Hi Paolo, how do you measure this?
I'm curious what ISO 'the low light setting' of Tim Dashwood is giving.
I framed my DSC chart, lit it with even light from two sources at the same distance/angle. Verified that the grey area was showing as a even fuzzy line in the center of the chart in the WFM.
Changed the aperture until the grey area fell on the 40%-45% marker. Adjusted my light meter to give me the same exposure. Read the resulting ISO rating. It was 200.

Now, as I said, there is no way of rating this solidly because the lens is not constant across the range but my measurement was not at the two extremes so, it can work as a reference. If you go wide you'll gain some sensitivity, probably around 220. If you zoom in you loose "film speed".
Also, I played with the Gamma settings a bit. If you're in a low light situation you might want to switch the Gamma from "Cinelike" to "Standard". Standard gives you more light but its has a less balanced, IMHO, response. The "X" shape of the grayscale from the chart looks a little bit shifted toward the top. The Cinelike gamma seems to be more even and able to handle highlights better.

I use 100% clip, maybe you are at 108% clip?

Yes, I use 108%.


FYI, the stock lens works out to be a T1.5 on the wide end, and a T2 on the far end.
Good to know, thank you.

That's assuming a lot. For example that I know what a Panny S35 is :)
Spiacente Paolo,

Panavision Super 35. Hahaha, that solves the DOF issue, that's for sure :)

I framed my DSC chart, lit it with even light from two sources at the same distance/angle. Verified that the grey area was showing as a even fuzzy line in the center of the chart in the WFM.


Paolo,

Just curious, are you using a spot (reflective) meter on the 18% gray, or an incident meter at the chart position?

In setting my exposure curve on a grad GS chart, I come up with a 320 ISO equivalent (wide open at 16mm zoom). FYI, the stock lens works out to be a T1.5 on the wide end, and a T2 on the far end.

I have found that using a "400 slide" meter setting gives a good exposure compromise for the HD100 on Caucasian faces.

1) When you say "400 slide" meter -- do you mean set an ASA of 400?

2) When one sets a CAMERA/FILM ASA and a Shutter-speed and the reads-out the Aperature, I assume that the reading is assuming the T-stop and F-stop are the same -- a ratio of 1:1. With Primes, this is a reasonable assumption. But with zooms it is not -- as you have measured.

So, in theory, one should set the Aperature slightly more OPEN than the meter reading. Correct? For example, 10% more open.

If that's true, shouldn't the ASA you dial into the meter be 10% LOWER. Thus, if you knew the TRUE ASA sensitivity of the camera, without lens, you would lower this value by an amount suitable for each zoom lens you installed.

HOWEVER, since you obtained the sensitivity with the stock lens using the Aperature indicated on this lens -- and the shutter-speed -- the ASA you got is already lower. So, the camera ASA can be computerd by:

Camera ASA = [ [ WIDE (F-stop/T-stop) + TELE (F-stop/T-stop) ] /2] X Meter ASA

So when you say ASA 320 -- have you already taken into account the AVERAGE difference (ratio) between F-stop and T-stop at Wide through Tele?

I ask, because as more lenses come to market -- we need the Camera's ASA. Then, each lens will have its own average F-stop/T-stop ratio.


Paolo -- why get as at the 40%-45% marker? What's wrong with shooting an AVERAGE brightly lit scene and getting the ASA there?


Last, I'm getting a bit confused because I think this discussion includes ASA values from two different lenses. Correct?

Paolo -- why get as at the 40%-45% marker? What's wrong with shooting an AVERAGE brightly lit scene and getting the ASA there?

I adapted the procedure from Scott Billup's "Digital Moviemaking" where Sean Fairburn describes how to do it. Specifically I placed the 18% gray of the DSC chart in that range because the line is rather thick and so it covers a range instead of just one value in the WFM. Fairburn says that that's where it should be and it actually makes sense when you look at the monitor. BTW, the camera's reading agreed with that too. The display on the LCD rated the scene as perfectly exposed. Not terribly important but a verification that the exposure was OK. It could be wrong but now I now that the reading from the camera is aligned.


Last, I'm getting a bit confused because I think this discussion includes ASA values from two different lenses. Correct?
I don't think so, I believe that we are talking about the stock lens. It's just that Enzo, like any self-respecing crazy Italian, can't stop making jokes ;)

1) When you say "400 slide" meter -- do you mean set an ASA of 400?
Yes.

