Click to watch the walkthrough video on VIMEO

Originally published in 3DW 13x

Can you generate rock textures in the GIMP for games?
Of course you can. One of the most common mistakes made when generating a rock texture, especially for games, is to forget differentiating between your diffuse and displacement maps – you need to go for the render of rock, not the look of rock, a mistake commonly made in many tutorials, as they add light and bumps to the diffuse texture itself. Adding a light into the texture will almost certainly cause it to render wrong in most cases, and unlike most organic matter, stone or rock displacement maps do not always follow the diffuse map’s colour scheme.

Create a new file at 1024×1024. If you look closely, a majority of rocks and stones will have a base colour, with a variety of hues and colours running through it. We’ll replicate this in the texture, starting with the base: In the Toolbox, set the foreground colour to RGB. 110, 116, 128, and background to RGB 89, 94, 104. Go to Filters>Render>Clouds>Plasma. Set the seed to 525692 and Turbulence to 1. Click OK. Go to Colors>Colorify and set the colour to RGB 89, 94, 104. Create a new layer, Layer>New, and go back to the Plasma filter: Set the slider to 3,5 and click OK. Go to Colors>Colorify, and set the colour to RGB. 110, 116, 128. Turn your texture almost completely grey by going to Colors>Maps>Gradient Maps. Go to Edit>Fade Gradient Maps, and drag the slider to 50. Click OK. Set the layer to Screen. Fill a new layer with Plasma, with seed at 3776291607 and Turbulence at 6. Click OK. In Colors> Colorify set the colour to RGB 122, 68, 15. Set the layer to Overlay or Screen, and Opacity to 55

We’ll add details next: Fill a new layer with a final round of Plasma with seed set to 525692 and slider dragged to7. Click OK. Desaturate via Colors>Desaturate> Lightness. Click OK.Go to Colors>Levels. Set the black Input to 160, and set the layer to Soft Light at about 25 opacity. Create a new layer, and fill it with black. Go to Filters>Noise>RGB Noise, uncheck correlated noise and independent RGB and set the sliders to 52. Click OK. Repeat. Set the layer to Soft Light, at 55 Opacity, and go to Colors>Levels. Set the white input level to 100 or (a little) less.
Create a final layer, and fill it with black. Go back to Color>Noise and fill the layer again. Set the layer to screen at 40 Opacity

 

Tip: Create a displacement-map via using and tweaking Filters>Map>Bumpmap mixed with Filters>Render>Lava set to Rough 50 and seed to 10+. Blend, and render.

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file or the tutorial. Using the content to add textures to your own (commercial) images is perfectly fine, though

Download Rock texture-templates for  the GIMP here!

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

 

Originally published in 3DW 142

Can you create lightprobes in the GIMP? Definitely:  They’re maybe a little finnicky to create, as it’s all about guides, shapes and layers, but having a basic, timesaving lightroom, tent or box-template will hopefully make up for this. And remember: You can use the following approach in Photoshop or any other 2D app as well.

Start by defining a grid for the ”walls”: Go to File>New Document. Set width to 3600 and and to 1800. Ensuring Show Rulers is checked in View, create horizontal guides on the following points: 450, 550, 900, 1250 and1350. Then create vertical guides on the 450, 1350, 1800, 2250, and 3150 points.

For the “floor” and “ceiling”, set the foreground colour to RGB 147 in the Toolbox. Create a new layer under the shapes., Click the Blend tool. Set Shape to Bilinear, and Repeat to Triangular Wave. Position your cursor where the 900 and 1800 points meet, and drag down to the 1350 horizontal.

For the ”walls”, create a new layer[Shift+CTR+N].. Using Rectangle Select in the Toolbox, create a rectangle on the 1350 and 2250 vertical and the 450, 1350 horizontal points. Fill it with white, and select the rectangle with the Fuzzy Select tool.

Go to Filters>Distort>Curve Bend. Check AutoPreview, and click Mirror. This will mirror the top curve to the bottom. Using the filter’s grid to keep your line straight, drag the middle point straigh upwards to about halfway up the first square, tops – aim for a soft slope within the 550 and 1350 horizontal guides. Click OK. Right click the Floating Selection layer, and select Anchor Layer. Duplicate the layer[Shift+CTR+D]. Select the curved rectangle, and fill it with RGB 128. Shrink it 15-30 pixels via Select>Shrink. Press delete and [Shift + CTR + A] to deselect. You now have a basic edge for the “walls” which you can manipulate and duplicate across the rest of the grid. Duplicate this and the curved box layers four times. Using the Move Tool from the toolbox, fit the copied layers into the grid-guide outline, making sure all shapes are aligned. Name them according to position Now that you have the basic shapes in place, you can tweak light- and sharpness. Let’s try one light source and soft edges: Select the ”walls” next to the middle layer, and fill them with RGB 164. Select the outermost ”wall”, and fill it with black. Set every edge layer to 50% Opacity and Filters>Blur>Gaussian Blur>20.

Save as a TIF, and use directly or use i.e HDRIshop to tweak exposures and save as a HDR-file

Expert Tip Add some fun extra lighting points on the “walls” by using the Filters>Light and Shadow> Supernova filter and tweaking the Radius, Spokes and Random slider settings.

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file or the tutorial. Using the content to add textures to your own (commercial) images is perfectly fine, though

Download Lightprobe-templates for  the GIMP here!

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

Originally published in 3DW 144

Even if you’ve dabbled just a little in 3D, you probably already know lighting is one of the hardest disciplines to master in 3D, and that using IBL – Image Based Lighting – can save you hours of setting up and tweaking complex lighting rigs. One of the fastest and easiest ways to bring realistic lighting to your render, using IBL boils down to achieving more with less, the way it works being genius in its simplicity: In most setups, you simply apply an image – also called a light probe or a radiance map – to a scene-encompassing (virtual) sphere, which then reflects the image back into the scene, providing light (luminance), overall colour-tone and reflections for your scene.

If you’re new to creating your own HDR-files, you’ve probably Googled a little, and know there is a plethora of books and online tutorials on how to set up a good IBL render. You’ll probably also know a search for shooting or create an HDR-file for yourself usually results in ending up on an art photography site. This is due to the fact the phrase HDR is interchangeable with artistic HDR-photography. What differentiates the two, despite having roughly the same work-process, is that HDR shot for CG needs to be as natural and effect-free as possible, as it needs to provide illumination, backplate and reflections for stills or animations. Artistic HDRs do not have these demands, and can hence play a lot more with tonal range (tone mapping) and motif.

