Not to be ungrateful, but I'm a bit unhappy with the progress on this program. I've waited a year for an update, and got none of what I'd hoped.
1. Multithreaded preview
2. Spherical Camera - (really, how hard is it to do this?)
3. more output formats (exr AND hdr would be nice)
4. more output GEO formats. LWO is not useful. OBJ, fbx, or anything that plays nice with the major apps like Maya and Max, and XSI
5. render passes, matte objects and a dedicated zdepth pass that is automated
I don't like jumping through hoops to convert this thing to that. Many man hours could be saved by inclusion of some simple stuff. Sorry if I sound angry... I am doing a project in TG again, after some time off, and am finding it to be very painful to deal with in a pipeline.
Andrew
exr is already there, why have hdr too ?
:)
Richard
by spherical camera you mean fish eye?
Because converting 6 90degree fov's is only done (for free) through tools that use hdr. Why not include more formats? That's sort of the point. Stop making users jump through hoops to convert one thing to another.
Why use obj when you have lwo? Because it makes life more livable;)
By spherical camera, I mean a camera that can output a 360 panorama in one shot. Other camera types would be nice though... fisheye, distortion controls... but I'd love to have an easier time exporting panos.
Andrew
Well you also have to take into account that size of staff working on the program. If you are comparing to vue or something you should know that Planetside dont have as much resources as e-on might have. Hence the price is much cheaper then vue. I think you are getting a lot more then what you paid with TG2.....
no offence but if you think that jumping through hoops to get what you want is something you dont like to do then I should say WAKE UP!!....:)
Quote from: MrHooper on December 02, 2010, 03:25:49 PM
Because converting 6 90degree fov's is only done (for free) through tools that use hdr.
Sorry but that's what "action script" was invented for ...
fbx would be good I agree
Render passes, hmm I see where your coming from, but leaving it open like this dose allow you to make other more exotic passes if you require, and adding a z-depth pass is pretty easy really ...
:)
Richard
why, I mean Andrew does have some valid points and wishes. Trouble is the wishlist is extremely long, and while for some, the few new featurs of 2.2 are *just* what they have been waiting for all year long, they might be less relevant features for Andrew and his priorities.
There was a "name your top 5 feature wishes" or so thread recently. I think it got a million replies and everyone was adding 5 more feature wishes. That's quite an intimidating expectation on the Planetside folks. To be realistic, it will only ever be a handful or two new features that get implemented. But how to decide what's both urgent and important at the same time, and should get a higher priority than the other thousand features on the wishlist? Everyone will give you a different answer.
I can well imagine that if one future update would focus on animation features, for me it would be an irrelevant update, personally, because I don't make animations, .... while others would be super happy.
On the upside, the application is constantly improving and I'm sure we will continue to see new, ingenious and revolutionary features added from time to time, that make TG2 so unique in the industry - despite, or maybe because of it's sharp focus on making awesome procedural landscapes. Well, if you'd ask me :-)
Cheers,
Frank
What Andrew would like with z-depth passes is that it is generated along with the final render, but I figure you already understood that, Richard.
It's pretty easy to do in a second render, but it takes a lot more time and the result isn't flawless.
It's probably a matter of strategy and Oshyan explained this before here.
First 64-bit and proper animation tools are of high priority.
What happens after that I can't tell, no idea.
A logical step, in my eyes, would be to work on FBX like features.
After that we would have a stable (already is), powerful (64-bit), well animate-able program with good exchange possibilities.
I think from that point on things like spherical cameras, renderpasses etc. should have priority.
This is strictly only my idea of how it could go though. Luckily I don't have to decide, too many things I'd like to have :)
Like dhaval said it's difficult for PS to develop as fast paced as Vue, as their resources are tiny compared to E-on's.
Frank approaches this from another perspective, but I think he's right as well.
Fact for me too is that it is improving, especially if you look where we started initially.
Vue evolves faster because of aforementioned reasons, but the fact that they "borrow" ideas from TG2 says enough about how unique it is :)
Perhaps before I gripe, I should say that I do love this addicting program. I gripe because I love. I mean, there are shots I do with it where people (like to president of the company) walk by my desk and say "WOW". Yes, I like the program;)
So, I don't mean to make people defensive. sorry if I'm a bit aggressive.
