Laurence,

Thanks for the help!

What I'm trying to do is use the full width of the HDV capture so I can pan across the video in post production. This way, I can -- to some degree -- change composition in post. Thus, doing the HD->SD conversion in-camera is not an option. Also, rendering to SD will just give me the middle of the frame, and what I want is to get the left side sometimes, the middle other times, etc. I need to pan/crop each event to do this. I simply need to know the "secret formula" for the project settings, pan/crop settings, and render settings so that I end up with the best possible quality.

Just got an FX1 myself and was wondering the same thing. You can crop the image pretty easy by clicking the event pan/crop button, right click inside the position box, and click Match Output Aspect. Hope thats what you were looking for.

Nick

Someone correct me if there is a better way, but here's what I do:
Capture your footage from your FX1 using Connect HD, if you have it. If not, capture the M2T file and convert it to an intermediary like Cineforms. This will give you a lot better editing experience.
Now load your footage into a Standard Def timeline.
On each clip you throw onto your timeline, go to the Pan/Crop tool. Right-click on that window and tell it to conform to Output Aspect and VOILLA! You have 4:3! Simple as that. Now you can position the 4:3 box anywhere on the picture you want. Can even pan and scan if you want. Just be careful about your keyframes or you WILL be panning without meaning to.

Now load your footage into a Standard Def timeline.

mjroddy, thanks. That's exactly what I was looking for. My only remaining question is whether this gives the best possible quality. My concern has to do with what point the video gets downsampled. I know, for instance, that in an SD project, if you include a high-res still, and then use track motion to zoom in, the still is down-sampled to SD prior to the track motion zoom so you end up with a really lousy looking zoom. By contrast, if you zoom that image in pan/crop, this feature in Vegas samples after the zoom, resulting in MUCH higher quality. My concern with setting the project properties to SD is whether that would cause the entire project to be resampled at SD prior to the cropping in the pan/crop, thus resulting in far fewer pixels, and a resulting loss in quality.

This question is for people smarter than me, but I believe you will be good using the mothod described above. The reaon I think this is because you can crop deper into that HD footage and it will stll look great - until you hit a certain point. I think you can zoom into it something like 60% and still have geat looking quality. I haven't experimented with that, but to me, this means that the HDV quality is there while the footage is on the time line. If it was downsampled, you couldn't zoom into it. That's my logic, anyway.

Yes, Pan Crop uses the full resolution.

John, not sure if you're using the original m2t files on the timeline - if so I expect editing is anything but fluid. As you probably know the alternatives are Cineform Intermediates, which are used right through, or widescreen DV avi proxies, which are replaced by the original m2t files before output. By far the easiest way to do this is Gearshift, but it can be done manually in V6.

John, not sure if you're using the original m2t files on the timeline - if so I expect editing is anything but fluid. As you probably know the alternatives are Cineform Intermediates, which are used right through, or widescreen DV avi proxies, which are replaced by the original m2t files before output. By far the easiest way to do this is Gearshift, but it can be done manually in V6

Peter, thanks for that input. Yes I'm definitely aware of the need to use proxies in order to make the editing smooth and easy. However, I was trying to just "get my feet wet" doing some basic things first, in anticipation of my first project next week. Setting up Cineform was an extra step that I wasn't quite ready for, and I was willing to put up with the sluggishness of the m2t files (not much different than MPEG-2 on the timeline), at least for the few small tests I'm doing.

I understand from recent posts that Pan/Crop produces better results than track motion. I use pan/crop and it works great. Haven't used track motion yet.

I set my project properties to the final render size, also, ie. DV or DV Widescreen. I am pretty sure this has no effect prior to the final output.

I am also sold on the intermediates. It takes a while, since I don't have the full version ConnectHD, but I just set it up prior to going to bed at night for large clips.

Welcome to HD, I think you will be impressed. I am.

John,
From what little I understand, you do NOT really want to edit the m2t files if you plan on doing any compositing, colour correcting, etc. Since m2t files are long GOP, it's a bad idea and will result in bad ... um... results. It's explained pretty well on other posts, if you wanna search other posts by me (I think it was called "HDV Explained" or something - sorry). Use the intermediary for bests results. That's what I understand, anyway.

John, use Pan Crop which uses the full resolution of the video. Of course, you know the drill about cropping and panning a high rez picture. I find usually a 1:1 ratio of pixels the best for thru put and stops artifact issues if you do not have to zoom. But with 16:9 HDV area, its hard not too when going to SD format. Have fun!

