Idea: Terragen Visual Reference Sheets

[Tangled-Universe]

I agree with Jon (Hetzen).
For instance, your example showing hard layer steepness could perfectly be submitted to the strata and outcrops shader's wiki.
However, even better would be one image covering all settings of the shader. I made it quite a while ago and it should be somewhere here, but I haven't found it yet.

[PCook]

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

Hetzen; The TVRS's are intended for a wiki. The idea is to create a reference where Terragen users can quickly lookup (in an online wiki) what the visual affect a particular setting would be.

TVRS's wouldn't be documentation par sa, in the sense that text would be secondary. Rather it would be a visual reference that is a) easy to create and post and b) fast to locate and use. TVRS's capitalize on visual recognition of change in the sample image.

-Pat

[PCook]

your example showing hard layer steepness could perfectly be submitted to the strata and outcrops shader's wiki

Under normal circumstances, that would be a good point. However, I think we want to avoid trying to combine with Planetside's wiki work, mainly because we need to be sensitive to the fact that the Planetside staff can't spend as much time on documentation given only three people and other priorities (at this time). If we were jamming in images into the current wiki, it would force the Planetside staff to have to moderate the wiki changes and I'm sure they would feel the pain. But more important is that the TVRS wiki would need to be unfettered by text, allowing many TVRS to get posted. However, Planetside would most surely reference one of more TVRS's or even copy them into their documentation. But they need to do that work as they are able.

even better would be one image covering all settings of the shader

Also a good point. But the TVRS's each need to be simple - highly focused on one setting. However, your suggestion would be implemented by creating a number of TVRS's as a set that is related to a particular shader.

-Pat

[neon22]

Taking your TVRS idea and just rendering a simple strata node with steepness (as in your original image).
The default output looks like this:
Seems like it could be good for this purpose...? Just a name change to the file would be good.

[PCook]

Thanks neon22 for running the test. Yes, your app could work for TVRS. A couple barriers seem to exist, please correct me if I'm wrong.

First, the installation of the various components is beyond most (you really should consider making an installation package). Second is that it's much more than needed for TVRS - will the user get lost in it's capabilities? Third, from a previous post you indicated that it may not work with the most current TG version - will a user need to install an older TG version?

You didn't build it for TRVS's, obviously. But I wonder if those persons willing to install your app and its supporting software will do so just for TVRS work. It just seems to me to be more technical than necessary for creating TVRS, which really should be simple and fast to make, else we'll not have many people making them.

-Pat

[neon22]

Your app could work for TVRS. A couple barriers seem to exist:
First, the installation of the various components is beyond most (you really should consider making an installation package).

I'd like to, and I see there is something called cx-freeze which might help me package it. But seriously its four steps. I didn't think it would be that hard. I guess the test was that it was too hard, or the app just isn't that useful so the barrier was too high for people to be bothered. All good information

Second is that it's much more than needed for TVRS - will the user get lost in it's capabilities?

Well its basically designed to do exactly what you propose for TVRS.
I.e. pick a a shader node, a parameter to vary, its range and how many steps.
(OK you can also choose to vary two or three params at the same time.)

It also allows you to tile if your image is really big (less of an issue these days but there wasn't a 64 bit TG2 version when I wrote it - and its still useful for load balancing if you have several machines.)
And it does panoramas with one button. Seemed like a useful thing to add...

The creeping elegance part would be trying to get to remote rendering on a network - but its not enabled at this point anyway.

Third, from a previous post you indicated that it may not work with the most current TG version - will a user need to install an older TG version?

Actually I used it today with current TG2 so it works fine but not if your tgd has animation in it. Then you lose
So I need to fix that - and add the 6 new nodes and the new parameters scattered through.

As a side Note - as at 2.3 there were 74 animatable nodes with 885 args.
Cheers...

