Skip to content

Technical

Tips, Tricks and Technical Information

DHDR/HDR YUV RAW Video Format

hdrfileformat.txt Take this text file and create a nice html output I can insert in a joomla article on www.drastic.tv It will replace this old one https://www.drastic.tv/support-59/supporttipstechnical/57-dhdr-hdr-yuv-raw-video-format edit Thought for 21.8 seconds Here is a complete HTML document for your Joomla article, replacing the old DHDR HDR file format page. It presents the header specification with clear sections, tables, and code samples.

Drastic HDR/DHDR formatDrastic software can read many file formats which it does not natively understand.  Formats that have a fixed video frame size and are logically organized can be described using the industry standard HDR text file format (NOTE:  Not the still image HDR).  This page describes the file format. 

Drastic Technologies · Technical Note

DHDR / HDR Raw Video File Format

A plain-text header that describes fixed-size raw video frames together with their associated audio and timecode, used by the dtx_HdrYuv.d1x plugin for videoQC, Net-X-Code Server, DTMediaRead, and QuickClip / VVW.

Format: *.dthdr / *.hdr In use for 25+ years © 1995–2026 Drastic Technologies Ltd.

1 Overview

This document describes the header (*.dthdr / *.hdr) file format for use with the Drastic Technologies dtx_HdrYuv.d1x plugin for MediaReactor and QuickClip / VVW. This format has been in use for over 25 years in various forms as a *.hdr. By default, Drastic software now writes and reads a *.dthdr (as well as still reading the .hdr). The *.dthdr is the same format with more capabilities, including better audio, split long file and time code support.

Herein are outlined the most common descriptors and their meanings as they have been standardized in our software. An updated version of this document is available at drastictech.com/hdrfileformat.txt.

The basic concept of the hdr file is a text file that describes a simple video stream in which each frame is the same size, and its associated video and audio information. On disk, there will be 2 or more files, normally with the same starting name:

  • E:\TestRecord\Testing.dthdr
  • E:\TestRecord\Testing.yuv
  • E:\TestRecord\Testing.wav

In this example the yuv file contains all the captured frames (as opposed to individually numbered yuv files), the wav file contains the matched audio and the hdr file describes them as specified below. When loading these files into QuickClip or MediaReactor, the hdr file should be chosen to make sure all files are picked up properly.

The basic format of the hdr file consists of a series of settings, each taking two text lines. The first line is the label and is always preceded by a percentage sign (%) and the second line is the value of the setting:

% Image Size (This line describes what the next line specifies)
486 720     (This contains the values - NTSC 720 x 486)
Tip: the value line contains only the value — everything after the value on that line is treated as a comment by most readers, but comments and other information may be lost if the file is re-written by QuickClip or VVW.

↑ Back to top

2 Settings

Color Format

% Color Format
YUV420

This setting describes the format of the video frames within the file. The current possible formats are listed below.

Planar YUV 4:2:2 — high bit depth

Token Description
yuv422_16 4:2:2 planar YCbCr in I420 order, 16 bits per component
yuv422_16le Little endian 4:2:2 planar YCbCr in I420 order, 16 bits per component
yvu422_16 4:2:2 planar YCbCr in YV12 order, 16 bits per component
yvu422_16le Little endian 4:2:2 planar YCbCr in YV12 order, 16 bits per component
yuv422_10 4:2:2 planar YCbCr in I420 order, 10 bits per component in 16 bit container
yuv422_10le Little endian 4:2:2 planar YCbCr in I420 order, 10 bits per component in 16 bit container
yvu422_10 4:2:2 planar YCbCr in YV12 order, 10 bits per component in 16 bit container
yvu422_10le Little endian 4:2:2 planar YCbCr in YV12 order, 10 bits per component in 16 bit container
yuv422_12 4:2:2 planar YCbCr in I420 order, 12 bits per component in 16 bit container
yuv422_12le Little endian 4:2:2 planar YCbCr in I420 order, 12 bits per component in 16 bit container
yvu422_12 4:2:2 planar YCbCr in YV12 order, 12 bits per component in 16 bit container
yvu422_12le Little endian 4:2:2 planar YCbCr in YV12 order, 12 bits per component in 16 bit container
yuv422_14 4:2:2 planar YCbCr in I420 order, 14 bits per component in 16 bit container
yuv422_14le Little endian 4:2:2 planar YCbCr in I420 order, 14 bits per component in 16 bit container
yvu422_14 4:2:2 planar YCbCr in YV12 order, 14 bits per component in 16 bit container
yvu422_14le Little endian 4:2:2 planar YCbCr in YV12 order, 14 bits per component in 16 bit container

