[Moogs]

So I convinced my boss to let me pursue a couple of DVD productions to use as revenue boosters for the seminars and group instruction sessions that we schedule each year.

I want to convey the finer points of digital imaging, using a combination of DV footage, screen videos and a voice-over track. During the splash sequence of the DVD we will have some Motion graphics sequences, as well as during chapter transitions and the credits sequence. And of course there will be a soundtrack, to fill the audio during transition periods.

I would like to do a DVD project with chapter menus and the usual items. My reasoning is that this will be a more user-friendly experience than having a Mac video and PC video (on CD or DVD), with the possibility of requiring a QT install, etc. It should work just like a normal commercial DVD. Pop it in your player and it automatically scales everything to fit the screen, you make your menu selection with the DVD controller and that's all it takes to get started. Agree?

Some of my concerns:

• I am considering 4:3 and my plan (to accommodate people with 15" monitors) is to make the screen videos about 1000x750, and then scale them back to 80% (800x600), so that the software GUI being demoed is still legible / recognizable, but there's no funky cut-off on the right and bottom. The palettes will be there (making it look like a "full Desktop"), it will just be a bit smaller scale. Good idea?

Do most PCs with DVD players (because of when they became popular) have screens with a higher resolution like 1024x768? Almost everyone we sell to will have a PC I'm thinking, but I really don't know much about the hardware demographics on that side of the planet. Or do the morons who install windows on their PCs in the factory still default everything to 800x600 (where people don't even realize they can go higher)?

• I will need to compress the QuickTime screen video (currently testing with the QT Animation codec on High quality, at about 10 fps) to MPEG 2. I assume I can do this with either DVDSP or FCP, but what I wonder about is the quality. Is it better to use H.264 (high quality) for the initial screen capture and then compress to MP2, in terms of maintaining sharpness and color... or does it not matter since it's just screen video? I don't want the fuzzy effect I see on other DVDs of this kind.

•Is it possible in FCP or DVDSP to insert instructions that cause QT (or whatever Windows uses to run DVDs) to freeze the video at a certain point, and then start up again on cue, without having a huge number of duplicate frames (in the middle) to mimic the same effect?

I'd like to have the screen video pause at different points and then have another video segment superimposed which scales to cover almost the full video screen from one corner (showing imagery or bullet points for example), and then scale back down into the corner and disappear, allowing the video to continue on from that point.

• I presume SoundTrack Pro comes with all the loops Apple sells (from the Jampacks) and more, right? No need to buy them as extras?


I bet a lot of you have seen these kinds of DVD instructional videos before. What do you like about them / not like about them? Give me some ideas for making a great production. One thing I want to avoid is monotony, which almost all of them are guilty of, but I also want to avoid gratuitous special effects.

I thought for example, about having a small border area around the video (which would come from a Motion project) that has a slow color shift effect. Kind of like the aurora borealis but with very light tones. Just something to make it look sharp, but is not so distracting or wild looking that it pulls your eyes away from the video / menus.

• Sub-chapters. I have several sections which will require more than one screen video sequence, but all of which are directly related and sort of sequential in the real world. Is there a way to have navigational controls outside of the DVD player that would allow you to reply sub-chapters if you didn't get all the details, etc?

Any and all ideas appreciated.

[709]

• I am considering 4:3 and my plan (to accommodate people with 15" monitors) is to make the screen videos about 1000x750, and then scale them back to 80% (800x600), so that the software GUI being demoed is still legible / recognizable, but there's no funky cut-off on the right and bottom. The palettes will be there (making it look like a "full Desktop"), it will just be a bit smaller scale. Good idea?

Any video on a Standard Definition DVD is going to have to end up at 720x480. No two ways about it. If you want to scale a 1000x750 4:3 image, scale it down to 720x540. You'll then have to vertically scale that to 720x480.

An exception to this is if you're capturing NTSC footage. You'll wan't to crop (not scale) that from it's native 720x486 to 720x480. Make sure you move it either up or down 1 pixel, otherwise you'll end up swapping the fields.

If you're absolutely sure this is *only* going to be played on desktops, you can put your text pretty much anywhere you like. If there's a chance it will be played on a CRT TV set you'll have to take into account what's know as 'Safe Title' and 'Safe Action'. Photoshop has a default template that will automatically give you guides for these.

Basically, the outer guides are 'Safe Action', meaning on most TVs your motion will probably show up and not be cut off by the edges of the CRT. The inner guides are 'Safe Title', meaning any text within these guides will show up on most CRTs. Obviously, this will cramp your working space a bit if you've got lots of type.

