small application for rendering out the same scene while selectively varying specific settings values. The scenes were output automatically and tiled together into a finished image grid (with what setting was changed as text below the image) so that the user could inspect the differences visually.

I always thought this was a great way to get a feel for the effect settings changes. And, if done across a range of parameters could allow a visual understanding of what controls do.

Unfortunately, I have no idea who this was, or how long ago! (I just spent 15 minutes searching and looking through old threads and came up empty - but I remember a grid of clouds with settings differences between them).

I only bring this up because, if this application could be found (or re-created ... there are programmers on the boards here and TGD files are just XML, and TGD can render from the command line.) then perhaps a group of users could be persuaded to supply some CPU cycles to help render out some of these images and that way a large visual library could be created and uploaded to the wiki relatively quickly.

Some coordination would be needed ... so that people didn't waste time rendering the same settings.

Just a random bit of brain storming.

~Micheal

PCook

I wholeheartedly agree with you Cam that the project must be managed. But I don't believe that one person should be developing the content because, as I reasoned, it will take much too long to have a useable reference. As with many projects that involve many people, a leader will emerge. But, it's too soon to decide management. Is the project even viable amongst the community?

You make a good point Cam, that many Terragen users are artists. But I think what we need confidence that there will be those who will contribute to reference material, including the artists amongst us. The idea is invite such people, give them guidance, and let them contribute into a system that is well defined. And, those that do make a TVRS enjoy a structured learning of at least that setting they chose to do a TVRS on – nothing like taking a close look at that setting to better understand it yourself.

I think it's also important to remember that this particular contribution requires the lowest amount of time to prepare, and doesn't require deep technical knowledge. A tutorial requires considerable investment of time. A book is a huge time investment. But a TVRS could be made by even those who don't have a deep knowledge of the setting because they can visually see how the setting affects the image, which is all that is really required for a TVRS. Once the TVRS is posted, others can provide or expand upon the technical explanation in text. As such, even hobbyist Terragenians can get reference materials into the system whereas, as it is now, one needs to be fairly skilled at Terragen to publish help material.

As for Planetside needing to do this, I also agree with you Cam. But, the Planetside forum has been lit up for at least 3 years with discussions of documentation needs, with some resorting to criticism of Planetside. But Planetside has made its decision how to handle documentation which has been clearly stated at various times in the forum. It doesn't appear to me that Planetside is going to take on such a project as this, nor does it appear it has the resources to do even if it wanted to. So, it seems to me that either the community is going to do this, or it's not going to get done.

-Pat

PCook

#8
Excellent points Micheal! A fellow by the name of Ryan Grobins from Australia did something similar:

http://www.motionmagnetic.com/a_terragen2/atmosphere_examples/atmosphere_examples.html

It's curious that his website at http://www.motionmagnetic.com/index.html# doesn't seem to link to the examples page at the link above, which I obtained from a Google search using the keywords "terragen settings".

And you are right Micheal, the TGD is in XML format. So it would be feasible to write software that "parses" the XML of a TGD to ferret out the setting that was changed. Thinking out loud, a user would make a setting, save the project, execute a command of the app to capture the XML of the TGD, then make a setting change, save the project, execute another command to capture the XML of the TGD, then repeat. The app would then examine what XML element was changed and isolate that one setting with its values. With the changed XML element identified, the app would then run the Terragen from the command line to produce a render of each of the three setting values. The app would then capture the TG screens and get them into a format ready for the wiki submission, and possibly even automatically upload the screen captures. That could work!

The user would still need to prepare a scene that was intended for the simplest possible demonstration of the target setting. And the user would still need to make the setting adjustment within a range that was meaningful, because the minimum and the maximum settings will not always be practical for illustrating the effect of a setting. And, you don't want to show a large complicated scene because the setting effect is not likely to be visually evident in a big scene. And the contributor still needs to know if a setting needs a TVRS or not. So, the human hand will still be needed, although the business of capturing the content for the wiki could be automated to some degree.

Even though an app could be developed to do the above, I wonder if its value would be all that significant given that the human still needs to be involved to setup a simplified scene that would focus on the scene element that the setting would affect. And, such an app would need to be distributed and supported. It would be possible for the entire TGD to be uploaded to a website for analysis, or in this case three incrementally saved TGD's each with the changed setting. Someone from Planetside might want to comment on whether uploading TGD's like that would be a licensing issue.

