Sekonic - L358 Calibration

Joined
Jun 1, 2007
Messages
880
Location
U.S.A
Does anyone know how to calabrate a Sekonic - L358 with either LightRoom or Bridge - CS5?

Thanks
Gary
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
Yes, take a photo of an 18% grey target using the (manual) exposure indicated by your L358 and then open it in PS. Using the color sampler tool determine the level of the target image. Since the target is neutral grey it should indicate a nominal value of 128. It won't be perfect but as long as it's within about 8 or so it's ok. If more than that, the camera is overexposing. If much less, it's underexposing.

In order to correct the difference between what the L358 indicates as the correct exposure (which you set in the camera) and what your camera actually does using that exposure you can calibrate the L358 in 1/10 stops using its calibrate feature as follows, per the L358 user manual...

"1. To enter the calibration setting of the meter it must first be turned off.

Press the power button on while holding down the ISO 1 and ISO 2 buttons simultaneously; the screen will display CAL 0.0 (for calibration).

2. The calibration setting can be changed by rotating the Jog Wheel while pressing and holding down the ISO 1 and ISO 2 button simultaneously. A range of +/- 1.0 EV in 1/10 stop increments is possible for calibration. The calibration setting is not displayed on the main screen once it is set."

In this way you can calibrate the L358 in order to compensate for what your camera is doing verses what the L358 indicates as the correct exposure. For example, if your camera is underexposing you can add positive compensation using the L358 calibration feature. Thereafter, when setting the L358 exposure in your camera the target image will sample close to 128 which indicates you have the correct exposure.

I hope that makes sense. :smile:
 
Joined
Jun 1, 2007
Messages
880
Location
U.S.A
Yes except.....Is this light room or bridge? In bridge I can find what I think is the color sampler tool. However, when I do that it is three numbers. They are kind of scattered.

Also I did buy a 18% grey card today at a local camera shop. They had two different brands. There was a huge difference to me in the color/brightness of the two. I choose the darker one.

Does this test shot have to be lit by flash? or can you use ambient.

Also when calibrated. When I take a meter reading will I still set my camera to match what is on the meter?




Thanks so much.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
I used Photoshop as an example since you said "CS5". The idea is to sample the reference target which should be as close to 128 as possible if your camera is exposing properly using the exposure indicated by the L358. Any program that can do that will work.

For example, use the 'Info' tab in PS and watch the RGB values. They should each be as close to 128 as possible. If so, then the grey target exposure is technically correct. But remember, technically correct may not always be what YOU want or what looks good for a typical full color (non test) image.
 
Last edited by a moderator:
O

Oldman

Guest
Yes, take a photo of an 18% grey target using the (manual) exposure indicated by your L358 and then open it in PS. Using the color sampler tool determine the level of the target image. Since the target is neutral grey it should indicate a nominal value of 128. It won't be perfect but as long as it's within about 8 or so it's ok. If more than that, the camera is overexposing. If much less, it's underexposing.

In order to correct the difference between what the L358 indicates as the correct exposure (which you set in the camera) and what your camera actually does using that exposure you can calibrate the L358 in 1/10 stops using its calibrate feature as follows, per the L358 user manual...

"1. To enter the calibration setting of the meter it must first be turned off.

Press the power button on while holding down the ISO 1 and ISO 2 buttons simultaneously; the screen will display CAL 0.0 (for calibration).

2. The calibration setting can be changed by rotating the Jog Wheel while pressing and holding down the ISO 1 and ISO 2 button simultaneously. A range of +/- 1.0 EV in 1/10 stop increments is possible for calibration. The calibration setting is not displayed on the main screen once it is set."

In this way you can calibrate the L358 in order to compensate for what your camera is doing verses what the L358 indicates as the correct exposure. For example, if your camera is underexposing you can add positive compensation using the L358 calibration feature. Thereafter, when setting the L358 exposure in your camera the target image will sample close to 128 which indicates you have the correct exposure.

I hope that makes sense. :smile:
Be sure to crop the photo so nothing but the gray card is showing. If you don't it will not be accurate.