Do most PCs with DVD players (because of when they became popular) have screens with a higher resolution like 1024x768?

Yes, but it doesn't matter for what you're doing. The DVD will be resized to fit the screen no matter what the resolution.

• I will need to compress the QuickTime screen video (currently testing with the QT Animation codec on High quality, at about 10 fps) to MPEG 2. I assume I can do this with either DVDSP or FCP, but what I wonder about is the quality. Is it better to use H.264 (high quality) for the initial screen capture and then compress to MP2, in terms of maintaining sharpness and color... or does it not matter since it's just screen video? I don't want the fuzzy effect I see on other DVDs of this kind.

Keep all your source video at it's native compression scheme until you compress it to MPEG-2. If you're creating motion graphic sequences the Animation codec on High quality will be just fine.

Why 10 fps? Either way you're going to have to get it to 29.97 eventually. I'd just design at that to begin with.

DVDSP does have it's own MPEG-2 codec. IMO it's at the lower end of the quality scale as far as MPEG-2 compressors go, but it should be fine for what you've got planned. Be sure to compress your stuff with the Two Pass VBR. I'd guess a Target Bitrate of 7 and a Max Bitrate of 8 would suffice. How much footage (ie: length of time including motion) are you planning on using?

•Is it possible in FCP or DVDSP to insert instructions that cause QT (or whatever Windows uses to run DVDs) to freeze the video at a certain point, and then start up again on cue, without having a huge number of duplicate frames (in the middle) to mimic the same effect?

Well, if I'm understanding what I think you want to do correctly, there's an easy way to do this without getting into chapter points and whatnot. Trim your video to end where you want it, and create a still image of the last frame of video. Use this still frame as the background to a menu. The transition will not be noticeable. Then, build an invisible button over the entire menu. A click anywhere on the screen will start playing the other piece of your trimmed video (of course, you have to tell the menu that).

I'd like to have the screen video pause at different points and then have another video segment superimposed which scales to cover almost the full video screen from one corner (showing imagery or bullet points for example), and then scale back down into the corner and disappear, allowing the video to continue on from that point.

'splain further please.

• I presume SoundTrack Pro comes with all the loops Apple sells (from the Jampacks) and more, right? No need to buy them as extras?

SoundTrack comes with a buttload of loops, but as far as I can tell the individual Jam Packs are not included. Not 100% sure on this one.

•I thought for example, about having a small border area around the video (which would come from a Motion project) that has a slow color shift effect. Kind of like the aurora borealis but with very light tones. Just something to make it look sharp, but is not so distracting or wild looking that it pulls your eyes away from the video / menus.

One thing to stay away from in the NTSC world is bright colors. A good rule of thumb is to stay away from full-strength primaries. A bright red will literally make a TV scream. If there's any way you could hook up a TV to preview your design on I'd recommend doing it.

[709]

Originally Posted by 709 Keep all your source video at it's native compression scheme until you compress it to MPEG-2. If you're creating motion graphic sequences the Animation codec on High quality will be just fine.

Actually, I should expand on this a bit.

Let's say you capture a program from DV. It's got a fade up from black and a fade down at the end. You don't futz with it at all. Keep this at it's native DV compression until you're ready to go to MPEG-2.

Let's say you capture a program from DV and bring it in to an NLE. You add a couple of effects, some titles, maybe a even a bug. If you have to export this for some reason (let's say you're going to After Effects or something) I'd export it at the above mentioned Animation setting. This way you're not double-compressing something. Animation is a 'lossless' codec. The file size will be much bigger, but you're not adding an extra layer of compression.

Of course, if you don't have to move your footage out of your NLE to some other program just leave it be...unless you're exporting your MPEG-2 directly out of QuickTime. Then output it at Animation. If you're using FCP you can export directly to MPEG-2.

[709]

Originally Posted by 709 Any video on a Standard Definition DVD is going to have to end up at 720x480. No two ways about it. If you want to scale a 1000x750 4:3 image, scale it down to 720x540. You'll then have to vertically compress that to 720x480.

Goddamn it. OK, one more thing you should know.

If you've got your motion graphics built at 720x540, QuickTime's MPEG-2 compressor will automagically resize it to 720x480 when you run it through. Personally, I prefer other tools to do the resizing, but I thought you should know that you *can* run a 4x3 through QT and have it come out usable.

One caveat. If you're going to rely on QuickTime to do your resizing, render your animation as progressive and not field rendered.

[Moogs]

Man. Thanks for the replies. I want to read through them carefully before I respond. I already have a couple questions about the native DVD resolution you're talking about. The still image thing sounds like a good idea. I will respond a little later once I have more time to address everything.

