Creating a Pan and Scan DVD. How???

[SaSi]

I think the title is self explanatory.

Is anyone of the more experienced people around able to explain the method of creating a pan and scan DVD?

Just the basic steps. When (which stage: encoding, authoring?) is the p&s info inserted to the streams, and how (which tool: the encoder, another tool, the authoring tool?) is it done.

I work with Scenarist and Mainconcept mainly. I can also play around with DVDMaestro.

[FulciLives]

STOP IT

All movies or films or TV shows or whatever SHOULD be watched in the OAR

Original Aspect Ratio

[SaSi]

OK, I agree. Totally...

However, Pan and Scan sounds like interesting and I'm curious by nature.

Just wish to know.

Sure, this ain't bad, is it ?

BTW, I recently rented a copy of the old classic, The Eagle has landed. It was a miserable production where the movie was cropped into 4:3 with the beginning titles and ending credits in anamorphic shown in 4:3, really stupid.

[Jolard]

First of all let me say I agree P&S is crap.

Now, as for actually destroying a movie by cutting huge amounts of it off, , you would need to do it either at the editing or encoding stage. For a real cheap version of this, you could simply use the cropping commands in TMPGEnc to chop of the sides to make it 4:3.

However to actually do a pan N scan properly (picking the most relevent sections of all scenes, scanning across the image if necessary etc) you would need to do it at the editing stage, in a program like Avid or Adobe Premier. And I bet it wouldn't be all that much fun. So it would be part of your original avi before you transcode.

[Vejita-sama]

I think we can all agree that P&S is crap. That's not the point/question. For a lot of us this is a hobby and we want to know/be able to do anything with digital media. Yes, even producded P&S DVDs. From an authoring standpoint it's a challenge/fun.

I mean what's the real point in making a motion menu for you home made DVDs? Why spend hours setting up 4-7 freeware tools each with their own GUIs (if any)? We mostly do this because it's fun.

For example, can I make an anamorphic DVD with a choice between full screen (P&S) or letterboxed (from one video source)? There's no reason to do this, but it's often fun to try/learn how

[JamesB69]

I believe he is asking how to insert Pan&Scan vectors into a 16:9 anamorphic dvd mpeg2. So then he could choose 16:9PS/LB in Maestro so each person can choose on their own player how to view the movie.

I do not know how to put these vectors into the video stream, but would love to learn if anybody else knows how.

[SaSi]

Interesting responses. From these, I gather that the original AVI must be "annotated" within Premier or similar with P&S vectors. Will look into it and dig Premier further. I thought Premier was about cut/paste and transitions.

[JamesB69]

anybody learned how to put inthe P&S vectors yet?

[gmatov]

SaSi,
I don't know if I'm qualified to be here, but, the one movie I converted to, I don't know, would full screen pass as p&s (?), since I hadn't watched it in LB yet, looked good in full screen, but all the opening credits were cut way down, only half a name showing, etc.

Maybe you are looking for a way to cut and edit or something, so that you choose which parts are full (I don't know if they are the same thing, sorry ), and which parts are LB.

Much as I prefer full screen, (only a 25" TV ), I don't think I'll try that again.

If this is incomprehensibe, just ignore me.

[SaSi]

anybody learned how to put inthe P&S vectors yet?

Haven't learned how yet, but now I know that it's something done before authoring and probably before encoding.

Still looking around for further info.

[housepig]

Sasi -

are you sure this is done by some coding method? I thought that pan & scan versions were actually re-edits done to the original source footage, not an on-the-fly playback choice.

and the more I think about it, the more I doubt it - if that was the case, then every dvd would have both versions - the p&s codes wouldn't add much headroom, as they'd be an instruction set, and the process of choosing and noting the panning and scanning would be what costs the most money.

so if the widescreen is there, buried under the p&s coding, a) wouldn't someone have found a hack for it to "unlock" the wide version, and b) wouldn't the cost of enabling the choice of p&s on a disc be offset by the amount of increased sales to people like me who will by a w/s version, but not a p&s?

