Just a thought about 35mm adapters and ISO ratings
- Thread starter cinebuddy
- Start date
PaPa
Major Contributor
but then you also have to consider the loss of light from the camear lens itself yes?
And how can you use a light meter for anything other than contrast ratios, if as dennis says, stopping down a few stops on the 35mm lens is much less in terms of light loss as if you were to stop the camera lens down. So the F stop on the reader would be inacurate yes?
With the loss of light with the adapter, lens, and camera f stop, i would be curious to see if my camera even rates at 64 asa.
And how can you use a light meter for anything other than contrast ratios, if as dennis says, stopping down a few stops on the 35mm lens is much less in terms of light loss as if you were to stop the camera lens down. So the F stop on the reader would be inacurate yes?
With the loss of light with the adapter, lens, and camera f stop, i would be curious to see if my camera even rates at 64 asa.
with and SGpro the HVX sits around 100.
Barry_Green
Moderator
Rating a camera at a certain ISO is a speculative prospect at best. For one thing, it all depends on what gamma curve you're using -- if you're talking about an HVX for example, HD Norm will give you a base ISO of 320, but CINE-D will be more like 250.
And you can't just say that:
And you can't just say that:
because it's very dependent on what camera was underneath it. It might be 100 on an HVX, and if so, it'll be ISO 25 on an HV20. The Brevis itself doesn't have any manner of ISO associated with it. It's the adapter + camera combination, and even then especially with a Brevis it'll depend also on which ground glass you have installed.
Ok then it rated the Brevis/Zeiss Glass/and HVX at 100. I understand the gamma curve part. But my question is does the glass on the front change effect the ISO rating? I might be asking this question incorrectly, i'm a light meter newbie. The reason I ask is because Ryan's test was without glass on the LE, now with a 50mm f1.4 on the LE HVX Black Press i'm getting f2.8 at ISO 160 at 24fps. Now i've confused myself.
Ryan Patrick O'Hara
Veteran
Very true. The ISO will differ based on many different settings. But also remember with the HVX and M2 combo there are TWO iris/aperture dials.
I found in order to find the ISO using an 18% gray card one needs to set the HVX iris and pretend it's fixed. That way you are now working with one aperture and can execute a normal lighting test and lighting curve chart.
Now my results will most likely report a lower ISO then yours because I do not keep my HVX wide open (unless I HAVE to) due to the obvious gain in edge to edge sharpness when HVX iris is shut down slightly.
With so many variables such as HVX iris, Lens speed, gamma settings, and other tweaks, it is best to do your own test with custom settings. This way you know exactly what your ISO is per project. Think of scene files like film stocks, they all have different ISO's.... except they don't usually vary as much.
Lenilenapi
Carbonite Member
I agree with Barry, ISO ratings in video are only ballpark aproximations at best. the fact that none of these adapters respond to light in a linear way throws it completely out the window. i would look for other tools to compare adapters.
You could ask how much an adapter loses on a specific video camera with its iris wide open and a 35mm lens set to f1.4 and f2.8 but much more general gets very dicey.
- Lenny Levy
You could ask how much an adapter loses on a specific video camera with its iris wide open and a 35mm lens set to f1.4 and f2.8 but much more general gets very dicey.
- Lenny Levy
Gotcha...understood....I know its approximate, but people still light with 35mm adapters and a light meter, so I still would like an answer to my question. Does the lens affect the ISO reading? Am I misunderstanding the results of the FreshDv tutorial in that the adapter/camera combo reading changes the moment you add glass to the front, and threby every time you put a different piece of glass on your ISO may change. If you key a person with a 2k at 24fps IS0 160 and your reading is f 4 with a 50mm 1.4 and you change the lens to 85mm 1.8 is that reading accurate now?
Lenilenapi
Carbonite Member
The answer is an absolute NO.
Nothing is accurate in that way with video or especially with these adapters. Depending on the adapter you may see no change at all between those two lenses. The meter is a tool for lighting but not a reliable exposure indicator like in film.
You can come up with aproximate ISO's but realize what the parameters are when your using them that's all. I use a light meter but i set exposure to a monitor.
- Lenny Levy
Nothing is accurate in that way with video or especially with these adapters. Depending on the adapter you may see no change at all between those two lenses. The meter is a tool for lighting but not a reliable exposure indicator like in film.
