I always shoot or scan MUCH larger than needed. That way you can pan all around and zoom in and out without losing quality. Keep things moving. You might even play witht the chroma keyer to allow some transparency to reveal underlying images in layers. Think in terms of how even still images are animated when you watch a documentary about times when there were still photos, but not movies. You often forget you are watching stills because they are constantly moving and changing layers.

Have Fun!

PS--editing with larger stills will have no effect on the eventual file size...they will all be scaled down to only what your selected aspect ratio is x 72dpi.

I just scan to 640X480 then batch matte them all to 720X480. By default VV3 retains the proper aspect ratio when doing this _and_ it makes for less worry about overscan being cut off on the sides.

I rarely find the need to scan to a higher res - I usually just don't zoom that far - but when I do I usually just double it to 1280X960. Most of the snapshots I get begin to look pretty bad that close anyway.

I have a TRV900 and take a digital picture of the actual and then download the picture files to VV3. If you shoot them in order, you can dump them on the timeline in the same order by selecting all pix, dragging the FIRST one to the timeline and they all go at once. If you apply a transition, that will happen at the same time also. I do this all the time and can dump 180 pictures with dissolves to the timeline in 6 seconds! Then you can go back and individually edit each picture for cropping or color correction or whatever. Easy, easy, easy.

thanks to all of you for your reply
i am talking about someone handing me 50 photographs (all different sizes) which were taken over the past 30 years
before i was shooting them with my camcorder, capturing etc.
then trimming them to 7 seconds on the time line
now-----from this moment on----i am going to just scan them
it will talk me less time and they will come out better quality
thanks to all
sb

655x480 is great is you're not doing any pan/scan, otherwise just scan them and drop them on the timeline. Then, for each photo, bring up Pan/Crop, right click in the image, and select "Match output aspect" and Vegas will create a pan rectangle that works just right for that photo, all set to zoom or pan without borders.
///d@

thanks again everyone
i really like that "match output aspect" tool

sb

Actually, 640/720 is .889 - not .9091. You're right, if VV3 mangles that ratio it's too small to notice.

But there are a couple facts that cannot be disputed:
1) A VGA screen is 640X480 and is square pixels. So I scan to this figure or close to it.
2) A DV frame is defined as 720X480. So if I don't do something the NLE will when it renders.

That's why SonicFoundry says to use 655x480 instead of 640x480.

>>1) A VGA screen is 640X480 and is square pixels. So I scan to this figure or close to it.

that statement doesn't make sense to me. why would the size of a VGA screen even enter the equation?

I've seen the VGA screen size referred to a bunch of times when discussing this topic, and I don't understand why? do you actually run at 640x480 and you are trying to make sure when you view your pictures they fill the edges of the monitor? otherwise, I can't think of a reason why anyone would scan at 640x480. if you scan at 655x480, your pictures will have a higher resolution and will stretch the correct amount in Vegas to account for the aspect ratio difference between pixels on a monitor vs. pixels on an NTSC TV.

640X480 is just an old standard PC frame size. It's the same ratio as 800X600 or 1024X768. You could use any of these and still maintain the proper aspect ratio of square pixels as seen by your scanner. Or you can just scan to >72dpi but then you'll still want to matte to 640X480 (or a multiple thereof) since the size of a TV screen is a fixed ratio.

It's still a mystery why the programmer at sofo reduces his 720 to 655 using an odd ratio like .90972. An even bigger mystery why so many people believe everything they read in a manual without question.

A lot of these weird aspect ratios come from timing considerations of the NTSC and PAL signals. 655 lines up with the waveform coming from NTSC analog video better than 640 does. There was a rather lengthy post explaining all this a while back. Try searching for "aspect ratio" and "timing" and you might find it.

As far as believing goes, i believe that a 655x480 image fills the frame better than a 640x480 image does because i see it happen on the screen ;)

Also if you want to get extra fussy don't forget to turn on 'safe areas' (click downward pointing arrow above preview window just right of blue overlay icon) to see what will likely get cropped off when viewed on a TV screen. It varies depending on TV model.

This has been covered a few times on this forum, but a short repeat:

D1/DV video use non-square pixels. If you want to use square pixel source material (as in computer graphics) in a video and not have it looked stretched, multiply the video frame width x destination format pixel aspect, and use that as the frame size in your graphics editor.

