Getting the right image ?

Re: Getting the right image ?

ALWAYS keep gain on 0.  it is only for use if you are not using any proper lighting, and do not have enough light, and dont care if your image looks like dog butt. gain will add noise, and chroma noise. and other ugly looking stuff to your image.

usein this order-> ND filters, iris, then shutter speed to control exposure.

if possible use only ND filters to control the amount of light  to the ccds. remove them or put them on in varying strengths

if you need to tweak the light from there a little,  use the iris (mostly you'll want to keep it open to retain the shallowest DOF. using the NDs first allows you to keep the same open iris setting. so again just use it to tweak it if you dont have the right ND, or its a little off)

then in an absolute must situation use the shutter speed to let less or more light in.. doing this will change the motion rendering of the images, so only do it as a last resort.
 
Re: Getting the right image ?

Iamloser. Cheers for that. One other thing if you don't mind...
What about white balance? If I'm in a situation where light is changing and I may have to shift the ND filter in and out several times during shooting, does it need to be reset each time I change the ND?

Ta.
 
Re: Getting the right image ?

Good job Andy!!!

nice post.... which reminds me to set the gain to low (or off)
 
Re: Getting the right image ?

On the "off color"faces: If WB is set correctly and the anbient light is not selective (sodium vapor, some fluorescent lights..) then also verify saturation and hue setting on the display device.
 
Re: Getting the right image ?

Andy great answers. Iamloser good explanation on exposure, nds, etc. You guys are the best!!!
 
Re: Getting the right image ?

Nice post Iamloser!

GAIN is an electronic amplifier that boosts the signal coming from the sensor. It makes the resulting image look brighter, as if there was more light in the environment. But unlike actually adding light to the environment, with GAIN there is no more information in the image. It can only work with whatever information is already present – whatever was detected by the sensor in the available light. To understand the results of using GAIN, you need to think about the qualities of the image before GAIN was applied.

Normally, you would use GAIN when there is not enough light in the environment. That means the original image, before GAIN was turned up, was underexposed. When the image is underexposed a lot of detail is lost because without enough light the sensor cannot distinguish between something dark and something black – and it all ends up black in the image. (This is called “crushing” the blacks). When GAIN is applied to this dark image, everything becomes brighter. But GAIN cannot restore the details that were lost due to underexposure. The result is an increase in noise in the image and an increase in artifacts. When the dark areas of the image were all black, it looked naturally dark. But when the dark areas are amplified with GAIN, tiny differences in blackness are exaggerated. And with enough GAIN, blocks of solid color show up that don’t look natural at all.

If you apply GAIN to an image that would otherwise be normally exposed, the problem is its effect on the highlights, as you discovered with the Zebra detection. Zebra puts a striped pattern on the LCD for any area of an image that is overexposed. And that is really handy because it can help you figure out how to position lights to get the best exposure. When GAIN is applied to an image that had normal exposure to begin with, areas that were just normally bright in the original image are amplified until they turn white. The details in these areas are lost. And with enough GAIN you get areas of solid white in the image that don’t look natural.

So when should you use GAIN?

As Iamloser pointed out, due to the risk of losing image quality, you will want to apply GAIN as a last resort when other methods of correcting the exposure are not effective. First, if there is enough light for a normal exposure, you want GAIN turned off (Low = +0dB). If there is not enough light for a normal exposure, then you will want to change the Neutral Density filter to allow more light. If there is still not enough light, and even with ND set to Off, the Auto Iris is still wide open and the image is still dark, then you can consider changing the shutter speed. Using a slower shutter speed, is going to let the available light strike the sensor for a longer period of time. That will increase the exposure. But it will also change how motion is represented. Moving objects will blur more at lower shutter speeds. Camera motion and slower moving objects that would have appeared sharp at default shutter speeds, acquire motion blur at lower shutter speeds. The slower the shutter speed, the greater the effect of motion blur.

At this point, your decision is probably driven by aesthetics. Which will be a better result for your purposes? A little motion blur due to shutter speed? A little noise in the image due to GAIN? Using them both together gives you some options. You might turn the GAIN up a little, but not slow down the shutter speed as much – make a compromise.

