You're hired!

Very nice. I agree about HD being more than a sharper image. It pulls hte viewer's eye to so many more areas of the screen, creating a seemingly faster pace. Voice over may serve a greater purpose with HD in that it may need to guide the viewer; keeping the focus from falling too much on detail and more on purpose/message or flow.

I bought the Sony HDR-FX1 primarily for the purpose of producing this video. A previous "Voice for Video" (in SD, of course) demo that I produced with Vegas easily helped to bring enough work to justify buying this camera.... And it's a really neat toy. (A friend just spent about the same amount for a set of Golf Clubs.)

If I were producing video professionally, It would certainly make sense to buy the Z1, the "Pro" version of this model, but since I will probably never have use for the added features of the Z1, this model certainly seemed adequate - especially when you consider my limited video shooting experience.

I majored in Radio/TV broadcasting at Long Beach state, about the same time that a kid named Speilberg was attending. (A friend tells me he was in our class, but I don't remember him.) The program back then had limitations, however, one thing that I did learn was something about lenses. We had some ancient (even for back then) black and white cameras with turret lenses - no zooms. This forced us to plan our scenes, and to choose which lens would work for each scene. Our "Studio" was about 30x20, and we were assigned 3-act plays... We learned about such things as depth of field, aperture settings, what a long lense will do to an image compared to a wide angle, etc. If you wanted to slowly move in to get a close-up, you'd plan a "dolly" shot. "Zoom" lenses were considered "evil".

So I bought my new toy, took it home and fought off the urge to stay up all night playing with it. Though I had worked with "Pro" cameras in the past, and played around with some of my customer's cameras, this is the first video camera that I've owned, which I would put in the "Professional" category. Sure, the consumer camcorders I owned would allow some degree of manual focusing, aperture setting (sort-of) etc. but it became really clear really fast that this camera would take a little getting used to- and a couple of hours with the instruction manual. Also, I had to get used to the idea that I'm shooting something which is quite a bit different from NTSC 720x480 (ok actually 360x480).

I remember my college days, working on 16mm films - that's what HD seems to be more like, perhaps even more like 35mm. In college, I found that 16mm was great - if you were shooting Kodachrome, and directly projecting without lab work -- as soon as you went to a lab print, the image lost something. When I was a student, I really wanted to do something in 35mm, but few students could afford that.

The HDV images I've seen, after editing, look more like 35mm than 16mm "release" prints. Yeah there's a lot that's missing, compared to 35 film, but for 3-grand and a shooting cost of practically zero it's good enough.

First thing I noticed about shooting with this camera is how critcal focusing is. Now, I understand why there's always a guy with a tape measure on the set of 35mm features. Next, you find that there will be lots of things you didn't notice while you were shooting that shows up in the pictures. A little piece of trash in a picture, which would be totally ignored on SD will glare at you on playback of the HD image. Also, lighting and shadows are much more pronounced -- in a positive way. It makes the image feel more 3-dimensional.

---

Capturing the footage took a while to get down. Vegas did the capturing just fine, but the process of capturing the "m2t" images, then re-rendering to .avi "intermedite" files was driving me crazy. I downloaded the Cineform ap, tried the 2-week demo, and sprang for the 2-hundred for the "Connect HD" ap. (In my opinion it's about $125.00 overpriced, but I'm probably wrong about that.)

My initial editing/rendering experiments showed that I could edit the actual HD "intermediate" footage on my 3-gHz machine without using "proxy" files, provided that the project was short. Whether I would want to use something like GearShift, would probably depend on how large my project was, but for my little show, I preferred to see a sort-of High-Def image on my 1280x1024 monitor, and put up with system sluggishness. Hopefully I'll be able to get one of those "dual-core" machines sometime in the near future.

