This is also my experience. Well-lit urban zones or monuments usually require EV4 @ 100 ISO. That corresponds to 1/15@f/2 @ 400 ISO. As Mark says, inevitably you will have both blocked shadows and some burned highlights, but the scene will be rendered as a night scene with a well described subject matter. If one wants to nail the exposure better the only* possibility is a spot light meter.
Originally Posted by rich815
The first picture, Saint Peter's clock, had such sharp lighting that even while using a spot light meter (which was not so spot as I was several hundreds of meters away so the 1░ angle covered a "large" area) I had to bracket some 0.5 EV to be sure not to burn the highlights. This is one of the few occasions, and the only one I remember, where I did bracket exposure. I remember I always was quite in the vicinity of EV4@100 ISO.
* Besides bracketing, that is.
The general experience I have now is that EV4@100 ISO is very likely to work well with slides. Walking around with a negative film I would use EV4@100 ISO without using any light meter.
Actually my point is that pushing is a choice (one of last resort for me), not a given. Detail need not be lost without thought and mid-tones need not be gritty unless you want them that way.
Originally Posted by Diapositivo
Street scenes at night, can be photographed quite nicely when film is used at it's ISO speed and developed normally.
That choice just means that we may need to adjust how we shoot, that we may need to carry a little extra gear, and even break the stereotype. Adding a touch of fill flash is a helpful thing used by pros, as are: bean bags, posts, walls, monopods and other camera supports.
Also spot metering isn't the only choice either. Yes for many shots, like your clock, it may be quite practical but where ever you can get into the scene, incident metering works just fine. The Jiffy Calc oneANT and Sirius talked about above works well too, the Ultimate Exposure Computer works nicely too.
Incident metering and night doesn't necessarily work very well and I personally would advice against it. In sun light the sun is so to speak always the same, one has a value for sunlit areas and a value for shadows.
With artificial light, and floodlights in particular, the sources of light is ambiguous and/or not reachable - think the dome of a church lit by floodlights, one cannot go up the dome to measure what's the light there - its distance from the subject is important for exposure and any kind of "averaging" can be very misleading because of the great brightness range.
Imagine a street lamp on a house, projecting light on the house (an awful lot, but diminishing very fast) and on the street. In this situation an incident light meter is basically useless. Besides, light sources at night are often in the frame and they must not be counted as far as exposure calculation is concerned, and with an incident light meter one ends up always calculating, in the exposure, also the brightness of the street lamps.
"Table" exposure is as we agree quite reliable in this situation.
Originally Posted by Diapositivo
Seriously, if you do as is normal for incident metering and stick the meter "against the subjects nose" or you are "in the same light" incident metering works perfectly.
If my subject is under a street light and I'm under the next one down the street, I can even meter right where I'm at because the light is equivalent.
If my subjects are going to be walking by a certain spot: on a sidewalk past a street light, store front, whatever; I can walk into the scene (like the one you describe) and meter "the spots I want" then walk back to my chosen vantage point.
As to distance from subject, that actually doesn't matter. Correct exposure for a given subject within a scene, and I'm using the word "correct" very loosely, is the same regardles of distance; your clock or church tower should get the same exposure at 5-meters or 50, as far as the film is concerned it's luminance doesn't change.
What does matter is the subjects size within the composition, that is a real factor in determining camera exposure. At 5-meters the clock may define the exposure fully, at 50-meters with the same lens in play there are more things in view to consider.
It depends on the composition, but generally in a night shot you might have, in the frame, one or several street lamps, or some details such as walls near the street lamp which are very bright, and then all shades of grey down to absolute night black.
The problem here is that although it is perfectly acceptable to have the light source itself out of the film's dynamic range, I personally consider that for the picture to be nice all the highlights which are not light sources (or specular reflections) should not be washed up.
My normal way of operation would be to measure those important highlights (the wall near the street lamp) and try to capture it correctly, letting the shadow fall into black where the film dynamic range dictates.
You can repeat the example above with a white marble fašade possibly lit by a spot flood lamp, a statue on top of a fountain lit by a spot flood lamp, that kind of stuff.
Sometimes historical buildings are lighted in a way which highlights certain architectural features. For instance some flood lights are pointed upward to make a chiaroscuro play with the tympani over the windows, or the lower decoration of the cornice.
In the pictures shown above, Saint Peter's clock and Saint Peter's square northern fountain, from the place where I was (Piazza Pio XII, that's actually in Italy, the place where I was was obviously less lighted than my subject) I couldn't walk to the fountain (Piazza San Pietro was off-limits) and couldn't walk inside the colonnade to try to measure the light with an incident light meter. I also could not fly to the clock which is lit by some spot lights and you have to be there to use an incident light meter. In that clock case using the light meter in the camera would give a useless average between black night and lighted architectural elements. A "table" exposure with some bracketing would have been the less-worse option.
To sum it up in one line, the bigger problem while taking pictures at night might be not your subject but the strongly lit wall behind it. Incident metering will not help you avoid washing it or understanding what might happen there.
I can propose an example which, being taken with a digital camera, it's even more telling of how disastrous can digital be in a high contrast situation:
As you can see there is a "white hole" in the picture which I personally find quite unfortunate. And unfortunately enough, that square has street lamps strategically fixed on buildings so that it is impossible to take a picture without a street lamp in the frame :(
Even if the picture is digital, it shows a situation where incident light metering would not help avoiding the problem. With a spot meter one could measure the wall near the lamp, place it on the "top" of the film curve, and examine where would the shadows of the fountain fall. Some recesses of the fountain would be sacrificed (falling into pure black) but the background would be much more gracefully rendered, with the tone of the building being preserved behind the fountain, and only the light source itself being blown up. Which means I have to go back there with film and a spot meter.
For the architecturally curious, the subject is the Fontana delle Tartarughe in Piazza Mattei:
I'd agree, yes. However the online JPEG version and the printed or full res scan version show quite a bit more detail, at least in the coat and some shadow areas. Lights are burned out to pure white but was not looking for much detail there. It's all choices. I was able to hand-hold that because I used Tri-X at 1600 in Diafine. With the luxury of a tripod I can use Acros at 100 (great reciprocity characteristics on that film BTW), expose differently and get MUCH better detail. Like so:
Originally Posted by markbarendt
Portero Hill, China Basin ContaxRX 35PCDistagon Acros @100 Diafine 5min 20C 08-2006 26 by rich8155 (Richard Sintchak), on Flickr
That was take at 11pm at night.
But for when a tripod is not possible or convenient or the subjects are moving then faster shutter speed and the trade-offs of accomodating such come into play. Pretty much Mark's point, just adding to it.... :-)
I said above that spot metering has its uses. I fully agree that if you can't get into the scene/sme light it can work fine.
Otherwise You seem to be making my case for me. :D
Yes we don't disagree in fact, I just wanted to point out that I find night pictures to present often the kind of situation where incident light metering is impractical.
This might be slightly off the topic because I see that it is mainly discussing B/W while I use slides. I have taken a few night shots of London street Christmas Lights lately with a hand-held camera using a Provia 400 film. Using a tripod in such crowded streets would have been rather difficult. There is evidence of camera shake, so I can't say the photos are exceptionally good. Here are three examples from my flickrs page, if anyone is interested
As I mainly use slides, this might be a slight deviation from the main discussion. i have taken a few shots of the London street Christmas Lights using a hand-held camera with a Provia 400 film. Using any form of camera support in such crowded streets at night would have been rather difficult. Camera shake is evident so I can't claim that the result is exceptionally good. But I can say that it can be acceptable. Here are three examples from my flickr page for anyone interested