There is still one other alternative. But people often forget it as they are trying to find the best camera settings. And that is: ADD LIGHT :). Just a little bit of light in the environment can alter the balance of the situation and make a completely different range of choices available.

Andy
 
Re: Getting the right image ?

versioncity1,

With respect to the Neutral Density filter, the filter is supposed to be a perfect shade of gray that reduces the strength of all colors of light equally. So you should be able to change the ND filter setting without setting the white balance again.

I am in the habit of setting the ND filter first, and then setting the white balance. So I usually set the white balance after any ND change, even though it might not be necessary.

Andy
 
Re: Getting the right image ?

starbuck & loser.... Big thanks for postings. Its very helpful to have your advice/knowledge; I'm just shooting anything at the moment to understand how it all comes together.......
Ta.
 
Re: Getting the right image ?

[quote author=andy_starbuck link=board=Events;num=1095676145;start=0#3 date=09/21/04 at 08:41:18]2. SELECTING THE SCENE FILE SETTING
.
.
.
Select (f1) or (f5).
[/quote]

Hi Andy,

No f6? May I ask why you prefer 24P versus 24PA?
 
Re: Getting the right image ?

I chose to introduce f5 to simplify and avoid the possible confusion in an introductiory tutorial.

24p Standard (f5) pulldown produces smooth motion. 24p Advanced (f6) can look jumpy and make someone think there is something wrong with the camera it they don't understand the method and purpose of 24p Advanced. I shoot primarily in variations of 24p Advanced.

Andy
 
Re: Getting the right image ?

With regards to the harsh red tones; I shot a wedding last night and at the reception the light really made everyone look sunburnt, even after WB. What I wound up doing is going in the settings and bumpimg up the color temp, reducing the color phase and color level until I was able to get closer to skin tone. It took a little work, which is hard to do on the fly, but the images looked fine in the end. It's just something you have to play with until thinks look like you want them to.
 
Re: Getting the right image ?

How many of you ever consider adjusting the Master Ped instead of gain when you really need that extra 1/2 a stop? Opinions please!!!
 
Re: Getting the right image ?

Super series of posts Andy.
Logical and well written, I wish everything was explained in this easy to digest method! :)


--Bop
 
Re: Getting the right image ?

Fantastic articles and not just for the DVX100.  I am a new owner of an XL2 and have found this thread to explain more about my camera and how to work with it than anything else on the web!

I do have a question though... I've read more than once about using an 18% grey card for seting up a camera and white balancing (as well as a light blue card and others), but I have no idea what it's really for or how to use it. Could someone explain it please?

Thanks!

MattC
 
Re: Getting the right image ?

The white card is used to set the white balance. The gray card is used to set the exposure.

I'll discuss the white card in this post and the gray card in the next post.

A white card is used as a reference point for setting the white balance. It is a tool for telling your camera how to interpret the color of the light in which you are shooting.

All lights produce a blend of different colors or frequencies. To our eyes, the white card will appear white under most common lighting conditions, because our visual system is very adaptive. So even if the color of the light is slightly yellow or slightly red, the card will still appear white to us. But the camera is not so adaptive. To the camera a slightly different color of light can completely change the camera's interpretation of color. If the white balance is not set properly, the resulting colors won't look like they did in the environment. They may not look real at all. Setting the white balance gives the camera the information it needs to represent colors more realistically. Setting the white balance tells the camera "this is white". If the color the camera captures happens to be slightly yellow or slightly blue, then the camera can figure out how to adjust its interpretation of the colors it senses so that the card appears pure white. And that means that all other colors in the environment will be more faithfully represented in resulting image.

As I mentioned before, your eyes are very adaptive. So a lot of things that look "white" to your eyes will actually have some color to them to the camera. If you set the white balance for one shot with a white T-shirt, and another time using a white piece of paper, and another time with a white-painted door, then the chances are that the white balances in each shot are going to be slightly different, because the "white" color of the shirt and the paper and the door will be different. And that means the colors will be slightly different in each shot. Using a white card provides consistency from shot to shot.