Dwight
 
Joined
Apr 3, 2006
Messages
5,608
Location
Texas
Page 26 of the 358 manual covers it. And here is a article from Sekonic about how to calibrate the light meter. They do not mention gray cards.
http://www.sekonic.com/Support/FAQs/Calibrating-your-flash-meter.aspx

Here is an article pointing out that all histogram data is gamma encoded anyway.
http://www.scantips.com/lights/gamma.html

Never a subject that anyone wants to understand, but real nevertheless. It specifically points out two facts - that the 128 value is not the middle of anything we can see, and that 18% graycards have no relationship to 128 on the histogram (the error may be coincidentally small, but it is simply a wrong idea which makes no sense - false science). It is good to know that with a reflected meter, it simply will not matter if if you use a white card, a black card, or an 18% card, you will get the same result (which is not the midpoint of the histogram).
 
Joined
Jan 15, 2010
Messages
3,874
Location
UK
To throw something else into the mix, it has always been my understanding that:

Light meters do not measure 18% grey. They are factory calibrated to ANSI standards. ANSI standards specify calibration based on Luminence not Reflectance. That Luminance factor equates to roughly a reflectance value of 12% grey (although not all manufacturers may follow this).

So if you are basing exposure on taking a reflected reading from an 18% grey card then you will need to compensate by opening up around 1/2 stop. It is a long time since I have seen the instructions but I am sure there is something about this on the Kodak grey card instructions

I believe the 18% value has been used in the print world the 18% said to be the approximate value of reflected light on a print at the half way point between black and white.

So anyone know where I can get a 12% grey card :smile:
EDIT: While the last comment was in jest I have just found that 12% grey cards are being manufactured Lastolite EzyBalance Grey 12%/White :redface:
"Lastolite have launched this product following a number of requests from photographers who are now using a 12% grey as their midtone reference when calibrating their camera before they start shooting. This compensates for the camera’s factory calibration settings and positions the midtone spike on the histogram in the centre position rather than slightly off to one side."
 
Last edited by a moderator:
Joined
Apr 3, 2006
Messages
5,608
Location
Texas
So anyone know where I can get a 12% grey card :smile:
It won't help. :smile: I do agree with 12% (Sekonic, Nikon and Canon do use K=12.5 for reflected meters). Kodak did say if metering their 18% card, to open 1/2 stop for average scenes. But it seems good to involve the obvious facts into the discussion.

We can only meter directly from a card if using a reflected meter, and we will get the same photo result metering a 12% card, or an 18% card, or a 90% white card, or a 5% black card. Simply does not matter which card we use. The light meter tries to make all of them come out the same middle gray.

These different cards (possibly also in different degrees of incident light) will of course give widely different reflections giving different camera exposure settings, but then when using those settings for a photo of that card, the reflective meter will try to make all of them come out the same one middle gray tone (some middle gray perceived by our logarithmic eyes and brains).
Here is an example of that, which any one can easily duplicate:
http://www.scantips.com/lights/metering.html#sun

Qualification: The huge difference is that if using those metered card readings for exposure of OTHER random and general average scenes, we would see better results from a 12% card, or opening 1/2 stop from an 18% card. Yes, the 12% card could help then. We do meter an exposure value. But what I am saying is that if photographing only the card, with the notion of resetting the meter so that this histogram spike is at its center point, then it could not matter less which card we use (different exposures, but same middle gray result). It is simply a wrong idea.

The big issue about a midpoint value in histograms is that any histogram we can ever see has been gamma encoded. Our RGB standards require gamma 2.2, and we never seem to realize that this of course means the data values have been modified by this formula. And then (theoretically in a perfect world) 18% does become 117, coincidentally at about 46% of 255. This is near enough the "center" that it badly confuses the troops. However this notion of "midpoint" is greatly complicated by gamma encoding moving the linear midpoint 128 to become 187 gamma, at about 73% of 255.
More here: http://www.scantips.com/lights/gamma.html

So since it will not matter which card you use... 18%, 12%, or a white card or a black card, and since the histogram data has been gamma encoded so that 128 moves to 187, then what is the logic of using an 18% card to change the calibration of our light meter? You can instead use a sheet of white copy paper, and get the same histogram result. It will still be gamma encoded however.

Modifying our light meters calibration so that a 18% gray card is at 50% on the histogram has been a false tale for years. But for entirely unrelated reasons, it is a coincidentally a small error. It can be then be compensated, so that it does not shut us down. But it is still certainly the wrong idea - no science behind it. Absolutely nothing says 18% ought to be midpoint of our gamma histograms. My bet is on Sekonic, they surely know much more about it than that.

Sekonic echos the ISO standards for calibrating meters, and of course never mentions gray cards. They say if you notice that all pictures in many different scene settings consistently come out a little off, then calibration can recenter those results for a better average. Said another way, if you can set a specific compensation in the camera and it is usually right, then you can set that same compensation in the meter instead.

