Light meters for DV camcorders

[Roger Berry]

This has been a really brilliant thread and I've tried out a lot of the suggestions and learned loads of things so thanks everyone. Like Carlos, I'm probably becoming an old bloke and finding it hard to accept new ideas but I'm starting to get there (BTW Carlos, my wife's sister's husband's mother is a Martinez; from Guadalajara near Madrid. Are you related? Probably not).

Anyhow, after reading all your comments, this is how I've got this thing figured. Hard as it is to accept, we don't need light meters any more. Set gain to zero and the stripes to about 90 and if you can't get zebras in your highlights at something like 1/50th at F8, there's probably not enough light to produce a nicely saturated image.

You've still go to get white balance and exposure right but that's another issue. I still wish I could use a light meter and know I was going to get the exposure right but I'm now convinced that's not the way it works.

Oh yes, one more thing; I reckon from the tests I've done the ASA/ISO rating for the XL1s is between 320 and 400... not that this matters any more.

Thanks again everyone.

[Helen Bach]

Carlos asked: 'As long as you can use the right ASA reference (where do you get that for different camera models?)...'

The same way that you get the rating for film? Do a test with your meter, your lenses etc. Instead of having to read densities with a densitometer, you read values in your NLE. "More details upon request".

Best,
Helen

[Carlos E. Martinez]

<<<-- Originally posted by Helen Bach :
The same way that you get the rating for film? Do a test with your meter, your lenses etc. Instead of having to read densities with a densitometer, you read values in your NLE. "More details upon request".-->>>

Well, I used to get my rating from the film manufacturer. As that's usually the "conservative" rating, we went up and down that value to work on the black resolution (deeper or thinner blacks). Using a densitometer for colour negative wasn't really something I cared for.

In any case my question was to see if someone had already seen or done some sort of table for some camera models. E.g.: the DVX100 is considered as having a "320 ASA" sensitivity.

The problems with "video-ASA" sensitivity is that you have to be "conservative", as the latitude is critical. Particularly in DV cameras.


Carlos

[Helen Bach]

Carlos,
'Using a densitometer for colour negative wasn't really something I cared for. '
Just out of interest, why?

Best,
Helen

[Carlos E. Martinez]

<<<-- Originally posted by Helen Bach : Carlos,
'Using a densitometer for colour negative wasn't really something I cared for. '
Just out of interest, why?

Well: one thing because that was the way I was taught cinematography. Projection tests, using a projector you did know, on a lab screen you were used to, were the real test.

A second one because where I started filmmaking, in Argentina, the DP took the film until first copy. So you had more control on what you were doing.

Densitometry made more sense in black & white days, where you could have a direct relationship because grays card and grays reflections from costumes and props were more readily measurable. It still makes sense in optical negative control.

But in color photography there are so many variables in the game that, in my opinion, it's very difficult to get the same accuracy even if you intend to.

Even color metering I think it's a waste of time. Going strictly by the book is a kind of precision that it's difficult to keep up with and gives too little to have fun with.

Of course nobody has to agree with me.


Carlos E. Martinez

[Jeff Donald]

Now days the colorist has more to do with the final look of the film than almost all the other factors. By trying to make a perfect exposure in the DV camera (film technique) critical amounts of data can be lost that the colorist can't recover. Applying film techniques won't always work today.

[Helen Bach]

Jeff,
You've got me puzzled with that post. Maybe we have a different interpretation of 'perfect'? I always though that the aim when producing the original, whether it be film or video, was to record the maximum amount for use in post.

Could you elaborate please?

Thanks,
Helen

[Jeff Donald]

Yes, very simply CCD's and CMOS chips behave in a very linear fashion. Camera designers can build in clip points or make them user adjustable etc. Panasonic has even gone so far as to incorporate these settings into downloadable files to customize the look of your camera, but that's for a different post.

How many stops of latitude is there in a DV file between white with no detail and black with no detail? I would say 5 to 6 stops for most cameras, especially in the under $4,000 price range. For the sake of this discussion and to make the math easier lets say 5 stops of light.

Our DV cameras record an 8 bit image. This provides us with 256 discrete tonal values (2^8). Most of us probably think that the 256 tonal values are divided equally between the 5 stops of exposure. But it isn't so.

When you reduce the exposure from maximum white by 1 stop it is a 50% reduction in light transmission. And with that transmission reduction goes 50% of your tonal values. That's right, 128 tonal values are in your top one stop of light. Put a clip on your whites (under expose) and half your detail is gone. The next stop is another 50% reduction, or 64 tonal values, another stop is 32 values, the 4th stop is 16 values and the last stop has only 8 tonal values. It's easy to see why under exposed video looks so bad. There's no tonal values in under exposed scenes.

