DV for TV
 

I got offered a job to do a commercial for national TV (which I took), but unfortunately I won't have my FX-1 in time (they need this thing next week). I told them that all I had access to was a DV (GL2) and that it may not look good against commercials shot with more sophisticated cameras. They don't care. They just want it done.

However, I care. This thing is going to have my name associated with it and I don't want it to look like crap.

The spot is 30 seconds and will be comprised of interior and exterior shots. It's a comical piece with one actor. Present day. I won't have any lighting equipment, but I am renting a boom mic. My question is, how should I set-up the camera to minimize its consumer-looking qualities? i.e. shutter speed, exposure, etc. Also, do you think I should try using Nattress or Film Effects to give it more of a film look or will that just be adding insult to injury? (I don't want it to look like a silent film)

I'm excited about this, but a little worried at the same time. I completely disclosed all my assets (or lack thereof) to these guys before accepting the job, and they were cool with it, but I have a feeling they had no idea what I was talking about. "Yeah, yeah...I'm sure it will be fine...just get it done."

Any suggestions would be much appreciated.

Sorry...I know this is the post-production forum, and I know I asked how to "set-up" the camera. What I meant was: How should I set-up the camera so that I have the most options in post? I think that is where the bulk of my work will reside. The GL2 doesn't have a lot of options in and of itself. Thanks.

Without the specifics of your creative concept it's hard to tell, but if you're doing color correction in post, it's typically best to keep the video clean and well exposed and do the trickery in post. Start with a clean, well exposed baseline and avoid any use of gain. I like using Magic Bullet for creating custom looks in post.

Shooting without good lighting is folly if you're really serious about having it look good. Now "lighting" need not be artificial lighting units, but using natural light to your advantage, but one major thing that makes DV look like amateur video when broadcast is when it's shot with lots of gain to make up for the lack of proper lighting. There are other factors, but this is one of the dead givaways of professional vs. amateur production. Contrast and separation between foreground and background (both lighting and depth of field contribute to this) is an important factor in high production values.

And then there's acting and production design and ... but that's getting off the topic.

The camera shouldn't really be a hinderance... if you check out the DV for the masses forum, some people are doing really good work with the GL2. The Silencer (see Simon Wyndham's posts) and Sundowning are two examples of things shot on the GL2.
Talent+experience makes a much bigger difference in how your footage turns out than what camera you shoot on.

You definitely DO need to light your footage well. Color correction is not the way to compensate for no/bad lighting.

Lighting is an art in itself and it would be hard for people to explain how to do it on an online forum. You can read about it, but you also need experience lighting setups and knowing how light 'works' and translates into what you want.

But if you have no lighting gear, here's what you could do:
A- Find a location that looks great. Going by eye alone can be a little misleading though... contrast is much higher on video than to your eye. If you just point your camera at something, you may not like all the nasty shadows everywhere.

B- If you don't like the shadows, use a reflector to get rid of them. A really easy reflector is to tape crinkled tinfoil onto a piece of cardboard. If you use four side of a box, you can make a reflector that folds up better. White foamcore (or other white surfaces) work too, but they're not as directional.

Not sure how you'd hold the reflector without a C-stand. ??Maybe use a table + gaffer or painter's tape. Just position it and tape it in place.??

If you look at Sundowning, I believe they did a lot of their lighting using a reflector.

C- So a lighting setup might be like this:
The strongest light source in the room (i.e. daylight streaming through curtains would give you beautiful soft light) would be the key/main light. One side of the talent's face will be lit a lot stronger than the other side.

Use a reflector to fill in the shadow on the dark side.

To get foreground/background seperation (i.e. you can clearly see the elements seperated), move the camera and the talent and the furniture around until the colors+brightness between the talent + the background is not the same. Make the background look nice by re-arranging stuff there.

2- I have a basic tutorial on how to do some color enhancement with Nattress Film Effects.
see http://www.glennchan.info/video/FCP/..._tutorial2.htm

Magic Bullet Editor's is also a good tool for some looks... you just go through em until you see something you like.

3- Camera setup:
NO gain. Make sure you have enough light that you don't have to use it.

Depth of field: If it's a close up, you can throw the background out of focus a little. Move the camera back + zoom in. Open up the iris as much as possible.
The shutter can lower the exposure (compensates for opening the iris). This may not work that well on fast motion... you'll get strobey motion. But some people like that (Gladiator for example).

Overall exposure: In my opinion, you should expose close to the final exposure you want. Color correction doesn't give you that much leeway to change things.
See http://www.glennchan.info/video/exposure/exposure.htm
Do your own test and figure out how the camera's zebras corresponds to what you like. An ok rule of thumb: With 100% zebras, expose to the point where highlight detail just blows out.

Hope that helps.

Guys -

Thanks so much for all of the knowledge. I am going to use some (if not all) of your suggestions in my shoot and editing.

