The first four parts of this series covered the decision to switch, the hidden costs, the migration process, and the rendering problems that show up six months in. If you have been following along, you are probably past the point where general workflow advice is useful. What you need now is specific numbers, specific parameters, and an honest account of what those settings actually do on G9 geometry — not on G8 geometry where most of the community knowledge was built. This article gets into dForce configuration from scratch on converted clothing, SubD levels and their interaction with HD morphs, the Iray Uber shader values that consistently produce better G9 skin results, and the G9 expression system at a level of detail that most tutorials skip entirely. If you have a question that this article does not answer, or you want to compare notes with other artists working through the same problems, there is a dedicated discussion thread where this series is being followed: join the conversation in the forum — technical questions tend to get better answers there than anywhere else because people can share screenshots and scene files. If you are still on the migration prep stage rather than the technical refinement stage, start with the migration guide before this one — the technical settings here assume you are already working in G9 regularly and hitting specific walls.
How to Rebuild dForce From Scratch on Converted G9 Clothing
When you run Auto-Fit on a dForce G8 outfit and load it on a G9 figure, the geometry transfers but the simulation is completely gone. What you have is a static mesh. You can fix this — it just requires actually setting up dForce from scratch rather than hoping the converter handles it. Here is what that process looks like with the specific settings that work well as starting points on G9.
Adding the modifier. Select the converted clothing item. Go to Edit → Object → Geometry → Add dForce Modifier: Dynamic Surface. This gives you a dForce modifier with default parameters that were not calibrated for the specific garment. Open the Simulation Settings pane. The three values you will adjust most often are Stiffness, Dampening, and Buckling Stiffness. Stiffness controls how much the fabric resists deformation: 0.1 to 0.3 for light fabrics like chiffon or thin cotton, 0.5 to 0.8 for heavier structured fabric, 0.7 to 0.9 for leather. Dampening affects how quickly motion settles — keep it between 0.2 and 0.35 for most garments. Buckling Stiffness is the one most people ignore and it is the one that matters most for structured garments: 0.4 to 0.6 keeps a tailored jacket’s silhouette intact while still allowing drape in unstructured areas.
Weight maps — the step most tutorials skip. If you run a simulation without painting weight maps, the entire garment simulates including areas that should be pinned to the figure — waistbands, collars, cuffs. The garment falls off or collapses where it should be flush. In the Weight Map pane with your clothing item selected, find the dForce Modifier weight map slot. Set a base weight of 1.0 across the whole garment first. Then paint 0.0 on attachment points. Transition zones between rigid and dynamic areas — the area below a structured collar, the top of a flowing skirt — should be 0.3 to 0.5 to avoid sharp simulation boundaries that look artificial. This is the work that makes a rebuilt dForce setup look professional rather than converted.
G9-specific collision offset. G9’s vertex density distribution differs from G8 at the torso and shoulders. The default collision offset in DAZ Studio (0.01) works on G8 but often produces fabric clipping at the chest and shoulder on G9 even with a correctly configured simulation. Increase the Collision Offset to 0.015 or 0.02 and re-run before assuming the weight map or stiffness values are wrong. Most persistent G9 torso collision issues are offset problems, not structural ones.
SubD Levels and HD Morphs: Where Your G9 Render Time Actually Goes
This is where a lot of G9 artists quietly bleed render time without knowing exactly why. Genesis 9 supports SubD levels independently of HD morph resolution, and the interaction between those two settings causes more wasted hours than almost any other configuration problem.
What SubD level actually does on G9. SubD increases mesh resolution at render time by subdividing the base mesh geometry. At level 1 the mesh doubles in polygon count. At level 2 it quadruples. G9’s base mesh is already denser than G8’s — particularly in the face — which means SubD 2 on a G9 character produces a significantly higher polygon count than the equivalent setting on G8. The practical rule: SubD 1 for any character occupying less than 25 percent of the frame. SubD 2 for hero characters at close to medium distance. SubD 0 for strict background fill characters where only silhouette matters. That distinction alone can cut render time by 30 to 40 percent on multi-character scenes with no visible quality change in the final image.
The HD morph trap. HD morphs on G9 require a minimum SubD level to display their detail. At SubD 0 or 1, HD morphs load but their high-frequency detail is not resolved — the mesh does not have enough polygons to express the fine geometry the HD morph is adding. The minimum to see HD morph detail is typically SubD 2, with full resolution requiring level 3 or higher depending on the morph. If you have HD morphs active and SubD set to 1, you are paying for the HD morph in scene complexity and memory without getting any benefit in the render. Either commit to SubD 2 for that character or disable the HD morphs entirely. Running HD morphs at SubD 1 is the single most common source of “my G9 renders are slower but do not look better” complaints.