Technically, you can use most image formats to light a scene, even art-photos. However, they won’t always provide you with a good result, so in most cases it pays to stick to file-formats specifically designed to handle a range of luminance, like HDR, or ILM’s (open) EXR. These are 32-bit files capable of storing a greater amount of luminance data than say, TIFs or JPGs.  Let’s take a closer look:

First presented by Paul Devebec in 1997, a practical explanation of what an HDR-format file is, would be that it’s an file-type specifically designed and developed to store a range of light -luminance – data. You acquire this data through photographing a scene or panorama via a series of exposures, ranging from dark to light (An exposure is the amount of light the camera’s shutter is set to slip through the lens at the moment the image is taken) The reason for taking multiple exposures, is the fact that most cameras are currently only able to catch one exposure per shot. This means that they only capture and store the luminance present in that shot. By shooting a series of shots of the same scene at a range of exposures, ranging from under-  to overexposed, merging them, and saving them as an HDR-file, the whole range of light (or lack thereof) from these exposures is retained in one file, which can then provide your scene with richer light and shadows.
The acronym itself in a nutshell, High Dynamic Range (Imaging), HDR (I) means that through catching more than one range (High) of exposures, and merging them, you store a range, rather than a single set of luminance (Dynamic Range) in a file. You also have Low or Standard Dynamic range images. These contain just one exposure, and hence, just that single set of luminance from the exposure. A series of L or SDRs can be merged into an HDR image, but you can never get the same range of luminance by saving a single L or SDR image as HDR image, even though single-shot-HDRs, as they are called, do get used now and then.

Preparing to shoot your own HDR: 

So how do you get your own HDR-files for CG? If you’re short on time, buy yourself a set from Dosch or HyperFocal. Or you can shoot or render your own, and the road from novice to intermediate isn’t that far. You’ll need some equipment, software, preparations, a little some practice, and a decision as to what kind of HDR-set you’d like to create: Perfect for beginners, mirrorballs are the easiest HDR images to shoot by far. Also called lightprobes by some, mirrorballs are series of centered shots taken of a highly reflective chrome ball at a range of exposures. It’s the reflection of the surroundings into the chromeball you want, so once you’re done shooting, you can merge the exposures, unwrap the mirrorball – just like a sphere, and the flattened image would show a full capture of the place you were at. The advantage of shooting mirrorballs for IBL, is that you can use even the most basic of cameras with manual exposure settings, they’re fairly fast to shoot, easy to reshoot, and don’t need anything but a stable place (preferably a tripod) to be placed on for the shot.

Compared to shooting it’s much bigger cousin, spherical panoramas, mirrorballs do have some drawbacks: No matter the amount of exposures you shoot, they cannot provide the resolution flexibility a spherical panorama can, so even if you shoot your mirrorball from several angles, you will be tied to your camera’s maximum resolution, which means you may want to avoid using mirrorballs for sharp backplates or reflections. Not only because of resolution issues, but also, depending on the quality of your ball, you may run in to  dent-distortion, noise from scratches, and often the issue of having to paint out the photographer from the image, and tidying up pinched edges.
More demanding by far, are spherical panoramas or maps. Also called lat/long maps or equirectangluar maps, they  are a 360×180 degree image sequence shot at different exposures, which are then stitched (merged, aligned and flattened), into a 360×180 (2:1) image. Panoramas can generate a massive amount of images during your shoots, but in return they provide the best basis for flexible, high resolution IBL sets from which you can extract luminance, backplate and reflection maps. Hence, they’re better suited for more demanding CG like i.e. matchmoving than a mirrorball. Their drawback is that they are time-consuming to shoot, more sensitive to errors, and may at times demand more postwork to make them work. But the results are usually worth it.
Mirrorball or spherical map, you’ll need the same basic equipment to start with: A camera, tripod, a chrome ball, and some memory cards will do just fine as a starter kit. Currently, most low-end and average DSLRs have sufficient functionality to get you started, but make sure your camera can let you do and control the following:

Aperture, which is the lens opening. The lens opening controls how much light gets into the image, and is measured in F-Stops. Aperture also controls the image’s depth of field, and thereby its focus. Learning to use it and setting it to manual will help you avoid your HDR-sets being shot with autofocus, risking your images will be shot at different depths of field, which can be hard to fix in post.
Also learn to set your exposure and ISO settings: Exposure means shutter speed, which controls how long light is let onto an image. It is measured in EVs – Exposure Values. You will want to control this manually, because you want all your images to have the same light and tone – the camera imposing its autocorrect or lightbalancing features on your image may completely ruin the set you are shootin
ISO controls the camera sensor’s sensitivity to light, meaning that i.e. bumping up the ISO setting will allow you to take brighter images in dim surroundings. ISO also controls noise. Keep the setting as low as you can, especially if you decide to save your images in a JPG format, as JPG is fairly noisy already.
Finally, also check how to control your White Balance, which controls the colour neutrality of the image, thereby helping you avoid yellow or blue tinted images.
To make shooting easier, check if your camera can do AutoExposureBracketing (AEB) and/or a burst function.  Spend some time exploring and practising these settings, it will pay off when you head out for your first shoot.

If you can get only one extra thing for your camera, make it a tripod. It’s worth its weight in gold, both preventing camera shake when shooting, but also ensuring the camera stays level and static:  When shooting HDR panoramas, you need the camera’s position to stay static (nodal). You can rotate the angles you’re shooting at, but the position of the camera itself must remain the same. This to avoid what is called parallax – parts of the image, normally the foreground, shifting in relation to the other images, something which can be very hard to fix in post. If you don’t have access to a tripod, hold your camera with both hands, and rotate yourself around the camera as you shoot the panorama, not the camera around you, which is our instinctive way of doing it. (Another thing you may want to practise a few times first)

If you want to start with mirrorballs, your best bet may be to get a gazing ball from a garden centre, a silver reflective Christmas ball, or an interior decoration set from Next or IKEA. If you’re going to shoot it on a tripod, make sure the ball has a hole in the bottom. Be aware materials and size matter: A stainless steel ball provides softer reflections than i.e. a painted glass or plastic ball, because of the way the steel is brushed. As for size, get small,   5-10 cm ones for indoor shots, and larger ones, like 25-35 cm for outdoors.
Shooting for HDR can generate a significant amount of images, so get some big and fast memory (SD) cards, especially if you decide to save your images as lossless RAW-format files, which can get very large, very fast. A compressed format such as JPG will only catch 256 levels of brightness in a scene per exposure, RAW catches about 4096. This means that in an HDR image based on a sequence of JPGs, your contrast details in dark and light areas will disappear rather fast, and the image may in addition introduce noise and artefacts into the render. This may not matter much if you’re out shooting on a misty or overcast day, with little contrast or shadows, but it makes a significant difference if you’re shooting HDR on a bright day or in a wooded scene with lots of details. You can of course save in JPGs, if you know your renders won’t need that many details in their luminance range, but go for RAW and lots of memory cards if you’re shooting on sunny days or in busy settings.