Excellent contribution to the discussion Micheal!

-Pat

mhall

Found it!

This is the tool I was thinking of:

http://www.planetside.co.uk/forums/index.php?topic=12721.0

The thread is only a year old. I was hoping to use it sometime, but never heard anything else on it. Most of the examples are in the atmosphere/clouds realm, but he does have one changing the blend on a stones shader ... I'm hoping you could do blends of most any property.

The author (neon22) is still active here, but it's been about a month since he posted.

Just came up with this:

http://neon.wireframe.biz/node/38

Looks like you need a number of other tools installed to use it.

~Micheal


mhall

Also, I got bored and spent a couple hours writing a simple program to allow a user to inspect a TGD, select a value and specify starting and ending values, along with a step count over which to modify the value.

It's working pretty well and outputs TGDs ... but no command line rendering. Also, it's not got much in the way of smarts. It will allow you to modify any value that can be converted to a number, but doesn't know if it should be limited to a particular range, or if the value is really a Boolean.

Still, kinda useful and I might play with it a bit more.

Hopefully, though, Neon22's is in working order and allows you to blend any value you want ... it looked pretty feature rich.

~Micheal

PCook

Good detective work Micheal!

I looked over neon22's Batch_TG2 and it seems very capable. It's an automated batch rendering app targeted at creating multiple renders from a range of a particular setting which would be useful in animations and panoramas, and possibly even creating HDRI's. The app is written in Python and requires several other third-party software apps to also be installed. However, it is something I'll try out for its intended purpose.

I can see how Batch_TG2 might be pressed into service to create TVRS's (Terragen Visual Reference Sheets), but the app is way overkill for TVRS's. And, one needs some technical knowledge to install it along with it supporting software. So, I wonder how many people would actually press it into use for rendering out educational content. It also doesn't work with the non-commercial version of TG2 because it requires the command-line capability - we don't want to marginalize the non-commercial TG users. There seems to be fairly significant barriers when it comes to using the tool for TVRS's.

But the introduction of Batch_TG2 to this discussion does raise a very important question. Would people be more likley to create TVRS's if they had a tool that automated the process? And, what would that tool look like?

-Pat

neon22

#12
Speak of the devil and he comes !!  :o haha

TGD_Batch has been rewritten a number of times and a year ago I migrated it to QT based UI, using pyside (a python wrapper of Qt).
I haven't updated it for the changes from 2.4.X - specifically for animation - which it did quite well. So it might not work so well with the latest tgd files. I'll get to it when I finish my current project. Sorry in advance...

Proper thread is here: http://forums.planetside.co.uk/index.php?topic=12769

You could animate a number of frames, each of which rendered a tile, which it would then stitch back together seamlessly.
It also makes panoramas. in a similar way. Yes, even animated panoramas.
But at the time imagemagick did not have openexr support built in so it could not stitch open EXR tiles back together so an HDRI pano could not be made automatically. Maybe this year imagemagick has openEXR support compiled in ?

I haven't tried it with the 2.4.X yet - but I did make a fix for hetzen using win7_64 bit. So it does run.
Its a simple procedure to load and get it working. My instructions are long and drawn out but the steps are few and simple.

Its all been modularised to support remote rendering and I started a Django website but lost focus a year ago to make a coomunity supported remote render mechanism to use spare CPU sycles on anyone's machine who cared to help out. Maybe I'll get back to it... I guess we're getting more users.

as for its suitablity to make TVRS's - looks like its well suited to it actually. Maybe you should try using it (on an older tgd file).
Cheers...

PCook

Thanks for joinging this discussion neon22. As a software developer myself, I can appreciate the amount of time you must have spent developing your application, and I'm sure it has served you well in your own Terragen work.

Given that there may be a tool to create TVRS's, my question to everyone following this thread is this:

Would Terragen users be more likley to create TVRS's if they had a tool that automated the process?

-Pat

Hetzen

I'd rather see these sorts of things within the Wiki tbh. Its a good idea and will certainly help flesh out the documentation.