[JamesB69]

"If the source video used for the disc is anamorphic (vertically expanded to fit a 16:9 aspect ratio onto a 4:3 video source) it may be viewed as a full-screen image (not letterboxed) on a wide-screen television. If the same content is played back on a standard 4:3 television, the user can select between vertically expanded (everyone looks tall and narrow), letterboxed (black bars appear at the top and bottom of the screen), or pan/scan (the image is full screen but the sides are cropped and the viewable area pans and scans the action). In order for the player to correctly pan/scan the MPEG-2 anamorphic image across the screen, the pan/scan vectors must be encoded into the data stream. Pan/scan vectors can either be created by a producer or imported with a source file, and are encoded into the data stream during authoring"

http://www.motionlink.com.au/pages/dvd_info/features_func.html

Now does it make sense? I still want to know how to add these vectors.

[dcsos]

Pan and Scan is done by hand during the FILM to TAPE transfer
This is the machine used to do ithttp://www.davsys.com/products.htm
Costs half a MILLION DOLLARS

skilled persons (colourists) watching the screen,
move a cursor on the contol console as the film is transferrred
to VIDEOTAPE(RANK CINETEL $250,000)
or HARD DRIVE (PHILIPSSPIRIT DATA CINE $$400,000)

Don't try this at home folks..
I think there are automated tools that do this in POST (after a full transfer is done..what this does is import data the DIRECTOR logged during the letterbox transfer and input it to an editing system * BUT WHY?)

Don't you want the director there, not some machine deciding what detail is in or out of safe action areas?
This sorta is like coloring Black and White films


*Avid does this post automation
Some feature films, transferred as 16x9, require a 4x3 Pan & Scan. We can import the telecine controller's shot change list to AVID Symphony for pan/scan data. This removes the need to the Symphony editor to manually locate each shot change.

ps there's no "LOCKED out" information or extra stuff on a 4x3 release that you're not seeing

[housepig]

ps there's no "LOCKED out" information or extra stuff on a 4x3 release that you're not seeing

I didn't think there was... I was actually using that as an argument against the idea that pan & scan was a set of playback instructions, rather than a reedited video track.

[Gazorgan]

You realize that do do a P/S you have to move a croping box through each frame of video. Some scenes it won't move, but other's it will be moving a lot (introducing more latteral motion to the viewed video).

For a full movie this is beyond the means of a single individual. You talking in excess of 170000 film frames for a 2 hour movie. And it's all subjective what to crop. Granted, proper software will go help (ie in fram 2000 x-offset is -20, in frame 2072 x offset is -30, use linear progression). It's done for certain movies shown on TV, other than that use letterbox.

[housepig]

For a full movie this is beyond the means of a single individual. You talking in excess of 170000 film frames for a 2 hour movie. And it's all subjective what to crop.

for full on, pro encoded results, I'd agree.

however, with some not-super-expensive prosumer video editing software, this stuff has become a lot easier.

most of the new editors (I'm thinking the latest versions of Vegas and Premiere) have added or streamlined their panning, cropping and motion tracking functions, because everyone wants to rock the Ken Burns "scrolling and zooming around the still pic" action.

So in Vegas, if I had a two-shot of two people talking in the front seat of a car, person A finishes talking. I set a keyframe, scrub through the video to where B starts talking, drag my pan over to it, set a keyframe. Vegas does all the motion inbetween my start point and my end point, so I don't have to go to every frame and re-frame it.

So it would still be super tedious, but not impossible.

[Cornucopia]

dcos, that's true if you're telecining for videotape or for simple 4:3 full-screen DVD (no choice given). If, however, you are starting with 16:9 and you want to retain 16:9 functionality for those who can make use of it on DVD, what you described would not be the way it would be done.