You’ll also need stitching software to bring it all together, and there are many applications to choose from. Look for a program or a combination of programs which can merge HDR exposures, import RAW,  stitch, tonemap, handle multi-row images and save in various formats and layouts. If you can’t afford any software, try using Hugin and PanoTools together with the GIMP – these three are all free and Open Source. For commercial software, you may want to try Mr. Devebecs sturdy golden oldie, HDRShop, or ptGUI, the commercial, and fuller featured version of Hugin, and of course there is always Photoshop and Autodesk’s Stitcher.

Shooting your scene 

Wherever you decide to shoot your images,  allow yourself the time to set up properly when you arrive. As your images need to be as crisp as possible to facilitate easier stitching,  ensure your focus, exposures, whitebalance, and ISO-settings are locked, so they’re all shot with the same parameters (bar  the variations in exposure). If you’re shooting a landscape, set the camera distance as far as possible – it eases stitching and gives you more detail to work with. Test your shooting angles to see if any issues show up. You may also want to double-check that you’ve turned off *any* autocorrecting or fix functions on your camera.  If you have company, ask him or her to always stand behind you whilst shooting to avoid cleaning up  extra shadows or reflections.

 

Unlike art HDR, which is often shot at three exposures, HDR for CG usually needs a little more.  Get better render results, by taking between five and nine, even though nine to fifteen exposures  aren’t unheard of.
When shooting a mirrorball, zoom in to make it fill your view as much as possible, especially the top and bottom edges. This will make cropping and merging easier. If pinching on the edges is a concern, try shooting the left and right at 90 degree angles – and whatever you do, make sure that your camera is the same height as the mirrorball and level, in order to avoid distorted or skewed shots.
When shooting spherical panoramas, remember to bring your memory cards:  On an average type camera, you will need to shoot a set of exposures of a bottom image at -90 degrees, followed by a full circle of exposures taken at -45 degrees, to cover lower ground and middle, one full circle at 0 degrees, to cover the middle, a full circle at +45 degrees to cover higher middle and lower sky, and a final set at +90 degrees to capture the sky. In order to make stitching easier, they will need to overlap at about 25-30 %, depending on your scene.  The angles are why you will need a multi-row stitcher application.
If you’ve lost track of where you are in your shoot. shoot again – it’s better to have one shot too many than one too little.

 

Pay attention to movement around you, even a gentle breeze can make trees or grass move when you least expect it, creating ghosts (movement artefacts) which need fixing in post. If you can, wait ’til the breeze dies down, and take the shot again, in order to avoid adding more work to the merging and stitching process. If it happens, consider taking a few extra clean shots for cleanup-clone-stamping in postwork later on, just to be safe.

On a bright, sunny day, pay attention to glare or halos – and consider taking a few extra shots of the sun with the sun covered, and a few extra uncovered – it will make postwork halo-cleanup easier.

 

Pay attention to people and property in your shots.  Most people will happily stand still a few seconds in order to help you with your shot, but there’s also another issue, legalities: This does not matter if you’re not going to share or sell your images, but it does if you do: If there are people in your shot, they will have to sign a release form permitting you to publish or sell the image with them in it. The same goes for private housing or property. This to safeguard privacy and property rights.  Note that some brokerages, like i.e. 3D Ocean will not accept your sets for sale if they have people or private property in them without a release form.

Stitching your scene 

This is where all your hard hopefully work pays off as a set of level, ghost free, evenly focused, exposured, and whitebalanced images, making stitching a piece of cake. Before you begin merging and stitching, however, make it easier on yourself by organising your sets into folders. It will ease the autoloading many applications offer.
Most panorama or mirrorball-unwrapping programs have similar workflows for loading, unwrapping, merging, tonemapping or stitching. Regardless of whether you shot a panorama or mirrorball, try running some manual merging, stitching and general tinkering before you use an application’s automated functions. It may slow you down a bit to begin with, but it can give you some solid hands-on experience, enabling you to make better sense of automation settings later in addition to providing you with a starter kit of troubleshooting skills.
Starting with the mirrorball, the easiest and cheapest way to unwrap it may be to just stay in Photoshop, with the help of either the Flexify plug-in from Flaming Pear or PanoTransforms if you have multiple angle shots (both cost about 20 USD). Mostly because Photoshop now has HDR functionality, and is less awkward on some functions than i.e. HDR Shop or PhotoMatix, like cropping – and your mirrorball needs to be finely cropped.

Using Flexify, the process is as follows: Merge your images using File>Automate>Merge to HDR, follow the instructions on screen, leave the Align Automatically checkbox alone, and merge your exposures. Crop the image as close as you can to the mirrorball’s edges. Go to Filters>Flexify, and select mirrorball under input, and equilinear as output. Click OK. Sometimes after an unwrap, you may see a lot of pinching on the sides. You can try fixing this the Christian Bloch way, by using the Offset filter a third into the image on the layers of your unwrap, moving the pinching to a more visible spot, and then just retouching it away utilizing parts of non-offset layers via overlays, layer masks, healing brushery and alpha mask-painting.

Panoramas are more demanding to work with: Despite Photoshop’s  recent increases in HDR-functionality, a dedicated multi-row panoramastitcher, like ptGui, PanoramaMaker, PanoWeaver or Stitcher, is still your best bet, as they have a larger, specialised toolkit and hence provide better results.  They’re  also easy to use, especially if you’ve got problem free photos.

Using ptGui, one of the most popular stitchers on the market as an example, the workflow is as follows (and similar to most other programs):  Click Load Images in the Project Assistant, and select your images. On loading them, you’ll be either be asked to provide EXIF information or bracketing parameters. Not stored by all cameras, EXIF is/are the camera and lens parameters stored in the images you’re importing. If you don’t know, enter the default parameters into the popup:  Rectilinear, 35 mm focal and 1x multiplier.
If it prompts about linking bracketed images, ptGui recognised a pattern of exposures on load, and wants to know if it should link them. Click “Enable HDR Mode but do not link images” – it gives you bigger freedom to work and troubleshoot, should you need it, as it will load all images, and not merge all images into one until you say so. After the images are loaded, which may take a while, they’ll display in the source image tab. Press CTRL + E to bring up the Panorama Editor, so you can see what’s happening.  Click Align images in the project tab. ptGui will now try to build your panorama for you.

