Okay, all you armchair photographers...

...take a look at this. For the following, I used Fuji NPH set at

box 400 ISO. My incident light meter indicated an exposure setting

of f11@1/250 sec. That is in keeping with the "sunny 16" rule as it

was a slightly overcast day in Arecibo, Puerto Rico yet quite

bright. The following photos were taken by laying the negative strip

on a light box and using a digital camera. The prints were scanned

on a flat bed scanner. No photoshopping was used except to resize

the photos. <P>

 

<center>

 

<img src="http://www.jdainis.com/arecfilm.jpg">

<BR>

 

The first frame (9) was the indicated exposure, f11, 1/250<BR>

The second frame (10) was overexposed 5 stops f4, 1/60<BR>

The third frame (11) was underexposed 3 stops f16, 1/1000<P>

The prints were made at a one hour WalMart Fuji Frontier. These are

the results:<P>

 

<img src="http://www.jdainis.com/arec9.jpg"><BR>

Normal exposure<P>

<img src="http://www.jdainis.com/arec10.jpg"><BR>

5 stops overexposed<P>

 

<img src="http://www.jdainis.com/arec11.jpg"><BR>

3 stops underexposed<P>

</center><P>

 

The color from the scan seems to be a bit bluer than the actual

photos in which the trees are true green. But, as I said, I made no

corrections in photoshop. I was rather surprised at the results.

They are not that bad.

An amazing demonstration,thanks.One thing that helped you a lot in this example is the cloudy day.The exposure latitude of any film,varies with the brightness range of the subject.You can overexpose a subject with a short brightness range more than a subject with a long brightness range before loss of sharpness,excessive grain,and long printing times cause trouble.On a bright day,you can expect slightly different exposure latitude limits due to the effect of the longer brightness range.

Excellent demonstration James.

 

5 over + 3 under = 8 stops total range used.

 

8 x 0.3 = 2.4 log E total latitude demonstrated in the film.

 

6 x 0.3 = 1.8 log E normal working range of a scene.

 

The film has excellent latitude and a long curve before you run into the toe and shoulder. Using that estimate, 2 stops of the over and under were on the straight line, and the rest was either on the toe or shoulder, or the film has an extended straight line.

 

If the density of the negatives are represented at the correct density, the film is probably conservatively rated as well.

 

It seems to me that is what you guesstimated in another post? I mentioned in the other thread that I had made exposures of 400 film at 1200 and 2400. That falls between your exposures and was done nearly 20 years ago, but it agrees in spirit with what you did. Fuji has a good film there. Wonder how EK would fare?

 

Now, lets see Steve do a series 10 stops over and under.

 

Ron Mowrey

Remember too that scanners are pretty good with underexposed negs, may not be so lucky with optical enlarger with -3 underexposed neg. (?)

Print film, particularly NPH, can take a brutal amount of over exposure. Fuji Reala and Portra NC can actually take more, and I've recovered VPS images shot 4-5 stops over exposed with only slight image loss. Note this was with conventional optical enlargers.

 

The large brightness range in this scene negates over dramatizing the under exposure latttide of print film. If anything, it serves to demonstrate the auto histogram capabilited of the Frontier to extract good information from a thin neg. Try it with general photo work, or use a traditional lab, and the last image will take a beating. A conventional lab would likely handle the middle image better.

An excellent demonstration. Especially useful for those who

obsess over whether their camera handles 1/3 stop

increments vs. 1/2 stop increments, or those who

argue that their camera with twenty-three segment

4-D matrix metering versus is so far superior to last

year's model with only twenty-one segment 3-D metering. Also good for

answering those who ask "I'm screwed -- I accidentally exposed

Portra 400 at 320 -- what can I tell the lab to fix it?".

<p>

Of course, the results could be quite different for different types

of film (particularly if you use slide film). The lab makes

a difference, too, and I'm impressed by the Frontier lab's

ability to pull detail out of the underexposed neg.