VRAM on multi-character scenes. G9 at SubD 2 with HD morphs active uses substantially more VRAM during Iray rendering than an equivalent G8 setup. A scene with four G9 characters all at SubD 2 with HD morphs active can easily exceed 8GB of VRAM and push Iray into CPU fallback mode — which is orders of magnitude slower. Drop background characters to SubD 0, disable HD morphs on anyone not in close frame, and re-render before assuming you need a hardware upgrade.
Genesis 9 Iray Skin Shader: Specific Values That Actually Work
The skin settings below are not pulled from a vendor preset. They are working values developed through iteration specifically on G9 geometry, tested across different skin texture sets and lighting conditions. Use them as a calibrated starting point — skin texture sets vary enough that adjustment will always be needed. The biggest mistake is porting G8 preset values directly and expecting equivalent results.
Diffuse and base colour. G9 native skin presets typically ship with Diffuse Weight at 1.0 and Diffuse Roughness between 0.0 and 0.1. Leave Diffuse Weight at 1.0. If the skin reads too flat under diffuse lighting, increase Diffuse Roughness toward 0.15 — this scatters the response and reduces the slightly slick look G9 skin can develop at certain light angles. Do not touch Base Colour Mix unless the vendor specifically instructs it.
Subsurface scattering: the parameters where G8 habits cause the most damage
Subsurface Weight: G8 presets often run this at 1.0. On G9, values between 0.7 and 0.85 produce more accurate skin response under direct lighting. At 1.0 on G9 the skin reads as translucent rather than as skin with subsurface depth.
Subsurface Radius: This is the value that produces the wax problem when ported from G8. G8 presets commonly use RGB radius values around [1.2, 0.6, 0.4] in centimetres. On G9, start at [0.85, 0.45, 0.30] and adjust from there. The red channel controls the deep scattering that gives skin its warmth — pull it too high and the skin glows warm in a way that reads as unrealistic under any light that is not itself warm.
Subsurface Direction (Anisotropy): Most G8 presets leave this at 0.0. G9 skin presets from 2023 onwards often set this to 0.1 to 0.2, which improves skin response under backlit conditions. Worth adding if you render in environments with strong natural backlighting.
Specular lobes and translucency: fixing the wet skin and glowing ear problems
Specular. G9 skin has more geometric micro-detail in the face, which means a single broad specular lobe (Specular 1 Weight around 0.5, Roughness around 0.4) can produce an almost wet-skin appearance on G9’s denser face geometry. Add a second tighter lobe: Specular 2 Weight at 0.15 to 0.2, Roughness at 0.15 to 0.2. This separates broad skin sheen from fine pore-level specular and produces a more accurate result on G9 face detail.
Translucency Weight. As covered in the previous article in this series, G9’s translucency response under rim and back lighting is more pronounced than G8’s at default values. The specific fix: Translucency Weight at 0.35 to 0.45 for indoor lighting setups. For outdoor or strongly backlit scenes, go lower — 0.25 to 0.35. Default G9 presets frequently ship with Translucency Weight at 0.5 or higher, which was calibrated for controlled promotional lighting. In complex environments it tends to read as unrealistic.
The G9 Expression System: Why Nothing From G8 Works and What to Do About It
The G9 expression system is architecturally different from G8, and the DAZ documentation explains it primarily through the interface rather than the underlying mechanism. Understanding the mechanism makes the practical workarounds much faster to reach.
Why G8 expression packs do not load. G9 expressions are built on a FACS-inspired morph system — Facial Action Coding System. Rather than named morphs like “Smile” or “Frown” that produce a complete expression in one step, G9 uses discrete action unit morphs controlling specific facial muscle groups: lip corners, cheek raise, brow inner raise, lid tightener, and so on. A smile is built by combining multiple action unit morphs at specific values. A G8 expression file contains morph target names that are specific to G8’s architecture — names that simply do not exist on G9’s rig. The file is not corrupt. It is written for a different system. There is no automatic translation between them. The upside: G9 expressions are composable. You can dial in a base expression and add secondary morphs for asymmetry and nuance without expressions fighting each other the way stacked G8 morphs sometimes did.