 

Sometimes you may be asked about controlpoints. These are the overlap points between the images (and the main reason for shooting with overlap). When it finds a match, ptGui will automatically add them, using them to align the images in the panorama, but if it can’t find any, it will ask you to define them, so it can continue building the panorama. Click Optimise when you’re done setting the controlpoints, to initialise the changes. Click Align in the project tab again if the panorama view doesn’t refresh. All that’s left now is to tonemap and tweak. In the Panorama Editor, you can choose your projection, and fine-tune image placement, or panorama-alignment via the menu and your mouse-buttons.  When you’re done, keeping the Panorama Editor open, go to the exposure/HDR view, check enable HDR-stitching, select the friendlier Exposure Fusion, and tweak your image’s tones, if needed. Press Optimise Now, when you’re done. In the Create Panorama View, select your output, and save.

 

Setting up your sIBL set

 

Once you’ve created your panorama, set it up as a sIBL-set for your render. Initially thought up by Christian Bloch and in continuous refinement together with other HDR-enthusiasts and developers at the hdrlabs.com site, sIBL stands for Smart IBL. And it is smart, as it takes a lot of work out of setting up IBL renders by generating backplate, reflection map, environment (illumination) map and light from a single HDR file, and then linking them together via a description file, making the components to the file sIBLings. If run through a loader script or the sIBL GUI, the sIBL set can be loaded into a (supported) renderer, as a complete, high quality, optimally lit  IBL setup, ready to render.

It’s easy to use as well:  First,  grab the free sIBL editor from the HDRlabs site at http://www.hdrlabs.com/sibl/software.html. Also grab the sIBL GUI if you’re on Maya, Max or XSI, or one of the Loader plugins if you’re on Lightwave, Modo or Cinema 4D.

If you don’t have an HDR panorama to work with, try downloading the Mountain Backgrounds free sample  from Dosch at http://www.doschdesign.com/samples.php

After installing and unzipping your files, open the editor, click x to leave the (empty) library screen, and go to File>Create New sIBL from HDR and load your own  file or MOUNT_02SX.hdr into the editor. As you can see, it automatically generates a thumbnail, environment map, and a reflection map. Click OK.  If you want to share your own sets, it pays to add some info by clicking on the generated thumbnail.

Once you’re done, click on the background image. Click Add light – it will display in the upper left of the image as a black dot. Click on the marker on the right of the light tab, and drag it so it covers the sun or the brightest spot in the image. Most times it will set the sun’s colour automatically, but if it doesn’t, set it to RGB 246, 247, 255. If needed, add more than one light at varying strengths if you’re in a scene with multiple lightsources.
When you’re done, go to File>Save sIBL as, and save your set. The save button is called Open, but it does save. If you use Maya, Max or XSI, start the sIBL GUI. If you use Modo. LW, or C4D, open the loader script to load your set, and render.

 Using 3DS Max as an example for the sIBL GUI, download the Mental Ray and Scanline templates on the template prompt. In the libraries tab, add your newly created set by clicking “Add Content”.  Select it, click the Export tab, and select your Max renderer on the left. Click Output Loader Script. You’ll get a message stating Loaderscript| ‘sIBl_3dsmax_import.ms’ output Done! As the sIBL GUI seemingly names all output scripts the same, click Open OutPut Folder, copy the file to your sIBL-folder.  In your max-scene, click  MaxScript>Run Script in Utilities, locate your script, and click render.

If you need a sky or landscape you can’t photograph, try rendering it. Most current 3D programs have the toolkit to facilitate this, but your best bet for good, fast results are landscape generators like Vue or Terragen. Vue, in particular eases panorama or skybox generation.

For a panorama, create a new scene. In Render>Render Options, set Render to Render to Screen. In Picture Size and Resolution, set Aspect Ratio: Free in the dropdown, and set Other to 720 x 360 Check Panoramic View, and Spherical Render, leaving the sliders alone. 

Now build your scene. If you’re new to Vue, try creating an infinite terrain via Object>Create>Procedural Terrain>Procedural Terrain Preset> Infinite Terrain, and selecting Flat Mountains.  Click the sphere in on the upper right to access the material editor, and select Rock and Grass. In the atmosphere menu, load an atmosphere which will generate at least a little contrast and shadows to make it more interesting, like Bisbee in the Spectral Sunlight folder. Turn off lensflares and other effects, should you pick a different atmosphere. As you’re going to render out in a fairly large size, you may want to avoid GI and AO skies for now, as you’re rendering it out in a larger size when you’re done.

If you’re using Vue Infinite or xStream, you can render, and  save your file as openEXR. One of the advantages of openEXR is that it saves all 32bit data from the render, so you can set and save out exposures from your image editor. If you’re on Vue Complete or other, save your file as HDR – which may demand some elbow grease, as the HDR format isnt’t always as friendly as EXR when it comes to saving out exposures, and you may experience grey files, or flat light. If you see your render look flat, try rendering out your probe at different exposures and merging them, using a stitching app or Photoshop. 

To render out exposures, Click on the Main Camera in the layer panel to bring up it’s properties. Set your exposure to -1.5, and render. Save your file as a TIF, with the exposure as part of the filename and render the next exposure, at -1.25. Render your way to +1.5 at 0.25 0r 0.5 intervals, and save every render. Open Photoshop, and go to Automate>Merge to HDR. Select the TIFs you just rendered out on the Source Files dialog, and click OK. The files will load and the Set EV Manually dialog will come up. Tweak your settings according to the exposure you rendered at (and easy to see, if it’s part of your filename), and click OK. The Merged dialog will come up with a preview of your merged file. Click Ok, Apply the lightprobe you made to your scene, and test if it works as intended. If it does, go back to Vue, and set your render settings to Final, or higher, set the image size to 3600 x 1800, and render out a backplate if you need it.

As with UV mapping, there are several layouts you can use when generating a lightprobe, depending on your needs or setup. The following are the most common ones:

Skybox

Also known as a cubemap or a lightcross, a skybox is a lightprobe cut or rendered to fit projection onto a six-sided cube, the viewpoint normally set to the center of the cube. These maps are commonly used in gaming engines due to the fact they  distort less and render more efficiently than spheres.

Mirrorball

Also known as a mirrored ball, or lightprobe, a mirrorball is an image of a chrome ball containing reflections of an env/pp ironment, and which is unwrapped to generate rectilinear map of that environment.

Angular map

An angular map, also called a lightprobe, is similar to the mirrorball, but has different mapping on the edges, providing you with better sampling and detailing on unwrap, but details look smaller.

Rectilinear

Also known as a panorama, spherical panorama,  latitude/longitude  or equirectangluar map, this format is most often a multirow map assembled to wrap around a sphere in 2:1 – 360×180 format.

Cylindrical

Best suited for single row panoramas, a cylindrical map has a 360 degree horizontal field of view, but a vertical practical limit of 120 degrees-

HDR and Photography Terminology for beginners

• Bracketing:
  • Also known as to bracket or exposure bracketing.  To shoot two or more images at different exposures of the same scenes.  Most DSLRs support this
• Aperture:
  • The the lens opening. It controls how much light gets into the image, and is measured in F-Stops.