I had wrongly assumed that my list was among the top requests. I know I made them some time ago. Most of them are actually pretty simple. Additional image formats, and spherical camera "seem" like easy ones. An obj or fbx exporter, maybe not quite easy... but not reinventing the wheels like 64-bit or multithreading. Wait'll you see my "other" list;)
I will pay more for better upgrades!
Andrew
Quote from: MrHooper on December 02, 2010, 02:59:46 PM
Not to be ungrateful, but I'm a bit unhappy with the progress on this program. I've waited a year for an update, and got none of what I'd hoped.
1. Multithreaded preview
2. Spherical Camera - (really, how hard is it to do this?)
3. more output formats (exr AND hdr would be nice)
4. more output GEO formats. LWO is not useful. OBJ, fbx, or anything that plays nice with the major apps like Maya and Max, and XSI
5. render passes, matte objects and a dedicated zdepth pass that is automated
I don't like jumping through hoops to convert this thing to that. Many man hours could be saved by inclusion of some simple stuff. Sorry if I sound angry... I am doing a project in TG again, after some time off, and am finding it to be very painful to deal with in a pipeline.
Andrew
To add to what has been said already, and please don't be offend, first up some one is late to the party in fact its over already and everyone's gone home already and second of all to me it sounds like a case of sour grapes to me!
What do I mean by the above its called prioritization meaning that critical things things like 64 bit compatibility and memory efficacy and other core functionality like software stability comes first bells and whistles and fancy doodads like increased interoperability such as FBX come latter.
Planetside are a business, meaning that there is a certain level of information they cannot release, as it is of a commercial and confidential nature, that if it where released before it where "Time to Market" ready: competitors of Planetside would gain commercial advantage from that information; this would not be a commercially or fiscally responsible move for Planetside to do, as would be the case with any other company.
I am sorry that your disappointed with 2.2 but it is the beast it is, this beast is sleeping right now, but it won't always be and that in time it will get teeth and claws then maybe the down voting can stop, but I'm not going hold my breath, all things considered as is 2.2 is a great offering, flawed yes but still I can see where a years worth of work went. Have Faith, and Peace be with you. 8)
Regards to you.
Cyber-Aangel
Quote from: Tangled-Universe on December 02, 2010, 03:50:22 PM
What Andrew would like with z-depth passes is that it is generated along with the final render, but I figure you already understood that, Richard.
It's pretty easy to do in a second render, but it takes a lot more time and the result isn't flawless.
Yep I knew that ;D
Its just not that hard to do and whilst a "ready rolled" solution might suite some people a "roll your own" is much more adaptable to a wider set of circumstances. I've never seen the Alpha build but judging from other apps DoF (that would be one main use of a z-depth render pass) is not an easy thing to achieve 100% realistically, there's always a compromise when effectively looking through blurred detail to see previously occluded detail. For myself one of the joys of Terragen is to be able to push the boundaries, to have a system that is open enough to let me try new solutions yet robust enough not to crash on me half the time rather than straight jacketing me into the manufactures pre-sets.
:)
Richard
Some of the things may already be in process. I think they (Panetside) released what they were sure was bug free. The wish list is very long...
Terragen 2.2 has been a wonderful release for me, starting with the faster render speed. Thanks!
>> Multithreaded preview
This would be great. In fact didn't a multi-threaded preview get added around the time the rest of the programme got multithreaded? I think it got removed because of problems, I guess they never found a way to get it to work.
Asside:
When I was a student I lived in two places at difference places at different times, one was a large flat with a studio room and great views, the other was a small (9`x6`) bedroom in a not very shared house. I hated the room, I loved the flat. I got almost nothing done in the flat with its large studio and great views, spent all my time gazing out the window, planning my next great work :) In the small room (which I hated, did I already mention that) I furiously burned the midnight oil, painting away into the wee small hours of the night and I created some work that I was happy with and some people were happy to give me money for :) (they were probably just being kind) The point is that restriction is not necessarily a bad thing, its useful, it forces you to find creative solutions to problems which can in turn inspire new creativity.
If Terragen had been the killer do everything landscape app that fulfilled everyone's ideals, it would not only be an impossibly complex tool but it would also lead to some very monotonous and boring work.
8)
Richard
Hi,
Not all the work over the time since the previous release has gone exclusively into what is in this current release. For example a lot of work has gone into getting a 64 bit version up and running. There are also a number of other things (new UI and nodes etc.) which will likely be going into the next release which were developed during the last release cycle.
