Skip to main content

Other Advanced Settings

Justin62628About 4 min

Output file naming abbreviation instructions

When outputting video files, the output file name has related abbreviations, and the meanings are as follows:

SLMSlowMotion slow motion output frame rate
16bit16bit high-precision output
QEQuickExtract quick frame splitting
HWHardware decode hardware decoding
DIDeInterlaced anti-interlacing enabled
DNDeNoise noise reduction enabled
GRGrain quick noise addition enabled
DPDepan quick de-jitter enabled
OCHDROneClickHDR one-click HDR
AI-PQAI-PQ HDR conversion
DBDeepDeband AI de-striping
VFIFVFI First Interpolation first frame interpolation then super resolution
DSDynamicSale enables dynamic optical flow
SScale uses a fixed optical flow scale
VFIFrame-filling model
FP16NVIDIA GPU half-precision mode
RRRIFE-Reversed reverse optical flow
RFERIFE-Foward Ensemble bidirectional optical flow
TTATTA mode number (starting from 0)
FDFrame Deduplication duplicate frame removal mode
SRSuperResolution super resolution model

Progress bar parameter instructions

During the task execution process, the progress parameters in the task output panel are as follows:

RThe number of frames currently rendered
CThe number of frames currently being frame-filled or processed
RPTReadProcessTime single-frame read processing time
WPTWriteProcessTime single-frame compression processing time
SThe most recently recognized transition
SCThe number of recognized transitions
TATTaskAcquireTime single frame-filling task acquisition time
PTProcessTime single frame-filling task processing time
QLQueueLength frame-filling task queue length
SRThe number of frames currently being super-resolved
SRTATSuperResolutionTaskAquireTime single super-resolution task acquisition time
SRPTSuperResolutionProcessTime single super-resolution task processing time
SRLSuperResolutionQueueLength super-resolution task queue length

Special instructions for progress bar parameters

  • In version 6.x and above, the "Decoding" and "Compression Pressure" panels in the output status panel will show the current decoding and compression pressure. When the decoding pressure and compression pressure are both 0, it means that there is currently no performance bottleneck and it is in the best state. The greater the values of the decoding pressure and compression pressure, the more serious the current bottleneck is. Note This value is inaccurate when using multiple frame-filling or super-resolution threads.

  • When only frame-filling without super-resolution, if the QL value remains below 10 for a long time, it indicates that the frame-splitting encounters a bottleneck. At this time, TAT>0.1. Please check whether the CPU usage is 100%. If so, please change the CPU software encoding compression parameters (such as changing the compression preset from slow to fast) or replace the CPU with更强 single-core performance. If QuickExtract is not enabled and the RPT value is large (>10s), please enable QuickExtract to reduce the CPU load. When enabling the high-precision workflow and CPU compression, if the CPU is not strong enough, QL=0 is a normal phenomenon.

  • When both super-resolution and frame-filling are performed, QL=0 and TAT>1 are normal phenomena. Try to ensure that SRL>0 and SRTAT to maximize the working efficiency of the graphics card, because SVFI performs super-resolution first and then frame-filling, for the same reason as above.

  • If the TAT (SRTAT) value is 0 and the QL (SRL) value is stable at a relatively high value, it means that the bottleneck of the working speed is mainly in the performance of the graphics card. This state can give full play to the performance of the graphics card, which is an ideal working situation.

Additional global settings

Note The current global settings correspond to version SVFI 6.15.9

This type of global settings is set in the global_advanced_settings.json file (referred to as GAS) in the software root directory and can be edited by opening it with a text editor. By default, this file needs to be created manually.

The following is an example of all configurations. The value of the key-value pair is the default value. You can directly copy the following content and edit it as your own GAS file.