You can come up with aproximate ISO's but realize what the parameters are when your using them that's all. I use a light meter but i set exposure to a monitor.
- Lenny Levy
PaPa
Major Contributor
the only use for a light meter for video and an adapter would be to check your contrast ratios.. which there in itself is very important, but nothing to exposure related settings, since, as its been said, one stop down on a 35mm lens -> ground glass does not equal an actual full stop of light loss.
Barry_Green
Moderator
They do? Never seen one on a set, never seen anyone try to do that...
(okay, not strictly true, I think Macgregor brought out his light meter once to set a background light to a certain ratio to the key, but other than that, no, I don't remember the last time I saw a light meter on a video set, much less a video set using an adapter).
You'd be much, much, much, much, much better off to use a waveform monitor than a light meter when dealing with stuff like this. A light meter is okay for roughing in or setting ratios, but for actually setting exposure? Not so useful.
Perfectly right here, roughing and ratios, but for actually setting exposure, not so accurate.
Lenilenapi
Carbonite Member
Barry I use it very ten when setting lights, especially for a larger shot or especially where people are walking through different areas or presetting when the cameras not available, but its always checked against a monitor for contrast and the monitor determines the actual f stop.
Lenny
Lenny
cardmaverick
Well-known member
The reason why you don't get "accurate" exposure with a light meter has so much to do with people using the wrong approach to setting it up, as well as not really understanding how to use them anyways.
NEVER base it off 18% gray when using video (or any digital camera for that matter). I thump this fact all the time here and no one ever seems to notice or listen.
The reason meters are based on 18% gray is because that is: 1.) an established standard density for comparison 2.) light meters, being designed for film originally, take into account that film has "iffy" boundaries exposure wise. Theres no hard white point in film nor is there a hard black point, and EVEN FILM STOCKS HAVE DIFFERENT CURVES. Thus they chose a density somewhere in the middle that all films could reproduce. Ever heard film DPs say a stock is 5 stops over and 3 stops under? They are referring to how many stops of range you get above and below 18% gray when you use the meter to reproduce 18% gray as 18% gray.
I guess what I'm saying is that the "video gamma curve variable argument" is total BS. Film has variable curves depending on what stock you use for shooting, as well as the stock its ultimately transferred to for projection in the theater. Theres a reason people shoot tests... to figure out the film stocks curves and generate an "exposure game plan" for their shoot. The same thing is done by DPs on major digital features.
If you base your rating off of green channel clipping at say 90 IRE, you will always be safe using your meter. You will get a nice strong exposure without loosing the highlights. If you switch your gamma curve and knee setting, well, guess what, you're basically using a new stock, so you should recalibrate, but I find that if you create a rating based off of the most contrasty settings, your pretty safe even when you change the knee, but its more optimal to just create a new rating.
As for this issue with the adapter...
I haven't done too much work with adapters, so I can only speculate, but here's my theoretical approach:
1.) Using the stock lens, figure out your settings, get an idea for what the curve looks like, etc.
2.) Put on the adapter. Once you have your camera aperture at its optimum for focus on the GG, lock it off and stick with it!
3.) With a fast lens on the camera open all the way, point it at a bright white wall, evenly light. If the waveform is blown out, adjust the lights, not the camera or lens. once you hit a sweet spot, like 90 IRE, you should be able to figure out the rating for the setup. Remember, f/stops are standard amounts of light transmission, t/stops are as well, but they are even more accurate due to the fact that they take into account not only aperture diameter and focal length, but light loss from lens elements themselves.
As for the issue of stopping down not having as much darkening "effect", I simply see this as like putting a diffusion filter onto the cameras stock lens, it simply lowers contrast, thus it changes your curve essentially, so....
5.) Go back over your camera settings, odds are your blacks aren't as deep anymore and the curve characteristics are now different. How many stops difference there is from say 70 IRE to 30 IRE, for example, will now be different.
Simply put, do your homework before a shoot! I just spent some time down south on location shooting with the HPX500. Earlier in the week during the equipment pack out and prep I set up the camera and my light meter. I never needed to rely on a monitor or waveform on set when adjusting my exposure, I was too busy to drag those around. Everything came out like I wanted, no burnt highlights, ect...
