OCIO and ACES in Houdini 22: Keep Karma Colours the Same on Every Machine
Karma converts textures into Houdini’s render working space using OCIO rules. If the render machine uses a different config or working space, the image changes:
- Config: Houdini 22 ships an ACES OCIO config (3.0) as a Houdini package
- Working space: OCIO Editor: default Linear Rec.709; switch to ACEScg if your pipeline uses it
- Textures: Source Color Space: sRGB, Raw or Auto (file rules)
- Data maps: displacement, bump and roughness load as Raw
- Farm: same config, same working space, same file rules on every machine
A render that looks right on your monitor and slightly off from the farm is almost always a colour configuration problem, not a renderer bug. Karma converts every texture into a working colour space when it loads it, and that conversion depends on the OCIO config, the render working space and the file rules on the machine that renders.
This guide explains how Houdini 22 handles colour, how textures are converted, which settings must match between your workstation and a render machine, and how to check them before a long render.
Checked against the SideFX Houdini 22 documentation (Color management in Houdini, Houdini 22 platform notes) and the OpenColorIO documentation, November 2026.
How Houdini 22 handles colour
Houdini ships a pre-built OCIO library and, according to SideFX, a slightly enhanced version of the ACES config file, delivered as a Houdini package; Houdini 22 ships OCIO config 3.0 for ACES and upgraded the OpenColorIO library to 2.5.0. The OCIO Editor in the Edit menu lets you view and edit Houdini’s config.
The key setting is the Render Working Space, the scene_linear role: a colour space with no gamma that defines the primaries everything is rendered in. In the OCIO Editor it defaults to Linear Rec.709 (sRGB); ACES pipelines switch it to ACEScg. Changing the view transform in the editor currently requires restarting Houdini.
The display transform does not convert any data: it only adds display gamma for viewing. The working space and the texture conversion rules are what change the rendered pixels.
Sources: Color management in Houdini, Houdini 22 platform notes.
How textures are converted
The USD UV Texture VOP has a Source Color Space parameter with three choices:
- sRGB explicitly converts the incoming data from sRGB to scene linear;
- Raw loads the data unchanged, for displacement, bump or roughness maps;
- Auto uses the OCIO file rules to decide the source colour space.
With Auto, files are converted according to the file_rules in the config. Houdini sets JPG, PNG and TIFF files to sRGB, for example. You can add Input File Rules in the OCIO Editor with a name, rule type and pattern; Houdini searches imported file names for those patterns, and rules higher in the list take priority.
For AOVs, Karma uses the data format defined on each RenderVar prim to decide whether it is colour or raw data; colour AOVs with three or more channels are written in the rule’s colour space. Utility AOVs such as normals and position stay untouched.
Sources: Color management in Houdini.
Why a render machine can produce different colours
Every setting above lives on the machine that renders. If your workstation uses a custom ACES config with ACEScg as the working space and a file rule for _acescg textures, but the render machine runs Houdini with its defaults, the same scene is converted differently and the render shifts.
Mixed-renderer pipelines need extra care: Redshift ships its own OCIO config (Maxon added one with ACES 2.0 support in Redshift 2026.8) and updated its OCIO library to 2.5.2 in 2026.9. If Karma and Redshift renders must match, point both at the same config.
Sources: Color management in Houdini, Redshift release notes.
Keeping colour identical on every machine
- Ship the config with the project. Keep the .ocio file (and any LUTs) in the project folder, not only on your workstation.
- Point every machine at it. The OpenColorIO library reads the config path from the
OCIOenvironment variable; in Houdini, confirm in the OCIO Editor which config is active and which working space it uses. - Validate it. OpenColorIO’s
ociochecktool loads a config and reports problems such as undefined colour spaces. - Set data maps to Raw so they never depend on file rules.
- Render one frame on the render machine and compare it with your approved frame before the range.
The same check catches displacement maps loaded as colour, a common reason for terrain or skin that looks too flat or too strong on another machine.
Sources: OpenColorIO: Using OCIO, Color management in Houdini.
SaaS or IaaS render farm for colour-managed Houdini work?