2) When one sets a CAMERA/FILM ASA and a Shutter-speed and the reads-out the Aperature, I assume that the reading is assuming the T-stop and F-stop are the same -- a ratio of 1:1. With Primes, this is a reasonable assumption. But with zooms it is not -- as you have measured.

So, in theory, one should set the Aperature slightly more OPEN than the meter reading. Correct? For example, 10% more open.
No. Too confusing. Since all two current lenses "officially" available for the HD100 are only calibrated in f stops, it's best to do the conversion on the meter and then transfer the setting to the camera as an f stop value.

Out of respect to Paolo, I just did a more comprehensive test using just a 18% gray card instead of a grad. The following are absolute values that can then be modified by the user to their own personal exposure preferences. These settings are based on the factory presets for HD/24/108% and HD/30/100%. I could not come up with any other method to arrive at the actual ISO sensitivity of the camera that would not be biased or affected by personal user settings.

To arrive at the absolute ISO settings for the HD100, I first had to factor in the difference between the f stop and T stop values. Since I know that the stock lens has a T stop range of 1.5 to 2.0, I set my meter between the 1/8 and 1/15 shutter speed marks for a 24 camera shutter, and at the 1/15 mark for a 30 camera shutter. The indicated f stop would then mirror the correct exposure for the median T stop of the lens.

For practical purposes, the stock lens can be said to have a median T stop of f/1.8.

Now, based on those shutter speed settings, and after doing a white balance using a 5600K light source, I shot a 18% gray card and passed it out to a vectorscope. A 18% gray value is normally said to be between 45 and 55 IRE (I used 55 IRE, as do most broadcast outlets). At 55 IRE on the vector, using my converted shutter speed settings on the meter, I came up with an absolute ISO value of 250 at a 24 shutter/108% clip and 200 for a 30 shutter at 100% clip.

Since my broadcast chain editors (I shoot for several broadcast outlets) all seem to like the HD100 output to be a little on the underexposed side, I would personally use a 320/ 400 ASA @ 1/15 shutter setting on my meter to arrive at the correct ISO / f stop for my uses (at a 30 camera shutter setting 100% white clip, using my personal color and camera settings). Deviating from the factory default settings will most likely result in a change to the exposure curve / ISO rating, that's why I used the factory defaults to arrive at my "absolute" HD100 ISO rating.

So when you say ASA 320 -- have you already taken into account the AVERAGE difference (ratio) between F-stop and T-stop at Wide through Tele?
Yes.

I ask, because as more lenses come to market -- we need the Camera's ASA. Then, each lens will have its own average F-stop/T-stop ratio.
Agreed.

Steve, I'm posting the manner at which I arrived at the absolute ISO setting for the HD100, because you are obviously a very technical person. For myself and Paolo, who are just simple Italians, a simpler answer would be to set your incident meter at between the 1/8 and 1/15 shutter mark @ 250 ISO for a 24 camera shutter/108% clip and 1/15 shutter@ 200 ISO for a 30 camera shutter/100% clip (using the stock lens), or:

24 camera shutter/108%clip = 1/12 shutter @ 250 ISO on the meter = corrected f stop reading for the stock lens.

30 camera shutter/100%clip = 1/15 shutter @ 200 ISO on the meter = corrected f stop reading for the stock lens.

...and then modify the above values based on your own personal user and exposure preference camera settings.

Two asides, the built in meter appears to be calibrated for a 12% gray card using the factory default settings, and using the factory defaults for the HD/30/108% setting, the ISO works out to be about 225.

You guys are so far ahead of me that I'm going to try to repeat back what I think I understand:

1) "These settings are based on the factory presets for HD/24/108% and HD/30/100%."

I had never noticed that the IRE limit was different between 24p (to film) and 30p (for broadcast video.) It makes complete sense. Thank you!



2) "To arrive at the absolute ISO settings for the HD100, I first had to factor in the difference between the f stop and T stop values."

OK.


3) "I know that the stock lens has a T stop range of 1.5 to 2.0, ... . For practical purposes, the stock lens can be said to have a median T stop of f/1.8."

The stock lenses F stop range is from f/1.4 to f/1.8. Thus, the average F-stop is f/1.6. Correct?