For example, imagine that your subject is wearing a red sweater. And imagine that you will be editing together scenes shot at different times or in different locations. If you've set the white balance with different objects having slightly different hues of white, then the sweater in each shot will have a slightly different color and it won't edit together well. The color might be a bit orange in one shot, a bit brown in anohter. So you will have to do more color correction in post to make the shots flow smoothly together. If on the other hand, you used the same white card to set the balance in each shot, then the red color of the sweater will be more consistent from shot to shot and will require less color correction in post.

To use a white card, you place it in the environment where you will be shooting. If the lighting is uneven, then placement of the white card becomes more important. You probably want the white card close to the most important subject. If it is a person then they might hold the card up or even hold it near their face. In this way you are setting the white balance based on the color of light at that particular location. After the card is positioned, you zoom in on the card until it fills most or all of the screen. It doesn't matter whether it is in focus or not, because the white balance just depends on color information. After zooming in, you press the button to set the white balance.

White cards can be purchased from photography and camera supply stores. The white color on a photographic white card has been manufactured in a controlled way to ensure that it is very close to pure white and not some hue of white.

You wouldn't normally set the white balance with a grey card. You mentioned setting the white balance with a slightly blue-tinted card. If you use a slightly tinted card to set the white balance then the colors will be off. And sometimes people use this method of in-camera coloring as a way to get a special effect. Other people prefer to capture the colors as authentically as possible and then make any color effects changes in post. Still others use in-camera settings to alter the color interpretation of the camera.

Andy
 
Re: Getting the right image ?

The 18% gray card is used for setting exposure.

When you use the Auto Iris on your camera to find the exposure, it takes the average of all the pixels on the screen and tries to make that average 18% gray by changing the Iris setting. Now that works pretty well most of the time. But what if the pixels on the screen don't average 18% gray? What if there is a dark object in the frame that makes the average much darker? Then the Auto Iris is going to try to adjust that to 18% gray and may overexpose the image. Or what if there is a big white object in the frame? Then the Auto Iris is going to close down too much and underexpose the image.

If you want consistency of exposure, then you would set the exposure using a standard object -- the 18% gray card. Like the white card, you place it in a location in the scene that has typical lighting. And you zoom in on the card to fill most or all of the screen. At this point the Auto Iris will be giving you the correct exposure for the gray card. Turn off Auto Iris and the exposure will hold at that setting. (If you leave Auto Iris engaged, then the exposure is going to change as you zoom back out from the gray card).

Of course that is not all there is to setting exposure. Using the gray card just establishes a standard setting -- a starting place. At that point you may have too many highlights in the frame. And you might want to use the Zebra setting to preview the areas that will be overexposed, and then manually reduce the exposure to minimize the burned-out areas. (Or move the lights). You might also need to go the other way and increase the exposure. For example, in the DVX100a manual it says that the CineLike color processing darkens the image somewhat, and you might want to overexpose by half a stop to get a brighter final image.

Hope this is all of value.

Andy
 
Re: Getting the right image ?

Andy you are my new hero!!

Ok, so for exposure I've been doing the following: I keep my zebra bars on 90% and I start out with the most open iris I can (usually 1.6) I then step down until the zebra bars go away, if the don't, or if I had to step waaaay down, I add an ND Filter. Right now I'm just using the ND filters that are built into the lenses. Is this an appropriate method or should I get an 18% grey card and start from there as you just outlined?

You advise is of GREAT value. You know (and I'll probably make this a separate post at some point) it's almost a shame that this place is known as a DVX100 site. I find that this site has the best all around information on dv movie making of anywhere on the net, and I frequent more than a few sites. I really hope that there is more XL2 inclusion on this site for right now it is my best resource.

Matt
 
Re: Getting the right image ?

The only difference between the DVX100a and the earlier DVX100 in terms of the above, and may I add, quite excellent tutorial is the DVX100 does not have auto focus in progressive mode (24p or 30p).
 
Re: Getting the right image ?

I'd breifly like to share my expereince of what I've been shooting over the last two weeks as I used this discussions advice prior to shooting.
I only very recently bought my DVX and then started shooting footage of a band who is on tour at the moment.
Prior to the band walking on stage I set the white balance under any available white light at the venue (often quite tricky finding one) sometimes a stage light, but often the mixing desk work light. So nothing very accurate...
Then set my camera to fully manual and generally shot wide open or at 2.8. Left the gain at zero and used 24fps settings.
So far I've only watched footage back through any available TV and I have to say the images look very good. I was slightly worried about my white balancing because of the constantly changing lights onstage but so far I can't see any major problem.
Next week I get to put it all through a pro edit suite to grade and cut. I'll let you know what the editor says....
 