Planar YUV 4:4:4 — high bit depth

Token Description
yuv444_10 4:4:4 planar YCbCr in I420 order, 10 bits per component in 16 bit container
yuv444_10le Little endian 4:4:4 planar YCbCr in I420 order, 10 bits per component in 16 bit container
yvu444_10 4:4:4 planar YCbCr in YV12 order, 10 bits per component in 16 bit container
yvu444_10le Little endian 4:4:4 planar YCbCr in YV12 order, 10 bits per component in 16 bit container
yuv444_12 4:4:4 planar YCbCr in I420 order, 12 bits per component in 16 bit container
yuv444_12le Little endian 4:4:4 planar YCbCr in I420 order, 12 bits per component in 16 bit container
yvu444_12 4:4:4 planar YCbCr in YV12 order, 12 bits per component in 16 bit container
yvu444_12le Little endian 4:4:4 planar YCbCr in YV12 order, 12 bits per component in 16 bit container
yuv444_14 4:4:4 planar YCbCr in I420 order, 14 bits per component in 16 bit container
yuv444_14le Little endian 4:4:4 planar YCbCr in I420 order, 14 bits per component in 16 bit container
yvu444_14 4:4:4 planar YCbCr in YV12 order, 14 bits per component in 16 bit container
yvu444_14le Little endian 4:4:4 planar YCbCr in YV12 order, 14 bits per component in 16 bit container

Planar YUV 4:2:0 — high bit depth

Token Description
yuv420_10le Little endian 4:2:0 planar YCbCr in I420 order, 10 bits per component, 16 bit container
yvu420_10le Little endian 4:2:0 planar YCbCr in YV12 order, 10 bits per component, 16 bit container
yuv420_10 4:2:0 planar YCbCr in I420 order, 10 bits per component, 16 bit container
yvu420_10 4:2:0 planar YCbCr in YV12 order, 10 bits per component, 16 bit container
yuv420_12le Little endian 4:2:0 planar YCbCr in I420 order, 12 bits per component, 16 bit container
yvu420_12le Little endian 4:2:0 planar YCbCr in YV12 order, 12 bits per component, 16 bit container
yuv420_12 4:2:0 planar YCbCr in I420 order, 12 bits per component, 16 bit container
yvu420_12 4:2:0 planar YCbCr in YV12 order, 12 bits per component, 16 bit container

Planar / semi-planar 8 bit YUV

Token Description
YUV420 Planar YCbCr in IYUV/I420 order (ST/Thompson/MPEG Groups)
YV12 Planar YCbCr in YV12 order
NV12 Planar YCbCr in I420 order, interleaved chroma (NVIDIA)
NV21 Planar YCrCb in YV12 order, interleaved chroma (NVIDIA)
YUV422P Planar YCbCr with 4:2:2 sub sampling (Sony)
YVU422P Planar YCbCr with 4:2:2 sub sampling (chroma inverted)

Interleaved / packed YUV

Token Description
V210 Interleaved 10 bit YCbCr v210 format (standard QuickClip 10 bit YCbCr)
YUV4224_10 Interleaved 10 bit YCbCr v210 format with alpha/key channel
YUV4224 Interleaved 8 bit YCbCr yuv2/UYVY format with alpha/key channel
YUV422_fields Separate fields of 4:2:2 YCbCr (Crescent)
YUV422 Interleaved 8 bit YCbCr UYVY format (standard QuickClip 8 bit YCbCr), alias uyvy422
yuvi422_16 Interleaved 16 bit YCbCr UYVY format
yuvi422_16le Little endian interleaved 16 bit YCbCr UYVY format
UYVY422 YUV 4:2:2 interleaved 8 bit packed as U Y V Y U Y V Y …
YUY2_16 YUV 4:2:2 interleaved 16 bit packed as Y U Y V Y U Y V …
YUY2_16le Little endian YUV 4:2:2 interleaved 16 bit packed as Y U Y V Y U Y V …
YUY2 YUV 4:2:2 interleaved 8 bit packed as Y U Y V Y U Y V …