[709]

No problem. I'm off to dinner, but I'll be back in a few hours.

[curiousuburb]

The elegant "still image + invisible button" technique may have a major downside... an invisible button.

Users who don't click will assume the DVD is broken and its author is an idiot...
even if you tell them they have to click, some won't click and you'll have a dead-end on disc.
Bad.

Ghosting a 15% opacity "click to continue" into the still image might still not be clear enough, but might be the minimum idiot-proofing... consider the blatantly obvious button... TV viewers with DVD remotes instead of mice might be equally confused, depending on the button assignment.

As an alternative, DVDSP has timeout settings which can delay end-jump from menus to next clip, or you could jump to slideshow (with timed refresh as fallback but still manually navigable) then back to video.

Or you could use transitions from the menu with their own timer and end jump settings.
<--click for tutorial/review

Use the built-in Simulator to test all your buttons and jumps until there are no dead ends.

I'm going to echo much of 709's excellent advice on most other points.

Keep source uncompressed or highest quality possible for as long as possible.
That said, I'd generally stay away from H.264 as a distribution codec due to high overhead requirements on the decompressing end.

As for the coloured/morphing/blinking/distracting frame... usually thumbs down.
If your job is to convey the finer points of digital imaging, use quality images, not blinking ones. Communicate your content rather than show off compositing for its own sake or as pixel filler.
Note the last line of your own quote below.

Originally Posted by Moogs I bet a lot of you have seen these kinds of DVD instructional videos before. What do you like about them / not like about them? Give me some ideas for making a great production. One thing I want to avoid is monotony, which almost all of them are guilty of, but I also want to avoid gratuitous special effects.

From the Two Birds with One Stone department:
Inside DVD SP3

For the price, a decent tour of what's possible (tips to help with the actual DVD SP production) and as a bonus, a sample of training DVD product in a form you might find worth mimicking or avoiding, depending on your tastes. Decent overview, technical section on MPEG specs and compression, an interface or the option to watch the videos externally in QT, and sample assets for following along.

Things I didn't like about it? Cheesy sample assets, teleprompter-reading, lackluster instructor who seemed to be filling time in places, too much talking head in general, not enough web linkage.

But as an instructor and trainer, I may be extra demanding about presentation skills.

[Moogs]

Thanks burb: I'll check out those links. I'm not sure I follow your suggestion regarding the transitions and whether they require an invisible menu but let me get to what 709 was saying for a bit while it's fresh in my mind....

709, Hope you fueled up because you're gonna have to burn some calories reading this post (war and peace, etc).

Any video on a Standard Definition DVD is going to have to end up at 720x480. No two ways about it. If you want to scale a 1000x750 4:3 image, scale it down to 720x540. You'll then have to vertically scale that to 720x480.

This is odd because of the instructional DVDs I was testing out earlier in the week, several seemed to use 650x507 when I did a screen capture of the actual video area inside the DVD widow ("normal" size). That doesn't match the aspect ratio of 720x480, or 800x600. Any idea why they would use a goofy format like that?

I figured the DVD player would scale things automatically but I thought I'd have control over the aspect ratio. There must be some way because the resolution you describe is 1.5:1 while widescreen (which many DVDs use, right?) is 1.77:1 or there-bouts. Can / should I do a 16:9 variant instead? I could stack the palettes side to side at right, instead of vertically down the screen. I'm using Snapz for the screen video so I can set it to any ratio and resolution I want.

Keep all your source video at it's native compression scheme until you compress it to MPEG-2. If you're creating motion graphic sequences the Animation codec on High quality will be just fine. Why 10 fps? Either way you're going to have to get it to 29.97 eventually. I'd just design at that to begin with.

Sounds good regarding waiting till the last thing to compress the video.

I used 10 fps because that is the default and it seemed smooth enough (though maybe not). I wanted to keep file sizes down but I suppose I won't be running into any barriers at 4.6 usable GB per disc. Although at 29.97 the file size is going to be 3x larger, so about 72MB/minute of animation by my calculations. I guess the real thing to watch will be the audio size since the speaking will be more or less constant, while the screen videos will only be a minute or two at a time most likely.

I guess this brings up another concern. How to do the voice-over. Either get the script written and read it at a measured pace, then try to match the screen video to its cadence / speed of instruction. Or the reverse, make the screen video so it proceeds without being snail-like, then record the voice over as you run it on screen. I'd err on the side of the first method.

I suppose too that it should be a separate AIF track (acceptable format for DVDSP?) that I just match up to the cue points in the video (a la Director), rather than trying to make each movie with it's own audio track via Snapz (not the ideal way to record I'm guessing).