Re: Getting the right image ?

MattC,

What works, works. One thing I’ve learned reading DVXuser.com is that there are often several different approaches that produce good results. Sometimes different methods offer different kinds of control, prioritize certain qualities over others, or will work under different conditions. The method you described sounds fine to me. I have a gray card but I don’t use it very often. Sometimes I’ll just use the gray card to see where I end up and whether there is any difference between that exposure and by setting the exposure using the camera alone. If I see that I am very obviously filling the frame with a large dark or light object, then I may go to the gray card to make sure I am getting a good exposure.

I usually move from bigger to smaller differences. So I’ll start with Auto Iris engaged and just change the ND filter as the camera assistant suggests. (And maybe boost the GAIN if the ND filter is off and the image is still too dark). Then I’ll put the iris in manual mode. Use zebra to detect areas of overexposure, and reduce the hot spots by manually changing the iris. And that’s it. Using this method I’ve noticed that I really have maybe one f-stop in each direction (depending on the light) that will produce acceptable images, and at that point I have to decide which setting looks better for what I’m shooting. Whether I use the gray card or not, I still end up with only about two or three alternative settings that will work, unless I start moving / adding / and changing lights.

You can open up a whole range of workable exposure settings by changing the shutter speed. Changing the shutter speed gives you control over the duration of the exposure. So while the iris and the ND filters determine the intensity of light reaching the sensors, the shutter speed controls how long the sensors are exposed to the light. And together they determine the total exposure.

If you use “slow shutter” to increase the period of exposure, then you will need to reduce the size of the iris to bring the total exposure back to an acceptable level. If, on the other hand, you increase the shutter speed, then you will need to open the iris to compensate. But if you start experimenting with shutter speed, you’ll soon discover that exposure is not the only characteristic that is affected. There is a trade-off between motion blur and depth blur.

Using the method you described, you start at 1.6 and step down until the zebra bars go away. So you are favoring a wider iris and better depth blur. If you find that the light is so bright that you have to step down the iris further than you would prefer, because the depth blur is diminished or lost, then you could try changing the shutter speed to provide more options. So you might use a faster shutter speed. That would decrease the total exposure. So you would have to open up the iris more to compensate. And that wider iris would provide better depth blur.

Now I mentioned that there is a trade-off between motion blur and depth blur. If the shutter speed is faster, then objects in the scene will have to be moving a lot faster before they begin to blur in the frame. So the improved depth blur comes at the cost of decreased motion blur. And if you used a slower shutter speed, and compensated for the longer duration with a smaller iris, then objects moving at a much slower speed would blur in the frame, and the extent of the blur (or smear) would be much greater. But the smaller iris would also result in less depth blur. So improved motion blur comes at the cost of decreased depth blur.

One thing to keep in mind if you start to experiment with shutter speed and exposure is that the motion blur that is created at 24 fps with the standard shutter speed is one of the characteristics that makes the video look like film. The motion blur in 24 fps video is similar to the motion blur in 24 fps film, and that helps convey the illusion of film to the viewer. When you change the shutter speed you alter this delicate balance and change the aesthetic qualities of what you are producing. For example, I’ve noticed that at 24 fps, the normal motion of arms swinging when a person walks has some motion blur to it. Changing the shutter speed will either reduce or increase this blur, making it look different than film.

The nice thing about experimenting with shutter speed is that the Auto Iris on the DVX100a will still help you set the exposure properly. So you can change the shutter speed and then engage the Auto Iris, and it will adjust the Iris for the total exposure.

I’m not suggesting that you change the shutter speed for any particular purpose. I am only suggesting that if you want more options to control the exposure on your camera, you might want to experiment with shutter speed. That can add one more method for controlling exposure into your repertoire. And it might enable you to get a particular aesthetic look or to shoot under particular conditions that otherwise might have been limiting.

Andy
 
You must log in or register to reply here.
Back
Top