When actually editing the json file, the value corresponding to the non-existent key uses the default value:

  "is_cudnn_enabled": true,
  "trt_max_workspace_size": 1073741824,
  "calib_sr_len": 10,
  "is_cuda_graph_enabled": false,
  "is_torchscript_enabled": false,
  "is_cuda_loading_lazy": true,
  "is_torch_trt_io32": false,
  "show_trt_debug": false,

  "debug_scdet": false,
  "excluded_filename_tags": [""],
  "fmnet_model": "final",
  "deep_deband_model": "deepdeband-f.pth",
  "sisr_state_dict_key": "params",

  "pure_scene_len_threshold": 20,

  "use_minterpolate": false,
  "vfi_use_pad": false,
  "forw_dedup_use_pad": false,
  "forw_dedup_use_correction": false,
  "use_full_vspipe_seg": true,

  "is_softsplat_stable": false,
  "is_softsplat_gaussian": false,
  "is_evict_flicker": false,
  "use_xformers": true,
  "encoded_audio_format": "aac",
  "encoded_audio_bitrate": "640k"
Key nameExplanation
is_cudnn_enabledWhether to enable cudnn acceleration
trt_max_workspace_sizeThe maximum workspace size of TensorRT, in bits. When using vspipe for super-resolution, it does not take effect, and you need to manually change the workspace size of
calib_sr_lenThe number of calibrations when TensorRT uses int8 quantization
is_cuda_graph_enabledWhether to enable CUDA Graph acceleration for supported models
is_torchscript_enabledWhether to enable TorchScript mode for supported models
is_cuda_loading_lazyWhether TensorRT enables lazy loading mode, which can VRAM usage
is_torch_trt_io32Whether torch trt uses fp32 as trt input and output. After enabling, the sawtooth caused by rife trt after the previous de-duplication can be significantly reduced
show_trt_debugWhether to display TensorRT debugging information
debug_scdetWhether to display debugging information for scene detection
excluded_filename_tagsExcluded file name tags. Supported tags are ["fps", "hdr", "vfi_model", "vfi_tta", "vfi_fd", "sr_model"]. The corresponding content of the included tags will not be displayed in the output file name
fmnet_modelThe FMNet model used, located in models/restorer/FMNet/models
deep_deband_modelThe DeepDeband model used, located in models/restorer/DeepDeband/models
sisr_state_dict_keyThe weight key of the custom SISR model used, applicable to ATD, SAFMN, SPAN, and Compact model structures
use_minterpolateWhether to use the minterpolate filter when reading videos with ffmpeg, sometimes it can be enabled to correct the number of ending frames
pure_scene_len_thresholdPure scene length threshold
vfi_use_padWhether to use pad instead of interpolate for input resolution preprocessing in the frame-filling process
forw_dedup_use_padWhether to use pad instead of interpolate for input resolution preprocessing in the forward de-duplication
forw_dedup_use_correctionWhether to use correction in forward de-duplication
use_full_vspipe_segAutomatically split the output chunk when using the full vspipe workflow to support the task progress reply
is_softsplat_stableWhether to enable SoftSplat stable support
is_softsplat_gaussianWhether to enable SoftSplat Gaussian kernel support
is_evict_flickerWhether to enable flicker removal
use_xformersDefault to using xformer to accelerate the inference of certain models
encoded_audio_formatOutput audio format
encoded_audio_bitrateOutput audio bitrate


Main page

Ctrl+1Open the settings panel
Ctrl+2Open the preview panel
Ctrl+3Open the output status panel
Ctrl+4Open the Steam settings panel
Ctrl+5Open the preference settings panel
BackspaceGo back one page
Ctrl+BOpen the basic settings panel
Ctrl+AOpen the advanced settings panel
Ctrl+PShow/Hide advanced settings panel
Ctrl+SSave the current settings
Ctrl+Alt+L or Ctrl+Shift+LLoad default settings
Ctrl+DDuplicate the current project
DeleteDelete the current project
Ctrl+DeleteClear the project list
Ctrl+IImport project
Ctrl+ReturnStart frame-filling
Ctrl+EStart compression

Task status page

Ctrl+EndStop frame-filling/compression
Ctrl+SpacePause frame-filling/compression

Mask drawing page


Player component

RightFast forward
LeftFast backward
Last update:
Contributors: Justin62628