How much footage (ie: length of time including motion) are you planning on using?

Haven't figured that out yet. Each screen video will be of differing length but I suspect each demo will be between 90 seconds and 3 minutes depending on what it is, and there will be between 15 and 25 demos per DVD, and maybe a total of a minute of Motion time (10 seconds for an intro, 3-5 seconds between chapters / after selecting a chapter from the menu, 20 seconds or so for the credits... something like that.

The DVD that has DV footage will consist of (guessing here because we haven't yet shot it) about 20-30 minutes worth, combined with maybe 7 or 8 demos and maybe some slightly more elaborate Motion stuff.

'splain further please. [regarding superimposed thingy]

Well, I'll try to asplain things better but if my head asplodes it be your fault.

Here's the deal. Picture our 16:9 video or whatever format it ends up being in progress. Screen movie of software process nears a stopping point, voice-over then segues into some "things to keep in mind" (just making this up but you get the idea).

As the narrator says "here are some things to keep in mind before you go any further", the screen video stops (or pauses as I envisioned it). At the bottom left corner a small box appears and quickly scales to top right, covering say 95% of our 16:9 video area. As the box finishes scaling, the narrator says "don't forget: you can't do operation A until condition B is met" and (maybe do all this using Motion?) a text blurb flies in from the left with a sparkling bullet point or whatever. Narrator says the next thing, next blurb flies in from the right, and so on.

When all the "things to remember" are done, the narrator segues back "let's go back and see how this works", and as he does so our bullet point box scales back in the direction from which it came and quickly disappears. Narrator says "let's move the mouse to y" and the video starts up again. My question was how to do that without replicating the last frame before the freeze, a ga-gillion times.

And the button could work except I don't want the user to have to do anything to make this flow correctly.


Regarding the compression: sounds like I should keep everything at the high quality animation compression setting until I'm ready to "compile" the project with FCP or DVDSP? The video footage is either going to come from a Sony DV cam or a D-Snap, which actually records to MPEG2 and is normally not usable on a Mac without FCP or the QT MP2 Component add-on thing. In former case I'm not sure what kind of compression settings or what the camera uses.

[Moogs]

Burb: you might be misunderstanding my idea about the Motion-based "aurora" effect. Actually my idea is a little bit like the screensave Flurry now that I think of it, just not so tentacle-like.

Essentially, picture our 16:9 video space or whatever it is... maybe something slightly smaller to accommodate this idea. Now picture the outer 10% of that space is black, and from time to time you have these very low intensity, slowly morphing "auroras" that kind of appear at random in different parts of this "border".

Inside the video area itself, of course we're using high quality images for our screen demos and such. The aurora idea might actually be a cool thing to use when someone pauses the video, but I don't know how that would be separate from the rest of the DVD content. You couldn't set something up so that when the movie is playing that black border area is just black, but when someone hits Pause on their DVD player console, the aurora effect starts up, or could you?

That's probably something that requires a more Director-like approach, which I don't have the option of for now. Don't want to mess with Lingo. Forgotten too much of it to want to crack open the manuals / dictionaries again. In general, I am very curious how (or if) I could maintain control over the video portions with the DVD controller, and then have the Motion effects be triggered by different events that are only indirectly in the user's control.

For example think of the action menus in Hollywood DVDs. It's sort of like if you mouse over a menu with a movie clip running inside it, sometimes it will pause... but the background of the menu itself keeps having whatever dynamic thing it has going. Sort of like the falling characters in the Matrix DVDs (if I remember right). So on my thing, the video would pause and that would actually trigger the effects that populate the border region of the video...

...almost like a visual cue to the user that their video is paused and/or something to add visual polish when everything is paused.

[curiousuburb]

Originally Posted by Moogs Thanks burb: I'll check out those links. I'm not sure I follow your suggestion regarding the transitions and whether they require an invisible menu but let me get to what 709 was saying for a bit while it's fresh in my mind....

I probably wasn't very clear myself <* squints at prior post *>. Lemme sleep on that.

I think I can answer a couple of these.

Originally Posted by Moogs 709, Hope you fueled up because you're gonna have to burn some calories reading this post (war and peace, etc). This is odd because of the instructional DVDs I was testing out earlier in the week, several seemed to use 650x507 when I did a screen capture of the actual video area inside the DVD widow ("normal" size). That doesn't match the aspect ratio of 720x480, or 800x600. Any idea why they would use a goofy format like that?