[Tangled-Universe]

your example showing hard layer steepness could perfectly be submitted to the strata and outcrops shader's wiki

Under normal circumstances, that would be a good point. However, I think we want to avoid trying to combine with Planetside's wiki work, mainly because we need to be sensitive to the fact that the Planetside staff can't spend as much time on documentation given only three people and other priorities (at this time). If we were jamming in images into the current wiki, it would force the Planetside staff to have to moderate the wiki changes and I'm sure they would feel the pain. But more important is that the TVRS wiki would need to be unfettered by text, allowing many TVRS to get posted. However, Planetside would most surely reference one of more TVRS's or even copy them into their documentation. But they need to do that work as they are able.

even better would be one image covering all settings of the shader

Also a good point. But the TVRS's each need to be simple - highly focused on one setting. However, your suggestion would be implemented by creating a number of TVRS's as a set that is related to a particular shader.

-Pat

As far as I know Planetside encourages us users to update the wiki with references, examples or any kind of useful information.

I guess you're worried by the quality, but that's only a matter of using the right methods to determine what certain settings in a node actually do and also make a systematic report of them.
Your concern about quality, if it is, isn't very valid since you want to make these TVRS's anyway. As it seems the information at the moment isn't sufficient enough anyway so I don't see any reason not to append it with extra information. As I just said it is encouraged by Planetside.

I have the feeling you like your TVRS idea so much + automating it potentially that you may eventually miss your goal or get passed it.
Why shouldn't a TVRS contain just one image containing all the settings?
Just look at this image, it isn't covering the whole story about the strata and outcrops shader, but with 1 or 2 extra ugly drawings I could easily explain hard layer depth, hard layer spacing and the tilt settings all in 1 big image;
http://www.planetside.co.uk/forums/index.php?topic=8415.0
It's not that hard to cover most of the story in one or up to a handful of images.

The way you propose it now I'm afraid it will end up with a enormous pile of images and information and I'm not sure if the message/information will really be absorped anymore then?
Especially inter-related settings are impossible to be dealt with using the approach you're proposing. The pile of images will only get larger and things will get confusing pretty quickly and easily.

I'd recommend using the least amount of images possible, explaining individual settings and only mention if one setting can effect the other.
It's up to the user then to apply that basic knowledge and refine it further through experimentation and experience.
I guess that's how it works with everything else too besides TG2.

Cheers,
Martin

[neon22]

Yes - interestingly my TGD_batch program does not produce useful results for a complex node like the strata node.
The interrelationships between the parameters is not at all evident when trying to make, say, a triaxial blend between steepness, altitude, numoctaves.
The only way to work out what is going on is to experiment with all the params manually, or to find some doc like the excellent link in Tangled's post about how the node actually works.

This is one reason why I seldom use the Blend aspects of TGD_batch anymore even though it makes nice pictures
(but panos are super useful IMHO :-))

[PCook]

As far as I know Planetside encourages us users to update the wiki with references, examples or any kind of useful information.

No argument there. But my thinking was to avoid trying to follow PlanetSide's wiki content so that the TVRS's could be populated without being throttled by another's content. However, PlanetSide would be welcome to use any TVRS they choose to augment their content. Keep in mind also that PlanetSide must moderate wiki content, so PlanetSide would need to buy into this TVRS idea if it wanted to use such material. But TVRS's are not intended to augment PlanetSide content, although they could.

I guess you're worried by the quality

Quality is not the objective. The TVRS concept is intended to produce quick references to visually demonstrate isolated setting values. Quality is secondary to usefullness. However, if by "quality" you mean "accuracy", then yes, accuracy is essential.

I have the feeling you like your TVRS idea so much + automating it potentially that you may eventually miss your goal or get passed it.

I don't fully understand your point Martin. However, reading your statment literally; First, I don't "like" my TVRS idea as much as it was intended as a starting point for discussion. If the community doesn't like the idea, it will not likley support the project which will likely result in one person hammering out a bunch of these that will simply go unused. As for automation, that would be wonderful. But how to do that without asking people to become software technical experts is the question?

