python_vali¶

`GetNumGpus`()	Get number of available GPUs.
`GetNvencParams`()	Get list of params PyNvEncoder can be initialized with.
`SetFFMpegLogLevel`(arg0)	Set FFMpeg log level.
`PySurfaceResizer`	CUDA-accelerated Surface resizer.
`PySurfaceDownloader`	This class is used to copy Surface to numpy ndarray using CUDA DtoH memcpy.
`PySurfaceConverter`	CUDA-accelerated converter between different pixel formats.
`PyNvEncoder`	HW-accelerated video encoder which uses Nvenc.
`PyDecoder`	Video decoder class.
`PyFrameUploader`	This class is used to upload numpy array to Surface using CUDA HtoD memcpy.
`PyNvJpegEncoder`
`SeekContext`	Incapsulates information required by decoder to seek for a particular video frame.
`SurfacePlane`	Continious 2D chunk of memory stored in vRAM which represents single plane / channel of video frame.
`Surface`	Image stored in vRAM.

class python_vali._python_vali.ColorRange¶

Members:

MPEG : Narrow or MPEG color range. Doesn’t use full [0;255] range.

JPEG : Full of JPEG color range. Uses full [0;255] range.

UDEF : Undefined color range.

property name¶

class python_vali._python_vali.ColorSpace¶

Members:

BT_601 : BT.601 color space.

BT_709 : BT.709 color space.

UNSPEC : Unspecified color space.

property name¶

class python_vali._python_vali.ColorspaceConversionContext¶: Stores information required for accurate color conversion.

class python_vali._python_vali.CudaBuffer¶

General purpose data storage class in GPU memory.

Clone(self: python_vali._python_vali.CudaBuffer) → python_vali._python_vali.CudaBuffer¶

Deep copy = CUDA mem alloc + CUDA mem copy.

Return type:: python_vali.CudaBuffer

CopyFrom(*args, **kwargs)¶

Overloaded function.

CopyFrom(self: python_vali._python_vali.CudaBuffer, other: python_vali._python_vali.CudaBuffer, stream: int) -> None

Copy content of another CudaBuffer into this CudaBuffer

param other:

other CudaBuffer

param stream:

CUDA stream to use

rtype:

None
CopyFrom(self: python_vali._python_vali.CudaBuffer, other: python_vali._python_vali.CudaBuffer, gpu_id: int) -> None

Copy content of another CudaBuffer into this CudaBuffer

param other:

other CudaBuffer

param gpu_id:

GPU to use for memcopy

rtype:

None

property ElemSize¶

Get size of single element in bytes

Return type:: Int

property GpuMem¶

Get CUdeviceptr of memory allocation.

Return type:: Int

static Make(elem_size: int, num_elems: int, gpu_id: int) → python_vali._python_vali.CudaBuffer¶

Constructor method

Parameters:

elem_size – single buffer element size in bytes
num_elems – number of elements in buffer
gpu_id – GPU to use for memcopy

Return type:

python_vali.CudaBuffer

property NumElems¶

Get number of elements in buffer.

Return type:: Int

property RawMemSize¶

Get size of buffer in bytes.

Return type:: Int

class python_vali._python_vali.DLDeviceType¶

Members:

kDLCPU : CPU device.

kDLCUDA : CUDA GPU device.

kDLCUDAHost : Pinned CUDA CPU memory by cudaMallocHost.

kDLCUDAManaged : CUDA managed/unified memory allocated by cudaMallocManaged.

property name¶

class python_vali._python_vali.FfmpegLogLevel¶

Members:

PANIC : AV_LOG_PANIC

FATAL : AV_LOG_FATAL

ERROR : AV_LOG_ERROR

WARNING : AV_LOG_WARNING

INFO : AV_LOG_INFO

VERBOSE : AV_LOG_VERBOSE

DEBUG : AV_LOG_DEBUG

property name¶

python_vali._python_vali.GetNumGpus() → int¶: Get number of available GPUs.

