I've always left my photos at full resolution, and just let VMS handle it. But I have seen discussions somewhere, however, where people claimed it was better to use photoshop or something to downsample the photo to something closer to DV size (approx 655x480 for a "square pixel" digital photo). (Still, I'd just use the original photo as is, but I'm by no means an expert at this...)

If you have VMS 6.0...

After you add the photos to the timeline, right-click on them, select "Switches", and then enable "Reduce Interlace Flicker". If you don't do this, you'll find that high resolution photos often have a distracting "shimmering" effect anywhere that the photo has light colored, fine detail. This effect would only be apparent when viewing on a TV (i.e. on an interlaced display). You could try making a DVD first to see if this is necessary -- some photos don't have these issues. (And don't do the Reduce Interlace Flicker on any *video* segments. I did that by accident once, and it does some really weird line-to-line averaging of the image, which comes out pretty odd when you have any motion in the video...)

Also, you might consider selecting "Best" quality for rendering, rather than "Good". This is also a topic for discussion -- there's a thread over at www.camcorderinfo.com/bbs about this right now -- but I'm under the impression that the algorithms used for "Best" quality do a better job when rescaling an image -- converting a high-res photo down to DV/DVD resolution. This may be especially noticeable when panning or zooming on a photo, rather than just having a static shot. Be aware, however, that changing from the default "Good" setting to the "Best" setting may result in taking about 2-1/2 times as long to render.

And if you don't have VMS 6.0 -- still using 4.0 or earlier -- I highly recommend the upgrade to 6.0.

Tim L

The single biggest disadvantage of leaving the photos at their original size is that they take up a lot of memory. Vegas has all your photos open in RAM while rendering. If you don't have enough physical RAM then the photos will start spilling over into paged memory on the hard drive and rendering will slow down to a crawl. Some have even reported crashing. The general feeling is that you should resize your images to be no larger than necessary for the closest zooming you will be doing. For example, if you have a 4x6" photo and you want to zoom in on someone's face that is 1" high in the picture, you'll need 480dpi to have that face fill the video frame. If you also want to show the entire picture before zooming in, then 4" high x 480dpi is 1920. So anything larger than 2880x1920 would be wasted. If you're not zooming in at all then anything beyond 655x480 is wasted.

Concerning the shimmering effect and resizing ... the problem with resizing is that some of the pixels have to be thrown away somehow. Unfortunately the choice of which get thrown away can sometimes result in fine details interfering with each other. Try this thought experiment: imagine a checkerboard with 1.25" squares and you are viewing it through 1" pegboard (if you don't know what pegboard is, there's a good picture here). The checkerboard represents the original image and the holes in the pegboard represent the few pixels we're keeping as we resize the image. Now, looking through the holes across horizontally, the first two may show white from the square behind them. The next hole will show black, but only that one hole. The next one is over a white square again. That makes it look like there is only half as much black as white even though in the original image there was exactly as much black as white. Now let's pan a little bit to the right and suddenly the second hole is black, so we now have one white and two black dots. As we continue panning this effect an produce a shimmer as we alternately have more white or more black pixels. Some checkerboards have red or gold lines dividing the squares, so we may even see occasional flashes of color between the white and black dots.

Now imagine this happening with the millions of finely detailed pixels in your original image as Vegas pans the low resolution video frame across them. The effect can be quite shocking and annoying. At the "good" setting Vegas does try to minimize the effect by blending adjacent vertical and horizontal pixels together to smooth out the image. With the 'best" setting Vegas averages diagonally adjacent pixels as well, using a weighting algorithm to calculate which pixel is closest to the middle compared with other details around it to give a smoother image. But, if your original resolution is much higher than the frame resolution, the averaging can still produce results that vary as different groups of pixels are averaged from frame to frame. This produces the shimmering effect.

Resizing the image in advance to the smallest possible resolution that still allows for the desired zoom level helps reduce the resampling and blending that must be done by Vegas and helps create a smoother image. It also means that Vegas has to do a lot less work when previewing and rendering, so it will run much faster as well.