DPX RGB / ABGR

Token Description
DPXRGBLEFILL DPX 10 bit RGB, little endian, filled
DPXRGBLE DPX 10 bit RGB, little endian, padded
DPXRGBFILL DPX 10 bit RGB, big endian, filled
DPXRGB DPX 10 bit RGB, big endian, padded
DPXABGRLEFILL DPX 10 bit ABGR, little endian, filled
DPXABGRLE DPX 10 bit ABGR, little endian, padded
DPXABGRFILL DPX 10 bit ABGR, big endian, filled
DPXABGR DPX 10 bit ABGR, big endian, padded

Planar RGB

Token Description
PRGB 8 bit × 3 planar RGB
PRGBA 8 bit × 4 planar RGBA
PBGR 8 bit × 3 planar BGR
PABGR 8 bit × 4 planar ABGR
PBGRA 8 bit × 4 planar BGRA
PARGB 8 bit × 4 planar ARGB

Interleaved RGB

Token Description
RGBA 32 bit interleaved RGB (TIFF)
ARGB 32 bit interleaved RGB (Mac)
BGRA 32 bit interleaved RGB (Windows BMP/TGA)
BGR 24 bit interleaved RGB (Windows BMP/TGA)
RGB 24 bit interleaved RGB (TIFF)
TIFF24 24 bit interleaved RGB TIFF ordering
TIFF32 32 bit TIFF (same as RGBA)

Special, stereo, Bayer and compressed formats

Token Description
FULLDUAL 10 bit dual frame YCbCr (stereo)
Grey 8 bit grey/gray plane of video data, alias Gray
DVSD DVSD 'dv/dif' stream 4:2:0 or 4:1:1 8 bit SD
DV25 DV25 'dv/dif' stream 4:2:0 8 bit SD
DV50 DVCPro 50 stream 4:2:2 8 bit SD, alias avdv
DV100 DVCPro HD/DV-100 stream 4:2:2 8 bit HD, alias avd1
AVCI100 Panasonic AVCi 100
DNXHD Avid DNxHD stream, alias av10, avd0~avd9
IMX30 Sony IMX MPEG 30 Mbit stream 4:2:2 8 bit
IMX40 Sony IMX MPEG 40 Mbit stream 4:2:2 8 bit
IMX50 Sony IMX MPEG 50 Mbit stream 4:2:2 8 bit
STEREO8 Dual 8 bit YCbCr interleaved streams (one after another)
STEREO10 Dual 10 bit YCbCr interleaved streams (one after another)
STEREO30 Dual 10 bit RGB DPX interleaved streams (one after another)
STEREO32 Dual 8 bit BGRA (Windows order) interleaved streams (one after another)
stereo12bggr Dual 12 bit raw Bayer (one after another)
stereo12rggb Dual 12 bit raw Bayer (one after another)
stereo12gbrg Dual 12 bit raw Bayer (one after another)
stereo12grbg Dual 12 bit raw Bayer (one after another)
bggrbe12 Bayer pattern, 12 bits per component packed big endian
rggbbe12 Bayer pattern, 12 bits per component packed big endian
gbrgbe12 Bayer pattern, 12 bits per component packed big endian
grbgbe12 Bayer pattern, 12 bits per component packed big endian
bggr12 Bayer pattern, 12 bits per component packed little endian
rggb12 Bayer pattern, 12 bits per component packed little endian
gbrg12 Bayer pattern, 12 bits per component packed little endian
grbg12 Bayer pattern, 12 bits per component packed little endian
ARRIBAYERDLRAW12 ARRI dual link raw 12 bit Bayer packed into YCbCr 10

