The World Of High Definition Tape Formats

**rbinck** · 02-09-2006, 12:25 PM

Our member 1080PsF has submitted a very nice essay on Recording Formats for inclusion on my blog. Anyone interested in HD tape recording should check this essay out:

The World Of High Definition Tape Formats

I would put it here also, but the forum editor does not like tables.

maicaw · 02-09-2006, 11:03 PM

Quote:

Originally Posted by rbinck

Anyone interested in HD tape recording should check this essay out:
The World Of High Definition Tape Formats

Don't confuse the HD Tape formats which include 45 extra "blanking lines" with each frame - the Broadcast HD formats include none for 720p and 8 extra "dummy lines" (not the same as NTSC blanking lines [VBI]) - for a total of 1088 in the case of 1080i as explained below---- per ATSC --http://www.ATSC.ORG/standards/practices/a_54a.pdf ATSC --
Guide to Use of the ATSC DTV Standard 4 December 2003 -page27

Quote:

5.2.7 ----Number of Lines Encoded
The video coding system requires that the coded picture area has a number of lines that is a multiple of 32 for an interlaced format, and a multiple of 16 for a non-interlaced format. This means that for encoding the 1080-line format, a coder must actually deal with 1088 lines (1088 = 32 x 34). The extra eight lines are in effect “dummy” lines having no content, and the coder designers will choose dummy data that simplifies the implementation. The extra eight lines are always the last eight lines of the encoded image. These dummy lines do not carry useful information, but add little to the data required for transmission.

The following production formats standards are not part of the consumer HDTV formats and do include the extra pixels only used in production - there are multitude of other formats also used by professional videographers in studio and post-production when required.

Quote:

5.2.1.1 Possible Video Inputs
While not required by the Digital Television Standard, there are certain digital television production standards, shown in Table 5.2, that define video formats that relate to compression formats specified by the Standard.
Table 5.2 Standardized Video Input Formats
Video Standard Active Lines Active Samples/ Line Picture Rate
SMPTE 274M-1998 1080 --------- 1920 --------- 24P, 30P, 60I
SMPTE 296M-2001 720--------- 1280--------- 24P, 30P, 60P
SMPTE 293M-2003 483--------- 720 --------- 60P
ITU-R BT. 601-5 483 -------- 720---------- 60I
The compression formats may be derived from one or more appropriate video input formats.It may be anticipated that additional video production standards will be developed in the futurethat extend the number of possible input formats.5.2.1.2 Sampling Rates
For the 1080-line format, with 1125 total lines per frame and 2200 total samples per line, the sampling frequency will be 74.25 MHz for the 30.00 frames per second (fps) frame rate. For the 720-line format, with 750 total lines per frame and 1650 total samples per line, the sampling frequency will be 74.25 MHz for the 60.00 fps frame rate. For the 480-line format using 704
pixels, with 525 total lines per frame and 858 total samples per line, the sampling frequency will be 13.5 MHz for the 59.94 Hz field rate. Note that both 59.94 fps and 60.00 fps are acceptable as frame or field rates for the system. For both the 1080- and 720-line formats, other frame rates, specifically 23.976, 24.00, 29.97, and 30.00 fps rates are acceptable as input to the system. The sample frequency will be either 74.25 MHz (for 24.00 and 30.00 fps) or 74.25/1.001 MHz for the other rates. The number of ATSC Guide to Use of the ATSC DTV Standard 4 December 2003 25 total samples per line is the same for either of the paired picture rates. See SMPTE 274M and SMPTE 296M.
The six frame rates noted are the only allowed frame rates for the Digital Television Standard. In this document, references to 24 fps include both 23.976 and 24.00 fps, references to 30 fps include both 29.97 and 30.00 fps, and references to 60 fps include both 59.94 and 60.00 fps. For the 480-line format, there may be 704 or 640 pixels in the active line. The interlaced
formats are based on ITU-R BT. 601-5; the progressive formats are based on SMPTE 294M. If the input is based on ITU-R BT. 601-5 or SMPTE 294M, it will have 483 or more active lines with 720 pixels in the active line. Only 480 of these active lines are encoded. The lines to be encoded should be lines 23–262 and 286–525 for 480I and lines 45–524 for 480P, as specified in
SMPTE Recommended Practice RP-202, “Video Alignment for MPEG Coding.” Only 704 of the 720 pixels are used for encoding; the first eight and the last eight are dropped. The 480-line, 640 pixel picture format is not related to any current video production format. It does correspond to the IBM VGA graphics format and may be used with ITU-R BT. 601-5 sources by using appropriate resampling techniques.

