Anamorphic composition and render

[ghoraa]

Hi Guys,

I have been scouring the web for this but unfortunately did not seem to get a good understanding so I have to resort to posting my question and will be extremely thankful if you could help out.

Basically, I have to deliver a video for broadcast. Here are their format requirements:

output format: Quicktime.mov IMX50 anamoprhic, Pal, Video encode: D-10/ IMX, Audio Encode: PCM
frame: rate: 25
frame rate mode: Progressive
resize: Active Picture: w- 720 h- 576, Frame Size: w- 720 h-608

Encode profile 50mbps Mpeg2 4:2:2@ML

Audio Profile GOP Structure I-Picture Only

This is my first time of delivering for broadcast and im panicking.

My composition settings are: HDTV 1920x1080, 25fps, Square Pixels.
When I set my composition settings to DV Pal Widescreen my image gets squashed to 4:3…
Could you please have a look at the attached pictures whether my composition settings are correct (picture 3) .
The composition that I get is 4:3, should I render it as it is (picture 1) or scale it so it is widescreen (picture 2).

I do not fully understand what anamorphic is so could you please advise which one of these will be correct according to the resolution delivery requirements mentioned earlier. Which one of these is the correct 720x576 anamorphic and will be good for delivery?

Thank you

[smrpix]

Picture 1 is correct. Anamorphic (literally greek "without shape") means non-square or "squished" is required. In picture 3 your resolution should be full.

[ghoraa]

Thank very much you for your reply.
Does it meen that it will display squished on the non-widescreen TV's and stretch out to normal size on the widescreen ones?

Thank you

[smrpix]

Not if they broadcast it correctly. That's the broadcaster's responsibility.

edit: and despite my use of "squished" earlier, "squeezed" is the more correct term.

[ghoraa]

Thank you very much. Picture 1 it is then. Thank you for clarifying this to me as my head is already hurting from reading about anamorphic and not understanding it

[smrpix]

Always looking for an excuse to link to one of my favorite sites: http://www.widescreenmuseum.com/widescreen/squeeze.htm

Good Luck.

[ghoraa]

another quick question if you have a spare minute. what does : Active Picture: w- 720 h- 576, Frame Size: w- 720 h-608 mean. Does it mean i need to add some extra pixels to the composition to make it h-608?

[ghoraa]

thank you for the link, a very good and straightforward explanation of the anamorphic.

[Cornucopia]

No, it means the person providing that spec may know less about formats than you do.

Unless I'm seriously mistaken, that part is a typo or mishmash of other sections of the spec. Notice that they also included "audio profile gop structure i picture only". Ahem! Since when does audio follow a gop structure, much less have pictures?!!

You need to talk to a real techie/engineer there, not some marketing person. And let the tech know that this is the kind of BS that is being suggested to their vendors.

Just look up the standard IMX 50Mbps format specs and go with those.

Scott

[Cornucopia]

Followup: The IMX tape formats will capture beyond the standard ITU/DVD MPEG D1 sizes, because they want to retain as much of the analog signal as possible.
Analog has 525 lines (NTSC) or 625 lines (PAL). This includes both the active picture area and the VBI (vertical blanking interval).
IMX only SHOWS the active picture, which is standard D1 (480 or 576 lines, respectively), but it records both active and some, though not all, of the vbi into the digitized frame. This frame size is 512 or 608 lines, respectively. That means that of the possible 22.5 or 24.5 possible vbi lines (above & below), IMX only captures 16 lines above & below. So closed captioning, timecode, and similar metadata can be retained, while the rest of the (missing) lines are just generated from scratch on output to analog.
But that says nothing about how IMX tape interacts with digital mpeg files...
One could say to create the exact frame size by making a standard D1 and just pad with black in the vbi (for all intents and purposes, you would be "letterboxing", though it should not affect the active picture's DAR), giving you an extra large frame. You could just as easily say to give only the standard D1 size and let the hardware generate the remainder.
Similarly in the other direction, some systems might get a 512/608 SDI data stream and strip out all but the D1 active picute data (since you have mod16 borders it is fairly trivial to remove). Other systems might retain the whole 512/608 stream and rely on the playback/edit software/hardware to remove the vbi.

Since I'm fairly certain from your post that you don't have embedded vbi metadata, it doesn't make sense to generate that larger picture size if you don't have to. However, you might have to.
That's why you really need to talk to an engineer. Same thing about the audio - do they really want all 8 channels accounted for from the spec?

Scott