Deprecated

Token Description
RGB_10 30 bit DPX/Cineon ordering
RGB_10LE 30 bit little endian DPX/Cineon ordering

No legacy alias — use the descriptive names below

Token Use instead
YUV411 use "YUV 4:1:1 8-bit"
YUV411P use "YUV 4:1:1 8-bit"
YUV410 use "YUV 4:1:0 8-bit"
YUV410P use "YUV 4:1:0 8-bit"
YUVA444 use "YUVA 4:4:4 8-bit" or "AYUV 4:4:4 8-bit packed"
UYV444 use "YUV 4:4:4 8-bit packed"
n/a MJPEG Motion JPEG (not supported)

YUView and pyuv descriptive names

The reader also accepts the descriptive raw pixel format names used by YUView 2.x and the short names used by pyuv 0.7. These value lines carry the whole layout, so no additional descriptors are needed. Names are case insensitive.

YUView style YUV

<order> <subsampling> <bits>-bit [BE|LE] [packed|packed-B] [(IL)] [Cx<n>] [Cy<n>]

<order>        YUV, YVU, YUVA, YVUA          planar YCbCr
               UYVY, VYUY, YUYV/YUY2, YVYU   packed 4:2:2
               AYUV, YUVA, VUYA              packed 4:4:4 with alpha
<subsampling>  4:4:4, 4:4:0, 4:2:2, 4:2:0, 4:1:1, 4:1:0, 4:0:0
<bits>         8, 9, 10, 12, 14, 16
BE / LE        byte order for samples wider than 8 bits (default LE)
packed         each sample is stored in the next whole byte/word
packed-B       samples are bit packed without byte alignment
(IL)           UV (and alpha) samples share one interleaved plane
Cx/Cy<n>       chroma siting. Accepted for compatibility; the reader
               uses the default siting for the selected subsampling.

Examples

YUV 4:4:4 8-bit
YVU 4:2:2 10-bit BE
YUV 4:2:0 10-bit LE
UYVY 4:2:2 8-bit packed
YUYV 4:2:2 10-bit packed-B
YUVA 4:4:4 12-bit
YUV 4:0:0 8-bit

YUView style RGB

<channels> <bits>bit [planar] [BE]

<channels>     RGB, RBG, GRB, GBR, BRG or BGR, optionally with A first
               or last (ARGB, RGBA, ABGR, BGRA, ...).
planar         three or four separate channel planes instead of packed
               pixels.
BE             big endian samples for formats wider than 8 bits.

Examples

RGB 8bit
RGBA 8bit
BGR 10bit planar
ABGR 12bit BE

pyuv / libyuv short names (8 bit)

P444 P422 P420 P411 P410 P400   planar YUV; P400 is luma only
I420 I422 I444 I411             accepted as planar in that subsampling
I410                            i410 is also a libyuv 10-bit 4:4:4 name;
                                use "YUV 4:1:0 8-bit" for the rare
                                PYUV 4:1:0 8-bit layout
IYUV                            planar 4:2:0
UYVY YUY2 YVYU                  packed 4:2:2

Common short aliases

Microsoft / Media Foundation

Token Layout
P010, P012, P016 4:2:0 semi-planar, 10 / 12 / 16 bit
P210, P212, P216 4:2:2 semi-planar, 10 / 12 / 16 bit
P410, P412, P416 4:4:4 semi-planar, 10 / 12 / 16 bit
Y210, Y216 packed 4:2:2, 10 / 16 bit (16 bit words)
Y410 packed 4:4:4, 10 bit + 2 bit alpha (one DWORD/pixel)
Y416 packed 4:4:4, 16 bit + alpha (two DWORDs/pixel)

QuickTime uncompressed Y'CbCr

Token Layout
V210 4:2:2 10 bit, packed (same as v210)
V216 4:2:2 10/12/14/16 bit, packed 16 bit words
V308 4:4:4 8 bit, packed as Cr Y Cb
V408 4:4:4:4 8 bit, packed as Cb Y Cr A
V410 4:4:4 10 bit, packed into one DWORD (2 zero low bits)