4) Therefore, we know the loss of light is the difference between 1.6 and 1.8. If, for example, based on a meter reading, I were to set an exposure of 1.6 -- the shot will be slightly underexposed. Correct?


5) So I want to bias the meter so that it indicates f/1.6 when the exposure should be f/1.8. That means the bias must be toward slight OVER-exposure. In fact, about 1/4 stop over-exposed. Correct?


6) Which means the meter's ASA setting should be reduced OR the meter's shutter-speed should be INCREASED. Either will cause, for the same amount of light, the iris value to be smaller. Correct?


7) "I set my meter between the 1/8 and 1/15 shutter speed marks for a 24 camera shutter, and at the 1/15 mark for a 30 camera shutter. The indicated f stop would then mirror the correct exposure for the median T stop of the lens."

Here's where I'm doubly lost!!!!


8)(A) You DECREASED rather than INCREASED the shutter-speed which would bias toward under-sposure. This ie reverse of what I think needs to be done get the correct exposure. (B) You mention a 1/30th shutter-speed. But, 30p uses a 1/60th shuttter-speed. (24p uses a 1/48th shutter-speed).


9) "Now, based on those shutter speed settings, and after doing a white balance using a 5600K light source, I shot a 18% gray card and passed it out to a vectorscope. A 18% gray value is normally said to be between 45 and 55 IRE (I used 55 IRE, as do most broadcast outlets)."

OK


10) "Since my broadcast chain editors (I shoot for several broadcast outlets) all seem to like the HD100 output to be a little on the underexposed side, I would personally use a 320/ 400 ASA rather than 200/250."

By increasing the ASA you do indeed reduce the exposure for a given amount of light.


11) "That's why I used the factory defaults to arrive at my "absolute" HD100 ISO rating."

I agree fully.


12) A simpler answer would be to set your incident meter ... .

Would it be any different for a reflective meter?

Ouch... there is a lot of bad math & procedure going on here.

1. to calibrate the camera's ISO, you should be using a standard B&W chip chart with the 10 steps of black/white. the reason is that if you set an iris based on a single value, you may well be clipping highlights and shadows, or not making a full signal in the real world.

2.the cameras black level, white clip, gamma and knee all come into play. The most important are black level and knee, since they can significantly change the correct exposure in the real world. In any attempt to derive an ISO rating, you should be using your normal shooting settings. You could also try the factory plain settings for a worst case ISO setting.

3. Shutter - the proper shutter for 24P is 1/48th, or 24FPS on you light meter. For 30P, I prefer to shoot @30th, but techincally you should be using 1/60th. For me 1/60th is just regular video looking, so I opt for the slower shutter speed, which is a one stop gain. very usefull in low light. Again effective ISO is in big part based your shooting habits/settings.

4. the change of T stop in the lens is close to a stop. averaging to F1.8 or F1.7 means you will be a 1/2 stop over at the long end, and 1/2stop under at the long end, and thats a big deal with video. The effective exposure is at the particular focal length the lens is at *not* the theoretical average because its not the averaged setting thats hitting the sensors, its the actual shooting iris. FWIW, most of the fall off *should* be happening at the very last 20% of the lens, most of it before should be pretty close.

to measure the cameras ISO you should point the camera at the chip chart and frame it up, *white balance* the camera because that does make a difference, and then look at a *waveform* monitor, not a vectorscope which is used to measure color levels/vectors, not overall video levels.If you are using a vectorscope to set levels, you are off by about 20% if there is color in the image. on a pure gray scale, it might be close provided the camera was white balanced.

Looking at the waveform, you should have 100/108IRE for white, and black at 7.5 or 0 depending on your video standard and delivery format. at that point, you can now take out your light meter, and read the light at the chart. Your shutter should match the camera. Now adjust the ISO until the lightmeter's F value matches the camera's. That ISO is what the camera is shooting at. You can now take the light meter and use it for prelighting and know it will be very close to what the camera will be happy with.


I have used this technique for years - I've worked as a LD ( and engineer) in a couple of TV stations where I did not have camera's around when lighting, yet had to get lit. Once I had my meter calibrated to the camera's, I would consistanly light sets, and find the cameras irised with +-1/2 stop of my meter reading. The main difference for the reading is based on if the set was high-key, shadowdy, or average ( typical interior ) and the camera was adjusted a little for the best look. 1/3 a stop is a visible difference in light to a video camera, so when I lit sets, my keys were always dead on to +- .1 stops, more than that and you had a problem.