Once those tonal values are gone (not captured on tape) there is little the colorist can do to bring them back. So how do I expose for video? Expose as hot as possible (over expose) without clipping the signal. Then in post pull the video into the bottom stop to create shadows. This requires a great deal of rendering, but will provide the greatest amount of detail (tonal value) for the colorist to work with. This technique is called Expose to the Right.

[Helen Bach]

Jeff,
Having read your post I see that there is no difference in the way that film and video are treated. To sum it up: know your medium. Obviously you can't expose video exactly the same way as you expose film, but you can apply exactly the same principles: understand where a light value will fall on the response curve of the medium. As I suspected, it is just that we have different interpretations of 'perfect' and 'film technique'.

My earlier post suggested that one could treat a video camera the same way as film, but replace the densitometer with the numerical value display in an NLE (for example). That will reveal exactly the relationship that you have described - though you may get some surprises vis-a-vis true linearity. It is very well worth doing in practice.

Best,
Helen

[Carlos E. Martinez]

<<<-- Originally posted by Helen Bach : Jeff,
Having read your post I see that there is no difference in the way that film and video are treated. To sum it up: know your medium. -->>>

You took the words off my mouth.

At first I was puzzled at Jeff's statements. Then I realized that it was really an attitude toward image handling, which I think is what I have always believed you should have. It gives coherence to what we do.

Video, more than film, is how you deal with the knees. I'd say it's only that. In film that's a problem that practically stopped being so.

In my mind I always compare video to old reversal 16mm film (7243?), which in the '70s we strived to mix with 7252 in lower light. Except for the grain (which video doesn't have), the 5 or 6 stops latitude was adamant. In those times we flashed the film as a way to lower contrast and improve grain, which is very much like working the black level in video. Though in video we can see the results at once. But the latitude limitations are similar.

When I started working in video, in the mid '80s, coming from film, I handled video as film. Opened the stop, burning the whites, used fog or low contrast filters, etc. Tubes were not too forgiving with overexpose, but it worked as a low budget "lighting" tecnique. It sounds quite similar to what Jeff says he does.

My only concern is "overexpose without clipping", as Jeff claims, which I don't think is possible without some gamma or contrast handling. Graduated NDs are a great tool for that, and I don't know why are not part of most basic DV kits. Pola screens not only lower your stop (which shooting at 1/60 is hard to keep low) but will handle the contrasts in a different way, also allowing some overexpose.

Sometimes the only way to keep contrast under control is to reframe.

Helen: can you ellaborate more on that camera stop/NLE numerical values routine? That seems interesting.


Carlos

[Ken Tanaka]

Not too get far off-topic here, but this may be a point worth noting as a follow-up to the immediately previous posts.

I've just gotten one of Tiffen's "Ultracon" series of filters (in my case, a 4x4). In brief, this graded series of filters (ex: Ultracon 1, Ultracon 2, ... up to 4) is designed to give you a bit more latitude when shooting high-contrast scenes such as brightly lit sunlighting or dim scenes with bright highlights / light sources. That is, they give you an extra stop or so in exposing your dark areas before you clip your highlights. My preliminary (informal) experiments with the Ultracon 2 suggest that these fellows really do seem to perform this magic as advertised.

They are not cheap...good filters never are. But they may be a very valuable tool to have if you're using a camera (as most of us are) that only affords 5 stops of latitude.

Forgive me if these Ultracons are old-hat to you. I just learned of them recently and had the impression that they are a newer product.

[Carlos E. Martinez]

<<<-- Originally posted by Ken Tanaka : Not too get far off-topic here, but this may be a point worth noting as a follow-up to the immediately previous posts.

I've just gotten one of Tiffen's "Ultracon" series of filters (in my case, a 4x4). In brief, this graded series of filters (ex: Ultracon 1, Ultracon 2, ... up to 4) is designed to give you a bit more latitude when shooting high-contrast scenes such as brightly lit sunlighting or dim scenes with bright highlights / light sources. That is, they give you an extra stop or so in exposing your dark areas before you clip your highlights. My preliminary (informal) experiments with the Ultracon 2 suggest that these fellows really do seem to perform this magic as advertised.

They are not cheap...good filters never are. But they may be a very valuable tool to have if you're using a camera (as most of us are) that only affords 5 stops of latitude.

Forgive me if these Ultracons are old-hat to you. I just learned of them recently and had the impression that they are a newer product. -->>>

I believe you are quite on topic. It would also be very interesting to have other opinions or experience stories on these and other graduted NDs.

Grads have not always been easy to find, except on pro sizes, but they are a precious tool to keep those contrasts under control and still get a beautiful image.

They should be part of every basic video kit, and I assume it probably is with people reading this lighting threads.


Carlos

[Charles Papert]

I've owned a set of 4 Ultracons for close to 10 years. While I think they can be very handy for particular types of high contrast, overall I think they may be a bit too sophisticated for the casual user. I say this without trying to be insulting, it's just that they are high-maintenance filters. They MUST be protected from stray light, meaning vigilant flagging via the mattebox or stand-mounted flags. As you pan off areas of high contrast, you will see the density of the shadows change. And they can occasionally cause a color shift.