Even when viewing the image on a 1280 monitor, it becomes clear that you're not working with SD video any more -- everything's different - every image affectis you in ways different from NTSC. Camera movement, tilts, pans & zooms, rather than being something that draws you into the program, is simply annoying. Any camera movement seems to need to be extremely gentle. However, unlike SD, a still picture is very watchable.

There is an accumulated body of knowledge that's certainly worth tapping into that's very applicable to shooting HiDef images. For close on a century people have been learning the science of shooting motion pictures and that knowledge is available from bodies such as the American Society of Cinematographers. Information such as maxixum panning and tracking rates. Now some of that isn't entirely accurate with HD video due to the higher frame rates, it's potentialy more restrictive than necessary however if you stay within the safe zones for 24fps you'll certainly be safe at 50i or 60i.
One other thing that many are now starting realise about HD is focus is critical, I'll admit I've had a few chuckles over this, so much hubris over how wonderful being able to shoot with a shallow DOF is, well that's all fine and dandy until you realise that you're reducing your margin of error when setting focus. Add to that the larger image size of HD and the slightest error in setting focus and it's going to stick out like dogs balls.
One of the most skilled jobs in cinematography is the focus puller, I know when you see them on the credits they don't rate a prominent place yet those guys can be critical, typically they'll have a close and possibly lifelong relationship with a camera operator or DOP.
One thing that I've not heard mentioned about HD that I suspect is also going to catch a few by surprise is accuracy of color rendition. SD TV is typically viewed on a small screen in a lit environment. What's on the screen isn't the primary color reference, if it shifts it doesn't have that much of an impact. But viewing HD on a large screen in a dark room what's on the screen is the primary reference source for our eyes. Very small shifts in color are I think therefore much more noticable. Reading articles written by DOPs shooting major 35mm productions, one thing that grabbed my attention is how subtle the filters used on the camera and lights are, I'd suspect using such subtle shifts would be unnoticable on SD video. That might be tempered by the much larger color space of film over video but none the less it's something that might be worth some attention.
Bob.

It took awhile to get the knack of getting HD video and Vegas to work together. I found that it's best to set "Project Properties" to match the output from the FX-1 camera, regardless of what the final project will be rendered at. That means using the 60i HDV project template. Using that setting produces the best display on the "Windows Secondary" display.

On my 3gHz machine, I can see realtime video with preview quality set to "draft (full) ". If I need to see better image quality, I can set preview to a higher quality setting, but there will be a reduced frame rate display.

Also, I found that the "Deinterlace Method" needs to be set to something other than "None", either "Blend Fields" or "Interpolate Fields", or all kinds of weird artifacts will show up in the final render - doesn't matter if I'm rendering to HD or SD.

24 Frames:

Some of my ad-agency clients shoot film, or more recently, 24FPS video. They claim that studies have shown that recall is better for some reason when commercials are shot at 24P. I don't know whether that's the case or not, but I do like the 24P look. I also think my V.O. work "sounds" better, (an illusion, of course) when accompanied by video that looks like 24P. I was excited when I saw the 24P "Cineframe" feature of the camera, because it produced something that really looked like film. I definately want some material with 24P on my demo!

Unfortunately, according to many "experts", the camera does not actually produce 24P, or anything like it. Sony uses some sort of magical video processing to create the "feel" of 24P. It works, and works well, but don't expect to get real 24P from the camera's Cineframe mode.

The Cineform folks claim that their capturing tool will convert 24P Cineframe to true 24P. It does that, but I found that the output is not smooth, with the feeling that frames are being removed. Cineform now recommends that you shoot 25P Cineframe (Which the U.S. version of the FX-1 will not do), and convert from there.

I remembered that Spot had suggested shooting at standard 60-I then converting to 24 by rendering at 24 from the Vegas timeline. I tried it and it works - and works very well. From now on, whenever I want 24-Frame feel, I'll shoot at 60i on the FX-1 and use Vegas to create the 24fps video.

Samples of test material rendered at 30/60, and also at 24 are available at:

(30FPS)


(24FPS)

-Travis