FFmpeg / libyuv planar names

yuv420p10le, yuv422p12le, yuv444p16le ...
i010, i012, i210, i212, i410, i412

Other common aliases

Token Layout
YUV444, YUV444P planar 4:4:4 8 bit
YV16 planar 4:2:2, Y then V then U
VYUY packed 4:2:2 V Y U Y
AYUV packed 4:4:4 A Y U V

Important bit packing differences

  • P010/P210/P410 and Y210/Y216 store 10/12 bit samples left justified in 16 bit words (the least significant bits are zero). The reader scales these correctly to 8 bit UYVY and 10 bit v210.
  • FFmpeg yuv*p10le/yuv*p12le and libyuv i0NN/i2NN/i4NN store the significant bits in the low bits of each 16 bit word.
  • Y410 DWORD: bits 0-9 U, bits 10-19 Y, bits 20-29 V, bits 30-31 A.
  • V410 DWORD: bits 0-1 zero, bits 2-11 Cb, bits 12-21 Y, bits 22-31 Cr.
  • Y416: four 16 bit words per pixel — U, Y, V, A.

Drastic filename shorthand vs Microsoft tokens

In Drastic-generated filenames, p<subsampling><bits> means a planar layout (for example _p210 = planar 4:2:2 10 bit and _p412 = planar 4:4:4 12 bit). Auto-create keeps that meaning. The explicit header tokens P010/P210/P410/P012/P212/P412/P416 follow the Microsoft semi-planar definition described above. For planar high-bit files use yuv422_10le, yuv444_12le, etc.

Video Line Pitch

The descriptive / P-series reader honors % Video Line Pitch as the luma line pitch. P010/P210 keep the same byte pitch for the interleaved UV plane; P410/P416 have full-width chroma, so their UV plane pitch is twice the luma pitch.

Fields

% Number of Fields per Image = 2 is supported for descriptive and P-series formats. Field 0 supplies even output lines and field 1 supplies odd output lines.

Writer

HdrWriteFile emits P010/P012/P016, P210/P212/P216, P410/P412/P416, Y210/Y216, Y410/Y416, V216/V308/V408/V410, YUV444P, YV16, VYUY and AYUV when the matching fcc is present.

Output negotiation

  • 8 bit YUV sources are presented to the application as UYVY (fcckViewGraphics8BitYUV).
  • YUV sources wider than 8 bits are presented as v210 (fcckViewGraphics10BitYUV); samples are scaled (or truncated) to 10 bits.
  • RGB and luma-only (4:0:0) sources are presented as 32 bit BGRA/RGB32, with >8 bit samples scaled to 8 bits. For 4:0:0 the luma value is replicated to R, G and B.

Image Size (NbRows, NbCols)

% Image Size (NbRows,NbCols)
486 720

The vertical and horizontal size of the video frame. Can be any size, but the following sizes are normal:

480x640, 480x704, 480x720, 486x720, 512x720, 576x704, 576x720, 608x720, 720x1280, 1080x1920, 1088x1920

Important: the "x" between the height and width is not supported. It must be a space in the hdr file, with the row (height) count first.

Number of Fields per Image optional

% Number of Fields per Image (optional)
2

Number of separate images that make up a field. For most images this would be 1, indicating one full size plane with progressive or interleaved lines. If there are two separate images that need to be interleaved, this would be set to 2. This value does not affect the Image Size above — that is always the absolute number of pixels vertically and horizontally.

Please note: this is the correct interpretation of this value. Previously we ignored this and recommended 2 where it should have been 1.

Signal Format Type

% Signal Format Type
interlaced

The recommended signal format output type: interlaced, progressive or segmented (psf). Default is progressive.

Number of Image

% Number of Image
1800

The number of individual frames in a file. If this value is larger than the actual number of frames available, it will be corrected — so if you are unsure, set it to a very large value (e.g. 4000000).