-in contrast to posts in this forum Note the use of 24, 60 and 30fps shorthand by the ATSC to refer to all the frame rates in their official document
Also reread these posts in light of this information http://www.highdefforum.com/showpost...8&postcount=35
and posts #46 to 55 of this thread http://www.highdefforum.com/showthre...847#post105847

1080PsF · 02-10-2006, 10:02 PM

Hey maicaw maybe you can help me with this.
I was reading the Recommended Practice: Guide to the Use of the ATSC Digital Television Standard, 4 December 2003, and I noticed a few things that I don’t understand.
BTW I never said that I knew anything about transmitting or the broadcast of a signal I only know about the recording side of video.

Quote:

5.2.2 Precision of Samples
Samples are typically obtained using analog-to-digital converter circuits with 10-bit precision. After studio processing, the various luminance and chrominance samples will typically be represented using 8 or 10 bits per sample for luminance and 8 bits per sample for each chrominance component. The limit of precision of the MPEG-2 Main Profile is 8 bits per sample for each of the luminance and chrominance components.

Does this mean that HD broadcast is 8-bit not 10-bit?

The reason that I ask is if it is 8-bit then this next statement doesn’t make and sense to me.

Quote:

4. OVERVIEW OF THE ATSC DIGITAL TELEVISION SYSTEM
The Digital Television Standard describes a system designed to transmit high quality video and audio and ancillary data over a single 6 MHz channel. The system can deliver about 19 Mbps in a 6 MHz terrestrial broadcasting channel and about 38 Mbps in a 6 MHz cable Television channel. This means that encoding HD video essence at 1.106 Gbps1 (highest rate progressive input) or 1.244 Gbps2 (highest rate interlaced picture input) requires a bit rate reduction by about a factor of 50 (when the overhead numbers are added, the rates become closer). To achieve this bit rate reduction, the system is designed to be efficient in utilizing available channel capacity by exploiting complex video and audio compression technology.

1 720 × 1280 × 60 × 2 × 10 = 1.105920 Gbps (the 2 represents the factor needed for 4:2:2 color sub-sampling, and the 10 is for 10-bit systems)
2 1080 × 1920 × 30 × 2 × 10 = 1.244160 Gbps (the 2 represents the factor needed for 4:2:2 color sub-sampling, and the 10 is for 10-bit systems)

If broadcast is only 8-bit as stated in 5.2.2 then these numbers would be a lot less.

720 x 1280 x 60 x 2 x 8 = 884,736,000 or 884.736 Mbps not 1.105920 Gbps

1080 x 1920 x 30 x 2 x 8 = 995,328,000 or 995.328 Mbps not 1.244160 Gbps

Quote:

5.2.7 Number of Lines Encoded
The video coding system requires that the coded picture area has a number of lines that is a multiple of 32 for an interlaced format, and a multiple of 16 for a non-interlaced format. This means that for encoding the 1080-line format, a coder must actually deal with 1088 lines (1088 = 32 x 34). The extra eight lines are in effect “dummy” lines having no content, and the coder designers will choose dummy data that simplifies the implementation. The extra eight lines are always the last eight lines of the encoded image. These dummy lines do not carry useful information, but add little to the data required for transmission.

Now this again would change the bit rate for the 1080i signal.

