GL2 footage not quite filling my NTSC monitor's screen

BradHigerd wrote on 1/17/2003, 1:58 PM
After downloading footage from a GL2 into a PC (P4, Vegas Video 3.0c, ATI 8500DV video card...), I viewed the footage on a 27" flatscreen Sony NTSC monitor. The footage looked very good, but it did not fill the entire screen. It had a hybrid aspect ratio somewhere between 4:3 and 16:9. I am still very new to the world of PC based NLE and MiniDV cameras, and therefore, I am not sure if this is normal for 720X480. If my calculations are correct, this footage would need to be about 720X525 to fill the entire screen.

In addition, I have seen some comments about de-interlacing and progreesive rendering in Vegas. I am interested in learning more.

Thanks.

Comments

Grazie wrote on 1/18/2003, 3:29 AM
Hiyah Brad! See I told you, you would get a swift reply [ DOAH!] . I see that Aaron Koolen has given you a reply on the SoWD site. Seeyah back there . . . Anyway is this your first view to this site? No - probably not eh?

Grazie
BradHigerd wrote on 1/18/2003, 10:28 AM
Yes, this is my first time here. About 3 months ago a friend convinced me to try Vegas along with all the other demos. After trying 5-6 different software options, Vegas was an easy choice. I just wish I had more proficiency with this program.

Unfortunately, work, school, and trying to produce a college television show have not allowed me enough time to learn how to use Vegas.

In what is likely a basic question, how do you move a group of video and audio clips together when you find out that you want to add something into the middle of a project? I'm used to Microsoft’s select procedures, but they don't seem to do the trick.

Thanks for your reply and please be patient with me as I learn my way around these forums.

Brad
Paul_Holmes wrote on 1/18/2003, 10:51 AM
Select all the clips using normal selection techniques, either Ctrl-click, or use the selection tool at the top to draw a rectangle around the clips you want selected, then press G, which groups them. Just remember that if you move them, then press U (ungroup), they will all move independantly from then on.

If you're dealing with one track only, you can right-click and choose "Select all events to the end" (something like that -- I'm not looking at the interface right now). You can then drag the first clip right or left and all following clips will do the same.
BradHigerd wrote on 1/18/2003, 11:06 AM
Thank you.
woodrose wrote on 1/18/2003, 12:31 PM
Grazie what/where is soWD

Thanks
murk wrote on 1/18/2003, 12:39 PM
This is an important question to figure where your aspect ratio is going awry.
BradHigerd wrote on 1/18/2003, 1:13 PM
Good question. My NTSC monitor is connected through the breakout block on an ATI 8500DV all-in-wonder card. The setup was analyzed by an ATI software writer who just happens to go to my church. Originally I had wired up a dual monitor system using the ATI card and an additional PCI video card. Things started going wacky quick. My computer (Windows XP) couldn't reconcile which monitor was which from moment to moment. My ATI contact told me that the problem has existed as long as he could remember, and that they couldn't completely overcome it (so far). In addition, Sonic Foundry's tech support reminded me that their product explicitly states that it does not support dual monitor.

They both suggested that I take my PCI video card out and use the TV monitor output from the ATI to run my NTSC monitor. It's not dual monitor, but it does allow me to see video as it looks on SD TV.

After following their advise, I had to resize my viewing area many times to get both monitors to look right. Both monitors display the same image, and it was tricky to get them to look right. Regardless, the black portions on the top and bottom are viewable on both monitors, SVGA and NTSC.

Sorry for the long answer. What do you think?
Grazie wrote on 1/18/2003, 1:36 PM
Son Of Watchdog [SoWG] - DVinfo.net with special references to Canon XM2/GL2 cammies:

http://www.dvinfo.net/conf/forumdisplay.php?s=743239a28c1f64582beb6bf54027f0b7&forumid=20

Enjoy

Grazie
tadpole wrote on 1/18/2003, 2:06 PM
brad -
Have you tried a firwire connection via your GL2 then S-video/RCA out to your TV monitor?
If you haven't tried this before, hook up GL2 to computer via firewire, turn GL2 on VCR mode, connect tv to GL2 via RCA/S-video cable, select correct input mode on TV.. then click on the "preview to external monitor" button on vegas preview window.

Of course this doesn't fix your problem, but curious to see if this method displays properly on your TV.

mikkie wrote on 1/18/2003, 3:59 PM
OK, the problem is in your hookup with the ATI card -- ATI doesn't in my experience produce an overscanned picture on it's own (AIW & 8500 on a couple of machines). Depending on the monitor, you have to go to the ATI advanced properties and manually enlarge the picture using the display tab. I generally try to allow 1/4" side to side extra on my PC's 17" monitor and this usually works well -- haven't had any complaints anyway.

For further info, the NTSC spec etc. originally calls for a picture larger then 640 x 480 because various TV manufacturers don't mask the outside of their picture tubes with the case to the same dimensions. Both of the picture dimensions are supposed to be larger, but the width is the most critical now days. Search for anything written by Adam Wilt & visit his site for more then you'll ever want to know about that end of things.

mike
mikkie wrote on 1/18/2003, 4:21 PM
I ignored your other ?

Again, check out Adam Wilt's site for more of the nitty gritty details...

Generally speaking, you want to deinterlace if your project's not going to be viewed on a regular TV -- if it'll only be watched on a progressive scan TV or PC monitor for example. Otherwise, leave the interlacing there -- on a PC most player software will automatically deinterlace it at least partially. Selecting progressive in Vegas will cause your video to be de-interlaced - ie: your PC monitor has a progressive display.