But using only any one metering may not be typical. Also good to realize that metering an 18% card has no significance. The reflective meter tries to make the camera settings show anything and everything as middle gray. Simply how reflective meters work.

Incident meters are a different story, but we cannot directly meter a card with an incident meter. If using an incident meter which meters the actual light source (instead of the metering the reflection from the subject), THEN a picture of a 12% card will be seen to come out darker than 18%, and 90% white will come out brighter than 18%.

But they all come out the same with a reflected meter.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
WayneF said:
Page 26 of the 358 manual covers it. And here is a article from Sekonic about how to calibrate the light meter. They do not mention gray cards.
That is the page I quoted, in part. However, it's not describing how to calibrate the meter to a camera. Thus, there is no reference to a grey card or any other calibration target. The instructons are simply on how to use the meters calibration feature.

My posts detailed how to use that calibration feature in order to calibrate the L358 or L758DR to a specific camera by using a known reference source, such as a grey card.

This has nothing to do with reflected meters such as those found in a camera. You are simply calibrating the Sekonic meter to a specific cameras exposure when using the Sekonic's indicated exposure settings.

WayneF said:
Modifying our light meters calibration so that a 18% gray card is at 50% on the histogram has been a false tale for years. But for entirely unrelated reasons, it is a coincidentally a small error. It can be then be compensated, so that it does not shut us down. But it is still certainly the wrong idea - no science behind it. Absolutely nothing says 18% ought to be midpoint of our gamma histograms. My bet is on Sekonic, they surely know much more about it than that.
At the risk of an appeal to authority argument, what I detailed is according to both the MAC Group and Sekonic Corporation. And using an 18% grey reference target is what they suggest when using the Sekonic calibration feature to match it to a specific camera.

The point is, to use a known reference source so that both the meter and the camera are in agreement with one another. The color of the target is not critical as you said, however, the 18% grey target is the industry standard.

I will reiterate...

If we presume the L358/L758DR indicates the correct exposure then we set the camera to that exposure while using manual exposure mode. We then take a photo of a known reference source. The camera should then correctly expose that target using the Sekonic's exposure settings. If it does not, then the camera is either over or underexposing when compared to the exposure the L358 or L758DR indicates. To correct for this difference we use the Sekonic's calibration feature as detailed in its user manual. Using this method allows the Sekonic meter and the camera to be within 1/10 of a stop of each other.

There are some who will disagree with this method and will instead compensate by adjusting the cameras exposure, only using the Sekonic's indicated exposure as a starting point. However, that is an iterative process based on personal experience and is not always repeatable.
 
Last edited by a moderator:
Joined
Jan 15, 2010
Messages
3,874
Location
UK
I see and understand what you are saying here and with one exception (although you have put in a some qualifications) I agree entirely.

Forgetting for the moment how good metering systems and camera manufacturers algorithms have become in producing fine results under a wide variety of lighting situations the fact remains that there is no real intelligence behind these systems.

So, as you say if you fill the frame with any one colour or black, white or grey and make an exposure the camera will of course make them all to come out the same density a mid grey. For anyone that doubts this It is so easy and quick to test – just go around taking images of white, black, grey and even coloured objects making sure you fill the frame and the results should be very similar densities.

The value of the grey card is to have a known reflectance value that can be introduced into the scene. If it is 12% then metering it directly and assuming this is what the manufacturers use then you do not have to make any changes to exposure. As you say if it is 18% then open ½ stop should get the same result. You could of course use any other shade of card, white or black for instance and then adjust the metered exposure by +2 or -2. Without any reference card I have often metered the palm of my hand making sure that the orientation of the palm is similar to the subject and knew that I would need to open up by 1 stop to get an accurate exposure (at least I used to with film!).

In the OP’s case I think the most important factor is to match the Sekonic meter response to that of the camera so he can just “set my camera to match what is on the meter?

I also suspect that the metering mode in camera will be important i.e. use either centre weighted or spot (I just do not know enough about how matrix does its magic yet).

My suspicion is that there is value in metering in camera a known value be it 18% or ideally 12% (as this is what the camera averages to?) and then this value should probably appear around the centre point of the camera histogram. If this is not the case then it may indicate the need for exposure compensation in camera.