Thanks for all the input on intermediates. Like others before me, I'll be darned if I can find HD Connect or Cineform anywhere in my Render As dialog or anywhere on the Sony download page. I've spent the last ten minutes searching this forum because I know others have had the same problem. The answer is somewhere on this forum ... I'll find it eventually.

[Edit] I found the posts. Apparently the word "Cineform" is not used anywhere in the Vegas UI. Why the heck does everyone tell you to use the "Cineform codec?" It's like someone telling me to use the "Delco" in my car rather than telling me to use the radio. Yeah, Delco makes it, but you won't find their label anywhere on the dashboard. Tell me to go to Render As, select the "Save As Type" Video for Windows, and then select the appropriate HDV Intermediate template, and I can understand what to do.

I'm also finding all the same problems that others have reported but which I hadn't paid much attention to, like lack of scene detection. This is a real killer for me because I do lots of sports, and stop/start the camera at every break in action. With basketball, I'll end up with several hundred clips. Editing one long file is completely unusable for me. Having to buy another program just to get that feature is also pretty silly, especially when it is going to cost several hundred dollars. I saw one post where someone tried to justify the stupidity of asking all of us to shell out $200 just to get scene detection by pointing out that we've all already spent huge amounts of money to acquire the HD camera etc. and therefore why would we have any problem spending another $200 for one more feature? I'd also like to preview the video in the capture application, but apparently Vegas can't do that either. I'm sure I can spend another $200 and get that feature too. I don't mind paying extra for a fancy stereo system in my car, but I sure don't expect to have to buy brakes for each wheel on an a la carte basis.

I just fired a lawyer who kept trying to justify his $450/hour charges by telling me that it was really only $225 after tax. But I digress ... ).

Bottom line is that I agree the HD stuff looks gorgeous but, as I feared before I plunked down the dough for the FX1, I don't think I have the 2-4x additional time to deal with all the extra time and, more important, this incredible additional -- and unnecessary -- complexity. I'm an engineer, but I'm getting old and cranky and no longer find it fun to deal with intermediate files, and extra renders, and the expense and hassle of having to buy more codecs, and jump through hoops in order to do something that I was able to do without ANY of that, just to get a sharper picture which I can't view on any TV set I currently own, and can't deliver to a mass audience for another four years (at least).

However, being stubborn, I am not totally ready to give up, after only two days, but I must say that for those of you who haven't made this switch, don't feel too bad about waiting on the sidelines a little longer. As I pointed out in another post yesterday, the DVD player first shipped in the US in March of 1997, but didn't overtake the VCR in sales for 4.5 years. I expect the penetration of these new HD players will take much longer, given the less than 100% penetration of HD sets, the lack of HD material (DVD had tons of SD material available that could be digitized quickly), and the competing formats (compared to only one DVD pressed disc format). In addition, DVD had a huge advantage compared to VHS, including menus, random access, much better picture, no degradation on repeated viewings, and much more. The only advantage of HD discs over DVD is better quality. I like quality or I wouldn't have spent the money on the FX1, but I can tell you with almost complete certainty that the switch to HD discs will be MUCH slower than from VHS to DVD. As another point, look at DVD Audio. It had other problems, but one of the main reasons it didn't succeed at all is that better audio quality was the ONLY advantage it offered, and that wasn't enough when the quality was pretty darn good to begin with.

OK, back to trying to figure out how to get this to work ...

John,

given the nice support, that you have spent to a lot of people in terms of scripts, I think there are enough people here willing to support you - as I do.
:)

Think about following workflow - maybe that fits better your needs:

1) capture long pieces of m2t files from the FX1. So about 20 or 30 minutes. The reason for doing that is, that up to now Vegas 6c does not support a real huge numbers of m2t files in the timeline. If you bring in a project 100 or 200 m2t files, the swap file will climb up and Vegas will crash. So, it is a good idea to work with larg pieces of your footage, but with a limited number of single m2t files only.

2) render the 30 minute pieces to the cineform intermediate. You can do that with Gearshift, or you can do that manually. Manually, you find the codec here:
- File/render as/video for windows (as file type)/template HDV 1080-60i intermediate, since you are in the NTCS world I think. Check that customize option, and you will see in "Video" the video format "CineForm HD Codec V 2.1"

3) go to the web, to

http://www.avcutty.de/english/index.htm

Here you find AV-Cutty, what is able to analyse the Cineform intermediate avi files - and what allows you to make an automatic szene detection and separation in single files for each of the scene.