1088 x 1920 x 30 x 2 x 8 = 1,002,700,800 or 1.0027 Gbps not 1.244160 Gbps

Quote:

5.1.1 MPEG-2 Levels and Profiles
The MPEG-2 specification is organized into a system of profiles and levels, so that applications can ensure interoperability by using equipment and processing that adhere to a common set of coding tools and parameters.3 The Digital Television Standard is based on the MPEG-2 Main Profile. The Main Profile includes three types of frames for prediction (I-frames, P-frames, and B-frames), and an organization of luminance and chrominance samples (designated 4:2:0) within the frame. The Main Profile does not include a scalable algorithm, where scalability implies that a subset of the compressed data can be decoded without decoding the entire data stream. The High Level includes formats with up to 1152 active lines and up to 1920 samples per active line, and for the Main Profile is limited to a compressed data rate of no more than 80 Mbps. The parameters specified by the Digital Television Standard represent specific choices within these constraints.

Is this statement true?

Because 4:2:0 means every other line has no chrominance just luminance, but in the footnote in section 4 it states that the system is 4:2:2.

Quote:

1 720 × 1280 × 60 × 2 × 10 = 1.105920 Gbps (the 2 represents the factor needed for 4:2:2 color sub-sampling, and the 10 is for 10-bit systems)
2 1080 × 1920 × 30 × 2 × 10 = 1.244160 Gbps (the 2 represents the factor needed for 4:2:2 color sub-sampling, and the 10 is for 10-bit systems)

Quote:

5.2.4 Film Mode
Material originated at 24 frames per second, such as that shot on film, is typically converted to 30 or 60 frame-per-second video for broadcast. In the case of 30 fps interlaced television, this means that each four frames of film are converted to ten fields, or five frames of video. In the case of 60 fps progressive-scan television, each four frames of film are converted into ten frames of video. This conversion is done using the so-called 3:2 pull-down sequence; prior to the introduction of 24P video equipment it was an inherent part of the telecine process. In the 3:2 pull-down, the first frame of film is converted to two pictures (frames or fields, depending on whether the output format is 60P or 30I respectively). The second frame is converted to three pictures, the third to two pictures and the fourth to three pictures. When describing the sequence, the film frames are conventionally labeled A, B, C and D; the video fields or frames 1-5 (interlaced) or 1-10 (progressive). In the interlaced case, the third field generated from film frame A is field 1 of Frame 3; the third field generated from film frame C is field 2 of Frame 5. Note that in the interlaced case, Frame 3 will contain video from film Frames B and C and Frame 4 will contain video from film Frames C and D. It is inefficient to code these sequences directly; not only is there a great deal of repeated information, but in interlace, Frames 3 and 4 each contain fields from two different film frames, so there may be motion differences between the two fields. MPEG therefore provides tools specifically for coding these sequences; these are top_field_first and repeat_first_field (see 13818-2, clauses 6.2.3.1 and 6.3.10) It is relatively straightforward for the encoder to detect the repeated frames in progressive scan video derived from 24 fps material. It is less straightforward to detect the repeated fields in interlaced video. Particularly with interlaced material, it is important that the 3:2 pull-down sequence be maintained; if it is not, encoder efficiency and picture quality may suffer. For this reason, it is becoming more common for material to be edited at 24 Hz before frame-rate conversion to 30I or 60P.

Now this is in my area and the statement is not true. The A frame is a 2, the B frame is a 3, the C frame is a 2, and the D frame is a 3.
It would look like this.

Film frames - each letter is a frame of film.

A * B * C * D and you repeat this sequence six times in one second you get 24 frames per second.

The 3:2 pull-down would make video fields or frames look like this.

AA BB BC CD DD and you repeat this sequence six times in one second you get 30 frames per second.

In 30-frame video each letter is a field, 2 fields equal a frame.
In 60-frame video each letter is a frame.

So in NDFTC zeros and fives are “A” frames.