<p>

This was standard exercise I remember doing decades ago,

with B&W film and a conventional darkroom. I think we

bracketed in two-stop increments from four stops

underexposed to six stops overexposed, and tried to

get the best possible print of each. It's good for

everyone to try something similar at least once,

to get a feel for what overexposure or underexposure looks like

and to get a feel for how precise one must be in measuring

exposure.

Your point is what ? That the film is a good film - indeed it is.

The comment that the conventional lab would handle the second shot better than the other 2 shots - I cannot agree with - The automated equipment in the conventional lab is very sophisticated equipment, and so is the frontier - the fact is that it is in the scanner and image processing. To understand something about the conventional lab you must also understand the process on the equipment - a full roll negative goes through a high volume (20k prints/hr) machine. There exists a technology some call it order uniformity - whereby the machine compensates its corrections on an order basis for both color and density, there is also some machines that have frame level correction and then there is a third that allows for frame level and order level in increments for both color and density. In effect the technical people of the lab control this response. None of these frames would be difficult to produce on such equipment and achieve similar if not better results.

By the way the color being a bit bluer from the scans - color Gamut is different on the monitor than the paper - you will never get a "True Match" - but if you scan and then print - the prints from your printer should match the print "in theorey" .. you will not be able to tell the difference - it all has to do with the color space.. if you work with Digital - use TIFF files.

So, why would I should I rate print film at other than box speed? I never saw much point in rating 400 film at 320. Since I'm now using cameras that read the DX code, I'm more worried about manually setting the ASA for one speed film and then fogetting DX is off if I change film speeds/types.

I recently had an opportunity for the same demonstration, tho' mine was through pure brain fart.

 

My cousin wanted some photos of her 15-year-old daughter dolled up in a black dress for a school party. I was switching between cameras using the only color negative film I had handy, Kodak Gold 100 (pretty good stuff, really).

 

All was well until I noticed that I'd underexposed several of the nicer poses by setting the aperture for the wrong reading on the flash, which was set to 400. With that black dress I figured there was no way we'd see any texture. The best we could hope for was that the minilab wouldn't make the prints too grainy.

 

Well, the two-stop underexposed shots came out pretty well. Not unacceptably grainy (looked more like Kodak 400 than 100) and some texture was still visible in her black dress.

 

Naturally the properly exposed photos killed. She's a criminally cute kid.

 

Some color negative films can be very forgiving.

"So, why would I should I rate print film at other than box speed?"

 

To give yourself more underexposure protection. Let's take HD400 as an example. It's data sheet says, "Satisfactory negatives result from exposures 2 stops under to 3 stops over the correct exposure." If you set your camera to rate this film at a lower EV (and you let the camera decide the exposure) then you are effectively giving yourself 3 stops underexposure.

 

In the typical "girlfriend standing under a tree with a clear sky in background" scene, it is my opinion that if you mindlessly use weighted average metering it is easier to underexpose the shadows than to overexpose the sky. When we see this picture, more relevant detail is lost in the shadows than is lost in the sky, so we decide the picture is poor. If we overexpose slightly we pick up a "zone" in the shadows. We usually don't care if we lose a "zone" in the sky in such a photo.

I never said I could print 10 stops over,I said from a tone reproduction standpoint only that many more stops of over exposure are built in to most negative films.I then said that I had seen T-Max 400 developed to a contrast index of 0.60 that had almost 10 stops beyond the diffuse-highlight region where the straightline continues without forming a shoulder.I also mentioned that 3 stops of overexposure is a practical limit,or else grain and density cause printing troubles.I found its best not to use any more of a film's over exposure latitude than is needed for adequate shadow detail.

Re. 3 stops underexposed - The Frontier's scanner and software have done a very

good job at extracting shadow detail which frankly I can't even see on the negative

(the trees, left hand corner). There is no way you could make a comparable print on

an analog system.

An excellent demonstration. Especially useful for those who obsess over whether their camera handles 1/3 stop increments vs. 1/2 stop increments.....

 

 

But for transparency shooting, such stuff IS necessary.