python_vali._python_vali.GetNvencParams() → Dict[str, str]¶: Get list of params PyNvEncoder can be initialized with.

class python_vali._python_vali.MotionVector¶

This class stores iformation about motion vector.

property dst_x¶: Absolute detination X position. Can be outside frame area.

property dst_y¶: Absolute detination Y position. Can be outside frame area.

property h¶: Macroblock height

property motion_scale¶: Motion prediction precision. E. g. 4 for quarter-pixel precision.

property motion_x¶: Motion vector X component.

property motion_y¶: Motion vector Y component.

property source¶: Where the current macroblock comes from; negative value when it comes from the past, positive value when it comes from the future.

property src_x¶: Absolute source X position. Can be outside frame area.

property src_y¶: Absolute source Y position. Can be outside frame area.

property w¶: Macroblock width.

class python_vali._python_vali.NV_ENC_CAPS¶

Members:

NUM_MAX_BFRAMES

SUPPORTED_RATECONTROL_MODES

SUPPORT_FIELD_ENCODING

SUPPORT_MONOCHROME

SUPPORT_FMO

SUPPORT_QPELMV

SUPPORT_BDIRECT_MODE

SUPPORT_CABAC

SUPPORT_ADAPTIVE_TRANSFORM

SUPPORT_STEREO_MVC

NUM_MAX_TEMPORAL_LAYERS

SUPPORT_HIERARCHICAL_PFRAMES

SUPPORT_HIERARCHICAL_BFRAMES

LEVEL_MAX

LEVEL_MIN

SEPARATE_COLOUR_PLANE

WIDTH_MAX

HEIGHT_MAX

SUPPORT_TEMPORAL_SVC

SUPPORT_DYN_RES_CHANGE

SUPPORT_DYN_BITRATE_CHANGE

SUPPORT_DYN_FORCE_CONSTQP

SUPPORT_DYN_RCMODE_CHANGE

SUPPORT_SUBFRAME_READBACK

SUPPORT_CONSTRAINED_ENCODING

SUPPORT_INTRA_REFRESH

SUPPORT_CUSTOM_VBV_BUF_SIZE

SUPPORT_DYNAMIC_SLICE_MODE

SUPPORT_REF_PIC_INVALIDATION

PREPROC_SUPPORT

ASYNC_ENCODE_SUPPORT

MB_NUM_MAX

MB_PER_SEC_MAX

SUPPORT_YUV444_ENCODE

SUPPORT_LOSSLESS_ENCODE

SUPPORT_SAO

SUPPORT_MEONLY_MODE

SUPPORT_LOOKAHEAD

SUPPORT_TEMPORAL_AQ

SUPPORT_10BIT_ENCODE

NUM_MAX_LTR_FRAMES

SUPPORT_WEIGHTED_PREDICTION

DYNAMIC_QUERY_ENCODER_CAPACITY

SUPPORT_BFRAME_REF_MODE

SUPPORT_EMPHASIS_LEVEL_MAP

WIDTH_MIN

HEIGHT_MIN

SUPPORT_MULTIPLE_REF_FRAMES

SUPPORT_ALPHA_LAYER_ENCODING

EXPOSED_COUNT

property name¶

class python_vali._python_vali.PacketData¶

Video frame attributes

property bsl¶: Amount of bytes decoder had to consume to decode corresp. packet. Useful to see when seeking for a previous key frame.

property dts¶: Decode timestamp.

property duration¶: Duration of a packet.

property key¶: 1 if frame is I frame, 0 otherwise.

property pos¶: Position of compressed packet in input bitstream.

property pts¶: Presentation timestamp.

class python_vali._python_vali.PixelFormat¶

Members:

Y : Grayscale.

RGB : Interleaved 8 bit RGB.

NV12 : Semi planar 8 bit: full resolution Y + quarter resolution interleaved UV.

YUV420 : Planar 8 bit: full resolution Y + quarter resolution U + quarter resolution V.

RGB_PLANAR : Planar 8 bit R+G+B.

