"I'm not sure why under-exposing would help?"

Let me explain our thinking, as muddled as it might seem:
We are of the belief that CRTs are dinosaurs rapidly on the way to extinction in the home consumer market. LCD, OLED, and Plasma make up an incoming tidal wave. Once a consumer has seen HDV on LCD, OLED, and/or Plasma said consumer will blame his aging CRT for all color saturation defects. Consequently, we are shooting and editing for the new consumer flat-panel technology (2in to 50in), not CRTs and not film movie screens. Unfortunately, the driving software for the new technology choses how the displays will, well, display, and they all seem to make different choices (ref: Vegas vs Nero). Now the consumer gets 0-100% ranges for brightness, contrast, tint etc that let the consumer do all sorts of wacky things. So the ultimate image *as viewed* is pretty much out of our control (other than our staying within legal broadcast limits).

Consequently, we under-expose to avoid blowing out details, and let the consumer have at it. On our editing equipment, this slightly-underexposed footage when displayed by Vegas gives us a rich tone, which when left uncorrected (as we do) does over-saturate the final image some. *This is intentional*, is our production style, because we've noticed that the home consumer prefers higher-intensity saturation products in that they will survive their effort to drive brightness and contrast to the screen's fullest potential.

We do the same with low end sound frequencies. We probably have a bit too much low end, but the consumers (both men and women) seem to prefer driving their butt-kickers to the point of shaking the floors. So we give them the data to do that if they wish.

Instead of fighting the flow, we go with it. It might be pure coincidence, but Vegas at factory settings does "seem a bit washed out", so we took advantage of that at the camera end (Sony's Z1Us).

Now our work flows through *without having to calibrate anything except the cameras* and the results pretty much look uniformly "good" on any non-CRT display, as well as digital projectors. As for how our footage looks on CRTs, well... think of that famous quote from Gone With The Wind (grin).

Thanks for an informative reply David. I'm thinking now I owe Spot an apology because I have said his recommended Z1 setting of +3 for color was too much. I'm going to set mine that way with 80 for Zebras and see what I get.

Your reply David stirs another question in my mind. You say you never color correct while I find I color correct 4 out of 5 of my shots. Allowing for the fact that your shooting is much better than mine (I'm not a full-time pro) this still seems a lot. Do you discard all your footage that is not 100% lighted right or exposed right to allow you not to color correct? I'd also like to know how you calibrate your Z1 to give you your result. I assume you do more than crank the color up and use the film gamma.

Michael

"You say you never color correct while I find I color correct 4 out of 5 of my shots."

Good point. We do color-correct occasionally to balance cuts from different cameras that happened to be shooting the same scene but accidentally at different settings.

"Discard footage?" At the end of 4 hours of shooting by 3 cameras we'll typically have about 45mins of combined lineal footage. We've never filled a tape, and some have as little as 30 seconds on them total. Out of the 45mins we will use the best 4-5 minutes. So, yes, footage gets "discarded" - but not because of color problems, just because some shots tell that part of the story better than others.

It helps to understand that we don't shoot weddings or corporate videos where the footage on the screen *must* match the exact dress color and skin tone of the star. We get to invent the colors, so to speak. For instance, when we shot elephants running through the park a few months ago, we set the cameras so they captured the richest greens possible. Neither Emily nor Ruthie are going to complain about their skin tone, as long as it still looks "gray", but an audience would notice wimpy-colored grass and leaves. In other scenes we have a 17yrold actress and an 85yrold actress side-by-side from pre-dawn to about 10am. Nobody questions the continuing change of skin-tone from gray to sunrise orange to pink to near-noon white in the 4-minute span of them jogging 5 miles. What would otherwise be considered wacky color actually serves to accentuate the passing of time.

Other examples: We shoot 90% of our *indoors* using available natural light, and leave the cameras on outdoor white balance. Our interior footaage is unusually warm unless we add some 5.5+k lights to cool the scene down. But that's the style to which we are accustomed and for which we are becoming known.