So if you could help me out I would appreciate it. Thanks in advance, Dan

maicaw · 02-11-2006, 12:29 AM

-I think I see some of the answers in the document - - and a dialog would be interesting to others besides you I assume -I'll try to decipher the document to my satisfaction - if you promise to stop the personal attacks - calling me a "pussy" and accusing me of "pulling facts out of my butt" in your vulgar and abusive PM's to me-Your move

1080PsF · 02-11-2006, 01:20 AM

I am sorry and I have stopped. BTW I only set you one PM. And I won't send anymore. When I asked before about dumbing it down for me I wasn't joking or being snide when you wright things it goes over my head a lot of times.

maicaw · 02-11-2006, 11:49 AM

Quote:

Originally Posted by 1080PsF

Hey maicaw maybe you can help me with this.
I was reading the Recommended Practice: Guide to the Use of the ATSC Digital Television Standard, 4 December 2003, and I noticed a few things that I don’t understand. ...So if you could help me out I would appreciate it. Thanks in advance, Dan

To begin with - others on this forum are better qualified than me to answer your questions about the limits of MPEG-2 as it relates to HDTV broadcasting - however I can offer my (simplified) concept and opinions nevertheless. Since 'Recommended Practice: Guide to the Use of the ATSC Digital Television Standard' is a tutorial - This source document which is referred to many times - including in your quoted passages - ISO/IEC 13818-2: 1995 (E) is the technical description of MPEG-2 - for this application - which in it's excruciating detail - may be more useful - the following sections make fairly easy reading and might be everything you want to know - my concepts would be based on simpler explanations from less technical sources

Quote:

Introduction...................................... .................................................. ........................... ................................ v
1 Purpose .................................................. .................................................. .................. ........................... v
2 Application .................................................. .................................................. .............. ......................... v
3 Profiles and levels............................................ .................................................. ............ ....................... v
4 The scalable and the non-scalable syntax .................................................. .......................................... vi
1 Scope .................................................. .................................................. .................... ................................... 1
2 Normative references........................................ .................................................. ...................................... 1
3 Definitions....................................... .................................................. ......................... ................................. 3
6 Video bitstream syntax and semantics .................................................. ................................................. 16
6.1 Structure of coded video data .................................................. .................................................. ...... 16
6.2 Video bitstream syntax .................................................. .................................................. ................ 29
6.3 Video bitstream semantics .................................................. .................................................. ........... 45
E.2 Permissible layer combinations...................................... .................................................. ..............209
Annex F .................................................. .................................................. .................................................2 34

some of the other sections are way beyond my math and software abilities -but I find them instructive anyway for understanding the amazing scope of the MPEG-2 technology to accurately transport so many useful auxiliary parameters - in addition to the "HD Video essence" -in one DTV bitstream while compressing it to 2% of the original

02-09-2006, 12:25 PM	#1
rbinck Administrator Join Date: Apr 2004 Location: Katy, Texas Posts: 12,338	The World Of High Definition Tape Formats Our member 1080PsF has submitted a very nice essay on Recording Formats for inclusion on my blog. Anyone interested in HD tape recording should check this essay out: The World Of High Definition Tape Formats I would put it here also, but the forum editor does not like tables. __________________ www.highdefinitionblog.com www.BlogOfHawaii.com

02-11-2006, 12:29 AM	#4
maicaw Crabtree's Bludgeon Join Date: Apr 2005 Posts: 2,001	-I think I see some of the answers in the document - - and a dialog would be interesting to others besides you I assume -I'll try to decipher the document to my satisfaction - if you promise to stop the personal attacks - calling me a "pussy" and accusing me of "pulling facts out of my butt" in your vulgar and abusive PM's to me-Your move Last edited by maicaw; 02-11-2006 at 12:52 AM.

02-11-2006, 01:20 AM	#5
1080PsF Compression Sucks Join Date: Jan 2006 Location: Monrovia, CA. Posts: 451	I am sorry and I have stopped. BTW I only set you one PM. And I won't send anymore. When I asked before about dumbing it down for me I wasn't joking or being snide when you wright things it goes over my head a lot of times.