AV-Cutty is freeware and can be testet as long as you want to do - but Andreas will be happy to get a small donation, if the tool works fine for you. And will send you a licence to avoid some small restrictions of the freeware version.

Andreas has prepared AV-Cutty to abe able to read both Cineform and also Canopus HQ intermediats - so that should work out fine. The scene separation will be done based on an optical analysis - but you are abel to preview that, before you write the single files to your harddisk.

Writting to your harddisk - that happens without any new rendering, so not loss in quality takes place at all.

4) Import now the scene-Cineform files to a new Vegas project - and you can use the bin structure as usual. As project setting, use now "HDV 1080 60i (1440x1080; 29,970 fps). And edit your project as you want.

5) For the final output, you can now decide about the format, as described earlier here. But frankly spoken, I think that in near future we will see HD-standalones - and plasma and lcds taking place more and more. So, I would not spend too much time with pan/crop the material back to 4:3, but would think about to stay with 16:9 in the future. And I would not stay with to SD-DVDs in the long run.

But that is up to you, off course.

Kind regards,
Wolfgang

Wolfgang you're right - this "crop and pan around the 16:9 HDV frame with a SD output frame without losing quality" is a temporary pleasure only - but it is a compensation for having to hang around waiting for HD output.

Well, we will see in 2006 the first standalones, that are ready to playback HD-content. And the first lcds in full HD quality are available now - even if more will follow.

So, from that side it is worthwile to think about the switch - and John has started so, by purchasing a FX1. The rest will follow.

Wolfgang,

Danke shön! Ich bin für ihre hilfe sehr dankbar.

It's been 35 years since I wrote that much German (three years in high school), so I hope I got it right.

From everything I've read, if I follow your procedure, and use Cineform intermediates, I don't need to substitute the m2t files prior to rendering because the Cineform intermediates contain the full quality.

The only remaining question after your excellent tutorial is whether I should use HD or SD for the project settings if my intent is to output to an SD DVD. I would think the performance would be faster if I use SD project settings (this is based on the "trick" I've used of setting project settings temporarily to 320x240 in order to improve preview speed).

I'm doing a project for delivery in two weeks. Obviously it will be SD DVD, and I'm delivering to 50 different people. Given this,I'm certainly not going to re-visit this project (a dance performance) in three years when some of them have HD playback. Also, my B-roll camera is my old DV camcorder. Thus, for the moment, I don't care about the HD delivery.

Once again, many thanks!

I set my project settings to SD. I haven't experimented with the preview speed. But I always switch back to m2t for final render. While the Cineform is great with very little generational loss, I just want to minimize the "middlemen".

John,
I can fully understand your position about those $200 for Cineform's codec. I also think Vegas 6 should provide a better workflow for HDV. But it doesn't.

Having spent the $200 myself (actually it was a promo for $149 that I bought one year ago) I can simply tell you it was well worth it. The workflow is so much improved with ConnectHD over Vegas 6.
You need just one step in one application to capture, convert and scene split the HDV material of your FX1 to AVIs. All this is no more complicated than capturing DV footage.
After that, working with those cineform coded avis in Vegas is just as simple, straitforward and almost as fast as working with good old DV.
Give it a try (there's a demo available that will work without further limitations for a couple of days).

Inside Vegas you always want to edit HDV material in it's native format (HDV 1080 - 60i) - even if you "just" intend to output to DVD. Do not crop. Leave everything set to 16:9 (unless you are going for special effects, of course).

For rendering you create yourself a modified version of the template "DVD NTSC".
All you change from the defaults is on the "Video" tab: "Field order: Interlaced, Top Field First" and "Aspect Ratio: 16:9 Display". Save this template.
Now you can create an widescreen MPEG2 file in SD-format with the best possible quality. This file can easily be imported in DVD-Architect or any other DVD-authoring software. You usually don't need any other filtering of the HDV material in Vegas.
Good luck.

John,

yu are welcome. And yes, you got it right.

No, you do not need to substitute the Cineform intermediate back, prior to render. That is the difference to proxy files, that must be substituted back - but not the intermediates like the Cineform or Canopus HQ codec.

In terms of preview quality and speed, I have done a small test - and have seen very similar playback capabilites, if I adjust the projekt settings to the original material (so 1080 60i in your case, as long as you render to such intermediates too) compared with 1080/2 settings. So, at least at my 3.2 P4 I see not really a difference. A better way would be to reduce the preview quality down, e.g. to "preview (auto)" instead "best (auto)".

