I have a Gossen Lunalight exposure meter. The basic unit has reflective (area and spot) and more importantly, incident light metering. The incident dome is not as complex shape as my Weston Master V but it has been invaluable for my slide exposure.
Now to my point ....
With all our discussion about histograms and exposure compensation, would it not be better to use an incident light meter which would completely avoid blown highlights?
It seems to me that blown highlights result either from incorrect metering, which should be taken care of with the in camera (especially spot mode)or handheld meter; or high dynamic range in the scene, which forces the photographer to make a compromise between shadow detail or highlight detail. Even with the Gossen, you will have to compensate in the latter case and decide whether to blow highlights or block up shadows. A meter is an unintelligent device.
F80 is right, the meter is just a reading, it's up to you to screw it up.
An incident meter is always the best place to start given the reality of the subject location (portrait/landscape etc).
However, it is your decision as the "intelligent" participant to adjust for the specific media, i.e. negative film, transparencies, or a specific CCD and its native quirks (specific films also have their own requirements). So yes, its up to you to balance the exposure within the dynamic range of your media using the appropriate settings based on you light meter (Zone system principles). and the tool used to qualify your judgement is your histogram!
"I got a good shot of the others ducking" W. Eugene Smith when recovering from a WWII shrapnel wound.
An incident light meter can be valuable, but as the others have pointed out, it is not a "magic bullet". It will give you an accurate reading of the amount of light falling on the scene, but it does nothing to help identify the range of brightness of the objects in the scene. If you have a scene with an extreme range of brightness, the brightest parts will likely exceed the dynamic range of your recording medium whichever metering tool you use.
I agree. An accurate incident meter would give you a good starting point to tweak your final exposures, as long as your image falls within the exposure latitude of the medium (slide or digital). Otherwise, the fun begins when we have to figure out how to best capture what we see by favoring either the shadows or the highlights.
OTOH, I have grown quite comfortable with the camera meter over the years. Nikon matrix is really amazing. My Gossen has become more of a flash meter than anything else nowadays. YMMV.
My point was relating to all the posts about blown highlights and applying the incident measurement would then ensure they are not blown. From that point the the photographer can decide whether to leave as is or move the exposure "bandwidth" down.
I'm not yet used to my D200 Matrix metering although I have some idea as my Canon T90 film camera has a manually selectable 2-9 point averaging mode.
>My point was relating to all the posts about blown >highlights and applying the incident measurement would then >ensure they are not blown.
As the answers have pointed out, an incident meter will not (and cannot) "ensure" that highlights are not blown. Using one will give the photographer more information, but he will still need to use his judgement to assess the range of light/dark tones in the scene and select a suitable exposure, if avoiding blowing out the highlights is what he wants to achieve.
However, if you can identify the brigthest point in a shot and use the camera's spot meter, add 2 EC and shoot it in RAW, you will have extracted the maximum exposure possible from that scene. No other gear is necessary.
For my use, I used to use a sekonic incedent meter and a pentax 1/21 spot meter with My f3s for shots where I didn't trust the in camera meter. Since I got the D70 I would "occaionally" use the 1/21 with my D200 I use the matrix with the function button on spot to "check" specific areas that frighten me.
"I got a good shot of the others ducking" W. Eugene Smith when recovering from a WWII shrapnel wound.
>However, if you can identify the brigthest point in a shot >and use the camera's spot meter, add 2 EC and shoot it in >RAW, you will have extracted the maximum exposure possible >from that scene. No other gear is necessary.
You are correct in what you say. The issue is that many might not identify the brightest spot but the incident light meter, in practice, always does.
>You are correct in what you say. The issue is that many >might not identify the brightest spot but the incident light >meter, in practice, always does.
Sorry, but I don't follow that. An incident meter, by definition, measures light from the source falling on the meter. It cannot tell you what the lightest object in the scene is, or how light that object is; for that you need a reflective meter, preferably (as Tony says) a spot meter.
>in the scene is, or how light that object is; for that you >need a reflective meter, preferably (as Tony says) a spot >meter.
The incident light meter does not care how dark or light the subject is even if the subject is pure white and in full sunlight. For my explanation lets consider reversal film. The incident reading should be taken with the dome pointing towards the camera (not to the Sun as I read in David's reply) With the ISO speed set it is already calibrated such that it's EV will just avoid absolutely clear film.
I have read that a ccd has a similar dynamic range to reversal film so I was assuming that the incident measurement would avoid blown highlights ("clear reversal film").
Of course I realise that the dynamic range of the senser is limited and just like when I use film I can increase the exposure if I want a subject in an exposure zone outside the dynamic range to be correctly exposed.