When we shoot our sitcom pilots, we will shoot in a florescent-light building with practically all-soft lighting. We'll probably use black cards to help create some facial shadow, or turn off a bank of lamps here or there. It will be the rare occurance when we set and strike lights, and those will probably be up just to create "colorpools" through which actors will travel to help with depth-illusion. We'll use high-k lamps and keep the white balance set to outdoors. Why? All our stories have people at windows and moving in and out of buildings all the time, with the cameras shooting both ways through windows and doors. We do have to adjust focus and iris, but never white balance. Well, almost never...

What all this does for us is to create a color style that is easily recognized as ours, and is pre-prepped for Vegas. As I mentioned, it looks harsh on Nero. Not everybody will like our look, but it works for us, makes the talent and crew happy with the results, and relieves some burden on the editing suite.

To put this into our overall business perspective, we *first* write and produce the music for our productions before we ever turn on a camera! We're 3 years into production on some scripts for which we've yet to purchase a single tape. We think in terms of 5+ years to get a production out. We might be unusual in this aspect for an independent company. The advantage this gives us is twofold: (1) we already know what we have to shoot for (the mood, pace, hit points, color effect, etc). and (2) when the footage is edited we don't have a budget/time problem getting the music that needs to go with it.

To be fair on your question, when I personally do stills for the web I spend enormous amounts of time getting each photo's color matched (or very far apart). That is because they are viewed adjacent to each other and the eye tends to make comparisons. A photo on the web that was taken from a shoot probably looks *very* different in color and tone than what you'd see on the screen. But then, nobody runs their browser simultaneously while running the movie. Of course, that's going to change in a few years when movies have hot links right in the story ("click on the milk carton, get one delivered right now - oh, and your fridge says you are low on eggs too...").

Comparisons distract from the message (or add to it, as the case might warrant), which is why most people chose to color correct. I'd be nuts to shoot a wedding using our style. No, I'd be dead and the jury would acquit the bride on "justifiable homicide." Similarly, if we were to be shooting a 30-second commercial that showed 15 shots of the same woman, we'd be tweeking lights, cameras and color wheels like Carnival Carneys in an effort to minimize any perceived change in skin color. That 30 seconds of finished footage probably took 2 days of constant shooting to acquire, and the lighting conditions (especially using natural light) would have varied dramatically.

But back to the very original point I was hoping to make in my previous post: When the end-user gets to vary the components of the output real-time, our business philosophy is that it's a better point of valor to give the customer as much data as possible, a default setting, and then just let them mess with the product as they see fit. Consequently, we no longer agonize internally over reaching superb color balance (and we don't shoot weddings). Dramatic and story-telling color, absolutely. Perfect sound, absolutely. But we know in our hearts that we'll never be the ones to make a movie as beautifully-color-crafted as, say, "A Good Year."

We strive for affordable, cute, funny, plausible action stories where characters, conflicts and situations are more important than color. We'll put the time and money into a rich sound track long before we go crazy over color.

Great David- thanks.

So you don't have a Z1 profile set that you stick with. You are tweaking, scene by scene, color level, and color phase, skin-tone, auto WB presets etc. I'll have to make time to try that.

Cheers,

By exposuring to put whites at 80% you're compressing the dynamic range, which is already very limited. While suiting a particular shooting style I think it undesirable. Of course it will certainly remove any needs for levels adjustments. I don't see this "washed out HDV", but the cameras deliver a range of 110% and you do need to adjust levels in post. Is this what you're seeing as "washed out"?

For me, no, that's not what is meant by "washed out."
I find that all of my "flat shots" (ie, I don't do anything in the camera to adjust it), need the saturation boosted as sell as contrast. Of course, I tend to like the "dramatic look," so I may not be part of The Norm.
I'm always amazed by colour correction. I take what I thought was a good shot, add the Saturation and Contrast (usually done with Colour Curves) and the resulting shot is stunning. Just have to be careful about keeping everything Colour Safe at that point.