Steve Oakley

Didn't Adam Wilt and the boys figure the ISO at just under 320?

Ouch... there is a lot of bad math & procedure going on here.

Steve Oakley

Steve, I did use a waveform to check the levels, I had used the vector first just to check where the 18% gray fell, and I had vector on the brain.

If you look at my original post, you will notice I used a grad to check the ISO of the camera, I come up with a 320 ISO using the grad.

This test was for Paolo: "Out of respect to Paolo, I just did a more comprehensive test using just a 18% gray card instead of a grad."

The stock lens is T1.5 at 16mm and T2.0 at 88mm with the falloff starting to kick in at about 55mm.

Shutter -- the proper shutter for 24P is 1/48th, or 24FPS on your light meter. For 30P, I prefer to shoot @30th, but techincally you should be using 1/60th. For me 1/60th is just regular videV looking, so I opt for the slower shutter speed, which is a one stop gain. very usefull in low light.

It's interesting that with the original JVC, I insisted that 1/60th was correct because 1/48th would be correct for 24p. Now I'm beginning to find 1/30th looks a lot less electronically "strobby" on the original JVC.

But, logically I can't justify why 1/48th looks good at 24fps, but a much slower shutter looks good at a slightly higher frame rate.

One possibility -- not often discussed (and OT) is that the nature of a film shutter's pass across a frame of film is different than the full frame shutter opening done in video.

If SOMEHOW the moving wedge shutter of the film camera's shutter yields it's own motion blur -- then the 1/48th number is not completely valid, The extra moving wedge blur might yield a true film shutter speed lower than 1/48th.

Alternately, the instant ON/OFF of an electronic shutter might make 1/60th LOOK more like a higher shutter-speed. For example, 1/96th. So, when we drop a video shutter by 2 to 1/30th -- the EFFECTIVE LOOK drops to 1/48th.

Either way, I'm convinced, that the electronic shutter -- even when set exactly to that of a film camera -- does not produce the same result. The electronic shutter seems to create more severe judder on motion.

Out of respect to Paolo, I just did a more comprehensive test using just a 18% gray card instead of a grad.
Thank you Enzo.
I won't even try to digest your numbers at this late hour (10:30pm) after a pretty intense day of shooting the first episode of "2nd Unit" :)
I just wanted to thank for your time and insight. I will take a look at your post tomorrow when I'll have the right energy to appreciate it :)

Didn't Adam Wilt and the boys figure the ISO at just under 320?
Yes, and I never meant to doubt his numbers, it's just that in his article he didn't mention the gamma type and value that he used. I assume that it was the factory settings. I wanted to check the value of TC3 because I noticed that the camera changes sensitivity to light based on the gamma settings. Try this for fun: Set gamma at "CineLike" -1 as per TC 3. Adjust the iris for what you consider correct exposure of the scene. Go back to the menu and switch to gamma "Standard" with the same -1 value. Your scene should become immediately much brighter.
Maybe Enzo or Steve or anybody else :) can explain what's happening when you do that. The way I interpret it is that the camera is more sensitive to light with Standard gamma and so the ISO speed depends on the type of gamma selected.
Or maybe I'm allucinating :)

You guys are so far ahead of me that I'm going to try to repeat back what I think I understand:
Eat more pasta! :)

The stock lenses F stop range is from f/1.4 to f/1.8. Thus, the average F-stop is f/1.6. Correct?
Steve, first of all we are talking two things. One is the quasi ISO of the camera (this is the film speed set on the meter). Two is the actual f stop to T stop conversion (we need a number to set the lens aperture to).

Since the lens is calibrated in f stops, we have no choice but to make the T stop conversion. The stock lens is f/1.4 to 1.8, but T1.5 to T2.0, so I called it T1.8 across the zoom range of the lens. If you shot wide all the time, then you could use T1.5, if you shot in the 60 - 88mm range most of the time, T2.0 or T1.9 would work. MY aim was to arrive at an f stop reading on the meter that would transfer directly to the lens using Paolo's 18% gray measurement system.

