Exposure issues

[Linc Kesler]

I've been having quite a lot of fun getting to know my GL2, and even though I had to put it to use fairly soon after getting it, I got very usable results (I'm doing oral history documentation in on-site locations). I have a background in still photography, so I'm familiar with the basics of exposure and lighting in that context, but I'm relatively new to video, and now I have some more detailed questions that I hope you people with more expertise will be kind enough to answer.

First, a question specific to the GL2. I think I understand all the auto modes and how they behave, and in my work sessions, I'm using manual, so that hand movements, etc. don't disrupt the exposure on the more critical facial areas. Generally speaking, I've been shooting about half a stop below the metered exposure (more on that below). But I've also gotten interested in the custom preset option, because I've found that a slight boost to color gain and a slight lowering of sharpness works best with my subjects' skin tones, and with the available light environments I've been working in. My qestion is about the "setup level" control. This control seems to behave in a general way like an exposure control, but as near as I can determine, it does not alter the three normal means of controling exposure (shutter, aperture, gain) at all. Is it, then, a purely electronic control? And if so, does it have any characteristics or consequences I should be aware of? So far, the setting I have been using (almost maxed to the - side), has looked good, similar to about a 1/2 stop less exposure, but I'm curious as to what is really happening. The change, when I click in CP, though not affecting the metered exposure or camera settings, does affect the zebra bars, getting them to the minimal highlights I think work best in most of my scenes.

That brings me to question 2. How do more expert people determine exposure? I know that experience and review of results are probably the best way to get the look you want, but do more expert users rely on the zebra bars to control the highlight exposure, the way a still photographer shooting slide film would do? Or do people attempt to meter a critical area and go with that (as an "average" or neutral meter reading?)? Or do people gage the look, once they have enough experience with the viewfinder or screen (or monitor) to have confidence in what they are seeing?

I'd appreciate any help.

One final note for anybody interested: I debated quite a bit whether to go with the GL2 or the VX2000 for this project, since low light is an issue. So far the video results even in contrasty low light (side light from a window) have been good--well within what I need. I went with the GL2 because of the long record of reported preamp noise problems with the Sony audio circuitry, esp in manual mode (which I thought I needed to handle the silences in interviews, and I do use manual now). The GL2 produces very clean audio with my setup (Sony corded lavaliers into a Studio 1 box), and the volume controls on the Studio 1 allow me to set the mic levels on the camera, and turn my mic down on the box when my subject is on a roll. That's nice, because I am close enough to the camera during filming (to keep subject eye contact close to the lens position) that I was picking up camera noise through my lavalier. So I guess my point is that this is a good camera when clean audio in a quiet setting using external mics is important. I'm really pleased with the results.

Thanks,
Linc Kesler

[Ken Tanaka]

Linc,
Thank you for passing along your experiences with your GL2 so far.

Re: exposure measurement, indeed I believe that most folks use the zebra display to determine exposure levels, especially now that the GL2 has variable zebra settings (like the XL1s). For example, many people use an 85 IRE setting to shoot faces with light complexions and 90 IRE for darker complexions. Etcetera. The first function I used on the GL2 was to set the Custom Key to activate/deactivate the zebra.

Some videographers do use light meters but the fact of the matter is that the whole camera is a giant light meter. (Yes, I know that shooters using big rigs with an eye toward film transfers need finer exposure measurement and control, but I honestly think that even this is alot of puffy b.s. )

Experiment.

[Jim McNally]

<<<-- Originally posted by Linc Kesler :

My qestion is about the "setup level" control. This control seems to behave in a general way like an exposure control, but as near as I can determine, it does not alter the three normal means of controling exposure (shutter, aperture, gain) at all. Is it, then, a purely electronic control? And if so, does it have any characteristics or consequences I should be aware of? -->>>

Linc, the setup on a camera is used to establish the black level of the camera. The equivalent of the audio VU meter is the video waveform. It measures the brightness of a scene with a scale that runs from 0 to 100. Black or the base of your video would show up at 7.5 on the scale if connected to a waveform. White would show up at 100. The 0-7.5 section is not normally used for video. 0 black can be used by graphics that require a "sub-black" level for keying purposes.

The setup adjustment allows you to play with where the base level of black falls. A contrasty scene can be made less contrasty by adjusting this setup. Conversely a dramatic scene can be made more so if you "crush" your blacks.