My standard exposure latitude test using <A HREF="http://www.photo.net/bboard/big-image?bboard_upload_id=6722684">

this scene (should be labeled EI 6400)</A> has about a

6 stop brightness range, depending on zoom or size of spotmeter,

which might be why I got only -2 latitude for NPH. On overexposure

I drew the line at +5 due to uncorrectable magentafication, which

was less of a problem with UC400. Ken S, I agree with Scott Eaton

that conventional labs have an easier time dealing with overly dense

negatives, especially color correction thereof, than digital process.

Note for instance that clouds in James's +5 shot are too warm.

NPH magentafication I encountered was primarily in scans.

Thanks for this effort James! However, as Bill said, it's not the scene I would have chosen for such test. But thanks all the same man. ...relaxing into my armchair now...

Thanks for the demonstration, James. I am very surprised at those results. I never had the bad luck to over- or underexpose a film so dramatically, not even in the times before DX coding of film cassettes. Nevertheless, the result are far from disastrous!

 

I always shoot colour print films at nominal speeds, measuring for the shadows, and am satisfied with the results. Only exception are low-contrast scenes, which I overexpose 0.5 - 1 stop. Old habit maybe, since the times of VPS.

On July 3rd, Steve Levine said:

 

"All of the Portra 400's can handle almost 10 stops of overexposure!"

 

He promised to post pictures and added that with his series of over and underexposure, similar to James examples above, we couldn't tell the difference.

 

Please go here: http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=008gt0

 

for the entire exchange.

 

I invite Steve to follow up on his promise.

 

Thanks in advance Steve. Eric and I are looking forward to it. I'm sure James would like to see them as well.

 

Now, I would like to add that there is a method by which the images of underexposed negatives can be catalytically enhanced to yield more information than what James has shown. This was the subject of the work I presented at a meeting of the International Congress of Photography. It would allow 3 - 5 stops or more of underexposure with results in imaging similar to a normal exposure. Of course, grain and sharpness would suffer. You don't get something for nothing. I have not tried this method in about 20 years, and it is kind of exotic, but if there is interest, I will describe it on the forum.

 

In addition, if anyone is interested, I can draw and post hypothetical curves to show what took place in James negatives to yield such good results and to explain my math above.

 

Ron Mowrey

<I>There exists a technology some call it order uniformity - whereby the machine compensates its corrections on an order basis for both color and density, there is also some machines that have frame level correction and then there is a third that allows for frame level and order level in increments for both color and density. </i><P>To cut to the chase, this is an overly complex way of saying that logic integrators on analog mini labs are smart enough to make corrections via global bias or each frame. Basically an upgrade from an Atari 2600 to a 5200 in a technology sense. Either is irrelevant since no optical mini-lab exists on the market today capable of accurate slope corrections beyond a 2-stop range in either direction. Even those integrators that are capable of a two stop slope, do so with little accuracy because mini-labs handle mostly amatuer print films, and amatuer print films are too F'ing erratic in terms of quality to even dream of establishing a 5 stop total slope set. 1 1/2 stops in either direction was the best I was ever able to accomplish with the most stable professional print films on the market like VPS and dedicated Kodak analyzers, and even then 1 1/2 stops was a stretch. Conventional mini labs are not capable of the process Q/C, nor material control that is required for this level of population deviation. Their common inability to get neutral or even consistent prints from B/W chromogenic films is a demonstration of this and their general <b>lack</b> of sophistication. If you can't produce grey correctly from frame to frame, don't tell me you can produce neutral prints along a far more difficult 5 stop slope.<P><I>

 