720 (DV NTSC frame width)

x

.9091 (DV pixel aspect ratio)

= 654.552. (Round it off to 655)

Height doesn't change. Use 655x480 for stills in NTSC DV projects and all will be well.

This would only be true if one assumes the .9091 figure. I have no idea what kind of calculator you're using, but on mine the ratio is more like .9 when one divides a VGA frame by a DV frame. (640/720 = .88889) (or 648/720 - see below)

Anyway, I really don't want to belabor the point. If you have proof, show it. We all agree on the definitions of the frame sizes(?); now we just have to figure out *why* the VV3 programmer only allows saving a DV frame as 655. I've never seen any reference to this in any NLE I've ever used nor in any link on the web anyone's ever supplied. The closest I can come is Xpress DV which, in addition to 640 allows one to save as 648X486 which I understand is a valid square pixel resolution. It offers no 655 (surprise).

And let's not try to justify this oddity with analog measurements. By the time we munge our pristine digital frames to a consumer's misadjusted TV the frame size is anyone's guess.

P.S. If you *really* want to get anal about this, try http://www.uwasa.fi/~f76998/video/conversion/
He never actually hits 655 or .9091 but by using complex fractions like 72/79 he sure manages to cause trouble! In the end he agrees with me though:
>Converting "computer" square pixels to "video" square pixels is a futile effort. You will not see the difference, anyway, and probably only lose some quality in the interpolation process.

The pixel aspect ratios as implemented in Vegas WORK in Vegas. SF engineering has been thru the math many times - hunt around these forums if you want to do some further reading.

Real world test:

Open a dv project in Vegas

Create a circle wipe between black and white generated media

Save a timeline snapshot of the circle (it'll save as 654x480 at project size)

Open it up in Photosh*p- is it round to the eye? yes.

Crop the circle right to the edges- square to within 1 pixel (meaning the circle is a circle and not an oval) ? yes.

Summersond,

When you capture with your trv900, I'm assuming you don't do any panning or zooming on those images because they are 720x480? I really like your method but find the need to do some motion work within the images and would like that option. I have never captured stills to a memory stick or floppy from the sony. Can you give me some more information on this... feel free to e-mail me at chenopup@hotmail.com if you want.

Thanks,

Cheno

Forgive me, but I'm new butI have to ask this; When you scan, crop or resize your .jpg images to the 655x480 or 640x480, is this sufficient resolution when you output your project to a DVD and show it on a TV? Is the limitation due to maximum lines of resolution on a TV so any higher resolution couldn't be used? Thanks.

Tim, precisely. Digital video is 480 lines high (or 576 for PAL). Televisions typically display 525 lines. Anything higher than this in the output is a waste.

>>>When you scan, crop or resize your .jpg images to the 655x480 or 640x480, is this sufficient resolution

655x480 is sufficient resolution for a TV. if you intend to zoom in or pan on your pics to make your slide show much more interesting, then use a higher resolution. personally I scan at a very high resolution so my pics are at least 1600x1200 to allow for lots of zooming and panning in Vegas.

the only reason 655x480 is recommended is if you are going to render to NTSC DV. your project will be 720x480. when you drop the 655x480 pic on your timeline, Vegas will automatically stretch the pic to fill the 720x480 frame, making your subject appear a little too fat on your PC screen, but since pixels on an NTSC TV are taller and skinnier than pixels on a PC screen, your subject will lose the extra weight when the final project is viewed on a TV.

it's just a matter of convenience. you can use any resolution pic and worry about the aspect ratio yourself. or if it's landscapes, you might not even care about things coming out a little to fat or skinny. in that case just make it look good and forget about this 655x480 business :)

if we are assuming that PC pixels are square, then I still don't see any reason to introduce 640x480 into the equation. a square is a square. all that needs to be done is find out the shape of an NTSC pixel using whatever formulas apply, and then divide the pixel's width by its height. there's the aspect ratio. no VGA screen size in that equation. that's if we assume PC screen pixels are square which would be an aspect ratio of "1".

I never said a circle saved from VV3 wouldn't look round on a PC. I was simply saying there seems to be no good reason why VV3 insists on giving us one peculiar size still *which no one else uses*, and then trying to justify it by picking an odd decimal multiplier apparently out of thin air.

I'm no expert; maybe VV3 is right and everyone else is wrong.

Just for laughs go over to the Avid forum and try to convince them of that. Write the author of the comprehensive article I quoted and see if you can convince him as well. Get back to us.