We can't do everything that everyone wants although we try our best :-). With limited resources we need to prioritise some features over other features. Some of the things you've mentioned are items we have planned as part of specific releases.
Regards,
Jo
Jo,
I hope the animation module is one of these things, since it has long been promised. But, I can wait. No push from me.
Let us know, though, please.
Quote from: jo on December 02, 2010, 07:35:02 PM
There are also a number of other things (new UI and nodes etc.) which will likely be going into the next release which were developed during the last release cycle.
I'll simply reiterate the basic premise, that I think wider customer adoption is being limited by being too closed a box. Helpful exports for geo and image data are critical for a pipeline. This isn't glamorous work, but it's the heart of working in a production. Pipelines can be built to integrate TG right now, no question. But each customer must sort of roll their own scripts to deal with lack of export. Purchase conversion software, hire an intern to deal with the data wrangling.... etc. But I'll stop here. You get the idea.
Andrew
MrHopper, Sir I must Interject,
You do, of course, realize that Terragen is not aimed at the geophysical and GIS market segments it is to physically inaccurate for that and dose not contain the precision or the data processing tools (Right now) nor the data IO formats to support this; this may change in future, but at the sole discretion of Planetside and no one else, yes users may make them aware of need, but that's about it.
Let us take Satellite data for example you have over fifty different file formats alone for satellite data and that is if your only interested in visible light data, but it gets far worse if you consider far and near inferred, spectral analysis and all the other kinds of data that can be collected (Mission design and Sensor provision dependent) many of these are proprietary any way or are designed to run on operating systems such as IREX.
Dependent on sensor load out, you can be talking about 100+ channels of data and that is even if you handle if off line rather than real-time, I cannot see why Terragen needs to handle it since this is not its market segment as far as I am aware.
GIS, has at least Fifty+ formats again many of these are proprietary, survey work produces vary large data sets alone, TG2 is not in the same market segment of World Builder, designed for things like Civic Planning and Golf Course Design, TG2 fits into the Market Segments of Digital Art Creation and Digital Asset (Also called Digital Content) Creation (DA[C]C).
What kind of pipeline did you have in mind? Because it sounds like you envision TG2 been adopted into a wider market band then is its primary market? TG2 has already been used in a number of motion pictures, shorts, and commercials and is likely to be so in the future, all good things come to those that wait. 8)
Regards to you.
Cyber-Angel
I would guess he's in the CG animation world with what he's said so far. That said, it is entirely possible to output all of what he's asking for right now, it's just more work than he/we would like.
I have a camera rig saved as a TGC I use to create HDRI sky box plates, and a preset PDF to attach the six sides together to port into Max. I've not really needed DoF, as what I've tended to need has motion blur on and doesn't really interfier with the foreground, but that being said, there's nothing stoping you setting up a TGC preset that has a distance shader attached to a surface layer to do this for you. As for an AO pass, again you can set this up with the environment light and a surface layer. Casting shadows can be done in your 3d app with an exported HF. etc.
What TG has over other landscape apps, is it's stability and 'real' look if you manage to master the tools at your disposal. It took me a good week of production time to suss out how to get camera data from a 3d app into TG, which now takes me minutes as I've worked out the process.
There are limits on what you can do, but I think the programs strengths far out way them, it's up to you to find a work around, or simply ask a question here. I would even challenge some of your top 5 as actually being low priority, as you can already do them. Although I would say importing obj sequences would be very usefull, especially if you're playing around with anything reflective.
Andrew, I just want to mention that we're getting a lot more interest from vfx production houses these days - something we've been working toward for a while - and we definitely recognize the needs of these users (including yourself) and want to address them as best we can. We do have most of the features you are interested in somewhere in our plans for the future.
In the meantime I think some documentation of the workflows people have developed would also help a lot. We would like to provide more coverage of the recommended approaches, but it's challenging because there are so many different applications and processes to potentially address. There are various threads around which do cover a few solutions, but consolidating this stuff in the wiki would be great. Perhaps an enterprising person can tackle that at some point. ;D
- Oshyan
I guess I could write something to help out. It'll be a week or so before I start on it.
Quote from: Hetzen on December 03, 2010, 03:39:24 PM
I guess I could write something to help out. It'll be a week or so before I start on it.
Looking forward to that ;)
I recognise most or all of these requests as ones that we plan to address. Oshyan's reply said most of what I'd like to say at this point. But I want to talk about the "matte objects" request.