Just look at this image, it isn't covering the whole story about the strata and outcrops shader, but with 1 or 2 extra ugly drawings I could easily explain hard layer depth, hard layer spacing and the tilt settings all in 1 big image;

Good point. But the idea of the TVRS's is to *not* attempt to cover the whole story, rather to isolate to a specific setting in each TVRS to show what that setting does. As such TVRS's become a fast wiki-based lookup of just a specific setting. TVRS are not intended, as I envison them, to be tutorials or node references - just setting referece points isolated to one setting, and not relationships of multiple settings. Then, using a wiki format, the user can look up the TVRS that concerns that one setting of interest to the user at that moment. TVRS's are, then, a fast reference *during the workflow*, not to study Terragen.

The way you propose it now I'm afraid it will end up with a enormous pile of images...

Not if each TVRS is focused on just one setting. The original idea was for each TVRS to show the affect of a particular setting in a lower, middle and higher setting, just enough to visually show the *affect* of the settings. In that sense, each TVRS is 3 images at most, stitched together to show the 3 selected setting levels.

I'd recommend using the least amount of images possible, explaining individual settings and only mention if one setting can effect the other. It's up to the user then to apply that basic knowledge and refine it further through experimentation and experience.

Agreed. Avoid relationships of settings. Avoid in-depth text explanations. Just give the user a fast visual as to what that setting looks like visually. This simplicity of TVRS's also allow TVRS to be constructed much faster, even by people who do not have indepth knowledge of the setting, thus more TVRS contributions, thus more value in a TVRS library.

I guess that's how it works with everything else too besides TG2.

Also agree. When we can reduce a thing to it's simpliest element, we can better understand it. Then, we can more easily merge that knowledge into the big picture.

Thanks for your points Martin. My question to the community is this: is this thread moving us closer to validating the TVRS idea?

-Pat

[cyphyr]

Great idea but I fear it simply won't work as you expect.
Terragen is an integrated system in that what each node does, and what each setting within each node effects, is governed by where that node is placed within the larger structure of the nodal tree. For example the steepness setting you're illustration your idea with is itself modulated by the patch size of the preceding "compute terrain", amongst many other settings.
I started a discussion a while ago about a "Terragen modus operandi" in which I was trying to bring forward an idea of a wider overview of how Terragen works. Some approved of the idea and some did not, but the general response has always been that if your idea is good enough and you get the ball rolling people will jump on board. However as has been mentioned most Terragen artists are busy making their latest master work and have little time for community projects.
Cheers
Richard

[Tangled-Universe]

Hi Pat,

With "like" I was referring to your enthusiasm and that it possibly may blind you from some disadvantages.
So keep up the enthusiasm Seemingly, from your reply, you're not suffering from tunnel-vision so eventually my concern seems invalid.

To answer your final question; I'm not sure.

To me it seems you're the only one with a clear idea on what you want and how it should be done, because I'm not getting it (yet).
For instance, you agree with me that the least amount of images must be used to cover explanations, yet you want to make a series of 3 images for every individual setting.
In my eyes these 2 don't go along really well together?

So at the moment I'm not seeing a real advantage of your TVRS over my suggestion to just support the wiki articles with some visual explanation by one to a handful of images in total.
I think it's really a personal matter of how people like their information presented. Clearly, I like it condensed as much as possible into the least number of images.
However, I can understand why you'd like to have every setting explained separately and I see some situations it will be better than my suggestion.

Maybe you can try make 1 or 2 and see how it works out for you but also see how people think of it? See if they think it is a really useful way of explaining stuff?
In a way I'm not very suitable for judging this I'm afraid when it comes to the red nodes.
Maybe I'm not the best person to discuss this with as pretty much anything would work for me since I have enough background.
Except for about a handful things they don't have many secrets for me anymore, so consequently I probably understand it much quicker than less experienced people.
However, for blue node logic (which I lack) I consider myself suitable as a reviewer on how useful information is. If you can make me understand blue nodes then you did a fine job

Cheers,
Martin

[Tangled-Universe]