Serena,
"By exposuring to put whites at 80% you're compressing the dynamic range, which is already very limited."

Correct. However, until I did that, the camera ops would blow out faces. That was unacceptable, so I had to find something simple as a guide. Zebra at 100% left too much margin for error. With the zebra pre-set to 80% I ask the iris to be opened only up to the point where zebra shows on the upper cheeks. So far, this has worked for all skin tones, all camera ops. Yes, it's a tradeoff, but getting a compressed dynamic range properly framed, tracked, focused, and containing all the data is preferable to one I cannot use. Unfortunately, I do not have the time or money to properly school the operators on the Z1Us. Moreover, we use a checklist for setup, not unlike a pre-flight pilot's list. Finally, these ops all use other cams on other shoots for other directors and themselves, and have individually requested that I not innundate them with camera setting requests that differ than those to which they are already accustomed.

Keep in mind I don't have you and Spot working the other two cameras, and that the bulk of my shooting with more than one camera is primarily shooting with crew and talent who have limited time budgets. Re-shoots are seldom possible (as would be the case with live-action documentaries or sports). Two takes or fewer is my goal.

I am blessed with actors who hit their marks and deliver their lines correctly the first time, and within character. I am blessed with crew that show up early, leave late, and learn to live with my style, anticipating my requests, and remain eager to return for the next shoot.

Part of the above success is because I'm willing to make color and range concessions that translate into good takes the first time and abbreviated shooting schedules. It's purely a business decision, and it's been working well with these Z1Us.

MH:
"So you don't have a Z1 profile set that you stick with. You are tweaking, scene by scene, color level, and color phase, skin-tone, auto WB presets etc"

Not scene-to-scene. See my previous post to Serena's observation. We've made same-camera presets for white balance and color. We turned off any skin tone option. We shoot 60fps (interlaced). We found a common setting for each "feature" that seems to work "universally" at our lattitude (between 40-44 North) and air quality. Things blue out a bit in the winter, get a tad orangie in the summer, but that's fine with us as long as the light is proper for the season in the scene. What is left for the cam ops is to pick the requested preset I'm calling for the scene, then pay sharp attention to focus, framing, iris, & zoom.

Now, if I see the color "out there" is not what I want, I'll add or subract actual light (almost always for color, not for intensity). I've been known to reschedule (or even kill) the shot and move to the next one - the only camera I'd re-program would be mine, and I never do that when I'm shooting tandem or triple. We've been known to wait for clouds to come, and we've waited for clouds to go. We've waited for the sun to rise and we've waited for the sun to set. We've said "Hey, it's raining cats and dogs and/or it's foggy - let's go shoot!"

As often as we can, we'll use old still photography tricks that are still free, just as we'll move a mic closer to get better, richer sound.

If you're running into problems with blown out highlights on faces this sounds more to me like a lighting issue than anything to do with the camera settings.
The HDV cameras do seem to blow out highlights worse than the camparable SD cameras and I do realise that at times there's little you can do about lighting. Still, that's where I'd be starting.

The argument for setting zebras to 80% is fine because of the circumstances and burnt-out highlights must be avoided (can't be fixed and look terrible). I do set zebra at 100% but am very careful to close aperture (and/or graded filters) to remove zebra bars showing. This does require constant monitoring and sensitivity to the issue, so well understand preference to under-expose. Just been doing post production where most highlights are burnt, so going to be offering zebra advice to the videographer who I expected to know better. Perhaps he should be setting 80%.

Bob,
Problem #1: The primary source of light is El Sol.

Problem #2: Camera ops open up the iris in an attempt to get shallow depth of field. To them, this is "proper" because in their other lives they live with aperture-dominant settings, so things like shutter speed or gain adjust to compensate. We shoot 0dB gain and 60fps locked. Oh, yes, I'll see that we need an ND or two, and call for that. Still, that iris is almost always too wide open.