Progressive refers to the scan lines in the picture tube, whether they progress from the 1st line to the second and so on, or whether they skip every other line, then go back and catch just the ones they missed on the second pass (interlaced). On (or rather in) a standard TV, there are actually 60 of these scans performed every second [to match our 60 hz power at the wall outlet if you like trivia], hence 60 fields per second whilest at 30 frames per second. To make it appear smoother, the 2 fields for one frame don't look the same -> the 2nd field is offset a little to better blend with the first field of the next frame. You can see this on a progressive screen (a PC monitor), so that's why people de-interlace, which more or less blurs the 2 fields into something closer to a regular static picture.

Now, there are exceptions, and I tried to keep this pretty general rather then write a book. ;) It gets involved when you start talking about the calculations for the actual 29.97 drop code & why it exists, more so if you take your camera into account & how it performs the orignal interlacing, still more so if you get into various interlacing and de-interlacing code, & then there's the matter of telecine and inverse telecine if your footage is at something other then 29.97, so think of this as only something to get you started.
mike
BradHigerd wrote on 1/18/2003, 10:53 PM
Mike,

Thanks for all the help.

As far as the aspect ratio, I think I have it figured out. The problem seems to be Window's Media Player. I took your suggestion and ran some diagnostic tests. Between the video produced from the “Print To Tape” function and an .avi file rendered this evening, it appears that I never had an aspect problem.

As far as the interlacing issue, I appreciate the advise. Everything we are trying to produce is expected to be viewed on a standard television. However, my follow-up question would be what are your thoughts on recording video in progressive (frame) mode? Is it useful, and if so, for what? I’m not fond of interlace artifacts, but who really is?

Thanks again,
Brad

mikkie wrote on 1/19/2003, 9:51 AM
Hi

I'm not sure I explained it properly, so a quick word on the ati video out first off... The signal a TV expects is larger then the 640 x 480 picture it shows in the center of the screen. ATI's output is pretty much a clone of your PC monitor's display -- if media player shows black bars top and bottom when you play your video full screen, that's what you'll get on the TV. Other software might play the video, stretching the picture to fill in those black lines. This is cool if you realize that the video is being stretched.

The thing to watch out for with the overscan I mentioned, is that your final project if rendered to tape might not fill every TV screen - DVD player embedded software would take care of this if that's the route you take. With tape this overscan does need to be there, & don't want you to have to find out the hard way. So even if the picture does fill *your* TV screen, you should enlarge it a bit beyond that to be safe on everyone else's. You can use the controls I mentioned -- optionally you can crop your video or resize it to 640 width (which would be the more accurrate when stretched back out. Trivia: the 500+ lines you mentioned is close to the actual video stream you're dealing with before your PC gets it's hands on it -- it's also a measurement/rateing of the quality of your camera.

Sorry about that, but I was afraid I didn't make this point strongly enough - this may not be a huge problem, but the cure is so easy, why not.

OK, the question you really asked was about progressive recording: Some cameras can record true progressive frames (can't remember on yours at the moment). Usually 24p is used instead of film, and is in a word - nice. The big question though is whether you're going to be able to keep anything progressive through to your final destination. If you can record 24p, and get it onto DVD, that might be the ideal way to go as the average DVD player will output 29.97 i to the connected TV, yet you can stay progressive on equipment that handles it. Buy a movie DVD and this is what you get 99.9 times out of 100.

Where you'll run into trouble is that the average TV doesn't understand anything else, so the equipment getting or storing the signal for the TV is going to be designed to work with video that's interlaced. Even going to DVD you might have problems getting progressive footage into & out of your DVD software depending on the individual apps. Even if you stay progressive all through editing, it could get interlaced on output by your card, &/or by the recording circuitry if going to tape.

When that happens, your output card or your deck or whatever will be responsible for converting your video to interlaced. If you shot at something other then 29.97, it'll convert that as well. Bad part is that this conversion might be lower quality then that provided by your camera shooting 29.97i in the first place, so you don't want to shoot yourself in the foot so-to-speak. Use the old axiom from the printing world, and plan your work from the very start based on it's final destination.

Finally, don't know if this matters or anything, but if/when you capture broadcast video or cable, much of the artifacts you see aren't from the interlacing as much as from the method they use to go from 24p film to 29.97i tv. There's not enough pictures to fill in the gaps, so they generate between frames that interlaced can look quite nasty. That's why in some circles inverse telicine to cut out these frames is so popular, included in winmedia & real encoders, that sort of thing. It doesn't look bad on your TV (which is the important thing), & it's not something you'll get with your camera -- it'll be interlaced, but not nearly as bad looking as captured broadcast.

hope that helped
mike
BradHigerd wrote on 1/21/2003, 4:24 PM
Hey Mike,

The video running through my Windows Media Player is compressed vertically. I'm not sure why I didn't see it earlier, but I think my desire to produce somewhat widescreen footage clouded my objectivity.

You wrote about interlace artifacts. I have a GL2 and recording on Normal Mode (standard interlace) produces a significant amount of this. It looks as though Canon's hybrid progressive mode might be the only way I can avoid this.

The interlace footage I have been working with in Vegas has given me some problems. I tried to deinterlace usng the Blended Fields and the Interpolation process, and the resulting rendered file played choppy (really staggered). I used the Properties settings and rendered the initial footage. I cannot seem to get usable video from the interlaced footage.

The "progressive" video produced from my GL2 is really some kind of hybrid between true progressive and 1/30 sec interpolation between odd and even horizontal lines. I'm an environmental engineer so I'm really only qualified to discuss what happens to the camera at the end of its useful life.

What I want to know is what is the best way to tackle this artifact problem on Vegas with the assumption that my video will be marketed on DVD and ultimately broadcast from a Beta tape.

Thanks for all your time,
Brad