Frames per Second

% Frames per Second
29

The frame rate of the video frames. This may be any value, but some values have special meanings:

Value Meaning
23 Means 23.98
24 Standard film
25 PAL / 25p / 50i
29 Means 29.97 — NTSC / 29.97p / 59.94i
30 NTSC NDF / 30p / 60i
50 50p
59 Means 59.94p (for 720p, 1080p, etc.)
60 Means 60p (for 720p, 1080p, etc.)

Header Offset optional

% Header Offset (optional)
0

How far into the file the series of video frames starts.

Video Alignment optional

% Video Alignment (optional)
512

Alignment of each frame within the stream. If the frames are back to back this value will be 1. It is normal to pad each frame to align it to a disk sector for higher speed reads and writes. This value is normally either 0 (which indicates 1), 512 (Windows), 4096 (SGI), or 16384 (JS).

Video Line Pitch optional

% Video Line Pitch (optional)
1440

The number of bytes in one line of video. Especially important for YCbCr 10 bit when frame size is 1280x720 or 2048x1080.

Video Offset optional

% Video Offset (optional)
0

How far into each frame, from the calculated start, the actual video starts.

Video Name optional

% Video Name (optional)
ThisFile.YUV

This is the name of the video file/stream associated with this header. Normally the video file has the same name as the hdr file, but with a .yuv extension. This allows the video name to be overridden to any file name.

Audio Name optional

% Audio Name (optional)
ThisFile.Wav

This is the name of an associated audio file. Normally QuickClip / MediaReactor look for sidebar wave or aiff files with the same name as the header file. This allows the audio file to be specified.

Audio Format

% Audio Format
Stereo

Channel layout of the associated audio file.

Audio Frequency

% Audio Frequency
48000

Audio sample rate in Hz.

Audio Block Align

% Audio Block Align
4

Bytes per block (sample frame) of audio: channels × bytes per sample.

Audio Bits Per Sample

% Audio Bits Per Sample
24

Bit depth of each individual audio sample.

Timecode optional

% Timecode (optional)
01:00:00:00

This is the time code value of the first frame. The time is assumed to run contiguously from this point.

Userbits optional

% Userbits (optional)
BAADF00D

This is the user bits returned for this stream. It is assumed it is the same for all video frames.

Start Frame optional

% Start Frame (optional)
100

This specifies the first video frame of the stream. By default this is 0, but in a circular file the last frame written may not be the last frame of the file. In the case above, if the start frame is 100, the last frame written was 99. The reader will read from 100 → end of file and then 0 → 99 automatically.

Max Segment optional

% Max Segment (optional)
1296000

This specifies how long one segment of a multi-segment record will be. Defaults to 12 hours (the value above is 12 hours in frames, NDF 30).

Please note: comments and other information may exist in the file, but if the file is re-written by QuickClip or VVW, then that information will be lost.

↑ Back to top

3 Calculations

Total Frame Pixels = <% Image Size (NbRows,)> * <% Image Size (,NbCols)>

(1) Frame Size = <size of video file> / <% Number of Image>

Frame size by color format

(2) Frame Size =
        YUV420    (<totalframepixels> + (<totalframepixels> / 2))
        YUV411P   (<totalframepixels> + (<totalframepixels> / 2))
        YUV422P   (<totalframepixels> * 2)
        YUV411    (<totalframepixels> + (<totalframepixels> / 2))
        YUV422    (<totalframepixels> * 2)
        YUV422_10 (<totalframepixels> * 16 / 6)
        YUVA444   (<totalframepixels> * 4)
        RGBA      (<totalframepixels> * 4)
        RGB       (<totalframepixels> * 3)
        RGB_10    (<totalframepixels> * 4)
        RGB_10LE  (<totalframepixels> * 4)

YUView / pyuv descriptive names

bps = 1 for 8 bit, 2 for 9..16 bit.