Holdouts, or mattes that punch a hole in the RGB and alpha channels, are possible simply by attaching a Constant Shader to your object and giving it colour = 0, alpha = 0. Whether you also want the object to cast shadows or be visible to reflection/refraction/GI rays, these can be set in the object parameters (although reflection/refraction/GI are all lumped under "other rays").
If you simply mean the alpha channel, this can be saved from the Render View as of v2.2, and the ability to save an alpha channel has always been available through the "Extra output images" option in the Render Settings, provided you render using "Render Sequence" button or render from the command line.
Let me know if you want some other kind of matte.
Matt
IMHO TG2 is developing pretty fast for the amount of manpower involved in this process.
Of course there are features someone wants to see implemented, but this will always be the case.
just to name two features: fractal erosion (as there's just no solution for this as far as i know [yes, Dmytry Lavrov got something there, but i talked to him a little bit and its basically a normal erosion algorithm that is applied while rendering and iterations limited to a certain amount of memory as far as i know.]) and a fractal blur function.
regarding output formats, you may take in account that every non-free output format added costs money, which will make TG2 more expensive. ;)
I'm just gonna slip a question in here.
Is the "preview window" gonna have multicore rendering support sometime soon?
- Terje
I think at least having the alpine shader fully pluggable would allow for faking certain types of erosion. Vue's eroded rocky mountain fractal has some interesting features as well.
The Alpine Shader is awesome. It maybe slow, but if used properly, it can give you all sorts of erosion effects. The key is not to think of it as bumps of mountains.
At a basic level, the faked erosion effects are simple. What can be done is opening up the disposition types as one example to be allowed to have other shaders connected to it. However, imagine every component being opened up. One had to realize that there are varying degrees of erosive geomorphologies along the same localized terrain. The alpine shader does sedimentary carry-out disposition but no fanning of the dispositions nor does it have a way to thin out and meander it's way downward slope. The rest of the shader does a basic fine-scale cutting based on fluvial effects but that is all. It's too much to simulate these effects but make it plugged and a lot of faked effects can at least get one step closer to realism. In fact, it would be cool to have several layers of effects all within the same shader, mixed by slope and elevation blends.
At a basic level, yes. You're right. But if I wanted to simulate lower detail, I'd maybe want to create a mask in something like GeoControl or WorldMachine2, that I'd apply to the fractal landscape in TG. I'd keep the infinite relief resolution of TG in the render, but have the external erosion simulation as a mask.
It's not perfect. But we often have to pick the camera viewpoint as our guide, and work accordingly. Otherwise we'd have to deal with a lot of off screen redundant calculations.
After reading the page on the TG2.2 release notes I would say that I am surprized how narrow the list of changes is. I was expecting the 64-bit edition myself. The release seems to address little more than clouds, shadows and lighting. If I had to place highest priority on any one new feature, it would be to add the ability to specify a spherically mapped mask for most procedural terrain parameters. Without this, you really can't design a planet with fractal detail that isn't uniformly covered by the same monotonous terrain. Science fiction is full of "desert planets", "ice planets" and "tropical forest planets" and such. But then there is Earth. It has all of the above and many more. Variety is probably the norm in the universe and not uniformity.
You can already apply a raster (image map) mask with spherical projection and use it to control coverage of multiple types of procedurals. Does that not address your needs?
As noted, 64 bit is coming in a future update and significant development toward that goal was done as we worked toward the 2.2 release.
- Oshyan
Well there you have it sports fans...just goes to show you can't please all the people all the time. This year ya get a single update (Fine update it is to don;t ya know) and what do ya know unhappy natives start coming out the woodwork, of all the lowdown, dirty, good for nothing deeds this man has ever seen, well I'll tell ya some thing for nothing I for one am might grateful for what I've got...see, but what some folks figure is that all that hard work, don't amount to a hill of beans...hang dang it.
Yip sir re, and if y'all don't mind me saying so...and if ya do so help me I'm saying it any ways, when to people work for a year like (and if you'll excuse me now, most kindly) like dogs; these no need to shoot em down like dogs, no sir.
If any of this offends, to bad, and in the words of my generation, go figure!
Quote from: Hetzen on December 04, 2010, 08:18:43 PM
At a basic level, yes. You're right. But if I wanted to simulate lower detail, I'd maybe want to create a mask in something like GeoControl or WorldMachine2, that I'd apply to the fractal landscape in TG. I'd keep the infinite relief resolution of TG in the render, but have the external erosion simulation as a mask.