Problem #3: Once *SOME* zebra shows on the Z1U screen, it is very easy to continue cranking open the iris until zebra seems to "go away again" - actually, the entire screen could be zebra-filled but the zebra routine tones itself down, and I have to admit that on a quick view of the screen it doesn't look over-exposed. I'd wish for a "PLEASE CLOSE THE IRIS" comment at any pre-set value of zebra showing. Or wireless remote holographic images floating in front of me so I could check their work like we do cams on jibs. But there comes a time when we simply have to trust the ops to do it correctly, and to assist and massage their egos in that direction they need a foolproof guide. 80% zebra works.

The Z1 has 2 levels of zebra. Using those should belt & braces the matter. I must admit I've never cranked open enough to make the zebras become invisible and didn't realise that happened. I've seen highlights visibly blow out.

Problem #1: The primary source of light is El Sol.
=========================================

Understand the problem, him and 'lit for stage NOT broadcast" is my personal bugbear.
El Sol can be tamed with scrims however although these things are cheap they can take a lot of manpower to rig, even more manpower to keep in place, and of course El Sol doesn't stay still.

One of my biggest clients who I do post work for manages to capture great looking footage of difficulty lit stage productions, something I've yet to master. Let's just say I'm only know realising the difficulties of the task.

Basically he zooms into the hottest part of the scene, sets exposure so nothing is blown out, zooms out and never touches exposure again unless there's a dramatic change in the lighting. This technique seems to work, yes at times a lot if not almost all of the frame is black. However nothing is blown out, noise levels are very, very low, colors look rich.
Of course this guy has that camera to his shoulder almost every day, I think that counts for a heck of a lot.

Bob.

That's good technique. One must expose for the brightest element (that is, make sure it isn't clipped).

Like David I lock in 0db gain and 60fps, yet I try to keep the iris wide for impressive dof and so would like more ND settings on the camera (I can't afford a matt-box yet nor do I have the luxury of a field monitor). I have being doing what Serena does and set zebras to 100 only just allowing them to show and I have noticed sometimes the difference between "seeing the zebras" and feeling sure "I'm not seeing the zebras" can be several f-stops. Also I find with the cameras limited range a bright sky ( all my work is outside and in the sunny Californian desert) sometimes just has to blow to get a reasonable shot of the subject, don't you agree?

To David I'd like to ask: If you don't tweak camera settings scene by scene, what did you mean by "calibrate" when you said "our work flows through without having to calibrate anything save the cameras." As you say your color choices are made in the camera and not in Vegas how do you do this without tweaking settings of color intensity and phase or WB scene by scene?

Yes, you have to make sure the camera is in full manual. I haven't observed several stops between zebras and no zebras and I would have said the difference between them just showing on highlights and not is small. Normally I close the aperture to well under and then open until zebras showing and close down until they're not; I can't recall noting the f-stop readings (I would have said maybe not more than half a stop). Of course there are times when the bightness range of the scene is such that less important stuff in the background has to be let go, but you don't want it in the foreground or within the area of attention (if I may call it that); white blobs are distracting. In cases of doubt I use a piece of white card in the same light. Should there be any difficulties in effectively using zebras then setting them lower provides more margin for error. If you are trying to balance DOF and exposure, correct exposure must have priority.

I could be quite wrong here but I think zebras are a pretty blunt instrument. Zebras show you when something hits, well whatever you set zebras at. Let's say 100% for the minute. Take a hyperthetical case, the entire scene is 50% grey. Should anything in that scene have any zebras showing, somehow I think not.
So getting a bit more real world, what if the scene contains nothing of a high luminous value, say all reds and yellows, again even with zebras at 80% I doubt you'd want them showing, you're going to over saturate the scene, it's not technically blown out but it looks like c**p.