BGR : Planar 8 bit R+G+B.

YUV444 : Planar 8 bit Y+U+V.

YUV444_10bit : 10 bit YUV444.

YUV420_10bit : 10 bit YUV420

UNDEFINED : Undefined pixel format, use to signal unsupported formats

RGB_32F : 32 bit float RGB.

RGB_32F_PLANAR : 32 bit float planar RGB

YUV422 : 8 bit planar: full resolution Y + half resolution U + half resolution V.

P10 : 10 bit NV12.

P12 : 12 bit NV12.

property name¶

class python_vali._python_vali.PyDecoder¶

Video decoder class.

property AvgFramerate¶: Return encoded video file average framerate.

property Bitrate¶: Return encoded video bitrate in bits per second.

property ColorRange¶

Get color range information stored in video file. Please not that some video containers may not store this information.

Returns:: color range information

property ColorSpace¶

Get color space information stored in video file. Please not that some video containers may not store this information.

Returns:: color space information

DecodeSingleFrame(*args, **kwargs)¶

Overloaded function.

DecodeSingleFrame(self: python_vali._python_vali.PyDecoder, frame: numpy.ndarray, seek_ctx: Optional[python_vali._python_vali.SeekContext] = None) -> Tuple[bool, python_vali._python_vali.TaskExecInfo]

Decode single video frame from input file. Only call this method for decoder without HW acceleration.

param frame:

decoded video frame

param pkt_data:

decoded video frame packet data, may be None

param seek_ctx:

seek context, may be None

return:

tuple, first element is True in case of success, False otherwise. Second elements is TaskExecInfo.
DecodeSingleFrame(self: python_vali._python_vali.PyDecoder, frame: numpy.ndarray, pkt_data: python_vali._python_vali.PacketData, seek_ctx: Optional[python_vali._python_vali.SeekContext] = None) -> Tuple[bool, python_vali._python_vali.TaskExecInfo]

Decode single video frame from input file. Only call this method for decoder without HW acceleration.

param frame:

decoded video frame

param pkt_data:

decoded video frame packet data, may be None

param seek_ctx:

seek context, may be None

return:

tuple, first element is True in case of success, False otherwise. Second elements is TaskExecInfo.

DecodeSingleSurface(*args, **kwargs)¶

Overloaded function.

DecodeSingleSurface(self: python_vali._python_vali.PyDecoder, surf: VPF::Surface, seek_ctx: Optional[python_vali._python_vali.SeekContext] = None) -> Tuple[bool, python_vali._python_vali.TaskExecInfo]

Decode single video surface from input file. Only call this method for HW-accelerated decoder.

param surf:

decoded video surface

param pkt_data:

decoded video surface packet data, may be None

param seek_ctx:

seek context, may be None

return:

tuple, first element is True in case of success, False otherwise. Second elements is TaskExecInfo.
DecodeSingleSurface(self: python_vali._python_vali.PyDecoder, surf: VPF::Surface, pkt_data: python_vali._python_vali.PacketData, seek_ctx: Optional[python_vali._python_vali.SeekContext] = None) -> Tuple[bool, python_vali._python_vali.TaskExecInfo]

Decode single video surface from input file. Only call this method for HW-accelerated decoder.

param surf:

decoded video surface

param pkt_data:

decoded video surface packet data, may be None

param seek_ctx:

seek context, may be None

return:

tuple, first element is True in case of success, False otherwise. Second elements is TaskExecInfo.

property Delay¶: Return encoded video delay.

property Duration¶: Return video duration time in seconds. May not be present.

property Format¶: Return encoded video file pixel format.

property Framerate¶: Return encoded video file framerate.

property GopSize¶: Return encoded video GOP size.

property Height¶: Return encoded video file height in pixels.

property HostFrameSize¶: Return amount of bytes needed to store decoded frame.

property IsAccelerated¶: Return true if decoder has HW acceleration support, false otherwise.

property IsVFR¶: Return true if video has variable framerate, false otherwise.

property Level¶: Return encoded video level coding parameter.

property Metadata¶: Return dictionary with video file metadata.

property MotionVectors¶

Return motion vectors of last decoded frame.: If there are no movion vectors it will return empty list.

