Don't use TIFF. Use PNG or JPG. Vegas uses an outside app to render TIFF, so it takes much longer.
Use tha sizes Vegas text generator use as default.
Tor

Hi

For the aspect ratio, have a look at Marty Hedlers Homepage:
http://www.martyhedler.com/homepage/Vegas_Tutorials.html

There you can find some helpful tutorials.

Stephan

Save your file as a .psd file, 720 x 486 The .psd (photoshop document) allows V4 to import any alpha channels and transparency. When you render your file- render as NTSC DV version of AVI or, if you are only using Photoshop- set your Vegas properties to "NTSC Standard (720x486, 29.970 fps)" and render the final as AVI- uncompressed. The file size will be rather large (approx 36 megs) for your 10 seconds of material. It should and will look fine- unless you are oversaturating any "red" colors!!!! Be careful- video hates red- especially if oversaturated

I make all my stills the same size Vegas saves them as when i save a video frame: 654x480. Never had/have ANY problems with pictures this size. I use them wiht video all the time and they match up perfectly. I use PSD just so that if I want I can change something in the picture and vegas will auto update the picture for me.

654x480x1.0PAR or 655x480x10 PAR will work fine.

When Vegas renders the project it will make these 720x480x0.9091PAR

The reason you might choose either 654 or 655 is that the real number should be more like 654.5-ish

People above are giving you a lot of numbers that don't apply

--654x486 only applies if you've set your project to this size. Normally, DV is 480 px high. NLEs that work in mjpeg formats use 486 pixels. The extra pixels represent retrace time which DV just ignores. You usually don't need to use 486px frame heights in your projects

--720x540 is a doozy because it's based on the assumption that the whole digitized frame will be a 4:3 image (which isn't true). A number of NLEs as well as After Effects make this assumption. Essentially, when they did the math the solved to get the desired result. But don't argue. if a program wants to use those numbers you just go along with it.

The correct number to use for DV and Vegas is 720x528. (To illustrate this, take a 655x480 image in Photoshop and resize the width to 720. The height will go to 528). However, after you've laid out all the titles in photoshop in a 720x528 frame you'll have to resize the image to 720x480 (squashing it), import it into Vegas, and then reset the clip's pixel aspect to 0.9091.

It's easier to just work in 654x480 stills. It's easier still to do the titles directly in Vegas.

All of this resizing of stills is done for reasons. You have to decide how much it's worth to you.

--Vegas assumes a still has a 1.0 PAR(pixel aspect ratio). When you render a project to DV Vegas will pad all the stills up to 720x480. This adds pixels to the stills but probably doesn't affect the look of titles.
--Initially, Vegas sizes media to fit the frame. It will size it up or down until one dimension fits. Because it takes PAR into account a 720x480 still will probably be scaled down to 655x437-fitting the width to the frame but with a 1.0 PAR. Normally, users go into the pan/crop tool and select "match source output" to make Vegas correct the image. I'd assume that vegas would then pad the image back out to 655x480-creating pixels. Then when you render it should pad the image again to 720x480.

The better way to do that correction is to right-click the clip in the media pool and reset it's PAR to 0.9091. No pixels added or subtracted and the render ought to go a little faster too.

The point is to get familair with ways to resize images for specific PARs. You want to, at least, be aware of when you're creating or discarding pixels and consider whether it'll affect your image quality. Often it doesn't but sometimes it's a concern.

Rob Mack

"People above are giving you a lot of numbers that don't apply

--654x486 only applies if you've set your project to this size. Normally, DV is 480 px high. NLEs that work in mjpeg formats use 486 pixels. The extra pixels represent retrace time which DV just ignores. You usually don't need to use 486px frame heights in your projects"

I beg to differ with the comment. First of all- I stated, if you are using photoshop- and not capturing DV video- set your project to "NTSC Standard (720x486, 29.970 fps)"---- a STANDARD video size! And render in uncompressed- this was good information and you just blew right by it... looks like ziggy's praise has gone ..... just kidding rmack- but please do not make a general statement that it appears we did not know what we were talking about! The information I gave was right on the money sir! :) Have a great night

Some of your problems are quite likely got more to do with the color range that you are using. If it looks OK on the PCs monitor but crappy on a TV then that's what's wrong.

There is heaps of stuff around about these issues, try looking on the web for character generators or similar.

Basically even if your colors are broadcast legal you still run into problems with the transitions in the color info on a composite signal, try putting 100% red on black, you'll see what i mean. Try to avoid adjoining colors on opposite sides of the color wheel, vectorscope is good for checking this, if you must use them try to have a small area of gradual transition between them. Actually reducing the resolution of the graphic selectively can make it look much better.

I second TorS suggestion about saving in PNG or JPG. PNG is a nice format because it can have an alpha layer and also you can save it as interlaced (which sometimes will eliminate the shaky image effect).

Also photoshop has a NTSC broadcast colors plugin that will bring your image into spec.

Good luck,

Franco

655 x 480 is the size of an NTSC frame. For digital photos, I feel comfortable using quite compressed jpegs. Text is another story however. Use PNG or photoshop formats so that the lines look sharp.

Laurence Kingston

Okay, I'll try to address the points and make myself clearer.

If you are just creating clips from stills then it makes very, very good sense to render them uncompressed. Yes, I blew by that but it's important. You want to keep your elements sharp until that final render.

You can make your stills any size you want and then scale them in Vegas. However, if you want them to fill the frame you should make them at project size-either actual project size or interpolated (square pixel) project size. Obviously your project size will depend on the settings you've chosen. That could be NTSC DV at 720x480, NTSC Standard at 720x486, or any of the other format templates.