This is getting a little pedantic - sorry about that!
The only part of your original post I was questioning was where you said that using an incident meter will completely avoid blown highlights. That is not true; the photographer still needs to assess the range of tones in the scene and use judgement to know whether to use the incident meter's reading or adjust it.
>"The issue is that many might not identify the brightest spot..."
Simple, look for reds and yellows first; then for whatever is most obvious (often sky and always clouds). We are usually looking for the red channel as this almost always clips first (even in skies). The one time this will trip you up and lead to lots of underexposure is under artificial lighting where an improper WB will fool you into thinking your red channel is clipped one or even two stops before it actually is. That's why you might want to do a WB preset even when you using RAW, to help your meter.
>"I have read that a ccd has a similar dynamic range to reversal film so I was assuming that the incident measurement would avoid blown highlights ('clear reversal film')."
CCD and CMOS sensors have a DR of about 10 stops, but SNR makes only about 8 of them useful. DR is more similar to color negative film but latitude is same as slide film (i.e., there is none unless you underexpose).
It can be. In the good old transparency film days I developed a technique, which in effect was pegging the highlights. The basis of this was I pointed the meter towards the light source, usually the sun, or sky in England, and reduced the exposure by one stop (Of course you could do that by halving the iso setting). The reason for this is that it avoided any variations of the angle of the meter, if it was pointing slightly upwards it would give a different reading to if it was pointing slightly downwards. This only is effective for outdoor lighting, in the studio the given value at actual iso works, and this is also the case if you are in a location with a light of bounced light, by the sea etc. So tests need to be carried out as follows. With Digital exposure response different to film but this makes no difference if you calibrate your meter. So the secret is to use the same technique and bracket, say by plus and minus a stop in 1/3 stop intervals per shots. This will give you a series of images which when examined should have your preferred exposure in the same place of the sequence every time. If it doesn't I have failed you and apologise for my misguidance. There is a slight problem with digital because it depends on if you are shooting RAW or jpeg as these files vary in exposure. Shoot a RAW with jpeg in camera and examine them in you viewer. Quite a surprise if you haven't done it before. However, once you have calibrated your camera to your meter and technique of using the meter you will always start at an exposure in the right region. you may have to alter it slightly but it can be the basis of a good exposure workflow. I used this technique with Canon (Film & digital), then Olympus digital and am about to do the same thing with my D200. It is kind of reassuring on a job when you set a value and the camera agrees with you. I still autobracket in awkward situations, especially as I am just putting files on a card and not reduceing my 36 exposures to 12. Even then 12 correct exposures was better than 36 incorrect ones. Addition. It's sunny today so I have just been out and run some quick jpeg tests on my D200. For me I would use iso64 in shadow areas and iso 80 in full sunlight areas. This is the incident reflector pointing at the sky in shadow areas and the sun in sunlight areas for files of camera iso 100 Hope this helps. David
I spent decades shooting chromes, walking around with a Gossen incident meter in my pocket. My early Nikon F's had the simple prisms without the TTL system. I'm glad to see the back of that business. (Yes, I still have a few of these meters.)
Listen, it's not the tools you use so much as knowing how to use them. Right now I'm shooting RAW and learning to control the amazing and complex D200 metering system.
The Weston Master invercone was shaped such that it accounted for angle and this is why it had to be pointed towards the camera. The Gossen dome is well "a dome" and your method might be equally suitable.
I do not understand what you mean by viewing when using RAW +jpg and me being surprised. When shooting in this mode I understood that we are seeing the RAW embedded jpg not the +jpg one?
If the meter needs to measure the light falling onto your subject, this poses problems when shooting distant landscapes !
Blown highlights and loss of nice fluffy clouds caused me headaches for ages until I realised that I actually had no understanding of light & color !
This is where reading books on zone systems, chromozone systems & spot metering helped me to understand light and dismantle a scene into component parts to allow me to get the range I wanted, fluffy clouds and detailed shaddows. I still have to brackets shots and blend in CS to get my dynamic range, but often by waiting for the right lighting conditions this can be avoided.
I think that blown highlights and problems with shadows can be overcome once you fully know your camera and understand it's short comings with respect to dynamic range. I am not saying Nikons Matrix metering is bad, just that not every scene fits the bill and knowing it has limitations helped me.
I too am an old Zone System fan. I find that it translates well to the digital format. I'm not saying everyone needs to learn the whole process (particularly the wet darkroom techniques), but I find within the Zone System everything necessary to avoid blown highlights using nothing but the spot meter and your brain. It requires a fair amount of thought and I'm as guilty of laziness as anyone else, but if you understand how exposure works, you understand the avoiding blown highlights is a matter of decision making, not instruments.