(This is garnered from Spruce and Minerva manuals and from experience)

Pan & Scan vectors are applied to an anamorphic (16:9) video during encoding to MPEG2. That's after telecine and before authoring. They are either incorporated directly into the encoder, or imported from a script. HARDLY ANY encoders are capable of doing this. From what I can gather, only Scenerist's hardware encoder and Sony's hardware encoder incorporate this function. This function then becomes a DVD player PLAYBACK function (since it is switchable, depending on what kind of monitor you have set).
This should NOT be used unless you're specifically able to follow through with all of the coding requirements. If P&S is selected for the encoding mode, and P&S vectors haven't been inputted correctly, the visible 4:3 window will be unpredictable. Some might show the center portion, some like Maestro's encoder will show from the top left corner.

HTH,
Scott

[JamesB69]

dcos, that's true if you're telecining for videotape or for simple 4:3 full-screen DVD (no choice given). If, however, you are starting with 16:9 and you want to retain 16:9 functionality for those who can make use of it on DVD, what you described would not be the way it would be done.

(This is garnered from Spruce and Minerva manuals and from experience)

Pan & Scan vectors are applied to an anamorphic (16:9) video during encoding to MPEG2. That's after telecine and before authoring. They are either incorporated directly into the encoder, or imported from a script. HARDLY ANY encoders are capable of doing this. From what I can gather, only Scenerist's hardware encoder and Sony's hardware encoder incorporate this function. This function then becomes a DVD player PLAYBACK function (since it is switchable, depending on what kind of monitor you have set).
This should NOT be used unless you're specifically able to follow through with all of the coding requirements. If P&S is selected for the encoding mode, and P&S vectors haven't been inputted correctly, the visible 4:3 window will be unpredictable. Some might show the center portion, some like Maestro's encoder will show from the top left corner.

HTH,
Scott

I was beginning to think I was the only one that understood it that way. http://www.pixeltools.com/MpegRepair.html

This tool called MpegRepair claims to put in the vectors after encoding.

Here is the quote from the site "Pan & Scan
For DVD producers, whose MPEG Encoder does not insert Pan & Scan info into their MPEG Stream, MPEGRepair can inspect the stream and, if requested, it can insert the Pan & Scan info into their MPEG stream without a need to re-encode."

Sounds like what we are looking for although doesnt look like they have a trial or demo available for download.

[dcsos]

MPEGRepair can inspect the stream and, if requested, it can insert the Pan & Scan info into their MPEG stream without a need to re-encode.

Yes it can re-insert the DATA generated by hand during the film to tape transfer..
That is, If your sytem costing a cool million* can insert PAN And SCAN vector into an MPEG stream, this tool can reinsert these vectors because it DESTROYS THEM in the process of "REPAIRING" your MPEG

*as CORNUCOPIA added, VIZARRO (sony) and SCENERIST support this in the MPEG-2 encoding stage.That'll cost ya less but not much!

[SaSi]

Hey guys,

I've been away for some time and doing some catching-up.

Interesting comments on the P&S question.

I've been doing some reading, the MPEG-2 specs actually, and this is what I learned.

The Pan and scan "trick" is actually about "vectors" alowing the decoder to display part of the 16:9 frame on a 4:3 display.

This is pure MPEG-2 functionality and requires the presence of "picture display extensions" in the stream. Typically, one such extension can follow each picture header, allowing the decoder to reposition the display window over the actual frame.

I've seen that CCE SP offers P&S info addition to the encoded stream, but have not tried to see if it just inserts empty extension blocks.

Manually coding the P&S info is one way that is certainly too tedious.

I am thinking that this can get automated in the following way:

P&S is supposed to show on display the "center of interest" of the film. Typically, this is where most o the action is. Typically, where most of the action is, this is where motion estimation vectors take their maximum values.

So, if a program scans the video stream, decodes the motion vectors for each frame and thus identifies the center of motion, it can easily generate a list of "center-of-motion" centers for the frames. This sequence, smoothed a bit to avoid jumping Panning, can be used as the information to be inserted. To fine-tune this, one could view the video with an appropriate editor and "move" the P&S window when it goes out of action (I expect this to happen in little action, semi-idle scenes).

I am doing some programming work decoding MPEG-2 video to discover what it is made of, and I might try to add this functionality, just for fun...