One question:
Do you mean that only the very brightest areas of the shot should be blown out i.e. white shirt as in your example? In other words, are you recommending for me to expose to the point where I do see zebra, but to minimize the amount by only allowing it to creep in on the very brightest spots in the shot?

Keep in mind that anything over 95% is not going to show any detail, and you want to preserve texture in all but your speculars (very bright spots like reflections on shiny objects) so a white shirt should still average well under 100 IRE so you see texture in the cloth. Avoid white, light grey shirts will read white, but will show more detail (washing white shirts in something like tea is an old trick to bring down the brightness). The goal is to keep the appearance of texture (staying under 95%) unless your intention is to have something to be totally blown-out without any detail, which flat and unnatural and a dead giveaway of amateur video.

Accurate exposure using a waveform monitor is ideal in situations (like a commercial) where you have time to fuss, zebras only show you the "high water mark," a waveform monitor shows you exact values throughout the scene. A laptop with DV Rack offers a versatile tool including a waveform for use on the set.

DV Rack also provides "black zebras" (which I wish cameras had) to show you where your shadows lie in the range from 0 to 100. Detail in the shadows is another important aspect of creating scenes that look more dimensional and real. Filling in the shadows with light so you see texture instead of pure black is another aspect of higher production values, of course, some areas of pure black and some areas of pure white help you perceive the scene as more dimensional, and how much is blown-out white and now much is totally black is a stylistic choice.

An an aside, one often overlooked feature of the Panasonic HVX200 and DVX100 cameras (which is trickle down technology from the Varicam) is the "marker feature" which has a spot meter giving you very exact IRE readings of a tiny portion of the scene, making exact exposure determination easier.

With many DV cameras, they'll actually record information above legal white level (a Y' value of 235 in digital units, or >100 IRE in equivalent analog terms). Most NLEs are able to access that information and bring it down into the legal range. If you do nothing, that information will just be clipped.

Some cameras will apply a knee to the highlights, which kind of "squishes" them to get additional detail. I don't know if that makes sense... but the color/exposure will behave differently in the knee region.

Anyways, point is... if you want to be nitpicky like me, you sort of do have detail above 100%. Although not much.

2- I would prefer using a broadcast monitor over a waveform monitor for setting exposure. This is if you want to favour aesthetics over engineering. You see right away what you're getting.

Don't use a consumer TV for setting exposure, some will make underexposure look ok.

Anyways, it's just a different way of doing things. If you don't have DVrack or a broadcast monitor, then do a test with your camera. Figure out how the zebras and how the LCD corresponds to exposure. Your camera probably has a reading for f-stops... shoot a scene and just keep going up a notch. Shoot it from underexposed to overexposed, and figure out how the zebras correspond to your exposure.

Well, there has been a change of plans (for the better?). The mic I am renting needs an XLR input and the GL2 I was going to use doesn't have an adapter. So, I am just going to rent a Z1 to 1) up the production quality, and 2) avoid any technical surprises. That said, a lot of your suggestions are still a huge help to me. Thanks so much. And if you have any more suggestions, please let me know...otherwise, I guess it's off to the HD forum.

GL2 for broadcast tv= yep it works

DV is not the best for commercial work, mainly because they dub those tapes over and over and over. You CAN use it for acquisition but edit it UNCOMPRESSED, not in the DV codec. If it is a national commercial, there should be budget for rental, why not rent something a little more robust?



ash =o)

Hello Ash,

What's the point??

If you acquire in DV... it IS DV. It's already compressed. It will be compressed for editing and after export.

Am I missing something here??

Thanks!

Yes, because once you add effects, color correction, etc. etc. it is RECOMPRESSED many times over. If you were not doing graphics, CC, effects, straight DV would be fine.


ash =o)

To the best of my knowledge, all NLEs work with material uncompressed internally.

I'd be interested if there were some sort of test that demonstrates the difference.

2- The one test I've seen does show a difference... this is because the DV deck used applies chroma interpolation on its SDI output. If a software DV decoder was used instead, some/many do not apply chroma interpolation. You could of course apply chroma interpolation yourself... i.e. FCP color smoothing or Nattress' G Nicer (which does a better job than normal chroma interpolation).

Hello Again Ash.

I can see the possibilities, although I didn't think that PPro treated filters and transitions that way.

NOTE: I would like to hear from someone at Adobe to confirm and explain that... or any other soources.


So then, how do you edit DV "Uncompressed"?

To capture the DV you setup a DV project, are you saying that once captured, start a new project (uncompressed) and import the clips?

How do you define a PPro-2 Uncompressed Project??

Thanks Again !!!

Ash, sorry, I disagree. If the editor is set as a DV timeline, your loss is minimal (0.2% at its worst per compression IIRC). You might be confusing non-native MPEG editors with DV or native editors?