This would keep Paolo from having to tell the 1st AC "4.5 plus 10%" :) I actually do all such conversions in my head, preferring to keep my meter at 1/48 unless I am under or overcranking.

8)(A) You DECREASED rather than INCREASED the shutter-speed which would bias toward under-sposure. This ie reverse of what I think needs to be done get the correct exposure. (B) You mention a 1/30th shutter-speed. But, 30p uses a 1/60th shuttter-speed. (24p uses a 1/48th shutter-speed).
You are correct. I counted down 1/3 - 1/2 instead of up, my bad. And yes, I set my meter at 1/48 for film and PAL video, and 1/60 for NTSC video (double the 1/24 and 1/30 speeds I used in my example).

Would it be any different for a reflective meter?
Not if you were shooting a 18% gray card. I only use a reflective meter to measure contrast ratios.

Let me set Paolo's system aside for a minute. By using a grad (instead of Paolo's solid 18% gray) and the factory defaults, I think I achieved a more accurate overall luminance reading for the base ISO camera setting.

Using my system, I came up with a quasi ISO setting for the HD100 of 320. However, Paolo (using his TC3 settings) came up with 200. Now we know that the gamma setting will change the overall exposure curve, so that may account for the difference in Paolo's ISO vs. mine.

Paolo's methodology, however was entirely correct. Exposure meters live in a 18% gray world.

Yes, and I never meant to doubt his numbers, it's just that in his article he didn't mention the gamma type and value that he used. I assume that it was the factory settings. I wanted to check the value of TC3 because I noticed that the camera changes sensitivity to light based on the gamma settings. Try this for fun: Set gamma at "CineLike" -1 as per TC 3. Adjust the iris for what you consider correct exposure of the scene. Go back to the menu and switch to gamma "Standard" with the same -1 value. Your scene should become immediately much brighter.
Maybe Enzo or Steve or anybody else :) can explain what's happening when you do that. The way I interpret it is that the camera is more sensitive to light with Standard gamma and so the ISO speed depends on the type of gamma selected.
Or maybe I'm allucinating :)
Yes, I see the need to describe the difference in ISO when settings are placed in the camera which change the gamma curve. If the TCV3 scene file is the max contrast the camera can produce then you should get a lower ISO value than stock settings. I struggle with this because in uncontrolled lighting (outdoor) the image get's shot to hell with overexposure and blooming (in the primary reds and secondary yellows) and the stock settings are more favorable for a more balanced look. It's odd because with a film camera, the film's speed is the constant in the equation and with the JVC, the ISO floats depending on what scene file you select. Solid work guys. It goes a long way in explaining where and why there is such a difference between scene files (not just color but the full gamut).

Panavision Super 35. Hahaha, that solves the DOF issue, that's for sure :)
Oh yeah! And the cost of the rental for the lens is probably higher than the cost of the HD100 :)

Paolo- I love the improvement with your TC3 settings, but... how would you modify it to really saturate reds and greens, without blowing out?

Eat more pasta! :)

In the USA most of us on diets because we don't drink enough good wine!

"You are correct. I counted down 1/3 - 1/2 instead of up, my bad."

OK -- now it all makes sense.


"Not if you were shooting a 18% gray card. I only use a reflective meter to measure contrast ratios."

1) If I were setting exposure with an 18% card/cloth I would use my meter in Reflective Mode to measure the light reflecting from the gray. Correct?

2) If I were setting exposure, I could also measure light falling on an AVERAGE scene using Incident Mode. Correct?

3) If I were computing contrast ratios -- which is for me the main reason to use a meter -- I would use my meter in Reflective Mode to measure the actual light reflecting from each area within the scene. Correct?

In all cases, I would set shutter-speed appropriately for 24p or 30p.

Now comes the question of which ASA to use.

1) Some say the ASA varies by WHICH gamma curve is selected. Can anyone demonstrate/calculate HOW Enzo's and Paolo's selected gamma setting actually would cause the difference between 200 and 320?

2) Some say the ASA varies by whether you have low or bight illumination. This is why some have claimed, the camera has several ASA values. I can't understand this.


Lastly, a dumb question. I take my old JVC HDV camera and point it at an average daylight scene. I set the camera and my lightmeter to 1/60th. I read the F-stop from the camera. Now I measure the EV of the scene. Using the F-stop and EV I find the ASA that matches. Why isn't that the camera's ASA?