In effect the technical people of the lab control this response. None of these frames would be difficult to produce on such equipment and achieve similar if not better results</i><P>Not sure if these remarks are based out of ignorance, or some stubborn defense of analog printing - perhaps a bit of both.<P>Optical mini labs are limited in their ability to make print corrections even more so than optical custom printing. They have the option of making your print darker, or make it lighter, or change the print's overall color - that's it. They cannot alter saturation or contrast in any respect, and if your neg is under exposed you are screwed and will get a grey print with little color saturation. Digital labs however can also adjust for saturation and contrast as well as non linear histogram tweaks which is why they are so desired over optical labs for amatuer film production. The Frontier can recover the small amount of shadow detail in the under neg by digitally enhancing it while I can guarantee an optical mini lab would have produced a pretty murky print. An optical minilab however would be able to handle the over neg much better because it has no film scanner with a limited dynamic range. When you optically print an over exposed color neg you simply expose the neg onto paper for as long as it takes to slam photons through it, reciprocity be damned.<P><I>I never said I could print 10 stops over,I said from a tone reproduction standpoint only that many more stops of over exposure are built in to most negative films.......I also mentioned that 3 stops of overexposure is a practical limit,or else grain and density cause printing troubles.</i><P>Steve is correct. I routinely handled professional print film from demanding clients and had to manually adjust and correct each frame with exacting color tolerances. Anybody in the professional lab biz will tell you that low contrast print films are capable of 2-3 stops of over exposure and far more than 10 stops of total dynamic range, but the printing software and papers are not capable of handling it, and the film starts to hurt beyond 2-stops over. 2 stops is the reasonable limit for lab software because beyond this you cannot get neutral prints because the print time gets insanely long and paper color shifts results beyond the ability of the software to compensate. I've printed entire weddings shot on 120 film well beyond 2 stops over-exposure, but it was far from pleasant, and the film was starting to lose proper tonal range. It also took 5x as long to print the entire order vs a correctly exposed wedding, and we charged the photog accordingly for being an idiot. My 14 year old cousin can get get properly exposed slides with her Pentax K-1000, so I feel little sympathy when a grown adult can't get accurate exposures with a modern SLR/TLR and print film.

 

 

>>>>>I feel little sympathy when a grown adult can't get accurate exposures with a modern SLR/TLR and print film----Scott <<<<<< Isn't this the scary truth?Catastrophic exposure errors,due to poor camera handling skills,does seem to be a popular malady here?

If you want to shoot weddings,learn to check your settings(fstop,shutter,flash)every damn frame!Also if you decide to carry more than a single speed of film,carry them seperately, and pay attention when you change rolls.

Steve Levine wrote in the Ron-referenced thread:

"All the Portra 400's can handle almost 10 stops of overexposure!

This doesnt mean that over exposing more than 3 stops is advisable,

it isnt." 'Twouldn't work on my camera, which only goes down to EI 6!

That's +6 stops; 9 with +3 EV I suppose. An Agfa minilab did produce

semi-acceptable prints at EI 6: better than most of the stuff that

comes out of disposable cameras with Max 800 anyway. And I metered

for the shadows, so bright spots may have been ~ +11 overexposed.

The reason I started doing latitude tests was a practical one:

to determine which film is best for whitewater rafting protography

amid shady forests.

I am not sure if I am the only one seeing this. The last scan, which is 3 stops underexposed has more details in its green foliage (i.e the lower 1/3 area) than the other two. The other scans include a shot which is 5 stops over!

 

I have a hard time understanding this phenomenon. I even for a moment don't doubt the exposure latitude of a print film. But foliage is anyways a less than normal brightness area. Check with a light meter, and see the foliage is a couple of stops below that ambient light reading. So if you expose it 3 stops under, you still get most details in the last scan above????

Lakhinder,

<BR>         I can't understand it either. Take a look at the clear frame area between the sprocket holes and the foliage area above it on the negative. There is almost no difference in density. (It is the same on the original negative.) To my way of thinking the foliage all should print black the same as the clear frame would.

Here is a slightly more standardized target. Reala 120 (100 ASA) film, shot at EI 100. I have a lot of respect for this film, and

here are some results.

 

(Shot on a bright sunny day. Light measured

with a light meter. The camera/lens combo were tested for accuracy

separately with Fuji Provia, so that factor is eliminated. Scanning

is done with an Imacon scanner. The files were converted to GIF,

and I do not know what artifacts will that bring in.)

 

These are not the original scans, but what

I could correct with adjusting the Curves. Though not very religiously

done here, I tried to make the lower left white block to similar brightness in each. That was not the only goal though.

 

You can see how at 5 stops under exposure some information can not be retrieved. 3 stop under is really not too bad -- some color shift is noticable of course. And then I have +2 and +4 stop over exposure.

Here is 3 stop under

Here is + 2 stop over exp.