You guys have a very sophisticated metering system in the D200 so why on earth are you using old fashioned meters. Better that we all try to learn how to use the cameras own systems. Surelt, if you use an external meter you will need to be in manual mose in order to 'key in' the speed and aperture. If you want to remain in the 'good old film days' then use the F100 instead. Ron
Ben, your article has one mistake and one point I would take exception to.
Spot metering works on whatever AF sensor you choose.
RAW is not simply a back-up to shooting JPEG that allows you to correct mistakes. Using ETTR we actually aim to push the exposure as much as possible, so the correct exposure for a RAW shot of the backpack would in fact be the closer to the 1/160s exposure than to the 1/250s exposure.
>Spot metering works on whatever AF sensor you choose.
Ah, that's very handy to know - thanks. I shall put that correction on the to-do list.
>RAW is not simply a back-up to shooting JPEG that allows you >to correct mistakes.
I wouldn't for a moment suggest that it is - that was simply the relevant point in that particular guide. You'll find I have a separate guide on RAW+JPEG.
>Using ETTR we actually aim to push the >exposure as much as possible, so the correct exposure for a >RAW shot of the backpack would in fact be the closer to the >1/160s exposure than to the 1/250s exposure.
I would, though, disagree there. I experimented with exposing to the right when it became all the rage a few years ago, and my own experience was that a correctly-exposed shot beat a corrected ETTR shot every time.
>"I would, though, disagree there. I experimented with exposing to the right when it became all the rage a few years ago, and my own experience was that a correctly-exposed shot beat a corrected ETTR shot every time."
This deserves more attention as it is a broadly blanket statement. You couldn't possibly be referring to color accuracy or SNR, but if you are please show me. If you are considering other factors like faster exposures are usually preferable to slower ones, and raising ISO to compensate for slower exposures will cause issues and there are notable break points in IQ at ISO 400 and again at ISO 800, then that would be a different discussion from what we are having here. Selecting an aperture might also be a consideration, but some of Nikon's exotic telephotos actually do better wide open and many lenses are nearing their peak performance at f/5.6 (I switched from f/8 to f/5.6 as the "go to" aperture when I went from the D70 to the D200 precisely because of the ISO 100 versus ISO 200 issue).
>>"I would, though, disagree there. I experimented with exposing to the right when it became all the rage a few years ago, and my own experience was that a correctly-exposed shot beat a corrected ETTR shot every time." >This deserves more attention
To be honest, I got involved in lengthy debates about it around three years ago and have no desire to revisit them.
All I asked for was some proof. I have many images taken properly exposed for JPEG and what I would consider properly exposed for RAW (ETTR), and the only difference I see after I have converted them is better shadow detail and less noise in the ETTR exposures. Since you seem to dismiss my technique as some sort of passé fade and claim that not using ETTR produces better results, then I would appreciate seeing your evidence because I would change my technique and advice if it convinced me.
Let's try to be more charitable, Ron. A meter is only a tool; the D200's meter is a better tool than many, but a handheld meter can still do some things that a camera meter cannot do. It's all about knowing when and how to use which tool.
And there are some very "new-fashioned" handheld meters out there...
>I do not understand what you mean by viewing when using RAW >+jpg and me being surprised. When shooting in this mode I >understood that we are seeing the RAW embedded jpg not the >+jpg one?
As others have already talked about this in great detail all I suggest is you take the same shot in "RAW" and jpeg and look at them in your photoshop browser. I put this point in so that if you did the test I outlined for RAW and jpeg files you may experience different looking files at a later date. From this you can evaluate your incident light meter technique for your own personal use. I found this exposure difference on my Canon 1D and Olympus E1. I have only just moved to Nikon and not yet looked at this technique closely just the quicK jpeg evaluation I did for my reply. I seldom shoot in RAW even though I am a working pro. Don't tell anyone David
Incident metering will avoid blow highlight and "place" every tone in the picture precisely where is supposed to me. The only limit to this would be in a scene where the DR exceeds the capability of the sensor.
>Incident metering will avoid blow highlight and "place" >every tone in the picture precisely where is supposed to me.
I disagree. Tones in an image are supposed to be where I intend them to be, not where and instrument says they should be. A specular highlight, is, almost by definition, 'blown'. Therefore the incident meter cannot place every tone where is is supposed to be _and_ avoid a blown highlight. It is up to me to decide if being blown and appearing the way I want it to are the same thing or not. That may be an extreme example but it illustrates the need for thought.
>The only limit to this would be in a scene where the DR >exceeds the capability of the sensor.