Out of respect to Paolo, I just did a more comprehensive test using just a 18% gray card instead of a grad.

Thank you my friend.


I came up with an absolute ISO value of 250 at a 24 shutter/108% clip and 200 for a 30 shutter at 100% clip.

Makes perfect sense. I just wanted to highlight, in case is not clear to everybody, that different shutter speed implies more/less light exposing the "frame" and that explains the difference in ISO rating. Thank you for testing the two different situations, my original test was only at 24fps. Your numbers give us a more complete picture.


For myself and Paolo, who are just simple Italians...

Hehehehe.

Thanks again Enzo, that was very useful and informative.

Paolo- I love the improvement with your TC3 settings, but... how would you modify it to really saturate reds and greens, without blowing out?
Thanks. If you want more reds please see the V2 on my web site. By changing the blue gain you end up with more reds. I wouldn't touch the green because it has side effects on the whole color matrix, just too complex. Also, if you like more or less saturation I would achieve that in post. Remember TC is not meant to be a look but a way of bringing the most information into your camera. If you want more saturated reds and green I would use the color correction of your compositer/NLE.

Thanks Paolo. The reason I ask, is because I am shooting a family of Cardinals in my back yard. Brilliant red against lush greens, and it just isn't capturing quite right. Of course, I also don't have any way to capture to the PC yet, or to view externally. I'm working on it.

Thanks Paolo. The reason I ask, is because I am shooting a family of Cardinals in my back yard. Brilliant red against lush greens, and it just isn't capturing quite right.
I see. For "lush green", if you're shooting under bright sunlight, I would use a polarizer. That should make the foliage become much better.
In absence of that use the built-in ND.

Use a polorizer... I should have thought of that. I still haven't received my UV filter yet, and I knew I should have added a Polorizer... (slaps forehead!) Thanks!

By the way, where is the sweet spot on the 16x lens?

By the way, where is the sweet spot on the 16x lens?
There is a post by Tim Dashwood about that, do a search in the past articles.
I think it's around f4.0 but Tim has a complete description of it.

Thanks again Paolo. It never hurts to get a second, or third opinion.

I just wanted to highlight, in case is not clear to everybody, that different shutter speed implies more/less light exposing the "frame" and that explains the difference in ISO rating.


It sure isn't clear to me.

A camera should, except for differences in gamma and perhaps maximum IRE level, have a constant ASA. This value we dial into a meter.

Now we set the shutter-speed on the camera and on the meter.

Next we measure the light (e.g., EV).

Then we dial in the EV and read-out the F-stop.

+++++++++++

Although the gamma curve and perhaps maximum IRE level are believed to alter camera ASA -- I haven't seen how the gamma curves value and clipping level can computationally convert one ASA to the other.

In short, it seems a good theory -- but I haven't seen anyone actually compute the two ASA values from the gamma curves and perhaps maximum IRE levels. I don't even know how to compute this.

But, one should be able to build a logicical agruement of how gamma curve A and a clipping level A should be able to move the "A" ASA lower or higher than the "B" ASA which uses gamma curve B and a clipping level B.

A camera should, except for differences in gamma and perhaps maximum IRE level, have a constant ASA. This value we dial into a meter.

Hmmm. A faster shutter speed brings in less light. This produces a different , lower, signal from the CCD. That's the best way I can describe it.
If you keep the frame rate constant, say 24fps, a 1/48 shutter speed is your baseline, say 200 ISO, at 1/96 you reduce the amount of light by one stop and you'll have 100 ISO.

Hmmm. A faster shutter speed brings in less light.

Yes, but that has nothing to do with ASA. The first step is to set ASA on your meter. It never changes for a given film nor should it for a given camera electronics.

Light-level is dialed in from the measure. Then shutter-speed vs Iris are recriprical.

I'm still back at the CLAIM a camera has different sensitivity values based upon the chosen gamma curve and/or clip setting. Gamma curves have a number that represents the curve. Like 1.8 or 2.0.

Does anyone know what this value means?

Does anyone know HOW this number would alter camera sensitivity? If the number is larger does it increase or decrease sensitivity. What is the equation for this?