Planar YUV, luma plane:      width * height * bps
4:4:4 chroma:                width * height * bps * 2
4:4:0 chroma:                width * ceil(height/2) * bps * 2
4:2:2 chroma:                ceil(width/2) * height * bps * 2
4:2:0 chroma:                ceil(width/2) * ceil(height/2) * bps * 2
4:1:1 chroma:                ceil(width/4) * height * bps * 2
4:1:0 chroma:                ceil(width/4) * ceil(height/4) * bps * 2
4:0:0 (luma only):           width * height * bps
Planar alpha plane:          width * height * bps
Packed 4:4:4 (3 channels):   width * height * 3 * bps
Packed 4:4:4 (4 channels):   width * height * 4 * bps
Packed 4:2:2 (4 values/2px): ceil(width/2) * height * 4 * bps
packed-B variants:           the same number of values, rounded up to
                             whole bytes per pixel or pixel pair
RGB packed:                  width * height * channels * bps
RGB planar:                  the same total, split into channel planes

Semi-planar P-series (Y plane followed by interleaved UV)

P010/P012/P016:  width*height*2 + ceil(width/2)*ceil(height/2)*4
P210/P212/P216:  width*height*2 + ceil(width/2)*height*4
P410/P412/P416:  width*height*2 + width*height*4

Packed high-bit formats

Y210/Y216:       ceil(width/2) * height * 4 * 2
Y410/V410:       width * height * 4
Y416:            width * height * 8
NOTE: 720x1280 in YUV422_10 is a special case. Please use a (1) calc for this.

Aligned frame size and seeking

(1) Aligned Frame Size = <size of video file> / <% Number of Image>

(2) Aligned Frame Size = (((<Frame Size> + <% Video Alignment> - 1) / <% Video Alignment>) * <% Video Alignment>

Seek to a frame offset = <required frame number> * <Aligned Frame Size>
NOTE: the math above MUST be integer to work properly.

↑ Back to top

4 Samples

calendar_01.hdr

-snip------------------------------------------------------------
% Color Format (YUV420, YUV422, YUV422P, YUV422_10, RGB, RGBA)
YUV422P  % Planar 422
% Image Size (NbRows,NbCols) (576 720, 486 720, 720 1280, 1080 1920)
576 720  % PAL Size
% Number of Fields per Image
1   % Both fields in single frame group
% Number of Image
12   % If greater than actual, reader will correct
% Frame per second
25   % PAL rate
%% Video Alignment
%512  % no alignment
%% Video Name
% optional e:\calendar_01.yuv
% Audio Format
%Stereo  % none, should be wav or aiff
% timecode
00:00:00;00 D
% userbits
FFFFFFFF
% Start Frame
0
-snip------------------------------------------------------------

table_tennis.hdr (MPEG test stream)

-snip------------------------------------------------------------
% Color Format
YUV420
% Image Size (NbRows,NbCols)
288 352
% Number of Fields per Image
1
% Number of Image
250
% Frame per second
25
-snip------------------------------------------------------------

↑ Back to top

5 Optional DNxHR Descriptors

The Color Format tokens dnrh, dnrs, dnrl, dnrq and dnrA identify DNxHR HQ, SQ, LB, HQX and 444 respectively. DNx recordings also write an optional descriptor:

% Video Frame Size
917504

This is the encoded byte count before Video Alignment padding. It defines the fixed frame stride for arbitrary DNxHR rasters without guessing from the uncompressed line pitch. Existing descriptors and legacy reads remain unchanged; readers may ignore unrecognized optional descriptors.

↑ Back to top

© 1995–2026 Drastic Technologies Ltd. All Rights Reserved.

www.drastictech.com · hdrfileformat.txt · support at drastictech.com · 416 255 5636

Image

Drastic™ develops professional video software for capture, conversion, streaming, analysis and remote collaboration. Backed by more than three decades of broadcast and post production experience, our standalone applications, automated workflow solutions and OEM developer tools help you work across formats and systems, verify audio and video quality, and move media efficiently from ingest to delivery.

Contact Us

Address: 

523 The Queensway, Suite 201

Toronto, ON

M8Y 1J7, Canada

Phone: +1 (416) 255 5636

Email: sales@drastictech.com

Fax: + 1 (416) 255 8780

Follow us

bluesky facebook

 

© Drastic Technologies Ltd. 2026