Maybe I'm wrong here but the more I think about this and get bitten in the butt on shoots and read what David is saying and dust off some old grey cells the more I think you simply cannot judge correct exposure using anything just in the camera. You have to factor in what the camera is looking at. But whoa that takes a lot of skill and judgement. So how to do it by the numbers. The way it's been done for nearly a century, by measuring how much light is hitting the subject.

The sprockets guys do this as a matter of course. They light to a stop. They don't light to a density on the film. I think this is where we've lost it. They measure the amount of light hitting the subject, not the amount of light reflected off the subject, to get the desired exposure.
Of course they are concerned by film density, they decide on what stop / light level gives the density that gives the 'look' that they want. Once they've got that from shooting tests they lock it in pretty much. Then they set the light level of the subject to give that reading on their light meters. If the subject changes this is correctly recorded on the film.
Davids problem (and probably many of us as well) is we're trying to get everything in every shot at some magic level. Reality doesn't work that way and cameras don't cope too well, When the talent walks into shadow, they get dark. Opening the iris etc to compensate is wrong but it's what zebras will tell you to do and it's what AE will try to do as well.

Generally agree. But the zebras don't tell you anything other than you're going to clip if you go there. You never have to fully expose up to 110%. In film you tend to light to a stop and incidence meters give you greater consistency than reflected light meters, but still you have to consider the reflectivity of the subject , the lighting ratio and you look after the blacks. You don't light drama and family comedy the same way. The thing you always do in film is know your film stocks and how each responds and which will produce the look you want, in combination with post.
So you start off with setting light and shadows, you know the characteristics of your camera and you've pre-planned post. Every detail you want in the final image has to be recorded and that governs how you expose the shot. The talent walking into dark gets dark, but unless you want them to disappear completely you'll see them against light from a distant streetlight, a window, or whatever. But of course you don't open the stop because the effect will be lost. An example that comes to mind is The Third Man, when Harry Lime is seen for the first time, in a dark doorway with just his shoes illuminated by the street light. Not very effective if you open up so you can see him standing there; certainly a case where you have to look after the blacks.
But Vic Milt knows a great deal more on this sort of question.

Bob,
"So getting a bit more real world, what if the scene contains nothing of a high luminous value, say all reds and yellows, again even with zebras at 80% I doubt you'd want them showing, you're going to over saturate the scene, it's not technically blown out but it looks like c**p."

Correct. That's why we have directors (grin). I'm operating one of the cameras most of the time. I call out "I'm on one ND, 2.5meter focus, 25 zoom, 3.4 f-stop. What do you folks have?" (I don't say "guys" because Chyane is manytimes our 3rd cam op.)

The "rule of thumb" that I presented previously is to place zebras "on the upper cheeks of the talent". It's a people-picture thing. When the subjects are not people all bets are off and I call a huddle (including the sound dude). Nevertheless, I try desperately to avoid *reprogramming* the cameras for color.

Yep, we shoot 50% grays (I did mention rain and fog, yes?) I agree with you, Bob, that wetware and experience are the best tools to apply (I did mention still photography tricks, yes?). On this subject I can only agree with you Bob (I usually do, but over the past years I've just read, not blathered, on this board).

"When the talent walks into shadows..." Uh, no, I didn't suggest we change the cameras, even the iris, during the shot. On the other hand, I notice I didn't say we lock the iris setting for the shot either. I did point out the reason we *ususally* stick with white balance for indoors is that we regularly shoot through windows and doors, and that our talent regularly moves through portals. This gives a reasonably-consistent color balance as we transition the shots, without having to reprogram the cameras. Remember, *our* interiors are normally shot with natural or 5.5k light.

We lock the settings for a shot. Talent moves in and out of light pools. Faces get darker and lighter just as happens in the real world. If we add light, it is usually color (not intensity) for the purpose of making light pools - which aids in creating the illusion of depth by directing the talent to move from pool to pool (the story does need motivation for the movement, of course).