As I believe that Photoshop uses a different method of generating histograms than those of the camera manufacturer I do not think this would be the best way to go.
At this stage then is the point IMO that the Sekonic should be calibrated to match the camera meter (at least for spot and centre weighted).
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
Tony W said:
In the OP’s case I think the most important factor is to match the Sekonic meter response to that of the camera so he can just “set my camera to match what is on the meter?
Yes, I believe that is what he was asking when he asked how to use the Sekonic calibration feature. And is what the calibration feature is for, using the method I described. The Sekonic manual simply describes how to use the calibration feature, not why you would want to or how.

It has nothing to do with the cameras reflected meter. I say that since reflected meters keep coming up in the discussion and can confuse the issue.
 
Joined
Jan 15, 2010
Messages
3,874
Location
UK
Yes, I believe that is what he was asking when he asked how to use the Sekonic calibration feature. And is what the calibration feature is for, using the method I described.

It has nothing to do with the cameras reflected meter. I say that since reflected meters keep coming up in the discussion and can confuse the issue.
The method you described (although I have no experience of the Sekonic) seems to me quite logical and correct. The only thing I am a little uncomfortable with is the differences that may exist between Photoshop histogram generation vs Nikon.

Edit: Just realised it is well past my time to turn in for the night
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
Understood. But you don't necessarily have to use PS. The goal is simply to determine whether or not an individual camera is exposing correctly when using the ambient light exposure indicated by the Sekonic meter.

If not, then the camera is at fault, although you calibrate the Sekonic meter to match that camera.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
Understood. But you don't necessarily have to use PS. The goal is simply to determine whether or not an individual camera is exposing correctly when using the ambient light exposure indicated by the Sekonic meter. If not, then the camera is at fault, although you calibrate the Sekonic meter to match that camera.

Some have asked, "Why would you change the meter by calibrating it when it's the camera that's not exposing correctly?" The answer is, because you can't calibrate the camera except by altering the exposure indicated by the handheld meter. And if you do that then using a handheld meter would be rather pointless.
 
Joined
Apr 3, 2006
Messages
5,608
Location
Texas
The idea is to sample the reference target which should be as close to 128 as possible if your camera is exposing properly using the exposure indicated by the L358.
You said this twice, it seems the heart of your system, but this was the specific part that cannot be right. For reasons explained, there simply is no possible relationship between an 18% gray card and 128 at the center of a histogram.

The gray card is left over from analog film days, when we had no better tools, and certainly no histograms. The gray card is NOT a digital concept. It can still have the same uses it always did (about judging the tone level of printed output), but it is not a digital tool. We can also use it as a neutral color for white balance, but it is pretty dark for that, and not spec'd to be specifically neutral, so now there are lighter tone "digital" gray cards which are much lighter than 18%. More expensive, but these are more carefully controlled to be neutral tone, for white balance. The 18% part is not a digital concept.

We called 18% cards as "middle gray", meaning we thought the human eyes logarithmic response perceived the 18% value as middle gray (when inside our heads). But who would know? :smile: The data was still 18%.

And 128 was the numerical midpoint of the histogram. One stop down is 50% of the light. However, this is of course only true in linear data, specifically only in RAW data at the sensor. However, we never see a linear histogram (no tools to show RAW), so everything we see has been gamma encoded (which now shows very different values. Specifically, 128 at 50% is moved to be about 187 at 73% in gamma data).

People did make a false connection, assuming the word middle must only have one meaning, and therefore, every instance of the word has exactly the same meaning and significance. Unfortunately, this is not remotely true. 18% is not 50%. And worse, we never see linear RAW histograms anyway, we only see gamma encoded histograms after RGB conversion. These are the simple facts.

Granted, some people do try to calibrate their lightmeters by centering a 18% card in their gamma histogram. Unfortunately, we still find articles promoting it even if it is the wrong idea. Fortunately, it is not a large error, because gamma does move it up near, to be a little below center. But unfortunately, 18% is not the middle of anything, and 128 is not one stop down in gamma data. And also, it is so easy to show that it does not matter which shade of gray card is used for this, since any shade from white to black works the same as 18%.


That is the page I quoted, in part. However, it's not describing how to calibrate the meter to a camera. Thus, there is no reference to a grey card or any other calibration target.
We do agree that there certainly is no mention of gray cards in the Sekonic meter calibration procedure.

If we presume the L358/L758DR indicates the correct exposure then we set the camera to that exposure while using manual exposure mode. We then take a photo of a known reference source. The camera should then correctly expose that target using the Sekonic's exposure settings. If it does not, then the camera is either over or underexposing when compared to the exposure the L358 or L758DR indicates.
That part seems OK, except histogram 128 is certainly not the determination of correct exposure of an 18% gray card. We would need some reason why that could be true, and all the facts show there is no possible relationship.
 