And pan/crop to get 4:3 - as suggested here - will reduce your preview quality again, as far as I have seen. But overall that could be a way how it works. So, if you can avoid 4:3, and got to 16:9, that would be another advantage, also from that side.

Or you film at least in a way, where you mark on the camcorder display the 4:3 area - the adjustment with keyframes is a painfull way, if you do not care about 4:3 when you shoot your movie.

As mdopp says: take care for rendering, that the material is upper field first - and adjust the template for that.

Wolfgang,

Boy, you sure make this easy.

For those following this, I did find the "magic number" to get an exact crop. I did it by trial and error, and then sat down and figured out the math. This exercise helped me better understand the often confusing issue of pixel aspect ratios; display aspect ratios; and resolution ratios.

Format aspect ratios (for NTSC)
HDV 1080i is 1440 x 1080 which is a ratio of 1.3333
DV (NTSC) is 720 x 480 which is a ratio of 1.50

Pixel aspect ratios
HDV is 1.3333
DV is 0.9091

The number 1.3333 shows up twice, but that same number refers to different things (they are related, but they are independent, different things).

The key thing I was looking for was what number to put into the pan/crop dialog box, for the Position Width and Height, in order to create exactly the correct aspect ratio that, when rendered to SD, would not result in any black bars at the sides or at the top/bottom. Obviously the 1440 number (the width) of the HD image must be increased in order to make the aspect ratio equal the 1.50 720x480 aspect. However, the difference in pixel aspect ratio means you can't just multiply the format aspect ratios (i.e., 1440 x (1.50 / 1.3333 = 1620). 1620 is not the right number. Instead, you also have to multiply by the pixel aspect ratios, since they are different for the different formats. Thus, the correct calculation is:

1440 x (1.50 / 1.333) x (0.9091 / 1.3333) = 1104.56

Thus, you unlock the "lock aspect ratio" button in the pan/crop dialog, and then enter 1104.56 for the position width.

I don't know whether being an engineer is a curse or a blessing, but there you have it, FWIW.

Boy?

Well, I assume nobody called me so, at least not the last 30 years!
:)))

And after that calculation: you let us believe, that you will not figure out the HD(V) secrets?

Actually, If I remember correctly, I simply select the "4x3" or "16x9" presets in the pan/crop, then resize accordingly while maintaining the aspect ratio.

Does your preview in pan/crop appear 16x9? If not, make sure your media pool m2t is set to 1.3333 aspect ratio.

I've never found the need to whip out a calculator for that yet.

And after that calculation: you let us believe, that you will not figure out the HD(V) secrets?

I'm working on it. I have come close to an interesting workflow using the Scenalyzer batch capture from the HDV->DV output from the camcorder. This gives me scene detection (that HDV capture doesn't provide) in less than 5 minutes (per tape), and also generates a proxy at the same time. However, the proxy is too low-res to use for much except quick cuts. I have a few other ideas I'm working on as well.

I am definitely not satisfied with the HDV workflow, and I am not yet wanting to spend another $200 for the Cineform HD Connect, although I certainly now appreciate what it can do.

I simply select the "4x3" or "16x9" presets in the pan/crop, then re-size accordingly while maintaining the aspect ratio.

This gets you close, but as my pappy told me, "Close only counts in horseshoes and hand grenades." You will find several posts (I think I've read every in this forum on this subject) where people were complaining about having a thin black border at the edge of the frame after cropping. I now believe that this was probably caused by using the presets.

There are also some interesting bugs that happen when you substitute a proxy, change project properties, and then pan/crop. They are pretty nasty, actually. I am working on getting a repeatable case that I can send to Sony.

Sony

I am determined to find a better way to do this, because HDV editing is a kludge. We certainly should not have to deal with all this manual fiddling in order to generate and use either proxies or intermediary files. If such files need to be used, they should be generated and incorporated into the project in an automatic way. Even scripts (like Gearshift) are too crude (no offense to the author -- scripts are extremely useful, especially that one, but they are, by their nature, not "elegant"). I want to edit and create; I don't want to install plug-ins, scripts, etc., much less have to write them.

Alan Cooper, who invented Visual Basic and then sold it to Microsoft, spent most of the 1990s going around to companies trying to advise them on how to make their software usable. In the process he wrote a wonderful book, The Inmates Are Running the Asylum : Why High Tech Products Drive Us Crazy and How to Restore the Sanity. Every product manager and every engineer at Sony needs to click on this link, read that book, and start practicing what it teaches.