<<<-- That brings me to question 2. How do more expert people determine exposure? I know that experience and review of results are probably the best way to get the look you want, but do more expert users rely on the zebra bars to control the highlight exposure, the way a still photographer shooting slide film would do? Or do people attempt to meter a critical area and go with that (as an "average" or neutral meter reading?)? Or do people gage the look, once they have enough experience with the viewfinder or screen (or monitor) to have confidence in what they are seeing? -->>>

I primarily always go by my zebra settings which I know are at 100%. I know others use them at other settings but for me I want to know the extremes of my scene and 100 units is as far as I can go. If I have them set at 85 I would risk underexposing elements of the scene that fall between 85 and 100. There is always something you can place in the scene to hit 100%. If your scene has a lot of darks in it, when you are doing your white balance, check your exposure for 100% at the same time and then leave it there.

Jim

[Jeff Donald]

Set-up (black level) has been discussed several times recently. One of the best threads on it is here http://www.dvinfo.net/conf/showthread.php?s=&threadid=2269 It translates video terminology into computer values (7.5 IRE vs. RGB). Remember the NTSC values (requirements) are only applicable if your programs are going to be broadcast. If your programs are going to tape (VHS distribution) or DVD you can extend the range of your video. This will give you more details in your shadows and highlights. You can use this to your advantage to make your material look more like film. By extending the black level from 7.5 IRE to 0 IRE you will have more detail in your shadows. The same is true of your highlights. I use set-up like any other tool. Sometimes I raise it and sometimes I lower it. It depends on the mood I am trying to convey.

Jeff

[Linc Kesler]

Thanks for the replies. Especially after reviewing the thread suggested by Ken, I think I have stumbled into a fairly complex issue. I don't think I'm close to getting the full picture here, but is this generally accurate as a summary?

1) The setup control establishes the black level. On the GL2, maybe at IRE 0 (or 7.5?). Setting the control to the (-) side lowers the black level towards "superblack" which is either below IRE 0, or IRE O (if neutral is IRE 7.5). Setting the control to the (+) side raises the black level relative to the neutral setting.

2) If the setup control does NOT change the white level, it would then operate, all other things being equal, as a kind of contrast control, since it would either compress the light level values into a smaller range at a higher setting, or stretch them over a wider range at a lower setting, thus effectively increasing the contrast.

3) The setting is particularly relevant for specific uses. USA broadcasting requires a narrower contrast range, since it cuts off black values below 7.5, making them all black (thus clipping any shadow detail that was there), and white values above 100 (is that the figure?), making them all white. Other applications (DVD) have more range, and may benefit from detail in the "superblack" range. Am I getting this right?

4) On the other hand, it does seem that the setup control in the GL2 affects the higher end values as well. When I lower the setup level, I see less zebra-striping on screen, and have less blocking of highlights in the resulting tape. So perhaps it affects the high end as well, and doesn't simply expand contrast at a lower setting by lowering the black threshold. If so, its operation and effects seem a little more complicated.

Am I getting this even half-way right? My current project is aimed mainly at DVD production, so I'm guessing my safest bet is to leave the setup on neutral, my zebra on 100, and not let anything I want to have highlight detail go to zebra. I realize that I have been using the setup as a kind of exposure control (reducing apparent highlight exposure, plus I like the look), but I may be setting myself up for some problems, or at least extra work, in post, and I'd be better just using the exposure controls. Does that make sense as an approach?

Finally, in the other thread, I was interested in the still film analogies for gain settings. I liked the distinction I think Ken made to controls "before" and "behind" the chips, with shutter and aperture "before" and gain "behind." In still photography, a good analogy for gain might be push and pull processing: push processing gives greater effective speed (higher ASA), but boosts grain ("noise"). Pull processing lowers speed, and can (less reliably) reduce grain. But am I also right in thinking that setup is another adjustment "behind" the chips, that it is a processing of the signal rather than an adjustment of the light creating the signal?

Thanks again for your help.

Linc Kesler

[Ken Tanaka]

<<-- Linc: But am I also right in thinking that setup is another adjustment "behind" the chips, that it is a processing of the signal rather than an adjustment of the light creating the signal? -->>

Yes, Linc. That is correct. Again, there is no direct analog in the film world. It's not really like the AE adjustment (which I liken to selecting ISO).

[Bill Ravens]

FWIW....