It's not perfect. But we often have to pick the camera viewpoint as our guide, and work accordingly. Otherwise we'd have to deal with a lot of off screen redundant calculations.
Indeed, there is always a temporary way around the problem. ;)
Quote from: PabloMack on December 04, 2010, 09:54:17 PM
Without this, you really can't design a planet with fractal detail that isn't uniformly covered by the same monotonous terrain.
i don't think so. i will upload a file with continents and earth-like distribution (in colors) of different landscapes later. its basically easy to get this with what we have already, just a matter of how much work you want to put in it. something like the "use y for altitude"-function combined with some fractals gets you some pretty good results.
Quote from: ChrisC on December 04, 2010, 05:33:19 PM
I think at least having the alpine shader fully pluggable would allow for faking certain types of erosion. Vue's eroded rocky mountain fractal has some interesting features as well.
hm, im not sure about this. i played with the alpine shader for some time trying to figure out what exactly it does and how i could influence it or rebuild it with another look from functions. try the alpine shader with only one octave and play around with it; its basically a voronoi with some pretty clever combinations and blendings for the next octave, but as its a voronoi not good for rocky mountains etc. but it just came to my mind that a billowy perlin as base could look good...
Quote from: Goms on December 05, 2010, 03:11:25 AM
i will upload a file with continents and earth-like distribution (in colors) of different landscapes later.
and there (http://forums.planetside.co.uk/index.php?topic=11284.0) it is :)
Quote from: Cyber-Angel on December 04, 2010, 10:36:49 PM
If an[y] of this offends, to- bad, and in the words of my generation, go figure!
None taken. Enjoyed your way of saying things...
That planet is a good starting point Goms.
I used to do this in Lightwave. Download a spherical image map of Earth, mars, etc. This will give an idea of how to paint the landmass for your custom planet. It is easier to get the projection to fit the sphere from the example. All that was necessary after that was to draw the mountains, plains, lakes, and rivers. It takes time to do it right. The last thing is to add the displacements and further detail using fractals and/or more detail image maps.
Quote from: Oshyan on December 04, 2010, 10:13:40 PM
You can already apply a raster (image map) mask with spherical projection and use it to control coverage of multiple types of procedurals. Does that not address your needs?- Oshyan
Have any of the "In Depth" sections of the user guide been completed? (Terrain, Shaders, Atmosphere, Lighting, Water, Objects, Cameras, Renderers, and Node Network)
The best one I can think of right now is Schmeerlap's Ben McDuff tutorial. http://forums.planetside.co.uk/index.php?topic=7733.0 It is probably the best explained and complete on the forum. I am planning on doing one soon explaining my method (madness) on terrain creation. It will include some of the functions I use, and why I am using them.
Much of the power in this program is not very well documented. I have picked up tons of information from the forum over the year or two I have been here. The best advise is to learn about the power fractal. I use them for nearly everything. Surface Layers and the distribution shader are also my favorite tools. When you are comfortable using them, start adding other nodes/shaders. The search function of the forum works well; but it really helps to be specific.
Follow Tangled-Universe, FrankB, Dune, Hetzen, Seth, Dandelo, Cypher, Ryan Archer, Oshyan, Jo, and Matt's posts. I have left out quite a few, sorry... These are the power users and some of the Planetside Staff. They have the best and most complete knowledge of the program.
im enjoying the new release very much after drama with my license key i finally got it up and running I agree 360 panorama output would be excellent but you can easily create a latlong with a few simple steps
the over all downfall of the program is lack of decent documentation for users I think planetside needs to get together with a group of experts and really knuckle down on the documentation ie video tutorials etc.
I also think planetside to keep up to speed in the market is really open up the sdk and let 3rd party devs tackle the smaller issues such as hdr output vector eps import as well as an fbx import module etc etc
so matt can work on the more important issues.
Keep up the good work planetside
Quote from: dhavalmistry on December 02, 2010, 03:33:54 PM
Well you also have to take into account that size of staff working on the program. If you are comparing to vue or something you should know that Planetside dont have as much resources as e-on might have. Hence the price is much cheaper then vue. I think you are getting a lot more then what you paid with TG2.....
no offence but if you think that jumping through hoops to get what you want is something you dont like to do then I should say WAKE UP!!....:)
I affirm 110% to dhavalmistry!!!