I gotta make that darn tutorial video....
NEVER base it off 18% gray when using video (or any digital camera for that matter). I thump this fact all the time here and no one ever seems to notice or listen.
The reason meters are based on 18% gray is because that is: 1.) an established standard density for comparison 2.) light meters, being designed for film originally, take into account that film has "iffy" boundaries exposure wise. Theres no hard white point in film nor is there a hard black point, and EVEN FILM STOCKS HAVE DIFFERENT CURVES. Thus they chose a density somewhere in the middle that all films could reproduce. Ever heard film DPs say a stock is 5 stops over and 3 stops under? They are referring to how many stops of range you get above and below 18% gray when you use the meter to reproduce 18% gray as 18% gray.
I guess what I'm saying is that the "video gamma curve variable argument" is total BS. Film has variable curves depending on what stock you use for shooting, as well as the stock its ultimately transferred to for projection in the theater. Theres a reason people shoot tests... to figure out the film stocks curves and generate an "exposure game plan" for their shoot. The same thing is done by DPs on major digital features.
If you base your rating off of green channel clipping at say 90 IRE, you will always be safe using your meter. You will get a nice strong exposure without loosing the highlights. If you switch your gamma curve and knee setting, well, guess what, you're basically using a new stock, so you should recalibrate, but I find that if you create a rating based off of the most contrasty settings, your pretty safe even when you change the knee, but its more optimal to just create a new rating.
As for this issue with the adapter...
I haven't done too much work with adapters, so I can only speculate, but here's my theoretical approach:
1.) Using the stock lens, figure out your settings, get an idea for what the curve looks like, etc.
2.) Put on the adapter. Once you have your camera aperture at its optimum for focus on the GG, lock it off and stick with it!
3.) With a fast lens on the camera open all the way, point it at a bright white wall, evenly light. If the waveform is blown out, adjust the lights, not the camera or lens. once you hit a sweet spot, like 90 IRE, you should be able to figure out the rating for the setup. Remember, f/stops are standard amounts of light transmission, t/stops are as well, but they are even more accurate due to the fact that they take into account not only aperture diameter and focal length, but light loss from lens elements themselves.
As for the issue of stopping down not having as much darkening "effect", I simply see this as like putting a diffusion filter onto the cameras stock lens, it simply lowers contrast, thus it changes your curve essentially, so....
5.) Go back over your camera settings, odds are your blacks aren't as deep anymore and the curve characteristics are now different. How many stops difference there is from say 70 IRE to 30 IRE, for example, will now be different.
Simply put, do your homework before a shoot! I just spent some time down south on location shooting with the HPX500. Earlier in the week during the equipment pack out and prep I set up the camera and my light meter. I never needed to rely on a monitor or waveform on set when adjusting my exposure, I was too busy to drag those around. Everything came out like I wanted, no burnt highlights, ect...
I gotta make that darn tutorial video....
Ryan Patrick O'Hara
Veteran
Why not use a 18% gray card and use it to find ISO? I do it all the time at film school and it works. I usually light the gray card until my in camera marker reads 40%... (40% because of video's ability to better represent the lower zones vs it's horrible upper zone ability where whites usually blow out much faster)
And because gamma settings and different scene files will have different ISO's, one just simply needs to have the settings on when one performs the ISO test!
And in reference to your curves, just make a curve chart. I have made a curve chart for the HVX scene file CineD, the Cinealta default, the Varicam Default, and two Cinelike (1&2) for the SDX. It's not hard.
So lets say I am working on a film and I have my scene file down before filming. I simply need to find the ISO by using an 18% gray card and exposing it at a certain f-stop to get a 40% reading on the marker. After which, without touching the fstop on camera, I will move the lights in toward the card (or just add more) every time I get exactly (to the tenth of) one stop more of light (on the card) I will write down the % the marker reads. Until I reach +100% I keep going. Same with below N (40%)
Here is a complete and made up sample of what a chart would then look like:
Stops %
+4 99+%
+3 90%
+2 82%
+1 65%
N 40%
-1 30%
-2 21%
-3 14%
-4 8%
-5 5%
And of course I now know the ISO of which to set the light meter.
