There are two ways to open the Render Gallery. You can either press Alt + ⇧ Shift + W to open a new floating window. When you click the New Tab button, you open a dropdown menu. From there, choose Solaris ▸ Render Gallery. What you can see now is a split window.

The upper part is the image viewer and provides most of the functionality of Houdini’s MPlay. The rendered images and snapshots are shown in the image strip, the gallery’s lower part. When you double-click an image in the image strip, it’ll be displayed in the image viewer.

In the viewer, you have several options like showing an image’s histogram, zooming in and out, or showing important render statistics if provided by the render delegate.

The image strip also has several buttons and the most important ones are Snapshot View and Background Render.

You can also open the Render Gallery from the Toolbox on the left of the viewport. There, click the Show/Hide snapshot strip button. The image strip will be attached to the viewport’s bottom.

For snapshots you don’t need a Karma mode and you can also create images from Vulkan or Storm. You can take the snapshot directly from the Render Gallery. You’ll get a 1:1 representation of the viewport, but you also save the current state of the LOP network. To revert to a certain network, right-click a snapshot in the Render Gallery and choose Revert Network to this Snapshot.

Snapshots consider the viewport’s aspect ratio and display mode.

Background rendering requires an active Karma mode and you can launch it from the Render Gallery. Houdini remains responsive while rendering, but it doesn’t react to viewport changes. As an in-process mode, background rendering will take Houdini down when it crashes. Note that this mode stores the LOP’s current state.

You can start clone rendering from the Clone Control Panel, but you’ll see the result in the Render Gallery. Clone rendering creates a 1:1 copy of the scene you want to render and responds to viewport changes. This approach requires more system resources, and if a scene takes ten minutes to prepare in Houdini, the clone process will also take ten minutes. As a separate process Houdini will not be affected when the clone rendering crashes.

Live rendering fills the gap between snapshots, background rendering and clones. The mode directly reacts to viewport changes. Live rendering requires fewer resources than clone rendering, because it uses less memory and the scene does not have to be processed again. When Houdini crashes, the live render process will die as well, and vice versa. As a true render solution, live rendering requires an active Karma render engine.

Unlike its counterparts, live rendering is tied to a separate node - the Live Render LOP. The node is also where you start the process and the result will be displayed in the Render Gallery. The node and its position in the network determine what is being rendered. This means that you can add multiple Live Render LOPs and take snapshots from different parts of a scene.

The Live Render LOP is the command center of the live rendering mode. The node utilizes render settings and camera data from upstream nodes. Please note that a Render Settings LOP and a Camera LOP are mandatory, but you can also define local overrides directly on or even inside the node. You also want to render an object, for example from a SOP Create LOP. The image below shows a minimum setup for the Live Render LOP.

Once you've created the network, let the Live Render LOP know where to find the scene’s Render Settings. By default, the path is /Render/rendersettings. You can also drag the rendersettings primitive from the Scene Graph Tree to the node’s parameter.

Then, choose a Render Delegate. Note that you don’t have to turn on the delegate in the viewport to get a live render. Houdini will perform the live rendering anyway and update the image when the viewport changes. However, we do recommend to choose the delegate you want to use for the final render.

Click the Start Render button. You can now see a green badge directly at the Live Render LOP node and the rendered thumbnail in the gallery. The icon makes the image distinguishable from snapshots and background renders. You can monitor how the thumbnail in the image strip reacts to your viewport changes.

With Stop Render, the rendering process will be interrupted and the gallery thumbnail will no longer react to changes in the viewport. To finally show the image in the gallery, double-click its thumbnail. There’s also a right-click menu available for each snapshot. Please visit the Render Gallery’s help for more information. When you click Start Render again, you’ll create a new gallery entry.

You can rename the thumbnails in the image strip. Just click the default name, for example liverender1 to make it editable. Then, enter a new name.

Overrides let you add changes to the live render without modifying the original network and its parameters. This is a convenient way if you want to try different camera perspectives or change image resolution.

The Override Camera parameter accepts a path to another camera. Again, you can drag any camera primitive from the Scene Graph Tree to the input field. The path will then be added automatically. If you need to change resolution independently from any upstream settings, open the Override Resolution dropdown. You can either use a percentage value from the original resolution or enter specific values.

A network is not limited to a single Live Render LOP, but in theory you can have as many nodes as you want. And, they can also render simultaneously. This, however, is a matter of resources. A more realistic scenario would be to have Live Render LOPs at different points of a network to subdivide a scene into parts. The part you're currently working on has an active live render session.

So you can separate objects from the background, render the empty background itself, or work on the scene’s lighting part. Later you can bring everything together in Copernicus. You can also create multiple Live Render LOPs and define different overrides.

You can start and stop a live render session directly from the File COP. From the Source dropdown menu, choose Live Render and point the Live Render LOP parameter to the Live Render LOP node. The most convenient method is to click the Open floating operator chooser button and choose the LOP from the panel. Note that a Live Render LOP is mandatory and it’s not possible to choose one of the upstream or downstream nodes.

Click Start to “remote control” the render process. Then you can return to Solaris and make your changes on the scene/viewport. When you go back to Copernicus, the File COP’s content will be updated automatically. If the render takes long, you can even monitor the render process directly in Copernicus. To do that, open a new floating panel with Alt + ⇧ Shift + W and click the button to add a new tab. From the dropdown menu, choose New Pane Tab Type ▸ Viewers ▸ Composite View. The viewer’s functionality is very similar to Houdini’s MPlay.

As in Solaris, the Stop button finalizes the render. To get a new image, click Start again.

It’s also possible to process the image from the Live Render LOP with other Copernicus nodes even when the render task is still running. The Viewing your output chapter explains how to get access to the final result.

There are two methods to write your gallery images to disk. One way is to save the image from the gallery’s image view that supports most of the features of MPlay.

1. Double-click the live render thumbnail in the image strip to open the image viewer. Or you open the Render Gallery tab above the viewport. When you click the active thumbnail, the live render will be displayed in the gallery viewer.

2. Right-click somewhere in the image viewer and choose Save Frame to open a Save As dialogue. There, fill in Filename, choose a Format and make the settings you consider important.

3. Click Save and write the image to disk.

You can also right-click the thumbnail in the image strip and choose Snapshot Render. This will “freeze” the live render’s current progress and saves it as an EXR file on disk. Now, right-click the new snapshot thumbnail. Load in File Browser, opens your operating system’s file browser and leads you directly to the saved EXR.