Returns:: list of motion vectors
Return type:: List[vali.MotionVector]

property NumFrames¶: Return number of video frames in encoded video file. Please note that some video containers doesn’t store this infomation.

property NumStreams¶

audio and video.

Type:: Return number of streams in video file. E. g. 2 streams

property Profile¶: Return encoded video profile coding parameter.

property StartTime¶: Return video start time in seconds.

property StreamIndex¶: Return number of current video stream in file. E. g. video stream has index 0, and audio stream has index 1. This method will return 0 then.

property Timebase¶: Return encoded video file time base.

property Width¶: Return encoded video file width in pixels.

class python_vali._python_vali.PyFrameConverter¶

libswscale converter between different pixel formats.

property Format¶: Get pixel format.

Run(self: python_vali._python_vali.PyFrameConverter, src: numpy.ndarray, dst: numpy.ndarray, cc_ctx: python_vali._python_vali.ColorspaceConversionContext) → Tuple[bool, python_vali._python_vali.TaskExecInfo]¶

Perform pixel format conversion.

Parameters:

src – input numpy ndarray, it must be of proper size for given format and resolution.
dst – output numpy ndarray, it may be resized to fit the converted frame.
cc_ctx – colorspace conversion context. Describes color space and color range used for conversion.

Returns:

tuple containing: success (Bool) True in case of success, False otherwise. info (TaskExecInfo) task execution information.

Return type:

tuple

class python_vali._python_vali.PyFrameUploader¶

This class is used to upload numpy array to Surface using CUDA HtoD memcpy.

Run(self: python_vali._python_vali.PyFrameUploader, src: numpy.ndarray, dst: VPF::Surface) → Tuple[bool, python_vali._python_vali.TaskExecInfo]¶

Blocking HtoD CUDA memcpy.

Parameters:

src (numpy.ndarray) – input numpy array
dst (Surface) – output surface

Returns:

tuple containing: success (Bool) True in case of success, False otherwise. info (TaskExecInfo) task execution information.

Return type:

tuple

class python_vali._python_vali.PyNvEncoder¶

HW-accelerated video encoder which uses Nvenc.

property Capabilities¶: Return dictionary with Nvenc capabilities.

EncodeSingleSurface(*args, **kwargs)¶

Overloaded function.

EncodeSingleSurface(self: python_vali._python_vali.PyNvEncoder, surface: VPF::Surface, packet: numpy.ndarray[numpy.uint8], sei: numpy.ndarray[numpy.uint8], sync: bool, append: bool) -> bool

Encode single Surface. Please not that this function may not return compressed video packet.

param surface:

raw input Surface

param packet:

output compressed packet

param sei:

unregistered user data SEI information to be attached to encoded bitstream

param sync:

run function in sync mode, will ensure encoded packet is returned when function returns

param append:

append encoded packet to input packet

return:

True in case of success, False otherwise.
EncodeSingleSurface(self: python_vali._python_vali.PyNvEncoder, surface: VPF::Surface, packet: numpy.ndarray[numpy.uint8], sei: numpy.ndarray[numpy.uint8], sync: bool) -> bool

Encode single Surface. Please not that this function may not return compressed video packet.

param surface:

raw input Surface

param packet:

output compressed packet

param sei:

unregistered user data SEI information to be attached to encoded bitstream

param sync:

run function in sync mode, will ensure encoded packet is returned when function returns

return:

True in case of success, False otherwise.
EncodeSingleSurface(self: python_vali._python_vali.PyNvEncoder, surface: VPF::Surface, packet: numpy.ndarray[numpy.uint8], sync: bool) -> bool

Encode single Surface. Please not that this function may not return compressed video packet.

param surface:

raw input Surface

param packet:

output compressed packet

param sync:

run function in sync mode, will ensure encoded packet is returned when function returns

return:

True in case of success, False otherwise.
EncodeSingleSurface(self: python_vali._python_vali.PyNvEncoder, surface: VPF::Surface, packet: numpy.ndarray[numpy.uint8], sei: numpy.ndarray[numpy.uint8]) -> bool

Encode single Surface. Please not that this function may not return compressed video packet.

param surface:

raw input Surface

param packet:

output compressed packet

param sei:

unregistered user data SEI information to be attached to encoded bitstream

return:

True in case of success, False otherwise.
EncodeSingleSurface(self: python_vali._python_vali.PyNvEncoder, surface: VPF::Surface, packet: numpy.ndarray[numpy.uint8]) -> bool

Encode single Surface. Please not that this function may not return compressed video packet.

param surface:

raw input Surface

param packet:

output compressed packet

return:

True in case of success, False otherwise.

Flush(self: python_vali._python_vali.PyNvEncoder, packets: numpy.ndarray[numpy.uint8]) → bool¶

Flush encoder. Use this method in the end of encoding session to obtain all remaining compressed frames.

Parameters:: packets – one or multiple compressed packets squashed together.
Returns:: True in case of success, False otherwise.

FlushSinglePacket(self: python_vali._python_vali.PyNvEncoder, packets: numpy.ndarray[numpy.uint8]) → bool¶

Flush encoder. Use this method in the end of encoding session to obtain single remaining compressed frame. TO flush encoder completely you need to call this method multiple times.

Parameters:: packets – single compressed packet.
Returns:: True in case of success, False otherwise.

property Format¶: Return encoded video stream pixel format.

property FrameSizeInBytes¶: This function is used to get the current frame size based on pixel format.

property Height¶: Return encoded video stream height in pixels.

Reconfigure(self: python_vali._python_vali.PyNvEncoder, settings: Dict[str, str], force_idr: bool = False, reset_encoder: bool = False, verbose: bool = False) → bool¶

DESC.

Parameters:

settings – Dictionary with nvenc settings
force_idr – force next encoded frame to be IDR key frame
reset_encoder – force encoder reset
verbose – output verbose information to log

Returns:

property Width¶: Return encoded video stream width in pixels.

class python_vali._python_vali.PySurfaceConverter¶

CUDA-accelerated converter between different pixel formats.

Run(self: python_vali._python_vali.PySurfaceConverter, src: VPF::Surface, dst: VPF::Surface, cc_ctx: Optional[python_vali._python_vali.ColorspaceConversionContext] = None) → Tuple[bool, python_vali._python_vali.TaskExecInfo]¶

Perform pixel format conversion.

Parameters:

src – input Surface. Must be of same format class instance was created with.
dst – output Surface. Must be of suitable format.
cc_ctx – colorspace conversion context. Describes color space and color range used for conversion. Optional parameter. If not given, VALI will automatically pick supported color conversion parameters.

Returns:

tuple containing: success (Bool) True in case of success, False otherwise. info (TaskExecInfo) task execution information.

Return type:

tuple

class python_vali._python_vali.PySurfaceDownloader¶

This class is used to copy Surface to numpy ndarray using CUDA DtoH memcpy.

Run(self: python_vali._python_vali.PySurfaceDownloader, src: VPF::Surface, dst: numpy.ndarray) → Tuple[bool, python_vali._python_vali.TaskExecInfo]¶

Perform DtoH memcpy.

Parameters:

src (Surface) – input Surface
dst (numpy.ndarray) – output numpy array

Returns:

tuple containing: success (Bool) True in case of success, False otherwise. info (TaskExecInfo) task execution information.

Return type:

tuple

class python_vali._python_vali.PySurfaceResizer¶

CUDA-accelerated Surface resizer.

Run(self: python_vali._python_vali.PySurfaceResizer, src: VPF::Surface, dst: VPF::Surface) → Tuple[bool, python_vali._python_vali.TaskExecInfo]¶

Resize input Surface.