• Tonemapping:
• Converting a true, 32 bit HDR image to a 16 or 8 bit image, and tweaking it in an HDR-application to either make it look realistic or… enthusiastically detailed and saturated.
• HDR file:
• A 32 bit image composed of merging two or more images taken at varying exposures. An HDR image is derived from an hdr file.
• HDR photography:
• A photography technique taking shots at several exposures, merging them, and tonemapping them into an image. Not always the same as HDR-for-CG photography, as the output demands are different.
• Single-Shot-HDR:
• An image of one exposure turned into an HDR. May not give the same optimal rendering results as “proper”, multiexposure HDR images
• Low dynamic range (LDR):
• An image containing one exposure, like jpg or tif. Several LDR images at varying exposures can be merged into an HDR image.

• F-Stop:
• The size of the lens opening. F-stop numbers work in reverse: A high F-Stop number, like i,e F/22, will mean a smaller lens opening and a sharper image. A low f-stop, like F/4, will mean a large opening, and a very short area of focus in the image before the rest turns into Depth of Field-blur.
• Exposure
• – shutter speed, which controls  how long the light is let onto the image
• EV:
• –  Exposures are measured in Exposure Values. In order to set your image to over-or under exposed, with the median being 0, you set your EV-settings to -1 to +1 – this will allow you three exposures, a base minimum for creating a HDR-image.
• ISO
• ISO is the sensor’s sensitivity to light. You up your ISO settings if you’re in a badly lit area, at the price of also introducing more noise into your image. Hence, always set it as low as possible.
• Stop:
• A phrase used to describe the interconnection between Aperture, ISO and Exposure, but it basically boils down to being another word for  adjusting Exposure:

• Parallax
• When nearby object move in relation to objects further back in your scene between shots.
• Stitching
• Aligning and arranging merged images into a panorama
• Merging
• Merging all the exposures you shot into a HDR format file

In order to make the most of your shots, you’ll need an application capable of merging, tonemapping, and stitching your shots in multiple rows, in addition to being able to output your work in several formats and layouts.

ptGUI

A favourite among photographers and CG-artists alike, ptGUI is Open Source stitcher Hugin’s very big brother. Able to stitch unlevel sets, and can output gigapixel images. Provides automated and detailed manual workflow, and is one of the best tools around

• o http://www.ptgui.com/-
• o – Personal license USD 120
• o – Professional USD 226

EasyPano PanoWeaver

A little more automated, and web-oriented than ptGUI, PanoWeaver offers some handy features, like tripod removal from your shots and immersive panorama generationgood for showcasing panoramas online in addition to using them for CG

http://www.easypano.com/panorama-software.html-

• Standard Edition    Price: $99.99/€99.99
• Professional Edition Price: $499.95/€499.95
• Batch Edition   Price: $999.90/€899.90

Autodesk Stitcher

Despite not having been updated since 2009, Stitcher is still a staple when it comes to stitching, and with good reason, as it can stitch almost anything, and fast. The only standalone stitcher providing Photoshop support.

http://usa.autodesk.com/adsk/servlet/pc/index?id=11390049&siteID=123112

• 400 USD

Kolor AutoPano

Offers anti-haze filter and can utilise Adobe’s freebie lens calibration tool to get rid of i.e vignettes. Based on SIFT, one of the most accurate analysers of similarities between images around. Very easy to use, and can stitch images together from multiple points of view.

http://www.kolor.com/

• 99 E standard version – Autopano
• o 199 E pr version – Autopano giga

Boxout 4 – HDR providers (283)

Dosch

With over 35 sets in their assortment, and one of the oldest HDR providers around, Dosch can take your renders from the desert to beach, summer to winter, night to… you get the picture.

http://www.doschdesign.com/products/hdri

Price: 119.00 USD / 99.00 EUR per 24 image set, set comes in skyboxes, mirrorballs, and panoramas.

3D Ocean

A relatively new 3D brokerage providing some excellent HDR sets by, among others, vray-wiz Peter Guthrie. Provides Smaller and hence cheaper sets than Dosch, currently has over 140 sets in the store.

http://www.3docean.net

• Price: From 15 to 25 USD per set, 45 to 150 per megapack. Sets normally come in equirectangulars.

HDR-Source

Having shot HDR over 10 years, HDR-Source by Charles Leo has seen HDR grow from an obscure format to the CG staple it is today. Provides both interior and exteriors, and currently has over 119 sets and 6 bundles in the store.

http://www.hdrsource.com/store/

• Price: From 14.95 USD per set, 149 USD per megapack. Sets normally come in equirectangulars, and are sIBL-ified

Aversis

Another experienced HDR provider, located in Belgium Avesis offers tutorials and models as well. Provides both interior and exteriors, and currently has over 50 sets and 9 bundles in the store.

http://www.aversis.be/hdri/index-individual.htm

• Price: From 10 E per set, 25-35 E per megapack. Sets normally come in equirectangulars, and are sIBL-ified

 

Dutch Skies 360°

You may already know Bob Groothuis’ work, as he’s donated several of of his Dutch Skies 360 0 to 3DW readers. Provides both interior and exteriors, and currently has over 50 sets in his store.

http://www.bobgroothuis.com/blog/category/dutch_skies_360_online_shop/

Price: From 9.99 USD per set, 109 – 335 USD per megapack. Sets normally come in equirectangulars

 

This post is under construction – better images and video are being added:

These days, vegetation and landscape generators generally have full texturing capabilities, population functionality, and the option to randomly generate and texture vegetation. This makes it a lot easier to generate and populate your digital landscapes within a single application, but it also makes it easy to just let the application do the work, particularly for vegetation texturing.
While it certainly helps to invest in an extensive toolkit to do all sorts of little jobs for you, the fact remains that most of the time you don’t really need most of it. Save yourself some money instead and retain more control over a texturing outcome, by honing your techniques in an image editor. Learning how to produce similar results to basic or advanced plug-in functions can save money and enable you to maintain more control in the creative process. It also gives you a better base of understanding as to what these various applications actually do – which in turn will help you create better vegetation:
Some of the most common mistakes I’ve seen made by novices and experts alike, stem from not planning and preparing before starting to texture, resulting in problems such as incorrect Bump and Normal maps, Halos, colour bleed often caused by a poor extraction, and unauthentic vegetation in a scene. These topics and more, such as plug-in free masking and moss-generation, will be tackled with some tried and tested texturing tips aimed at the budding digital gardener.