On the other hand, i wouldn't suggest opening up every single photograph you are using and resizing them to exactly the correct dimensions just because some of them are a few percent larger than necessary. This would create a lot of work for you, and every time you alter and resave an image some of the quality is lost. If you have lots of large image files then it's worth resizing. If you have a few that cause the shimmering effect then resizing just those may help. If all your pictures are 1500x1000 and you need them to be 1310x960, don't bother.

And for those of you still reading ... thanks for your patience. ;)

Hmm, that is interesting. I just finished a simple project where I created a DVD for my wife's parents that has 3 slideshows. Each slideshow is bewteen 5 and 10 minutes, and contains stills, video, and background music. All of the still images I imported to VMS 6 were 2048 x 1536, and I didn't resize them.

When I first burned the DVD and test it out on my tv, I saw a lot of flickering, mostly just on the stills. I mostly noticed it around the date/time stamp on the picture. The orange lettering would flicker real bad when doing a pan/zoom. I changed the quality to "best" and then enabled the "reduce interlace flicker" switch for the entire video track (stills and video). This helped, but I still see some flicker (it's not too bad tho).

So, are you saying that instead of doing what I did, I should do the following:

1. Resize all of the still images before importing to VMS. For example, I have a program that I can use to do a batch resize. I should resize to what resolution?

2. Once I create my timeline with stills and video, highlight all of the stills and enable the "reduce interlace flicker" switch. So, only enabling that switch for the stills, not the video

Is this what you would recommend?

BTW, before I created the DVD in DVDA, I did enable some type of setting that is supposed to eliminate some of the interlace flicker, and that helped tremendously as well.

dang chein . . . . (fumbling through the manual) . . .what page of the manual was that on? B-)

Manual? We don't need no stinkin' manual! ;)

The single biggest disadvantage of leaving the photos at their original size is that they take up a lot of memory. Vegas has all your photos open in RAM while rendering. If you don't have enough physical RAM then the photos will start spilling over into paged memory on the hard drive and rendering will slow down to a crawl.

Excellent point. A project I was working on a while back had both video and stills, and when previewing the project in VMS the stills definitely slowed things down. When I had a crossfade from one still to another -- especially if the timeline had just scrolled over to a new page, and I was using "Best" quality for preview -- I would frequently see the new photo not appear in the preview window until after the crossfade was over -- or sometimes the photo would barely appear before it was time to move on to the next one. I don't recall how big the files were, but I think I shot them at probably 4 - 4.5 MPixel resolution.

Maybe I need to be looking for a good (but cheap?) program to rescale the photos before I bring them into VMS. Seems like a nuisance and extra effort, but more and more evidence seems to be indicating it might be worthwhile.

Tim L

VMS having all photos open in RAM while rendering was surprising to me - so I made a test: I rendered a small 30 sec clip consisting of few videoclips (640x480), texts etc and a few 2 Megapixel stills. The PageFile grew, during the first 30 sec rendering, up to 800-880 MB and sayed there for the last minute of the rendering time.
Then I opened a 40 minutes project, consisting of the same type of ingredients - but with several times more of each - i.e. some hundred 2 Megapixel stills from the same cameras - but with the same crossfade-types and other effects, that I always use - now again the PageFile grew during the first 30 secs and stabilized at 770-850 MB. My RAM is 1 GB beeing nearly full exept for 100K during the rendering process.
The difference in the PF size between the two renderings is hardly significant - but - to my humble opinion it could show that VMS does not keep all several hundred 2 Megapixel stills open for the time of the full rendering - supposedly (and as I would have programmed it myself) it only has them open when they are directly in use in the rendering process.
Of course both tests had the same specifications for the finished result (800x600 etc) BTW I use 6.0a platinum.
If anybody wants to batch resize a good tool would be Irfanview (free).

I would be interested to see if anyone notices a difference (good or bad) when scaling down the size of the still images before importing to VMS. I think BreezeBrowser will batch resize images. I've used it for batch renaming files, but never batch resizing. I know it resizes, just not sure about batch resizing.

Irfanview can do batch conversion of stills with resizing.

/Olle