As Barry did mention, Video sets are usually without light meters. A waveform monitor is a great way to look at exposure. But I find the main reason the lightmeter is dead on video sets is the inability for people to set and use them. Plus, (I'm not knocking anyone) the combined laziness of looking at a preview monitor to judge exposure has also plagued the use of the light meter. I for one am a strong believer in the zone system and spot metering.
So as you can see, I'm not using the gray card for anything but a control for the test. It's all based off of my personal exposure of skin tone for 40%. Now is skintone 18% gray? No. But I know that and everyone has different skin tones. Lighter then 18% gray card and darker then 18% gray card. So I find this a suitable way to find exactly what I need.
Feel free to respond if I am misinformed. I'm always open to new ideas.
And because gamma settings and different scene files will have different ISO's, one just simply needs to have the settings on when one performs the ISO test!
And in reference to your curves, just make a curve chart. I have made a curve chart for the HVX scene file CineD, the Cinealta default, the Varicam Default, and two Cinelike (1&2) for the SDX. It's not hard.
So lets say I am working on a film and I have my scene file down before filming. I simply need to find the ISO by using an 18% gray card and exposing it at a certain f-stop to get a 40% reading on the marker. After which, without touching the fstop on camera, I will move the lights in toward the card (or just add more) every time I get exactly (to the tenth of) one stop more of light (on the card) I will write down the % the marker reads. Until I reach +100% I keep going. Same with below N (40%)
Here is a complete and made up sample of what a chart would then look like:
Stops %
+4 99+%
+3 90%
+2 82%
+1 65%
N 40%
-1 30%
-2 21%
-3 14%
-4 8%
-5 5%
And of course I now know the ISO of which to set the light meter.
As Barry did mention, Video sets are usually without light meters. A waveform monitor is a great way to look at exposure. But I find the main reason the lightmeter is dead on video sets is the inability for people to set and use them. Plus, (I'm not knocking anyone) the combined laziness of looking at a preview monitor to judge exposure has also plagued the use of the light meter. I for one am a strong believer in the zone system and spot metering.
So as you can see, I'm not using the gray card for anything but a control for the test. It's all based off of my personal exposure of skin tone for 40%. Now is skintone 18% gray? No. But I know that and everyone has different skin tones. Lighter then 18% gray card and darker then 18% gray card. So I find this a suitable way to find exactly what I need.
Feel free to respond if I am misinformed. I'm always open to new ideas.
Last edited:
cardmaverick
Well-known member
You certainly can use 18% gray based off a certain IRE value, but like you said, you need to create different profiles, etc.
The reason I don't normally support this practice is because...
Video DOES have an exposure limit, the highlight. And its pretty darn well defined, even more so when you can view the cameras linear output like Red.
Because you have a hard point to be under, I think its better to base the rating off of that point. It's also more user friendly to just find the brightest area in the scene with the spot meter and just use the f/stop it gives you.
When you base off of 18% gray it becomes a guessing game. Can you really find 18% gray in the scene that easily? Many people have a hard time doing this, further more you can have pools of light in a scene, making things more complex. Placing a gray card in the scene for metering is now kind of tricky. Do you put in the brighter spot or the darker spot, etc.
If you base off highlights, its super fast, simple and easy.
I'm not saying you can't use the meter the other way around, I'm just saying its a bit more challenging than it needs to be. Ironically, video's hard highlight point does have some beneficial use... :shocked:
I totally agree about video's "WYSIWYG" ability making many people very lazy, as well as very ill informed. Ever notice its always the "easy" technologies that spawn lots of crap?
As soon as cameras like RED become more prevalent, people will get back into light metering, because video isn't aways a WYSIWYG format like many believe it to be.
Barry_Green
Moderator
Video is exactly a WYSIWYG format. Red isn't a video camera, and neither is film, therefore your comments would apply to those. But with a properly-calibrated monitor, video is nothing *but* WYSIWYG.
cardmaverick
Well-known member
Two words:
Color Correction
Have you ever shot a show without any knowledge of post processing, only to get screwed by the colorist? Been there, not going back. I said "it isn't always". Meaning sometimes it is, sometimes it isn't. I've had to work both ways in the past.
PaPa
Major Contributor
how many zones do you alocate for video work? I know film its 0-10 5 being medium grey, but what aboud video for lack of latitude?