Not sure which discs you're referring to having seen, but some older interfaces designed in Director or other apps may have video windows driven by MIAW scripts set to scale resolution mathematically, so while it is odd to see odd numbers in the resolution specs, it can come about from coding idiosyncrasies, re-cropping to mask artifacting or subtitling, or other curious reasons relating to the tools used to build it (electric or organic).

Originally Posted by Moogs I figured the DVD player would scale things automatically but I thought I'd have control over the aspect ratio. There must be some way because the resolution you describe is 1.5:1 while widescreen (which many DVDs use, right?) is 1.77:1 or there-bouts. Can / should I do a 16:9 variant instead? I could stack the palettes side to side at right, instead of vertically down the screen. I'm using Snapz for the screen video so I can set it to any ratio and resolution I want.

16:9 instead of 4:3? Generally no. Aim for the biggest audience, not the bleeding edge.
(Unless you know for a fact the majority of your audience are 16:9)

Be aware that legibility of fonts and small details (including colour) may drop depending on the device (if you get folks viewing on TV). Do the rest get pan & scan? crop? squish? letterbox?

Originally Posted by Moogs Sounds good regarding waiting till the last thing to compress the video. I used 10 fps because that is the default and it seemed smooth enough (though maybe not). I wanted to keep file sizes down but I suppose I won't be running into any barriers at 4.6 usable GB per disc. Although at 29.97 the file size is going to be 3x larger, so about 72MB/minute of animation by my calculations. I guess the real thing to watch will be the audio size since the speaking will be more or less constant, while the screen videos will only be a minute or two at a time most likely. I guess this brings up another concern. How to do the voice-over. Either get the script written and read it at a measured pace, then try to match the screen video to its cadence / speed of instruction. Or the reverse, make the screen video so it proceeds without being snail-like, then record the voice over as you run it on screen. I'd err on the side of the first method. I suppose too that it should be a separate AIF track (acceptable format for DVDSP?) that I just match up to the cue points in the video (a la Director), rather than trying to make each movie with it's own audio track via Snapz (not the ideal way to record I'm guessing).

It is generally easier to do the audio first and get its speed and cadence right, then produce animation (in part because speech can only get tweaked about 20% before it sounds unnatural, but nobody calls you on unexpected frames in your video), but if you're dealing with narration of a captured process, the intuitive pace of the process is priority. Normally the critical factor is maintaining synch with the onscreen content. If you find you need longer discussion of a point, build in a still frame and hold it until your discussion can smoothly blend back to video.

AV for DVD SP needs to be elemental streams rather than multiplexed MPEG.

If your Snapz capture included your narration (and was via a decent mike, not the internal), FCP should spit out identical filenames for the audio and video elements - which DVD SP knows to import in pairs if dragged 'n' dropped into the appropriate window.

As far as cue points within the audio or additional markers, you might test a bit of each method to see if there are efficiencies for the overall project. AIFF import is fine (DVD spec wants 48KHz rather than 44, unless you're going DTS or up to 96) but DVD SP will re-encode as required, and you might find more flexibility with multiple small AIFFs (one per movie) than one long track you try to keep in sync.

I'd suggest doing the capture live in snapz to ballpark a natural speaking tone with which to time your demo. Extract the video track from the result. Add overlay sparkle highlights or arrows to the video (recompressing as little as possible, ideally keeping all the overlay stuff distinct as if an editable layer in motion rather than rerendering). Re-record the audio to the same timing using better mike/software. Match the newly improved audio with the newly improved video.

Originally Posted by moogs Haven't figured that out yet. Each screen video will be of differing length but I suspect each demo will be between 90 seconds and 3 minutes depending on what it is, and there will be between 15 and 25 demos per DVD, and maybe a total of a minute of Motion time (10 seconds for an intro, 3-5 seconds between chapters / after selecting a chapter from the menu, 20 seconds or so for the credits... something like that. The DVD that has DV footage will consist of (guessing here because we haven't yet shot it) about 20-30 minutes worth, combined with maybe 7 or 8 demos and maybe some slightly more elaborate Motion stuff.

Don't forget any time for rendered transitions (per button), slideshows, extra languages/angles.

Originally Posted by moogs Well, I'll try to asplain things better but if my head asplodes it be your fault. Here's the deal. Picture our 16:9 video or whatever format it ends up being in progress. Screen movie of software process nears a stopping point, voice-over then segues into some "things to keep in mind" (just making this up but you get the idea). As the narrator says "here are some things to keep in mind before you go any further", the screen video stops (or pauses as I envisioned it). At the bottom left corner a small box appears and quickly scales to top right, covering say 95% of our 16:9 video area. As the box finishes scaling, the narrator says "don't forget: you can't do operation A until condition B is met" and (maybe do all this using Motion?) a text blurb flies in from the left with a sparkling bullet point or whatever. Narrator says the next thing, next blurb flies in from the right, and so on. When all the "things to remember" are done, the narrator segues back "let's go back and see how this works", and as he does so our bullet point box scales back in the direction from which it came and quickly disappears. Narrator says "let's move the mouse to y" and the video starts up again. My question was how to do that without replicating the last frame before the freeze, a ga-gillion times.