That is indeed an exception, but I think for most people that is _the_ exposure issue and it is an issue that always _requires_ decision making and compromise. No single meter reading from any type of meter can reliably make that compromise the way you want it.
When taking a portraits under studio lights, and trying to capture a mood, I would still use an Incident meter (Sekonic, in my case) to come up with right amount of exposure when using D200, and set the camera to manual.
In a studio lights, I'd measure light falling on the subject from all the light sources, and not necessarily the combined light that camera would see. I use ratios to size up the light falling accordingly. In this case D200 metering would not help you setup your light.
For general photography, I would use 3D Matrix metering. D200 does not necessarily replace an Incident meter.
Since my remarks on this may have been lost in all the replies above, I just want to give a concise response. The answer to your question is no. Measuring incident light instead of reflective light has no bearing on your exposure. The time you spend taking incidental light measurements would be more effectively spent taking a shot of the scene and reviewing your camera's histogram.
Sorry I perceive your answer to be irrelevant to my OP. If a photographer wants to spend a few extra moments on an exposure decision then its around six hours shorter than the first recorded exposure (Niepce's Heliograph)
Many shots are lost in "a few extra moments". I see no advantage whatsoever to packing around a light meter and messing with it during a critical moment. What is it that you cannot see in your camera's histogram that you can see in a light meter?
Tony, certainly there is a place for the well considered exposure in DSLR photography. It might not be on the street or the football field but nothing wrong with it for stills, landscapes, portaits etc. But in essence I agree. In my mind, if one _needs_ to go to the lengths of carrying a hand held meter, a MF or 4x5 might be a better companion to it than a small camera packed with an array of expensive and very capable light meters as well as AF sensors. But as I said before, IMHO a true spot meter is all that one needs to thoroughly and completely evaluate and control exposure, and it is right there in the camera. Its only drawback is that one has to fully understand the principles of exposure and be able to apply them. For anyone who does not want or need to become that proficient, the matrix meter should be more than effective enough.
One of the few things Ken Rockwell ever said that I agree with is that he uses a DSLR to meter shots for his LF equipment. I have done the same; and yes, I do understand exposure enough to do all my shooting with Spot metering if I actually had to – or when it is advantageous to do so, and I know the difference too.
>Many shots are lost in "a few extra moments". I see no >advantage whatsoever to packing around a light meter and >messing with it during a critical moment. What is it that >you cannot see in your camera's histogram that you can see >in a light meter?
You are correct about the potential of missing a shot but then in that case the content is probably more important e.g. JFK on that day.
Re your last point, you are considering two different situations as you will have already taken the shot then spent time evaluating the histogram and possibly taking a corrective shot whilst the incident meter photographer has spent 5 seconds taking an incident reading before the shot and has no need to take another shot.
My OP was only proposing that the method would always avoid blown highlights. However I should have added that such a technique would only be valid after the specific incident meter e.g. in order of their introduction: Norwood/Weston/Sekonic/Gossen/Minolta, are correctly calibrated.
Here is a thread which I have just found which all might find of some interest.
As an early participant in this post and as someone who has followed it through its circumnavigation of topic.
It is probably safe to summarize that incedent metering is, in fact, an "ideal" measurement tool when all components are considered and compensated for, especially dynamic range of the scene. This is most evidenced in a controlled environment like portraiture, where the reading can take place at the subject location.
Where this highly accurate reading is not possible, as in a situation where one cannot physically get to the exact location of the subject, Spot metering a number of "spots" in a subject area of wide dynamic range will enable one to place the exposure where one desires it, ala Zone system.
Practical 35mm photography lends itself to the meters in hand which for convenience/lazyiness lends itself to the incredibly accurate and very useful combination of in-camera matrix and available spot meters.
"Measure twice cut once" is an old and very wise axiom but not always practical "Aim, aim, aim, aim and fire" can frequently cost you the "Decisive moment". which is why my camera is always at 1/250 f8 in my bag.
So in Tony's suggestion, he is using his very expensive and sophisticated light measuring tool (D200 with multiple readouts)in a way not unlike the polaroid backs that have been used for decades by sophisticated MF,LF users.
Me, I am lazy and I have found that I already carry way too much to want to add my old 1/21 spot and my Sekonic Studio pro meters.
My D200 has not let me down especially when compared to my F3s where I needed the other meters.
The link posted previous to this comment indicates a situation where experience with your tools is more relevent than which tool.
The well published photo Guru Bryan Peterson convinced me in a workshop a few years ago that much can be accomplished with Nikon in- camera metering if you learn how it works.
"I got a good shot of the others ducking" W. Eugene Smith when recovering from a WWII shrapnel wound.