Carlos, just to clarify Ken's use of "graded" to mean that the Ultracons are available in several grades i.e. strengths; the Ultracons are not graduated nor do they incorporate ND.

In terms of grads, they are fine filters but I'm not sure if I would consider them a basic filter. They have quite a few limitations, in that they are not recommended for shots that involve tilting as the effect will ride through the frame, nor if your subjects head is high enough in the shot that it crosses the horizon line, which is often the case. They are great for landscape and establishing shots however. The other thing that takes them, I think, out of the realm of a casual (i.e. budget-minded) shooter is that you really need a set of different strengths of ND as well as hard and soft grads (wide shots call for soft grads and telephoto shots for hard grads.), which can be up to six filters for starters. And then there's colored grads...I recently saved a murky, cloudy sky by using a hard ND6 grad in the bottom of the tray, allowing the sky to open up from gray to white, then adding a Sky Blue 2 grad to the top to insert some blue color into the sky.

Many readers of this forum are desperately seeking a quick form of alchemy--"how can I make DV look like film"--and always assuming or expecting that a simple filter or setting on the camera will achieve that. I just get a bit concerned that they will rush out and buy something and then be disappointed that it doesn't magically transform their footage into "The Matrix" or something.

Filters are a tricky investment for DV filmmakers, I think, in that they are generally looking to spend the least amount possible and buy a single filter rather than a set. As I'm sure from your experience you are aware, Carlos, one size doesn't fit all when it comes to many types of filters--simply altering your focal length will often require switching out filters to maintain the same look, since the effects of the filter will be more or less pronounced (as is the case with diffusion, demarcation lines with grads, etc). Of course, one can make great use of whatever one has and work with the results. If you can only buy one grade of filter in a class, that's generally better than nothing at all, but consistency from shot to shot may be compromised as a result.

[Carlos E. Martinez]

Charles,

I have the feeling that you may have a wrong idea about my experience or if I know what I am doing when I talk about the advantages of using grad NDs. Believe me I do know quite well what I want or what I am after.

Your explanation on how a graduated ND works is certainly on the spot, and I am very much aware of their limitations. In a way a graduated ND is very much like a split-diopter lens, in the sense that you have two separate areas that take care of two different fields.

Grads are much more easier to deal with, but you have to be careful when you move your camera. Pans are usually alright, but tilts can certainly be a problem. There's a separation area that you have to be aware of and be very conscious on how to deal with it.

Grads now come in several grades, but until recently you could only find them in .3, rarely in .6 grades. Now .9 is also available. For colour grads I certainly see a much more limited application for. They wouldn't be on my short list.

One thing puzzles me: why do you compare a budget minded shooter to a casual shooter? They are quite different in all respects.

What I say and stay by is that a set of at least three grad NDs may solve your shot when working in video, particularly in documentary situations which are difficult or impossible to control. More particularly in DV, where dealing with the contrasts can be even harder. They are also a must in professional video, where cameras can handle perhaps two stops more but still have the same knee problem. In fact, in "budget no object" video job I would certainly order a matte-box where I could have a very large grad filter that I could move in a straight line during the shot if necessary.

No doubt that to use a grad ND you have to know what you are doing, but that is the same with polarizers too. But grads are not so hard to learn how to properly use them. If you mean it's not the same as an UV it certainly is not. Fog and low contrast filters are also hard to work with.

My interest is not in making DV or video to look like film: is how to better shoot video that might be blown up to film. Quite a different matter. Because what I have to do is handle the video limitations so they work in my favour.

But I never tried Ultracons until now or other low contrast filters that do not take resolution away from video, which as I see it is the most important thing when blowing to film.

Carlos

[Charles Papert]

Carlos, my intention was not to question if you know what you are doing. I actually indicated in the last paragraph of my post that due to your experience which you laid out for us earlier, you seemed more than qualified to speak on this issue. The comment that begins "Many readers of this forum" was not aimed at yourself.

I probably should have been more careful than to equate "casual" with "budget-minded"--these days everyone has to be budget-minded. I was referring to what I observe to be the vast majority of the users of this forum, to whom the purchase of a single filter is a substantial investment. After all, we regular hear "I want to buy a matte box to make my camera look more cool; what's the absolute cheapest one I can get?"

In this vein, then, I would personally not recommend that grads should be considered part of a "basic video kit" as many readers here know it. As we agree, you need a set of grads to cover the essentials, which is a substantial investment if we assume a minimum filter size of 4x4 (which may not necessarily cover a wide angle adaptor). I do think they are valuable as I indicated, and I do use them myself. I also use a Preston Microforce with a digital motor as a zoom control, which delivers the most delicate and sensitive response available, and I consider it part of my own "basic" shooting kit; but I wouldn't bother recommending it for readers here due to the cost.