Parameters:

src – input Surface. Must be of same format class instance was created with.
dst – output Surface. Must be of same format class instance was created with.

Returns:

tuple containing: success (Bool) True in case of success, False otherwise. info (TaskExecInfo) task execution information.

Return type:

tuple

class python_vali._python_vali.SeekContext¶

Incapsulates information required by decoder to seek for a particular video frame.

property seek_frame¶: Number of frame we want to seek.

property seek_tssec¶: Timestamp we want to seek.

python_vali._python_vali.SetFFMpegLogLevel(arg0: python_vali._python_vali.FfmpegLogLevel) → None¶: Set FFMpeg log level.

class python_vali._python_vali.Surface¶

Image stored in vRAM. Consists of 1+ SurfacePlane(s).

Clone(self: python_vali._python_vali.Surface) → python_vali._python_vali.Surface¶: CUDA mem alloc + deep copy. Object returned is manager by Python interpreter.

property Format¶: Get pixel format

property Height¶: Height in pixels of plane 0.

property HostSize¶: Amount of memory in bytes which is needed for DtoH memcopy.

property IsEmpty¶: Tell if Surface plane has memory allocated or it’s empty inside.

property IsOwnMemory¶: Return True if Surface owns memory, False if it only references actual memory allocation but doesn’t own it.

static Make(*args, **kwargs)¶

Overloaded function.

Make(format: python_vali._python_vali.PixelFormat, width: int, height: int, gpu_id: int) -> python_vali._python_vali.Surface

Constructor method.

param format:

target pixel format

param width:

width in pixels

param height:

height in pixels

param gpu_id:

GPU to use
Make(format: python_vali._python_vali.PixelFormat, width: int, height: int, context: int) -> python_vali._python_vali.Surface

Constructor method.

param format:

target pixel format

param width:

width in pixels

param height:

height in pixels

param context:

CUDA contet to use

property NumPlanes¶: Number of SurfacePlanes

property Pitch¶: Pitch in bytes of plane 0.

property Planes¶

Get SurfacePlane reference

Parameters:: plane – SurfacePlane index

property Width¶: Width in pixels of plane 0.

static from_cai(dict: object, format: python_vali._python_vali.PixelFormat = <PixelFormat.RGB: 2>) → python_vali._python_vali.Surface¶

DLPack: Make Surface from CAI, don not own memory.

Parameters:

dict – dictionary which corresponds to CUDA Array Interface specs.
fmt – pixel format, by default python_vali.PixelFormat.RGB

Returns:

Surface

Return type:

python_vali.Surface

static from_dlpack(capsule: capsule, format: python_vali._python_vali.PixelFormat = <PixelFormat.RGB: 2>) → python_vali._python_vali.Surface¶

DLPack: Make Surface from dlpack, don not own memory.

Parameters:

capsule – capsule object with manager dltensor inside
fmt – pixel format, by default python_vali.PixelFormat.RGB

Returns:

Surface

Return type:

python_vali.Surface

class python_vali._python_vali.SurfacePlane¶

Continious 2D chunk of memory stored in vRAM which represents single plane / channel of video frame. It supports DLPack specification.

property ElemSize¶: Get element size in bytes

property GpuMem¶: Get pointer to SurfacePlane data

property Height¶: Get height in pixels

property HostFrameSize¶: Get amount of host memory needed to store this SurfacePlane

property Pitch¶: Get pitch in bytes

property Width¶: Get width in pixels

class python_vali._python_vali.TaskExecInfo¶

Members:

SUCCESS : Success

FAIL : Fail

END_OF_STREAM : End of file

MORE_DATA_NEEDED : More data needed to complete

BIT_DEPTH_NOT_SUPPORTED : Bit depth isn’t supported

INVALID_INPUT : Invalid input

UNSUPPORTED_FMT_CONV_PARAMS : Unsupported color conversion parameters

NOT_SUPPORTED : Unsupported feature

RES_CHANGE : Video resolution change happened

SRC_DST_SIZE_MISMATCH : Input and output size mismatch

property name¶