Tip 1: Affordable texture sourcing

If you’re just in need of some simple background or fill textures for painting over, or  just want to record something to see how it moves, consider using your cell-phone camera. To get yourself the best possible image, set the camera to the largest resolution it can handle, and avoid using the .jpg format, if possible. A common mistake when using your phone’s camera is to use the zoom extensively, which is a bad idea for texturing shots: A cellphone zoom will greatly reduce the quality of an already low-resolution image by pixellation, making it difficult to work with. Hold the phone rock-steady, and step in as close as you can, or your camera can handle without distortion. Avoid using a flash to reduce shine on the texture. After uploading, adjust the white-balance in your image editor, and if needed, sharpen it with an Unsharp Mask at a low Radius and Threshold.

Tip 2: Extracting complex vegetation

When extracting complex vegetation, such as a leafy tree from a background, you can usually generate a good alpha through copying the blue channel and darkening it until the image shows the Transparency map you want. However, a frequent problem is colour spill which becomes more visible the darker your background gets. It usually consists of remnants of colour from the image surrounding your extracted vegetation. To remedy this, select the extracted plant, duplicate it, select the colours spilling out, and on the duplicated layer generate a border mask from the selection, and lock pixel transparency if desired. Then clone from non-bleeding parts of the plant. This fixes the spill and can also make a tree’s foliage look denser. Duplicate the layer with the fixed colour spill below your current one, and blur it. Then update the alpha to match the tightened extraction to avoid halos

Tip 3: Clean your maps

When you scan a leaf, a common problem is over lighting on parts, manifesting as white or bright areas in the texture. Desaturating a scanned texture like this into a Bump map will give you unwanted and inaccurate light in the Bump map showing up as white or light grey areas. This will falsely render as a raised area in most renderers when the Bump map is applied to geometry. Using such a map as a basis for a Normal map in a Normal map generator like CrazyBump, Shadermap or Nvidia’s Normal Map plugin will make it worse, as apps like these put out what they’re programmed to process, which isn’t necessarily the same as you see. Avoid this situation by cleaning up and flattening incorrect overlighting before making your Bump map (and then your Normal Map) by using a High Pass filter and mixing it with some manual processing, like a desaturated overlay to maintain detail. Painting in details by hand will produce even better organic results such as stalks and veins which will be raised and cracks or even cells which will be lowered by the maps.

 

Tip 4: Detailing vegetation surfaces

Adding features and details to the surface of your plants will add richness and realism to your renders. When detailing bark, for instance,  remember there is a difference between moss and lichen. A plant, moss likes it shady and damp.  Lichen, a fungus, grows on anything and comes in many shapes and colours, ranging from grey, to ochre, to teal.  Unlike moss, lichen also thrives in extreme environments, which is why you’ll often find rocks and stones covered in in as far north as Bear Island.You can add either to your bark by rendering a fractalbased texture in a landscape generator like i.e. Terragen: Rendered bump and displacement maps in a texture can sometimes give you better results to work with than i.e phototextures alone, as it’s much easier to control and vary the details. Pointing your camera straight down from about an average male’s height, on a flat, white terrain (for easier blending), with flat light, add fractal layers in various ranges of coverage, using hues of green, grey, ochre or teal, adding grades of noise and displacement. When you’re done with the buildup, render it out, and add to your bark through an overlay or a blend layer  I did a quick walkthrough for how to fake moss or lichen on bark, using Terragen’s freebie classic version (even though 2.0 will work perfectly too, as will Vue) and Photoshop – After you’ve rendered out your moss; you can add it to your plant or tree surface texture using an Overlay or a Blend Layer.

Tip 5: Asset library

Save yourself a lot of time by creating a texture library of vegetation and foliage. Use it to create basic, recyclable vegetation textures such as barks, grass, leaves and petals in various shapes and colours with Transparency Maps. As you build up the collection it will become faster and easier to mix and match textures for greater variety and then plug them into alphas for implementation. Especially if you build up layers with subtle differences of tone and detailing, allowing you to save out only what you need as textures for your shaders by turning layers on or off. Overlays from scans or photos on basic textures can add even more fine detail, with the added bonus of softening potential colour and shine issues and decreasing the need for retouching. Avoid seams or tiling issues in the overlays by sending them through an offset or tiling filter like the offset filter in Photoshop or TextureMaker when you add them.

Tip 6: Age before beauty

And make sure you age gracefully; In addition to increasing scaling on a mature tree bark-texture, adding moss, lichen and some trunk conks, age it further, or make it  appear diseased by muting the colours in the texture, and adding in some loose bark, deformed growths, and a few larvae and beetles.

Tip 7: I can see your Halo

One of the easiest things to forget is keeping an eye on your Transparency Map. If you get halos in your render, which are those white edges outlining your texture, remove them by adding a similar colour layer under your texture, and crisping up your alpha by using Levels and Curves.  If you’re using a plant generator to render billboards, set the background colour in your rendering program to the average colour in the vegetation. This makes tweaking a lot easier, with the added bonus of painless tweaking of edge density in the alpha. This works especially well for coniferous vegetation

Tip 8: Rendering digital vegetation

Also – if you’re  rendering  billboards, in addition to setting the background colour to the average colour in the vegetation, remember to keep your light flat. You’re rendering billboards, not art (even though the billboard itself may be) so keep your light flat – use a skylight rather than a sun or set of spotlights if you want to be able to recycle it. It’s annoying  having to tweak light that’s not present in the scene out of the vegetation.

Tip 9: Blend options

Finesse your textures with Layer Blend Options. In Photoshop a Blend IF is perfect for blending barks, cracks, and moss, in addition to image tuning and will also give you great results for sharpening. Use black and white to set Transparency and Opacity on the layers, and use grey to fine tune your blends.

Tip 10: Texturing frost

Don’t set everything to white when you’re texturing frost, as frost tends to form on the surface first, and then into cracks. When texturing leaves, overlay a frost layer covering the entire leaf, and then add a further overlay of stem and veins for the most realistic result.

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file or the tutorial. Using the content to add textures to your own (commercial) images is perfectly fine, though
Download petal-files for the GIMP here!

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

  Originally published in 3DW 121

Creating a flower texture in the GIMP is easy; mix a little noise, a few gradients and some smudging, and you’re good to go Here’s how:

Open GIMPQNA_Petal_Start.xcf file from the disc or download.  Using  guides makes it easier to generate a path. Create a new layer: Layer>New and set the layer fill type to white. Add a transparent layer on top of it. Click the Path icon in the toolbox, and generate three anchor points by clicking down the middle vertical guide where the horizontal lines intersect.  Click Edit in the path tool, ensure polygonal  is unchecked,  and drag the middle anchor point to the leftmost guide. Click to activate the handles, and create a gentle curve. Select the bottom anchor point,and create a small curve, and move it to the middle vertical guide. Click Stroke Path, and set line width to 3. Duplicate the layer: SHIFT+ CTRL + D and flip it by going to Layers>Transform>Flip Horizontal. Right click the duplicated layer, and select Merge Down. You may have to paint in some areas to ensure the petal is fully closed for filling in.