Break it into three clips.
One is the 1st demo and ends with the full screen end of the demo.
Two is bridge, the shrinking demo, the text callouts, clippy the paper clip, obscure formulæ, the count from sesame street, whatever. First frame is full screen of demo 1. End frame is full screen start of demo 2.
Three is demo 2.

This preserves the video structure per lesson with unique clips for the review segments.

not sure if that works for you.

Originally Posted by moogs And the button could work except I don't want the user to have to do anything to make this flow correctly. Regarding the compression: sounds like I should keep everything at the high quality animation compression setting until I'm ready to "compile" the project with FCP or DVDSP? The video footage is either going to come from a Sony DV cam or a D-Snap, which actually records to MPEG2 and is normally not usable on a Mac without FCP or the QT MP2 Component add-on thing. In former case I'm not sure what kind of compression settings or what the camera uses.

Raw DV may be better than relying on the camera's internal compression, but you are correct in the general principle that everything should stay as high quality and uncompressed for as long in the chain as possible.


As to your 2nd post regarding the auroral overlays during paused DVD playback, I think the answer is no. If you were coding it inside Director for playback (which called the MPEG2) you should have enough control over the pause function to have it do whatever you want, including alpha transition loops. Pause on a DVD is just that... still image of the normal content. No special tricks supported by the spec other than Jacket picture which tends to be for multi-disc DVD players anyway.

DVD SP will allow buttons to be looped videos, menus to have looped video backgrounds, and the transition link I posted earlier should explain a few other dynamic options for interactivity once you've got groovy video to play with. I'd say render some aurora clips and play mad scientist to see how they work as menus or loops.

[709]

Sorry, I got tied up dealing with a sick pet last night. I'll be back after I fill the veins with my first pot of coffee.

curiousuburb offers some excellent advice as well. It's going to take some time to get through this all. At only 10 posts already this might end up being the longest one-page thread in the history of AN.

[Moogs]

Truly mammoth, thanks to yours and burb's expertise. I was reading up a good bit today and found some good stuff out but also triggered more questions. Sounds like in general I should compress / transcode everything in compressor before I ever import the files into DVDSP - audio, video, Motion clips, all of it. Just set different parameters for each in the batch window.

One thing that's sort of unclear to me after reading through some of Apple's PDFs: any reason to use LiveType for titles, given the advent of Motion and all it can do with text?

Also, I am thinking of using the screen format of 1440 x 960 for the demos, so that there is no goofy cut-off effect like those in the videos I saw (where basically the top left quarter of the screen was recorded only. Basically then compressor will scrunch it down to exactly half that size, hopefully retaining some of the menu text legibility, etc.

(BTW burb, the videos I referred to with the goofy dimensions? All were by reputable people. Scott Kelby, Will Crockett, etc. I was using a 23" ACD and the std DVD player... nothing fancy. At normal sizes those videos were not reducable to any of the ratios we're talking about. )

Another thing that caught my eye today was that it seems DVDSP can use (without transcoding it to 29.97 fps) DV at the native 24fps. This was a "WTF" moment for me because I thought everything got moved to 29.97, whether for TV or Computer screens. So if I can use 24 fps DV... why not 24 fps screen animations or 24 fps Motion video? *scratching head*

Also there were recommendations not to use interlaced video. Is there a way to control this at the source? Either on DV cameras or in Snapz? There are none that I can see / recall.

[Moogs]

Something D O O economics..... anyone.... Anyone?

[curiousuburb]

DVD SP has an internal compressor/codec package that can accept multiple sources at different frame rates (29.97, 25, 24fps) provided a given project is uniformly PAL or NTSC (don't mix clips... digitized clips at other frame rates can be converted on their way out).

There are potentially plenty of cases where animation can be generated at 24fps or other, lower frame rates (usually to lower file size versus the extra data required for 29.97fps) and it may be more efficient. Depends on your content. When in doubt, leave it out... if you can't see the difference, make it smaller.

Compressor as a standalone program has all the oomph of DVD SP's internal engine, but also offers the ability to Batch multiple renders at customizable settings for overnight handoff.