We use natural/available light, the major pros and cons are:
(P) Cheap, doesn't required a lighting truck or power source, doesn't cause the talent to sweat (look at the closeups of Santa's Elves in Santa Claus 3 - at the North Pole the kids are all sweating bullets!).
(C) Risk losing detail (or getting noise) in blacks, hard for the camera ops without years of b/w still photography experience to pick a good setting, Director has to direct the camera ops.

MH,
"If you don't tweak camera settings scene by scene, what did you mean by "calibrate" when you said "our work flows through without having to calibrate anything save the cameras." As you say your color choices are made in the camera and not in Vegas how do you do this without tweaking settings of color intensity and phase or WB scene by scene?"

You make a good student, asking pointed questions. A moment of background here might help: I was raised in So Cal 50s+60s, and understand the desert light (and shadows). My "calibrations" currently in our Z1Us probably would be changed were I shooting out there.

"How do you do this..." is answered by the needs of the scenes. We don't shoot weddings. Let's take weddings as an example then. The hair and skin color of the bride *must* stay identical from shot to shot, indoors to outdoors. The bride couldn't care less about Bob's observations of naturally-occuring changes in light in the real world. The day of the wedding is a fantasy world. The bride (or her mother) sets the rules. You betchum Red Ryder, I'd be setting the camera up specifically for the light of that one wedding. I'd have color-correcting gels to place over the two or three portable (florescent) lights, and a grip handling each light and we'd be using radios with ear buds and dressed in tuxes (or sheepskins or robes or whatever the bride's wedding theme is).

First, I'd change all my presets to match the range of lights and colors *for that day, for that environment* - and then as we moved room to room, indoors outdoors, I'd pick the correct preset. I *still* would have to adjust iris, ND, focus, zoom, framing, where I stand, light placement, and gels.

Second, I'd recognize that weddings are a couple of hours of sheer boredom (the buildup, the march, the service, the vows, the kiss) and then about 30 minutes of total chaos where nobody, even the bride gives a thought to us pesky-in-the-way camera folks. But of course that's where the money shots come in for video, and where we would be shooting run-n-gun. *NO TIME TO REPROGRAM THE CAMERAS*

So, all "calibration" is done beforehand by anticipating the needs of the story as placed in that particular environment. Pre-"Calibrate" the dressing area, the alter, the grand entrance, the exit, the escape, the limo, the reception, and the bride-groom-party photo ops.

But we don't do weddings. We shoot our own stories so that we can assemble shots that all have the relative same color characteristics, "calibrate" for that, and go to it. Our shot list might look like "Scene 32, 33, 35, 6, 7, 9, 48" The choice of shot order being dictated first by available talent and second by changes of costume.

Now, here's the really neat part. We write the scripts with shooting in mind. The first place we "calibrate" is in the script! Since we don't want the cast and crew to complain to the director, and the director in turn to fire the script writers, we (as script writers) think through the shots from a safety/sound/color/lighting/camera calibration perspective and write accordingly.

Note "safety" and "sound" come first and second on the calibration list.

Geezum, we try to leave those cameras alone during and between shots. With a Z1U if you accidentally poke nearly any button or the jog wheel while the shot is rolling, that shot in that camera is blown. Gone. And the op will almost *never* know it until the editor starts screaming.

"But Vic Milt knows a great deal more on this sort of question."

Correct.

In addition to all useful information being presented so far please be sure to take a look also at the simple StudioRGB versus ComputerRGB issue.

Vegas reads HDV m2t files in StudioRGB space, i.e. 16 to 235. Therefore you will need to set your preview monitor to StudioRGB as well (you already knew this, hugh ;-).

Now, when you render out to m2t or HD-mpeg2 StudioRGB is pretty fine.

However when you render out to VC1 (aka WMV-HD) you need to convert to ComputerRGB (0-255), otherwise the colors appear "washed out".
Probably you knew this as well (then just forget this post). But I was stumbling over this issue when I made my own first steps into HD.

Of course for serious work you should always control the output with the video scopes.

Martin