Building a working expression library on G9 without buying everything twice
The most efficient approach is to build a small set of named expression poses rather than relying on individual morph presets. Load your G9 character, dial in the combination of action unit morphs that produces the expression you need, and save it as a full pose preset — not a morph preset, but a pose preset, which records the morph dial values as pose channels. In one click it applies to any character sharing the same base rig. Start with five or six: neutral, three positive variants (subtle, moderate, strong), one uncertainty, one negative. You will need far fewer individual expression assets than on G8 because you can interpolate between saved poses using the dials rather than needing a separate asset for every variation.
Asymmetry. Symmetrical expressions look artificial at portrait distance under directional lighting. G9’s action unit morphs can be applied independently to left and right sides of the face — the left and right versions of most action units are exposed as separate dials. After dialling in a base expression, drop one corner of the mouth by 5 to 10 percent, raise one brow slightly, add a minor lid tightener on one side. It takes two minutes and makes a visible difference in every portrait render. This is the kind of detail that separates renders that read as alive from renders that read as generated.
Organising a Dual G8 and G9 Library So You Stop Losing Time to It
A practical problem nobody in this series has addressed yet: how do you actually organise DAZ Studio’s content library when you are running G8 and G9 assets side by side across several thousand items? The answer is not exciting, but ignoring it creates friction that compounds slowly over months until it is genuinely costing you render time every session.
Custom categories over folder structure. DAZ Studio’s Smart Content system uses metadata categories rather than folder structure. This means you can tag assets with G8/G9 compatibility without moving any files. In the Smart Content pane, create custom categories: “G9 Skin Tested”, “G9 Clothing Converted”, “G9 Rebuild Needed”, “G8 Only”. Drag assets in as you test them. Building the categories takes a few minutes per asset the first time. The payoff six months later is that you are not re-testing items you already tested, and you are not accidentally loading a G8-only expression pack on a G9 character and spending ten minutes wondering why nothing applies. The category system is underused by most artists because it requires upfront investment. It earns that investment back at a few hundred items and pays dividends beyond that.
Naming conventions for custom G9 presets. When you save custom G9 expressions, lighting presets, or skin calibration presets, prefix the file name with “G9_” explicitly. DAZ Studio does not separate G8 and G9 presets by default in the pose or material preset libraries — they are shown together, which means a library built over several years of G8 work is mixed indiscriminately with your new G9 saves. The prefix makes filtering trivial and prevents the recurring problem of loading a G8 material preset on a G9 character by mistake. It costs nothing and removes a low-grade irritation that otherwise happens two or three times a session.
G8 vs G9 Iray Skin Shader: Parameter Reference Table
| Shader Parameter | Typical G8 Preset Value | Recommended G9 Starting Point | Why It Differs |
|---|---|---|---|
| Subsurface Weight | 1.0 | 0.70 – 0.85 | G9 mesh density amplifies SSS response; full weight reads as translucent |
| SSS Radius R (cm) | 1.1 – 1.3 | 0.80 – 0.90 | Denser mesh resolves SSS at smaller radii; G8 values produce wax appearance |
| SSS Radius G (cm) | 0.55 – 0.65 | 0.40 – 0.50 | Same reason; green channel affects mid-tone warmth |
| SSS Radius B (cm) | 0.35 – 0.45 | 0.25 – 0.35 | Blue channel (surface scatter) needs less radius on denser mesh |
| Translucency Weight | 0.45 – 0.55 | 0.30 – 0.45 | G9’s translucency response under rim lighting is stronger at default values |
| Specular 1 Roughness | 0.40 – 0.50 | 0.35 – 0.45 | G9 face geometry resolves highlights more tightly; broad lobe reads slightly wet |
| Specular 2 Weight | Not commonly used | 0.15 – 0.20 | Second lobe separates pore-level specular from broad sheen on G9 detail |
| Specular 2 Roughness | Not commonly used | 0.15 – 0.20 | Tight lobe for fine surface specular; complements the broader primary lobe |
| Diffuse Roughness | 0.0 – 0.05 | 0.05 – 0.15 | Slight roughness reduces flat diffuse response on G9’s smoother base geometry |
| SubD Level (render) | 1 – 2 | 0 – 1 (background), 2 (hero portrait) | G9 base mesh denser than G8; SubD 2 on all characters causes VRAM problems in scenes |
Got a Specific G9 Problem? Continue the Discussion in the Forum
These settings are starting points calibrated on a range of G9 skin sets and scenes, but skin textures vary enough that your mileage will differ. The value of a specific number like “SSS Radius R at 0.85” depends heavily on the skin texture, the lighting temperature, and the render distance. What works well in a warm interior scene with soft lighting will look different in a cold exterior scene with hard directional light.