DVD SP4 now also supports distributed rendering akin to Shake QMaster, so not only can you batch, but the render farm can pick up the slack over the network overnight.

I'm going to be offline for a few days while relocating to N. Ireland.

I'll try and catch up to this thread early next week.

[Moogs]

Hey best of luck with the move. Ironically I am also moving this weekend so there you have it. After today or tomorrow morning I probably won't be on again for about four or five days *gasp*. How will we survive one wonders...

[Moogs]

Couple questions in the interim: I was tinkering with FCP 5 today and so far I love it. One question: any means of removing markers from the timeline without using the marker menu? There seems to be no means of selecting them with the arrow tool and dragging it off and/or right-clicking them and selecting delete.

Also I got the distinct impression from the DVDSP manual that Compressor yields higher quality results than its internal compression utility because you have more control over the parameters, etc. Also I think 709 mentioned that the DVDSP compression engine is not as high quality but anyway the question stands.

As it is my plan is to film, generate the motion stuff, generate the audio and then use compressor before any of them see DVDSP. No?

[Moogs]

All righty. After a week+ of being disconnected from the known universe that is the cyber-intra-web, I am back in action. My new residence actually pulls an extra 150-200 K/s than the old one. Burst speeds are mondo-quick.

So then, Mr. Burb and 709... I'm ready when you are.

[curiousuburb]

know the feeling, having just relocated myself...
now surfing the interweb with a view over the water to the Mull of Kintyre on a clear day...
still awaiting the broadband leprechauns... hopefully later this week.

[709]

And I'm up to my ears in invoicing. Christ. I hate paperwork.

Not to fear Moogs, I've got quite a few things written down and ready to post...give me a day.

[Moogs]

Sounds good boys... if you need a few days, no worries. We've waited this long, no harm in waiting a couple more.

[cooop]

Originally Posted by Moogs Couple questions in the interim: I was tinkering with FCP 5 today and so far I love it. One question: any means of removing markers from the timeline without using the marker menu? There seems to be no means of selecting them with the arrow tool and dragging it off and/or right-clicking them and selecting delete. Also I got the distinct impression from the DVDSP manual that Compressor yields higher quality results than its internal compression utility because you have more control over the parameters, etc. Also I think 709 mentioned that the DVDSP compression engine is not as high quality but anyway the question stands. As it is my plan is to film, generate the motion stuff, generate the audio and then use compressor before any of them see DVDSP. No?

Hi Moogs. I've got a couple of quick answers for you on these points. First, the easiest way to add a marker in FCP is simply pressing 'M' at any point in the timeline. The easiest way to remove the marker (without using the menu command, that is) is navigating to the marker (fairly easy when snapping is enabled) and pressing 'M' again. A dialog will appear that allows you to make the marker a chapter marker (among others) or delete it altogether.

I'm fairly new to Compressor myself, but I received the same impression from Apple's documentation: that is, since you have more control over individual parameters in Compressor, you're likely to get higher-quality output. However, DVD Studio Pro does let you select different encoding modes and bit rates in its Encoding preferences pane, so I think you can safely encode your media with DSP's QuickTime MPEG encoder for most projects. Others may have more to say about this, however.

Hope this helps.

[Moogs]

Originally Posted by cooop Hi Moogs. I've got a couple of quick answers for you... The easiest way to remove the marker is navigating to the marker and pressing 'M' again....

*slaps forehead*

You'd be amazed how many times I can figure something complex out on my own and then when it comes to stuff like this I can bang my head against a wall for an hour and still not figure it out.

Glad to hear you had the same impression about Compressor's capabilities relative to DVDSP.

You did indeed help. Jobu thanks you, and I thank you.... cheers.

[Moogs]

^
^
^
*punt*

[Moogs]

Next interim question: I will have several subtitles / bullet points that are woven through each chapter of the DVD, usually as visual filler for talking points that take place between actual human video or screen movies (Snapz). I gather I can add them to the DVDSP timeline in one of the many subtitle "channels", but what's the best application to make them with? Motion or LiveType?

What are the output files from those apps that I drag into the actual timeline.. QuickTime?

[Moogs]

Because there are not enough questions in this thread, I'd like to add:

You know the DVD tutorial that ships with the suite (Promotional piece for the Blue Horizon surf video)... I was paying close attention to how that thing itself was put together and saw some things I really like about the video clips (and they obviously all fit on one DVD).

Thing is I can't figure out how to replicate the effect. What I like is that when an app is first launched for the demo, you can see the entire thing, but then when they go to perform a specific operation with the cursor, the video not only follows the mouse (a la Snapz), but it selectively pans and zooms in and out at key points so you can read the text. And when they do, the text is nice and legible.