Great idea but I fear it simply won't work as you expect.
Terragen is an integrated system in that what each node does, and what each setting within each node effects, is governed by where that node is placed within the larger structure of the nodal tree. For example the steepness setting you're illustration your idea with is itself modulated by the patch size of the preceding "compute terrain", amongst many other settings.
I started a discussion a while ago about a "Terragen modus operandi" in which I was trying to bring forward an idea of a wider overview of how Terragen works. Some approved of the idea and some did not, but the general response has always been that if your idea is good enough and you get the ball rolling people will jump on board. However as has been mentioned most Terragen artists are busy making their latest master work and have little time for community projects.
Cheers
Richard

Yes, exactly my concern I epxressed a few posts ago.

I see the usefulness in some situations, but I think it's mostly because of above mentioned reasons why I also don't think it will work as expected.

However, he's also right that that if an idea is good and people are convinced then they will likely hop on and help out.
So I'm still happy to see us proven wrong.

I hope you understand that we do support your idea (very much!) since we all paid for this and still have to do with incomplete documentation, but Richard (I think) and I are trying to make you aware of the difficulties and scope of the task.
Imagine the difficulty and scope when considering PS couldn't do it themselves
The open ended nature Richard mentioned is debit to this and that's why his modus operandi discussion is such a good one.
Terragen is a kind of language. Nodes are words and the network is the grammar. If you don't do things in the right order then it doesn't make much sense.
Your TVRS would be the spelling of those words, but wouldn't cover the whole story.
My suggestion wouldn't also, except for showing the complete spelled out words.
The language you will need to learn by practice and lurking these forums.

The best tip I can give you is using the search function and use the "search by user" field and look what Matt and other experienced guys have to say.

[Kadri]

Talking is fine , no problem!
But talking if this works or not is not so much meaningful.
Make it ! Talk about how to make it better.
I like reading what all you write here and it is of course not up to me to say what you should write or not.
Threads like this come and go but nothing much does change.

The problem TG2 faces is that the user base is not so high that user generated content is high too.
Basically some people have to make this.
I wanted to make a tutorial about using texture maps in TG2 but making a tutorial is harder then i thought.
And i was already kinda aware of this  Although it is very easy considering other aspects of TG2 .
I will make it anyhow hopefully.

And if the user base would be high we would get professional made books tutorials etc. too anyway.

If you guys make a tool that is easy to use for such tutorials it could rise the probability that users use it for tutorials.

Not sure if this comes out as pessimistic .This is not my intend.
Make it , use it! If it has problems they could be fixed , changed .
If we only argue if this or that will be useful , i can see already another thread from 1-2 years latter linking here !

[yossam]

I agree with Kadri. I have seen I don't know how many threads about improving the documention, but nothing ever comes of it. I don't know if it's apathy on the part of the members or something else. I understand that Planetside has more than enough to do already. I for one will be glad to donate render time to this project or any other that will improve the ease of use for the users, current and future. I think that the user base would grow if the software were a little easier for beginners to understand. I have friends that are interested in the software from seeing what pics are posted here, but they are scared too death that they will not understand how to use it without getting frustrated.

Again count me in, for whatever use that I can be.

[PCook]

I for one will be glad to donate render time to this project or any other that will improve the ease of use for the users, current and future

Ahh, sweet progress. It's such a downer to focus on what can't be done or why a thing shouldn't be done. yossam brings hope.

I think that the user base would grow if the software were a little easier for beginners to understand.

Good observation. If beginners can get started easier, it's likley they will be helpful to others sooner, if at all. A good start get's a person's confidence up to help get through the tough times.

they are scared too death that they will not understand how to use it without getting frustrated

Just last night a friend told me that he had downloaded Terragen after seeing it on my PC two weeks ago and a couple images I managed to create. He said he didn't have even the slightest clue how to use it, and uninstalled it. What a shame because he was interested in Terragen. People can't keep getting scared off like this.

Do we have any other supporters (or dissenters) of TVRS's?

-Pat