The most useful thing after reading a technical article like this one is comparing results with artists who are working in the same conditions. The forum thread where this series is being discussed is the right place for that: post your render results, your parameter questions, and your specific G9 problems in the dedicated thread. Screenshots and before/after comparisons tend to produce faster and more accurate answers than text descriptions alone — and the discussion there reflects real working conditions rather than tutorial-optimized scenarios. If you find values that work significantly better than the starting points in the table above, share them there. This is the kind of knowledge that is genuinely more useful when it is collective.
Technical Questions That Come Up Repeatedly
My dForce simulation on a converted G9 outfit runs for a long time and then the garment explodes. What is causing it?
Almost always a collision offset or timeline issue, not a stiffness setting. First check that the Collision Offset is at least 0.015 for G9 — too low and the simulation detects false intersections and over-corrects. Second, check the simulation timeline: if your scene animation is longer than the default dForce simulation range, the simulation can destabilise in the later frames. Set the simulation range explicitly to match your actual pose or animation range. Third, if the garment has no weight maps, the entire mesh is fully dynamic including areas that should be anchored — paint 0.0 weight on attachment points first and re-run.
What is the correct SubD level for G9 during simulation? Should I simulate at high or low SubD?
Simulate at SubD 1, render at whatever level your scene requires. Running dForce simulation at SubD 2 increases simulation time and memory use significantly without producing meaningfully better drape results. The simulation geometry does not need to match the render geometry. Set SubD Render level to your target and SubD Simulation level to 1 in the figure’s mesh resolution settings.
G9’s FACS morphs are overwhelming. Is there a faster way to work with expressions than dialling individual action units?
Two options. First, most G9 expression packs sold by vendors are already built as combined FACS morph presets — they expose a single dial that mixes the underlying action units for you. Buying two or three well-reviewed G9 expression packs gives you a usable expression library without manually dialling action units for every expression. Second, build a small set of your own poses as described in the expressions section above — save the combined morph state as a pose preset and load it in one click. You do not have to use the raw FACS system directly unless you want the compositional control it offers.
I applied a G9 skin shader preset and now some areas of the figure have completely different shader settings from others. What happened?
G9 characters are divided into multiple material zones — face, lips, mouth, torso, limbs, fingernails, and others — each with separate shader settings. When you apply a skin preset, it should update all zones together if the preset was built correctly. If some zones are wrong, the preset was either built for a different character’s UV layout, or you applied a single-zone material instead of a full-character preset. Check that you are loading from the character’s specific material preset folder rather than a generic skin folder. The face zone in particular sometimes receives different shader defaults than the body because some vendors treat them as separate calibration problems.
My G9 portrait renders look softer than my G8 ones even at the same resolution and SubD level. Why?
Check whether Depth of Field is active in your camera settings. G9’s denser face geometry produces more micro-surface variation, which interacts with even slight DoF blur more visibly than G8’s smoother base mesh did. A DoF setting that was imperceptible on G8 can produce noticeable softness on G9 detail. Turn DoF off and re-render to check if that is the cause before looking at shader or SubD settings. The second possibility: your Iray sample count is the same as for G8, but G9 skin with SSS and HD morphs converges more slowly — it needs more samples to resolve cleanly. Try increasing samples by 30 percent and check whether the softness resolves.
Can I use G9 characters and G8 characters in the same Iray scene without compatibility problems?
Yes, without any technical issues. DAZ Studio handles mixed-generation scenes without conflict. The only practical consideration is SubD and memory management — see the VRAM section above. The material systems are separate per figure, so a G9 skin shader on a G9 character and a G8 skin shader on a G8 character in the same scene are independent and do not interfere with each other. The render engine processes them the same way regardless of generation.
The Full Genesis 8 to Genesis 9 Series
- Genesis 8 vs Genesis 9: Which Should DAZ Artists Use? — where the series starts (3D Shards Blog)
- The Truth DAZ Artists Don’t Tell You — hidden costs and ecosystem reality (3D Shards Blog)
- How to Switch Without Losing Your Library — the migration prep guide (3D Shards Blog)
- Six Months In: The Rendering Problems Nobody Warned You About — mid-migration troubleshooting (3D Shards Blog)
- Genesis 8 to G9 Migration Discussion Thread — ask questions and share results (3D Shards Forum)
- G9-native and dual-compatible asset collections — 3D Shards Store