When I tried creating some full screen video from a 23" cinema display, even at 50% scale, the size was ridiculous for a 30 or 40 second clip. Like 1.2GB or something at high color depth, which you need for demos that work on video or images. It seems to me they are also using a high color depth but somehow are getting these clips to both be high quality when zoomed in, but small enough in total to fit on one DVD.

Is there a tool out there better than Snapz for screen video? Any of the tools that ship with FCPS do that?

[curiousuburb]

I don't have the Blue Horizon content handy, so I can't speak to its methods per se, but from memory, I think it may have been built in smaller segments, which would allow for different encoding rates and tricks in each section to further optimize size.

Theoretically, you could use the "screen zoom" function in System Prefs -> Universal Access to zoom up text (which follows the mouse) and grab the screen with Snapz at a more 'reasonable' resolution (ie: less than 23" Cinema default res).

Not sure if Snapz is just calling the same function internally for its pan & zoom.

In principle, you could also do it as a still grab with iPhoto Ken Burns for the pan & zoom, but the duration of frames will impact your overall movie size - again, likely to be less than 1.2GB... but it depends on your requirements for legible detail... why have full frame capture if the only thing moving is the mouse cursor?

Compressing the mouse cursor moves on an otherwise still frame (without pan & zoom) is more akin to the crunch you could apply to talking head video, rather than forcing the codec to compensate for unnecessary FMV changes if only newscasters lips are moving anyway.

[Moogs]

Interesting. I tried to watch the screen video segments carefully. At first I figured they were all like 10 seconds or less, with the last frame just being repeated for a few seconds to make it sync with the voiceover / transitions better. But there were several that appeared to be 30 seconds or more.

To me it kind of looks like they captured their whole screen, but selectively panned and zoomed, as the mouse movement was taking place. I don't necessarily have to capture the whole 23" screen. In fact most people who use the DVD will have something between 15 and 17" most likely and in 4:3 format. The thing is, I do want to be able to show the entire interface of the app I'm using.

It doesn't have to be 100% magnification all the time, but if it's going to be a lot lower (probable), then I want to make sure when I zoom in on window controls (either manually using FCP to scale / zoom into different parts of the screen movie), the text on the window controls and palettes needs to be crisp and I can't have any dithering either when working with images.

[curiousuburb]

If I were capturing and doing a pan and zoom mouse movement on this forum page, for example, I would be using a much tougher codec, since I'm really only looking at about 16 colours (allowing for a bit of shading and anti-aliasing). Even the animation codec at 256 colours will probably be overkill in bit depth, but might be a starting point... that should help knock the file size down to more manageable levels.

1.2GB sounds insanely overkill for screencap to illustrate buttons and cursor actions.

Do you really need to zoom from full screen all the way down?
Can you fade from a still of your crisp images and interface in situ, and blend in the zoom from a "buttons/palettes only" grab at something like 600x400? The critical content still gets the gee-whiz mouse trail motion you want, without wasted bandwidth or disc space. YMMV

Depending on what your interface/content is composed of, try more aggressive settings until you lose the minimum spec crispness... when in doubt, throw it out. Do some tests and if you can't spot the difference, turf the larger file and test again. Repeat until you find the 'absolute minimum' your client/boss will accept. Save those settings as "Moogs minimum grab" for future use.

[Moogs]

Sound advice. As far as the interface elements themselves I could probably get by with thousands of colors easily, and I try to go somewhere between medium and high quality, but I could probably get away with medium. I guess there are all sorts of tricks in terms of overlaying still images on the part of the video that holds my image window, then when I click "OK" for whatever setting I'm using, replace it with the finished image after the change, etc.

I'm still interested in that effect they use though because it's obviously bandwidth friendly but it looks very cool / professional. If you get a chance to look at the Blue Horizon tutorial, the parts I'm talking about can be seen in the first minute or two of the FCP section of the DVD (after the new features segment). Watch from the time when the launch the app, through the first couple example exercises and see how they manage the video of the interface. Works really well IMHO. Sort of mixes the best of full screen capture and zoomed in, "follow-the-mouse inside a fixed viewport" type deal.

[Moogs]

Sound advice. As far as the interface elements themselves I could probably get by with thousands of colors easily, and I try to go somewhere between medium and high quality, but I could probably get away with medium. I guess there are all sorts of tricks in terms of overlaying still images on the part of the video that holds my image window, then when I click "OK" for whatever setting I'm using, replace it with the finished image after the change, etc.

[chop]