Flash XDR

[David Heath]

Originally Posted by Piotr Wozniacki

Just to correct your last sentence David - I guess you meant this:

" if an HD-SDI link will carry 1080i/60, it should also carry 1080p/30."

Am i right? :)

No, though there may be some debate depending which part of the world you're in, different territories have different standard setting bodies. In Europe the EBU defined a few years ago that standards should be written in the form ab/c, where a=no active lines, b=scan method (progressive, interlace, or psf) and c=no of FRAMES/sec. See http://www.ebu.ch/en/technical/trev/trev_home.html and click on article 301: Editorial: HDTV format wars, third paragraph.

They decided there was merit in the final no always corresponding to frames, and not frames for progressive systems, but fields for interlace. Hence the two numbers simply define the amount of information without reference to the scanning system.

I think the forward slash is supposed to denote that the new system is being referred to - hence old nomenclature is 1080i60, new standard is 1080i/30, but 1080i/60 infers a 1080 line system with 60 frames, 120 fields/sec.

So no, I could have put 1080i60, but I didn't mean 1080i/60. :-)

[David Lorente]

Originally Posted by Mike Thomann

...

Bandwidth calculation:
Resolution: 1920×1080 = 2200×1125 with blanking lines
Color depth: 10 bits/point/color plane = 30 bits/point
Sampling: 4:2:2 = 2/3 bit/point

(2200×1125 pixles) x (30 bit/point) x (30 fps) x (2/3 bit/point) = 1485000000 bps
1485000000 bps = 1.485 Gbps

That leaves exactly 0 kbps for 2 channels of uncompressed audio! So now what? Help please...

That's not how SDI works.

In analog interfaces the signal must carry the vertical and horizontal sync signals as well as blanking intervals. These lay before and after each active part of each line (active part is the portion of the signal which effectively contains picture signal).

But in digital interfaces, there is no need to digitize the sync and blanking portions of the signal. Their values are always known, and they are of no use while the video remains digital, so in order to save space, only the active samples are stored. If the video must be converted back to an analog interface, the sync and blanking will be generated and interleaved with the active video by the DAC.

But it is impractical to transmit just these active samples, because this would complicate the design of DACs, so it is more convenient to 'leave a gap' between the EAV (end of active video) of each line and the SAV (start of active video) of the following line. The 'duration' (or number of bits) of the gap correspond, as you are guessing, to the duration of the sync and blanking of each line.

And, of course, to leave something unused is against an engineer's religion, so they quickly developed methods to take profit of these gaps and fill them with ancillary data, that is, any kind of data that could fit in the remaining bandwidth, such as audio, captions, test signals, error correction, metadata, etc.

So, of these 1.485 Gb/s, just 1.244 are used for video. That leaves enough bandwidth to accomodate 48 channels of uncompressed 24 bits, 192 KHz digital audio!

[William Urschel]

Excuse me, off topic, another new device using drives instead of flash, and JPEG2000. The following publicity explicitly states that the "Elite HD" supports the EX-1!

http://www.ffv.com/releases/080325.htm

Except for info in that link, the device is apparently not available for view until NAB (as a prototype?).

We all knew something like these devices would be coming - as for myself, like one other responder to the above link, I may just wait for Cineform SOLID.

[Mike Schell]

Originally Posted by David Heath

No, though there may be some debate depending which part of the world you're in, different territories have different standard setting bodies. In Europe the EBU defined a few years ago that standards should be written in the form ab/c, where a=no active lines, b=scan method (progressive, interlace, or psf) and c=no of FRAMES/sec. See http://www.ebu.ch/en/technical/trev/trev_home.html and click on article 301: Editorial: HDTV format wars, third paragraph.

They decided there was merit in the final no always corresponding to frames, and not frames for progressive systems, but fields for interlace. Hence the two numbers simply define the amount of information without reference to the scanning system.

I think the forward slash is supposed to denote that the new system is being referred to - hence old nomenclature is 1080i60, new standard is 1080i/30, but 1080i/60 infers a 1080 line system with 60 frames, 120 fields/sec.

So no, I could have put 1080i60, but I didn't mean 1080i/60. :-)

Hi David-
This can be very confusing. However, most people specify the number of fields when describing interlaced formats and the number of frames in progressive formats. The Sony EX1 brochure specifies 1080/59.94i (59.94 fields per second) and 1080/29.97p (29.97 frames per second). That's the "standard" we followed on the Flash XDR specification. But. if in doubt it's always best to ask.