What are the values of the two gamma curves that have been used for the testing. For example, what is the value of CineGamma?

If one understands all this, it should be possible to compute an ASA at one gamma from the ASA at the other gamma. If one can compute -- then it validates the verbal explanation. Without a math proof, it's all just words to me.

Paolo,

We were doing some camera tests using TCV3 and Mr. Dashwood's Warm .scn files. I have not yet been able to color and grade the footage yet, but I'd like to get your input on something.

I'm really impressed with how your TCV3 renders colors and I would very much like to use it for the commercial we will be shooting soon. But one thing bothers me - we shot and will be shooting under 3200K lighting and implementing a hot window behind the subject, similar to the images I've included below. While white balancing the camera for 3200K and implementing TCV3, the background outside goes a little.... or a lot.... turquoise.

http://www.sect1.com/files/josh/take_one.jpg

Compare that to Mr. Dashwood's Warm .scn file, which has a much more pleasant bluish cast.

http://www.sect1.com/files/josh/take_two.jpg

When it comes time to shoot, I will be placing Rosco 3411 (3/4 CTO) on the window to drop the skylight down in color a bit, and obviously that will affect the color - I'm just not sure in what way because the gel has yet to arrive and I'm unable to test it.

I'm wondering if you've had a similar experience with mixed sources and the TCV3 settings and/or any advice to give me. Thanks, you've already been a great help with the work you've done on the .scn files.

Joshua.
This is a rather simple case to test. Go outside and white balance for that light and check the colors. You are mixing two different color temperatures here so I'm not surprised of the result. As you mentioned, adding gel to the windows and bringing the same color of light will solve the problem. Another approach that you can use is by using daylight balanced lights to light the interior scene. Both LitePanels and Kinoflows can be used for that.

Paolo,

Forgive me if I wasn't clear enough on my first post. I'd like the window to be cooler, but the hue that daylight is rendered while white balanced to 3200 is a little too turquoise under TCV3 for my tastes. I was looking to match the discrepancy in color temps more like Mr. Dashwood's Warm settings, where the blue is paler. Or would any attempt to alter the color matrix negate the affect of "true color?"

Thanks for your help.

TrueColor is just a starting point, I tweak it from time to time. For example, I had to alter the level of the blues in order to get some reds to display the way I wanted. The important thing is to do it in a controlled environment. All the settings for gain and rotation affects each other. Think of a rubber band stretched around six poles. If you move one pole the tension at the other points changes. In order to make a change to one parameter you need to look at the result on a Vectorscope.

Good luck.

Paolo, thank you very much. Guess I'll have to wait until the gels come in.

So what about the 250 w/ the 17x5 HD lens? Will your settings hold up to the new lens? I'm sure it's going to affect something...and what about the 250 in general? What would you expect to happen if I loaded them onto one?

Are these settings specific to the 100? I'm learning from you, man - thanks.

Paolo, thank you very much. Guess I'll have to wait until the gels come in.
I think your blues are closer than you realize. The one shot with TC3 is a lot more saturated, bringing up the "turqoise." However, I think the darker blue and the turquoise blue compared next to each other are about the same.

When you turn down the saturation of the TC3 image in Photoshop, or another image program, aren't the blues in the two pictures matching in hue better?

Paolo, your suggested settings bring spectacular improvement. My staff and I have EVERYTHING to learn about the technology of color in our cameras. That is, we know NOTHING about it. But by simply implementing your suggested settings for the GY-HD100U found in the watermarked chart at this location:
http://www.paolociccone.com/hd100-calibration-3.html
...we have improved our color from 'poor' to 'spectacular.'

We had been wondering why the color was not quite what we expected from this wonderful camera. I still don't know "why"...but I do know now what to do to fix it...thanks to you, Paolo.

Bravo, Paolo! È la gente come chi, che ripartiscono la loro ricchezza di conoscenza, che fa il la maggior parte per rendere a questo mondo un posto migliore, giornaliere.

Riconoscente, --Scott

Are the truecolor 3 settings written for the compression that occurs when you record to tape as opposed to doing direct capture to either something like the Firestore DR-HD100 or on to a laptop via Scopebox? Or would they be considered to be the optimum settings to record directly to disk as well? If not, has anyone done something similar for recording directly to disk?