Which farm type fits a colour-critical pipeline
Both can render ACES; they differ in whether your exact config runs on the render machine.
| Situation | Usually a better fit | Why |
|---|---|---|
| Default Houdini colour settings | SaaS (managed farm) | Nothing custom to carry over |
| Custom studio OCIO config and file rules | IaaS (remote workstation) | You install the config and check the OCIO Editor yourself |
| Karma and Redshift renders that must match | IaaS | Point both renderers at one config on the same machine |
| Colour-critical delivery (HDR, client LUTs) | IaaS | Compare test frames in your own viewer on the render machine |
| Short one-off render, standard look | SaaS | No setup time |
Rule of thumb: if you use Houdini’s defaults, any farm will match. With a custom config, file rules or mixed renderers, a machine you control runs exactly your colour setup.
Colour-managed rendering on an iRender server
On iRender you rent a whole Windows workstation, install your own Houdini build and sign in with your own SideFX license, so the scene opens exactly as on your desk:
- Upload the project together with its .ocio config and LUTs through iRender Drive.
- Start a 4x RTX 5090 server (32 GB per card), which we recommend for Karma XPU; 8x servers also render Karma XPU, but it does not get proportionally faster.
- Install your Houdini build. The Karma XPU kernels compile on the first render and the cache stays on your saved server image. Set the OCIO variable or package so Houdini loads your config, and check the working space in the OCIO Editor.
- Render one frame and compare it with your approved frame.
- Render the range with husk or the USD Render ROP.
Colour checklist
Use this as a final check before you start the full range on any machine.
Frequently asked questions
1. Which OCIO config does Houdini 22 use?
Houdini 22 ships an enhanced ACES config (OCIO config 3.0 for ACES) as a Houdini package, with the OpenColorIO 2.5.0 library.
2. What is the default render working space in Houdini?
The OCIO Editor’s Render Working Space defaults to Linear Rec.709 (sRGB). ACES pipelines change it to ACEScg.
3. How does Karma decide a texture’s colour space?
From the texture node’s Source Color Space: sRGB converts, Raw keeps the data, Auto follows the OCIO file rules (JPG, PNG and TIFF default to sRGB).
4. Why do my farm renders look different from my local renders?
Usually the render machine uses a different OCIO config, working space or file rules. Ship the config with the project and check the OCIO Editor there.
5. Which colour space should displacement maps use?
Raw, so the values are loaded unchanged.
6. How do I check an OCIO config is valid?
Run OpenColorIO’s ociocheck on it; it reports problems such as references to undefined colour spaces.
7. Does changing the view transform change my renders?
No. The display transform only adds display gamma for viewing; the working space and texture conversion decide the pixels.
8. Which OpenColorIO version does Houdini 22 use?
Houdini 22 upgraded OpenColorIO to 2.5.0 and ships OCIO config 3.0 for ACES.
Render your Houdini Karma XPU scenes on iRender
iRender is an IaaS GPU render farm: you rent a bare-metal server with 4x NVIDIA RTX 5090 (32 GB) GPUs, an AMD Ryzen™ Threadripper™ PRO CPU, 256 GB ECC RAM and a 2 TB NVMe SSD, connect by remote desktop and work on it like your own workstation, with full administrator access. You install Houdini and use your own SideFX license; your server image is saved between sessions, so Houdini, your packages and the Karma XPU kernel cache stay installed. For Karma XPU we recommend the 4-GPU servers.
Multi-GPU servers are billed per minute (1-GPU servers have a one-hour minimum), and sessions of 3 hours or more get 10% off. Our 4-GPU servers are liquid-cooled; in our own measurements the GPUs hold 60–70°C under sustained render load. Elite support is available 7 days a week by hotline, live chat, Skype, WhatsApp and email.
This video walks through rendering a Houdini Karma XPU scene on an iRender server in five steps:
New users get a 100% bonus on the first top-up (minimum $50). See the Karma XPU render farm page for server options, or download the iRender GPU app.
Sources: Color management in Houdini, Houdini 22 platform notes, OpenColorIO: Using OCIO, Redshift release notes, husk. Checked November 2026.
If Karma XPU renders are slow on your machine, iRender’s Karma XPU render farm gives you a dedicated remote server where you install Houdini, bring your own license and render on multiple RTX 5090 GPUs (for Karma XPU, 4 GPUs is usually the most efficient setup).