If you make your stills at interpolated size then you should make them at 655x480 for NTSC DV or 655x486 for NTSC Standard (Not DV25). This is the easy way in Vegas but you just need to be aware that when you render to 720px wide the 655px-wide image has to be resampled and pixels have to be added. There's really no way around it, changing an image from 1.0 PAR to 0.9091 PAR means that something has got to give in one way or another. If you were to start with a still that is 720x528 (or 720x534 for NTSC Standard) then some pixels will have to be discarded in the vertical dimension.

720x486x29.970 is one of the digital NTSC video standards. Media100 (a system that works with motion-jpeg files in 4:2:2 sample space) uses this frame size. The extra 6 lines, 4 at the top and two at the bottom, are black and represent retrace time.

You may want to use this template in your Vegas project. It could be useful if you are writing a file for use in another NLE that works in mjpeg/422 space. Or there may be other reasons.

NTSC DV25 is 720x480x29.970 in 4:1:1 sample space. That sample space will blur colors mightily at the edges of your nice photoshop graphics. You want to avoid this until the last render pass. So if you need to render a graphic you should do it uncompressed or chose a compressed and lossless codec.

DV25 doesn't sample the top 4 and bottom 2 lines because these aren't supposed to contain image information. When a DV25 signal is converted and output as an analog signal the 6 lines are there regardless of whether they are part of the DV25 signal. This is because the time still has to elapse between the end of one frame and the start of the next. DV25 just doesn't sample it. Other formats do sample during that time.

You should only be using the NTSC Standard template when you know you really need it. Usually you would be using the NTSC DV (or PAL) template.

If you set your project prefs for "NTSC Standard (720x486, 29.970 fps)" and then render out using the NTSC DV template you will find that your 720x486 frame has been squeezed down to 720x480. You will see black edges at the left and right. Your render times will be longer too. Also, be aware that the NTSC Standard template is "Upper Field First" while the NTSC DV template is "Lower Field First".

Finally, if you render your project to match the NTSC Standard template and then output it via 1394-Vegas has to recompress your rendered file just to get it out the pipe.

I'm not saying that any one project format is wrong-obviously they aren't. But they are desigened for specific types of output. NTSC Standard is not NTSC DV and the two don't really mix-although I'd say that using DV format clips in an NTSC Standard project will cause fewer headaches than doing it the other way around.

Rob Mack

First of all I have to apologize for volunteering so much info as to make things more confusing.

AZEdit is right that, if you need to render your titles or other graphics, you should do it uncompressed to preserve your sharp edges. This is absolutely good advice.

As far as image size goes, you should make your images match the interpolated square-pixel size of your project. For NTSC DV (DV25) that's 655x480. For NTSC Standard (which is not DV25) that's 655x486. To check for the proper size, set your preview window to one of the (full) options and then right-click the preview window and choose "Simulate Device Aspect Ratio". The dimensions listed as "Display:" will be what you should make your stills at in photoshop.

You can use a variety of formats for the stills but PNG or PSD will work very well.

Don't use 720x480 (or 486) unless you are prepared to do some juggling. Since Vegas assumes all still images have a 1.0 PAR it will scale the image down to 655x437. If you make Vegas correct it in any way other than resetting the clip's PAR then you'll make matters worse.

If you want to use 720x480 (or 486) then you'll want to lay out the text in either a 655x480 (or 6) image and then resize it to 720x480 (or 6) OR you can lay out the text in a 720x 528 (or 34) image and then resize it to 720x480 (or 6). If you do this you have to set the clip's PAR property to 0.9091 by right-clicking the clip in the media pool, selecting properties, and changing the PAR. You can make Vegas remember this PAR for this file type by clicking the diskette icon in that same properties dialog. From then on Vegas should set all new files of that type to that PAR.

It's a lot simpler to avoid this though.

Rob Mack

This began with Lamont creating a FBI warning in Photoshop and evolved into the technicalities of importing stills to Vegas. However useful that is (I'm in PAL-land, so most of these numbers are alien to me) I think this needs to be said over and over again:
You can create wonderful graphic images in Vegas' text generator and you can move them about or animate them with track motion, pan/crop or 3D LE. There are fonts out there for nearly everything, not just letters and numbers, and many of them free. You can scale them up and down without loosing the sharp edges, and they will behave in your renders. You can even find free(?) or at least cheap apps that will let you make truetype fonts from your own graphics, fonts that you can then use in Vegas' text generator.
I'm not dismissing this thread - I realize there are things that needs to be imported as images to Vegas. I just think Vegas' text generator too often is overlooked and even underestimated - possibly because it doesn't have so much instant flashiness as some apps does.
That said, I have absolutely no idea whan an FBI warning is and what it looks like. It does sound serious, though.
Tor

.psd files should be vary stable, and barring a few color combo's that should be avoided (as stated in other posts) there's no logical reason you should end up with bad gfx. My suggestion would be to go into your psd file and your text-layer, make sure your anti-aliasing is set to smooth, set the transparency level on the text level to about 95%. Now turn off all the other layers except the text layer, save the file as a psd, and import into VV. When you place it on the time line the text will show up over the default black screen and should look fine.

Let us know if this or any other info here works for ya

rmack- I like your posts...I will look into what you stated when I can and thanks for the feedback! Your feedback and comments are appreciated- and we all don't have to agree- but rather make educated decisions on what works best for the environment we are in. Have a great day!