There are no settings written for compression or not. Please refer to the article that I wrote about TrueColor, http://www.paolociccone.com
The configuration has been designed to be the most neutral, true to 1:1 representation that the HD100 can achieve. This is not changed by the compression. On the other hand capturing component out at 4:2:2 will allow you to preserve more color information and TrueColor is designed to avoid data loss.
BTW, while I initially designed TrueColor with the smaller DSC chart, I recently tested him with with the 28-Color and I'm happy to report that it actually held very well. I might tweak it a bit later but it's a very solid configuraration.

Scott, I don't think I replied to this before, I just saw it. Thank for the kind words.

I still don't know "why"...but I do know now what to do to fix it...thanks to you, Paolo.


Very few cameras come out of the factory correctly calibrated and "correctly" is a subjective term. Some people could argue that just having the camera reproducing reality is not really their target. I suspect that some of the defaults in these cameras depend on someone's idea of what is most likely to meet the taste of the target audience and sometimes these ideas can be totally off, as it is in the case of the HD100. In my way of shooting I like to have the least amount of "interference" from the camera so that I can set up the scene and light it create the mood that I want and if my eyes like what I see then the camera should save that image.


Bravo, Paolo! È la gente come chi, che ripartiscono la loro ricchezza di conoscenza, che fa il la maggior parte per rendere a questo mondo un posto migliore, giornaliere.


Il piacere e' tutto mio, Scott. :)

Steve...

I agree with your statement that ASA "should" be invariant. However, in my own tests on a Canon XL2, I noticed slight variations in the ASA value with changes in shutter speed. This can only be explained, in my mind, by theorizing that the sensitivity of the sensor varies with exposure in a non-linear fashion that isn't compensated by the camera's internal mapping. Another possibility is that the non-linearity is compensated in the light meter calibration. Testing with a separate light meter should answer this question.

Hi, is the true color setting ok for the Hd 201/250 ??

Hi, is the true color setting ok for the Hd 201/250 ??

There is a version that I made for the HD200, see my website: http://www.paolociccone.com

[QUOTE=Paolo Ciccone;483025]Hey Enzo, e tu lavori a Hollywood e Cinecitta' ? ;)

"If you do broadcast work normally, TC might be reproducing "illegal" levels. Of course you can reduce them in post but you're settings require less manipulation. I'm aiming at making the HD100 as close as possible to film so I didn't check the values against broadcast levels."

Please forgive me if I get this wrong. I am looking to set up my HD201 with the TC3 colors. I think these settings might be what I am looking for. You say that the TC may represent illegal levels. In what way? I am shooting for broadcast quite often but am not used to setting up my camera, tough Eric Gulbransen has given me some confidence I can. Should I not do this then?

I am also trying to match my HVX200 to the 201 and I am looking for filmlike colors. Thanks in advance.

By the way, this forum is great.
Svein Rune

Svein, it you have a HD200/250, you should use my TrueColor HD200 either version 1.0 or 2.0, the color matrix for the HD200 is completely different from the HD100 and the settings don't apply.
My comment about the "illegal values" means that the camera can produce digital values that are beyond the range of SD broadcast TV. They might be. It depends on what and how you shoot it.

Paolo.
Thanks. That was the quickest response ever. Wow.

Are these the settings you refer to?

Master black -2 Color Gain Normal
Black Stretch1 Color Matrix Standard
Detail MIN R Gain 3
White Clip 108% R Rotation 4
Knee Manual G Gain 2
Level 90% G Rotation Normal
Cinelike OFF B Gain 3
Gamma CineLike B Rotation -3
Gamma Level -1

Svein Rune

Hello Paolo
I have a question
i'm using true color V3.0 (love it)
should i have Motion Smooth on?

thank you
Joe

Joe, Motion Smooth should be yanked out of the camera, if possible
Turn it off, for the love of God man, turn it off ;)

Hello Paolo
i turn it off
i don't know how it got on i must have done it by mistake
once again you help me
Thank you
Joe

Paolo.
Thanks. That was the quickest response ever. Wow.

Are these the settings you refer to?

Nope. TrueColor HD200 is at http://paolociccone.com/TrueColor-HD250-part2.html
Alternatively I have another configuration adjusted for Nikkor lenses but that could work for the stocker as well: http://www.dvinfo.net/conf/showthread.php?p=807147#post807147