SDXL Settings Pipe (JPS)
The leaner settings unpacker, no refiner fields
- sdxl_settings
- image_res
- width
- height
- res_factor
- sampler_name
- scheduler
- steps
- cfg
- cfg_rescale
- clip_skip
- filename
Same job as SDXL Basic Settings Pipe (JPS) - unpack a bundled BASIC_PIPE of SDXL generation settings into individual, wireable values - but a leaner set of fields. This one carries image_res, width, height, res_factor, sampler_name, scheduler, steps, cfg, cfg_rescale, clip_skip, and filename. Compared against SDXL Basic Settings Pipe's output list, it's missing cfg_refiner, ascore_refiner, and step_split, and it names its step count steps rather than steps_total - the signature of a settings source built without a refiner stage in mind, rather than JPS's flagship Basic Settings menu specifically.
The pack ships several different SDXL settings menu nodes beyond the handful covered elsewhere in this series (its README also lists a plain "Generation Settings" menu among others). This Pipe node's job is generic across whichever of them produces an sdxl_settings-typed pipe - same unpacking convention, just paired with a different, leaner upstream settings source than Basic Settings.
How it works
Straightforward unpack: one BASIC_PIPE in, eleven typed values out. image_res/width/height resolve whatever settings dropdown chose the resolution into real numbers you can wire into EmptyLatentImage. res_factor is there for downstream upscale math - pair it directly with Multiply Int Int (JPS) if you're computing a 2x/4x target dimension. The rest - sampler_name, scheduler, steps, cfg, cfg_rescale, clip_skip, filename - go wherever your sampling chain and save nodes live.
The inputs and outputs that matter
sdxl_settings(BASIC_PIPE, required) - from whichever JPS settings menu produced it.- Outputs:
image_res(INT),width(INT),height(INT),res_factor(INT),sampler_name,scheduler,steps(INT),cfg(FLOAT),cfg_rescale(FLOAT),clip_skip(INT),filename(STRING).
How to install it
ComfyUI Manager: search "JPS Custom Nodes for ComfyUI." Manually:
cd ComfyUI/custom_nodes
git clone https://github.com/JPS-GER/ComfyUI_JPS-Nodes.git
Restart. Nothing extra needed beyond the base pack.
Common issues & troubleshooting
Missing outputs you expected from another JPS pipe node. If you were expecting cfg_refiner, ascore_refiner, or step_split and don't see them here, that's not a bug - this pipe unpacks a leaner, non-refiner-oriented settings source. If your workflow genuinely needs those refiner-specific fields, you want SDXL Basic Settings Pipe (JPS) paired with SDXL Basic Settings (JPS) instead.
Values look wrong. As with any Pipe node in this pack, this one only reads what an upstream settings menu produced - trace the problem back to whichever node built the sdxl_settings pipe rather than debugging here.
Not sure which upstream settings node this pairs with. Match by the pipe's type name (sdxl_settings) rather than by node title - several of JPS's settings menus can feed a pipe of this shape, so check what's actually plugged into this node's input rather than assuming.
Inputs (1)
| Name | Type | Default | Description |
|---|---|---|---|
| sdxl_settings | BASIC_PIPE | — |
Outputs (11)
| Name | Type | Description |
|---|---|---|
| image_res | INT | — |
| width | INT | — |
| height | INT | — |
| res_factor | INT | — |
| sampler_name | euler,euler_cfg_pp,euler_ancestral,euler_ancestral_cfg_pp,heun,heunpp2,dpm_2,dpm_2_ancestral,lms,dpm_fast,dpm_adaptive,dpmpp_2s_ancestral,dpmpp_2s_ancestral_cfg_pp,dpmpp_sde,dpmpp_sde_gpu,dpmpp_2m,dpmpp_2m_cfg_pp,dpmpp_2m_sde,dpmpp_2m_sde_gpu,dpmpp_3m_sde,dpmpp_3m_sde_gpu,ddpm,lcm,ipndm,ipndm_v,deis,res_multistep,res_multistep_cfg_pp,res_multistep_ancestral,res_multistep_ancestral_cfg_pp,gradient_estimation,ddim,uni_pc,uni_pc_bh2 | — |
| scheduler | normal,karras,exponential,sgm_uniform,simple,ddim_uniform,beta,linear_quadratic,kl_optimal | — |
| steps | INT | — |
| cfg | FLOAT | — |
| cfg_rescale | FLOAT | — |
| clip_skip | INT | — |
| filename | STRING | — |