Select the outside of the petal, and invert the selection: CTRL + I. Add a new, transparent layer, and name it BaseColour. Set foreground colour to R234, G111, B10, , the background colour to, R179, G29,  B13. Click the gradient icon in the toolbox, and set Shape to Shaped Spherical, check the dithering and adaptive supersampling checkboxes, and drag both its sliders all the way to the right.Fill the selection. Add a new transparent layer and call it Noise. Fill the petal with white, and go to Filters>Noise>Spread. Set the spread to 470. Blur it: Filters>Blur>Gaussian Blur at 10. Set the layer to overlay. Repeat the previous step, and set the  Gaussian blur to 5. Create a new layer, and call it Roughed_Edges.

Fill the selection with Render>Clouds>Plasma with Random seed at  1036897181 and turbulendce at 7. Go to Colours>Desaturate, then  to Filters>Blur>Motion Blur at zoom, x256, y 508, and length 23. Set the layer to overlay, and it’s opacity to 75. Set the foreground colour to R132,G53, B6, and the background to a R121, G26, B12. Select the paintbrush, set it to sparkle brush and check “use Colour from Gradient”. Varying the scale, draw some lines upwards to generate shape and contrast, and set the layer to overlay.

Tip: You can use the same technique for leaves using similar greens, but use the Shaped Angular gradient, and fade the RoughedEdges layer to about 25%.

If you’re into digital vegetation, you may also like my “Working with Digital Vegetation”-article.

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file or the tutorial. Using the content to add textures to your own (commercial) images is perfectly fine, though

Download petal-files for the GIMP here!

 

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

Originally published in 3D World issue 113

Watch a six minute, no sound walkthrough on Vimeo!

Step 1:

If you want to add falling snow to an image or render, remember that realistic falling snow is soft-edged, and that the bigger the flakes, the softer and more indistinct their edges will be. Hard edges tend to denote rain, hail, sleet or dandruff. Go to File >New and create a new image at 512×512 pixels. Click Advanced Options, and select Foreground (black by default) in the Fill With dropdown.

Step 2:

Create a brush (or a layer if you don’t want to save as brush) for the snow closest to a viewer’s eye first: Go to Layer>New. Set it to screen. Fill it with black. Go to Filters>Render>Clouds>Solid Noise, and set it as follows: Detail 10, Randomize Seed checked, Tileable checked, size set to x16 and y 16, and details to 10. Go to Colours>Curves, and drag the curve to around x180 and y to 95

Step 3:

Add a little more detailing by adding some noise. Go to Filters>Noise> RGB Noise. Uncheck Correlated Noise and Independent RGB. Grab the R slider, and drag it to about 0.70 – the Green and Blue sliders will follow. Click OK. Go to Colours>Brightess and Contrast, and slide the contrast slider all the way to the right. Go to Filters>Blur>Gaussian Blur and set it to between 7 and 15 horizontal and vertical.

Step 4:

Go to Colours> Levels and set the left slider to 50, the middle to 1, and the right one to 153.  Click OK.  You may have to tweak Contrast or Curves a little to make the flakes brighter. Go to Filters>Blur> Motion Blur, and set it to Linear, with a 10-20 length, and an angle of 45 to denote motion. Tweak with Contrast or Curves again as needed

Step 5:

If you want to save this as a brush, set the layer to screen, and delete the layer under. Go to File>Save As…, and click Select File Type. Select GIMP brush, and save the brush in GIMP 2.0sharegimp2.0brushes as snow_front. Click on the paintbrush icon in the Toolbox, and the Refresh brushes icon in the lower right to refresh the brushes. Change mode as needed when using the brush.

Step 6:

For background falling snow, create a new document as in Step 1. Create a new layer, and fill it with black. Go to Filters>Noise> RGB Noise. Using the settings from step 3, drag the sliders to 1, and then use Filters>Noise>Gaussian Blur and set both horizontal and vertical to 5. Now go to Colours>Brightness–Contrast, and set the contrast between 30 and 50, depending on how dense you want the flakes. Use Filters>Motion Blur to soften them a little, at length 5 and angle 45. Tweak the contrast as needed, and make them “pop” a little by using Colours>Curves upwards, at 167 x 160. To simulate proper snowy weather, create several brushes or layers with varying density. Remember to make your sky dark, low, and soft if you are aiming for a heavy snowfall.

Quick Tip: For stylized snowflakes, go to Filters>Render>Nature>IFS Fractal – try creating a few new triangles and arranging them in a circle, with one in the middle

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file. Using the content to add snow to your own (commercial) images is perfectly fine, though

Download asphalt for the GIMP here!

 

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

Go to File>New. Set size to 3000×3000, and Foreground (black) under advanced options. Open 3DW_QnA_Planet_Undertexture.jpg and 3DW_QnA_Planet_Overtexture.jpg. Select all and copy each, and paste them into the file you just created via Edit>Paste As>New Layer. Set the Over_texture to Grain Extract, and merge it down via Layer>Merge Down.

Click the merged layer in the layer-dock, and go to Filters>Map>Map Object. Select Sphere in the dropdown, check Transparent Background, and, under the Light-tab, select No Light in Lightsource type. Click OK. For the shadow, create a new layer, and set it to white. Go to Filters>Re-show Map to Object. In the Light tab, Set Lightsource type to Point light. and lightsource colour to black. Set the X and Y Directional vectors to -1,2 and the Z to 2. Under Material, set all values but both Diffuses to 0. Set Diffuse to 1 under Intensity levels, and to 0,3 under Reflectivity. Click OK Set the layer to Hard Light. This will create a realistic planetary shadow

 

Add a planetary shadow hue by rightclicking the Planet-layer in the layer dock, and clicking Alpha to Selection.

Create a new layer, set it to soft light, and fill it with R45, G140, B200. Rightclick the layer and click Add Layer mask. Leave it at White, and click OK. Click the blend tool in the toolbox, and reset the fore and background colours to black and white, and set the gradient to FG to BG. Drag blend direction from the upper middle of the planet shadow diagonally to the right, just outside the selection With the selection still active, create a new, transparent layer under the planet shadow. Fill it with the same colour as the shadow hue, R45, G140, B200.

Create a new layer. With the selection still active, Select>Shrink and set it between 40-80. Fill the selection with black. Unselect, and run Filter>Gaussian Blur at 300. Merge the layer down, and set it to screen. Set opacity somewhere between 40 and 60. Generate a new selection again, and create a new layer. Set colour to R45, G140, B200 again. Click the paintbrush in the toolbox. Select Circle Fuzzy 11 and set scale to 2,3 Uncheck Pressure/Hardness under Brush Dynamics. In the select menu, Select >Select to Path. Then go to Edit>Stroke Selection. Check Stroke with a Paint Tool and emulate brush dynamics. Click Ok. Sset the layer to overlay, Give it a a gaussian blur between 40 and 80. Set opacity to 80, and you have a basic planet ready for detailing

 

Tip: For a more visible outer glow, blur, tweak and set the texture to screen and then increase the size of your planet shadow a little to make the glow line up properly.