Joined
Apr 3, 2006
Messages
5,608
Location
Texas
My suspicion is that there is value in metering in camera a known value be it 18% or ideally 12% (as this is what the camera averages to?) and then this value should probably appear around the centre point of the camera histogram. If this is not the case then it may indicate the need for exposure compensation in camera.
We agree on many things, on each of your several points, except more qualification could help this one.

The reflected meter does try to provide an exposure which will make any scene average to some "middle gray".

All indications are that this tone is about 12%, for any reflected scene.

Of course, it does not appear at 12%. The reason is that the histogram we see is gamma encoded, so 12% ideally in a theoretical perfect world, would be about 38% on the histogram. The original 50% center point has been moved from 128 at 50% up to about 187 at 73%, in a perfect world. But the world is not perfect, and the digital camera is doing its own things inside there, contrast and white balance and brightness adjustments, which shifts tones a little, in an unknown way.

Usually we see a broad distribution of tones, instead of a narrow peak. When we do see a narrow peak, it is usually a little below center, but at times, we may see it a little above center. So the histogram will never show numerical precision of these things - the camera does manipulate things. The basic concept is there however.


As I believe that Photoshop uses a different method of generating histograms than those of the camera manufacturer I do not think this would be the best way to go.
Right, there are samples of each method shown at http://www.scantips.com/lights/gamma.html (in the yellow box along the right hand side). When the camera is showing its one single luminance value, it is showing theoretical computed grayscale luminosity values, which have their own meaning. The camera can however also show the three actual individual RGB curves, which Adobe typically shows overlaid in place.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
WayneF said:
That part seems OK, except histogram 128 is certainly not the determination of correct exposure of an 18% gray card. We would need some reason why that could be true, and all the facts show there is no possible relationship.
Wayne, it isn't 'my system'. It's one many people use and as I said is the method MAC Group/Sekonic USA suggest using. Including the 128 RGB value with a tolerance of +/- 8, , when calibrating the meter.

Remember, it's an ambient exposure, not a reflected exposure as determined by the camera. In addition, the reference target should be evenly illuminated within 1/10 of a stop, ambient. Again, as determined by the Sekonic meter.

Although there is nothing wrong with them, I purposely avoided or limited theoretical discussion or mention of reflected measurements in this case since they can sometimes confuse the issue and not serve to answer the question.

If someone wants to know why something does what it does and the science behind it, that's fine. In fact, I'm all about in depth understanding of things. And that is something you're very adept at explaining. But sometimes a simple step-by-step "how do I do this" serves the purpose. It's unusual for me to say this, since there was a time when I was involved in another form of instruction whereby the simple 'see this, do that' method of teaching would likely get one killed. It was only by teaching others how to think did they learn to survive and in fact dominate. I only say this since it seems we're butting heads and coming at the original question from two different angles.

In this case, however, I am acting as more of a messenger and only relaying what I've been told and what has worked for me and many others on an almost daily basis.

I'll end with these questions: what, exactly, is correct exposure defined as? That is, from the standpoint of an ambient light meter, not a reflected meter or ones creative intent. Simply put, use a handheld meter and set the camera as it suggests. Now take a shot. If the image is either under or over exposed then which exposure is correct: the handheld meters exposure settings or the image the camera produced using them? How do you know? And lastly, once again, the manual tells us how to calibrate the meter. But why, exactly, do we need to do so and how does one go about it if not the way I relayed?
 
Last edited by a moderator:
Joined
Apr 3, 2006
Messages
5,608
Location
Texas
Wayne, it isn't 'my system'. It's one many people use and as I said is the way MAC Group/Sekonic USA suggest doing it, including the 128 RGB value or as close to it as one can get when calibrating the meter.
Can you point out where Sekonic says this?
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
In writing? No, but I've spoken to them many times. MAC Group knows me quite well. I both use extensively and am involved in beta testing for several products they distribute. Again, i'm just acting as a messenger, in case you want me to show you the math. :wink:

Edit: The only thing in writing is how to calibrate the meter. Again, as shown in the user manual on page 26. "How to" in this case meaning what buttons to press and what knobs to turn. But it doesn't tell you why, what it means to calibrate it or exactly how to go about it. That's what I'm here for. :biggrin:
 
Top Bottom