DSE, (umm, Sir... Mr. Spot)

(Okay, I'm really, really reluctant to question anything the great master of all things Vegas says :-), but humbly, here goes...)

It seems to me that it would be better (or maybe just easier) to think in terms of *vertical* resolution for the still photos, and let the horizontal resolution fall where it may. For NTSC, with no zoom, it seems I'd want a photo 480 pixels high -- so I'd have a pixel for every visible scan line in the DV video. Then let the horizontal stuff fall where it may.

Now, with a 3:2 ratio photo aspect, your suggestion of using a 720 x "whatever" photo ends up with a vertical resolution of 480 pixels anyway... so the point is kind of moot...

But... for a PAL user, they would probably want to start with 576 or so pixels high (if no zoom), and let the horizontal pixels work out accordingly. Again, this gives them a vertical resolution at least equal to their PAL DV resolution. Otherwise, their "720xWhatever" photo would only have 480 pixels vertical resolution, and therefore would have to be up-sampled to fill the screen vertically. (I think...)

Anyway, for me, its just easier to think of vertical resolution as being more of a common ground between photos and DV. A 720x480 DV image and a 655x480 photo will be the same size, so its simpler for me to think in terms of the vertical resolution, and let the horizontal numbers fall where they may.

Okay, so that's my relative "noobie" way of looking at it.

(I humbly defer now to your great wisdom, and will gladly accept a smack-down if you need to put me back in my place...)

Tim L

PS: I have a post below dealing with my experiences with using high-res photos in VMS (4.5mpixel, 2592x1728), and coming to the conclusion that you *should* resample them to something smaller before bringing them into VMS.

Link: Still Photo Sizes, Good vs Best Render

Nothing to put you in your place, and I can't really argue your logic. This isn't something that should, or does require a lot of thought. You can do it the "Sony way" and think horizontally, or you can do it the "Adobe way" and think vertically. Generally, you'd want to consider the horizontal, no doubt, since you *may* not fill the frame, but it also would take into consideration whether the image was acquired landscape or portrait.
Since I don't ever mess with photos for purposes of resizing before dropping into Vegas (but my life was strong with AE long before Vegas existed), I tend to think the Adobe way when necessary. (Which is rarely....)
Vegas will correctly manage the PAR as well, which is another reason it's somewhat silly, IMO, to mess with the files prior to import, unless you're concerned about disk space.

I'm still fairly new to all this. In my first "real" project a couple months ago, I just pulled my 4.5 mpixel photos (2592x1728) into VMS and used them in their original size. I was doing slow zooms on them (maybe starting at about 100% of the photo and zooming in to about 80% of the photo) over a 4-5 second or so period. I ended up gettings lots of really distracting image things going on (kind of like flickers, and waves), and at that time found that rendering on the "Best" setting (instead of "Good") helped things look considerably better. But of course, rendering with Best takes a lot more time.

Recently, I revisited that project, and did the experiments that I linked to in my other post. I've found that, at least for this project, resampling the photos to 900x600 (because I'm zooming in on them) and rendering just on the Good setting produces excellent results -- on par with the "Best" render when I had the original high-resolution photos -- but will much faster rendering times. In addition, editing and previewing should go a lot more smoothly with a timeline full of 400KB photos compared to a timeline full of 2MB photos. (I had a lot of stills in this project.)

I know the files are pretty big (about 18MB each), but for anyone who has the time and a decent internet connection, follow the link in my other post (to my savefile.com page) and download the "Church-Good" and "Church-Good-900x600" mpeg files. You can easily see for yourselves the huge visual improvement in the project that used the downsized 900x600 photos. (At DSL speeds, these files will probably take about 3-5 minutes each to download.)

I don't mean to keep belaboring this point, but I'm posting here partly out of guilt, because in the past when people asked about this topic ("how should I resize my photos to use them in my video?"), I would always post a reply that amounted to "Don't worry about it. Just bring them into VMS as they are and VMS will take care of everything".

But now I have seen the light! I have been converted! As a result of the tests I did, its clear to me that at least in some cases, you can get significantly better results by downsizing hi-res photos before bring them into VMS. Again, this only helps if you are zooming or panning on the photo. I would expect a purely static photo to look the same whether it was brought into VMS as a high-res photo or as a downsized version.

Tim L