DV standards mandate that "black" is at 0 IRE. On tests with my XL1s, neutral setting on the setup level corresponds to 7.5 IRE.
-2 on setup corresponds to 0 IRE. I would recommend setting the setup level to 0 IRE. Anything higher than that and you're throwing away latitude and reducing contrast on the overall image. When the signal gets transcoded to DV format, the IRE gets set to zero anyway, no matter what you (arbitrarily) decided to limit it to during capture.
For the more technically minded, this is all explained at Adam Wilt's website.

[Jeff Donald]

The setting of the setup (black level) is something I've been experimenting with recently. Your comments, Bill, kind of confirm my suspicion. To maintain the maximum amount of detail in the image shoot everything at -2 setup. Then in post, add setup or brightness and contrast as needed. You can always take away detail in an image, but what's not there is hard to create.

Jeff

[Linc Kesler]

I've had a chance to look at Adam Wilt's explanation of setup at
http://www.adamwilt.com/DV.html
and, if I'm understanding him, he's arguing that establishing a "pedestal" black level floor of 7.5 IRE for American broadcast TV is something best done in outputting DV to analog in post-production. He's arguing that in the digital world, there is no reason for a setup level, and that the DV spec more properly uses the full range of values.

So why have a setting on the camera? Perhaps it is so that footage designed to be output as analog from the camera can be adjusted to meet the broadcast spec. He's arguing that introducing a pedestal (like 7.5) at the capture stage is wasting bandwith and needlessly producing washed out blacks in DV editing and applications. So, is the logical conclusion then to set the setup level on the camera to its lowest (most negative) value, especially if the intended output is to NLE and eventually DVD?

It would be nice to know what the various setup levels on the camera actually do (do they correspond to 0, 7.5 at neutral, and some higher IRE value on tape, for instance?). And why does the setup affect highlight readings, if it primarily affects black level?

I hope I am not being annoying to anyone in picking at this issue, which I will probably realize later is not of much consequence, but it's because I am a newbie committing non-repeatable project footage to tape before I can really explore editing that I'm trying to avoid a mistake I'll have to try to compensate for later. I appreciate your patience if this is not a very interesting line of inquiry, and the help people have already given.

Linc Kesler

[Bill Ravens]

Linc....

No apology is necessary as far as I'm concerned. Several years ago, I was asking the very same questions. It's the only way I know of to come to an understanding of what I'm doing in a consistent, professional and high quality way. My own search ror understanding allowed me to make some intelligent choices about my method. Hopefully, I can give you enough info that you can make your own choices.

At any rate, the old NTSC standard established a maximum black at 7.5 IRE in order to give some margin above negative IRE values where the NTSC signal carried information like a sync pulse. If the video signal was too low, it would interfere with the sync pulse. With the advent of DV, there is no need for a sync pulse. Nevertheless, I worry that writing video information at levels below 0 IRE might somehow interfere with the NTSC standard somewhere along the line. I therefore opted to limit my pedestal to 0 IRE. In testing with my XL1s, I was unable to actually lower the black level below 0 IRE...but, I may have been limited by the calibration on my video signal analyzer.

Anyway...probably more info than you wanted to hear.

Good luck.

[Jeff Donald]

Why have a set-up adjustment on a digital camera? If I shoot everything at 0 IRE my contrast range may be greater than what I want. Different film stock is used to get different results. Some film is grainier, some has more contrast etc. I find the adjustment of the set-up to be similar to a contrast adjustment. If I was shooting a very moody scene I might want less detail in my shadows. If I adjust the set-up I can cause the shadows to block up (less detail). I could shoot it with less set-up and add the contrast in post. By using filters or proc amp controls in a NLE I could change the set-up or contrast (some NLE's have both). However, I would also have to render the scene after applying the filter. Rendering takes time and can lower the quality of the finished result. I've always followed the notion to shoot it right and fix the problems in post. However, I can think of some scenes or situations (sports, theater) where adjusting the set-up to 0 IRE would be best. I think for most people the greater contrast range (0 IRE) will be the best and most flexible setting.

Jeff

[Brad Higerd]

What would be the best situations to shoot with the setup set at 0 or above?

[Jeff Donald]

If the material is to be broadcast a set-up of 7.5 IRE is required. DV equipment pretty much takes care of that for you. This thread will help understand the issues and has a good link to Adam wilt's site for further reading. If that doesn't cover your questions or they remain unanswered, post back.

Jeff