Most of you who know me, know I love space art, and can spend hours tinkering with anything from an asteroid to a nebula. I also love looking at space art. Unfortunately, cruising the various scifi galleries can sometimes be a little frustrating, as things sometimes look plain wrong. This in itself isn’t much of an issue – we’re all at different levels of mastering graphics, and I’ve published some abominations of my own over the years. However, the one thing which tends to get my personal goat up, is the endless stream of images tagged “surreal space scene” – which nine out of ten times are a cop-out for ignorant or poor visual execution.

Hence, in the name of putting my money where my mouth is, I thought I’d share my own planet tutorial, whilst also sharing some additional tips on space scene creation; The drawback of writing QnAs for i.e. 3D World is the fact you usually have a wordcount – 500 words or less to do them in. Which doesn’t often leave room for further tips or musings

Space scene creation tips for The GIMP or Photoshop:

Watch your bumps!

• When you have a whole planet in view, it’s very likely you won’t see fine detail or the bumps of mountain ranges. In order for a bump-detailed mountain range to be seen from far away in outer orbit, it’d need to be tens of thousands of kilometers tall.
• If you’re still going for bump – your planet needs to be humongous, so keep an eye on your sizes and perspective

NEVER trust a plugin!

• Nebulas are fun. They’re even more fun when they look like nebulas and not as if they came straight out of a Flaming Pear or the Universe Image Creator thingie. Trusting software instead of your eyes is a common mistake, and you should trust your eyes over a plugin. Just because it’s software, it doesn’t mean it gets it a look or a concept 100% right. Also – If you’re going to use fractals, do spend some time on making it look like a nebula and not straight out of Apophysis – as a rule of thumb, it’s a good thing when people can’t spot which application you’ve created your spacescape in. Check out the reference list below for some good overviews over how nebulae should look

Step away from that LunarCell – NOW!

• You know you’re yelling at deaf ears when even BioWare use LunarCell for their planetary grid (at least, the planets look like they’re straight out of LunarCell). If you truly want your planets to look unique, put the effort in – The tutorial on these pages is just a tiny fraction of what you can achieve with some good, base maps and some elbow-grease.

Work your textures

• When it comes to what kinds of maps to use for planet generation, or how to generate one, you have several options: You can create one from a photograph of rock and/or stone – but you’ll need to be careful with the levels of detail, as it’s all about size – again, if you show a whole planet in your image, you are very far out – and your texture needs to reflect that. This means that i.e cracks will need to be either resized, or blurred, in order for it to look realistic. A better way to do this, is to generate your own planetary texture via using multiple sets of fractal noise in greyscale, and then colourising them via a colour overlay. Or, better yet – render out a texture in Vue or Terragen!

Mind your starfields

• As a general rule, Greg Martin’s tutorial for generating stars still stands. However, keep in mind that variety makes or breaks a starfield. One thing to avoid in particular, is the “break up starfield by render clouds”-timesaving-trick – most of the time it will look crap. The Creative Mac people have an old, but still pretty damn good set of Photoshop tutorials with brushes by Dave Nagel – the brushes are worth checking out, and very worth using when creating your starfields.

And last but not least…

• I know you’ve probably heard it before, but keep practising. I’ve been on sickleave a while, and haven’t drawn any scifi for a while, and damn, I really needed to work at remembering stuff that used to be second nature. So whatever you do – keep at it

Reference links

- currently undergoing reconstruction

Download Conditions and terms of use

By downloading the file you agree to not sell, re-distribute or republish any of the content contained in the file or the tutorial. Using the content to add textures to your own (commercial) images is perfectly fine, though

Grab the Planet Tutorial for the GIMP files here (~85 Mb)

No exceptions without written permission. Files are copyright Cirstyn Bech-Yagher. Feel free to contact me regarding this:-)

Related Posts

Part 1: Creating the Wall

Acropora is one of those apps which let you stumble upon handy, new ways of doing things. For instance I found out by accident when reviewing Acropora for 3D World, that  it takes about five seconds  to generate ruined walls and other architectural bits in this application, and I found it pretty handy for game-props and quick background stuff. Here’s a quick little tutorial:

1. Open Acropora or create a new document; File>New.
2. In Voxel Tools, click Create Region, and draw a rectangle on the Primary surface, and then click Cone under Geometric modifiers. you’ll get the fastest results using the cone-modifier, but cube also works, just remember to dial down the strength setting considerably.
3. Once you’ve added the cone modifier, go to Noise modifiers, and click Volumetric noise. As you can see, instant ruined wall.

1. Now for the tweaking, if needed – the settings should be pretty default; Click Volumetric Noise in the tree under Primary Surface, and set the properties like this:

1. I like layering things so I can have fine-tune control over whatever it is I am doing, so I added another set of volumetric noise, and set it to:

   

• For variations, go to the Volumetric noise setup quad in the lower left of the screen, and start dragging the x-slider (the top one).
• Also try adding a Crust from Miscellaneous or a Hilly one from Terrain modifiers. You may want to bump up the voxel-space by clicking the Increase Voxel Densitybutton in the top bar, but too dense a mesh makes the wall look way, way too fractalised, which is an effect you want to avoid, as it isn’t pretty. Personally, I wouldn’t go higher than 256x256x256, tops.

• Export when you feel you’re done.

Part 2: Exporting and UV Mapping

 

I tried to extract the mesh alone via using the Extract region function under Voxels tool, but I didn’t like the way it worked, so I tidied it up by exporting as object,  ( File>Export Primary Surface) and opening it in Max by just selecting all the polys I didn’t want on the plane and deleting them.

After that I took the mesh into UVLayout, for mapping. One problem I had was unwelded and loose tris – the blue lines in the image below

Uvlayout lets you fix that by  checking the weld and clean boxes under Load Options:

I got the best mapping  results in UVLayout by just cutting along the top and sides, dividing the wall into top, front, back, leftside and rightside:

I never trust just chequered maps, so I plugged my own, below  into About UVLayout>Preferences>Map. You get it onscreen by pressing  and cycling T when you’re in the 3D view.

When you’re done with the UVmap, find two stone textures and i.e a rock or earth texture, and mix them up ’til you get a result resembling this:

 - And render out:

Insert your text here

The product was reviewed on a purchased copy. There was a lot of (chatty) contact with the reviewer and developer