HP5+ over exposed by 4 stops, will my recovery plan work ?

Hi folks,

I was rather annoyed at myself to discover that I over exposed a 120 roll of HP5+ by a whopping 4 stops ! The good news is that I think my highlights will not be compressed at all but I wanted to run my solution past you.

The SBR was 7 stops from Zone II to VIII inclusive which would have otherwise lead to Normal (N) development time. In short I plan on reducing the film development time to be that for N-1 which in my process will lower the gamma from 0.5 to 0.4

Below you can see my analysis, but I am amazed that I can still keep my scene in the straight line portion of the curve by forcing a minor reduction to its CI/gamma/slope. I will just print using 1 grade harder MG filter. What am I going to lose out on here ? This seems too easy to recover from.

** warning that a bit more theory and maths will now follow... **

I have analysed the HD/characteristic curve of HP5+ and it is linear out to at least a density of 2.1 log units which has basically been my saving grace here. Normally a neg exposed for N development in my process would result in a density ranging between 0.3 and 1.35 on the neg. [gamma=(1.35-0.3)/(7x0.3)=0.5]. 1.35 is nowhere near 2.1 but 4 stops over exposing would put the density right on 2.1, however if I reduce the gamma to 0.4 and calculate the range for Zone II to IX, you will see it comfortably makes it in. I am calculating from zone II up to Zone IX (i.e. an 8 stop range) as I want to include even the last nuance of detail in my highlights. So here is the formula:



Now because D_zone_IX_end<2.1 , I should be OK. In fact because I still have a bit of wiggle room I could develop to N-0.5 which would give D_zone_IX_end = 1.92 Developing for N would put D_zone_IX_end right at a density of 2.1 and permit no room for any error in dev time/temperature/age/minor film exposure variations etc.


regards
Peter

I'd go normal too.

Thanks Michael and Mark.

Michael, have a look at figure 4 here. It shows graphically what I thought happens to the curve's shoulder with increasing development times. Basically the shoulder moves up and to the left of the HD curve (Dmax of the film eventually causes the shoulder to clip as dev time is increased). Two other examples. Dead Link Removedshowing the HD curve for TriX and another example using AZO paper (which I assume could be applied to film in general) here in chart 3 . Both show the shoulder moving up and to the left with increasing development time. If I draw a vertical line say mid way along the straight line portion off all those curves and then start moving to the right (up) each curve, one will traverse a longer stretch of the straight line portion on the traces with less development. Note that I'm not seeking to shift the exposure range down the curve, I always lock the exposure range to the horizontal axis. Less straight line portion this means less over exposure latitude.

Finally on p. 230-232 of WBM 2e there are also some pertinent notes on Exposure Latitude. Near the end of p.231, we read "when in doubt it is better to err on the side of under-development allowing for more exposure latitude".

regards
Peter

In general, you take a normal scene and develop N-1 and use a higher paper grade or filter to print.

That's a general reciprocity idea.

You figured that you are on the straight line. So you could develop Normal and use the same paper grade or filter as if you had exposed normally. The contrast should be the same as you would have normally gotten. You would develop less if you expected a longer than normal scale subject would have shadows still at 0.3 and highlights at 2.1, but yours isn't going to have a thin shadow.

You will have a density 1.05 to 2.1 if you develop normally, and the only difference from a perfect normal negative that you would print for 16 seconds is that you will print this one for 40 seconds.

I am generalizing however. You will have more grain. To reduce that, your original plan to develop N-1 is fine.

Thanks for your observations Bill. I didn't know that increased exposure results in more grain.
A max density up to 2.1 is right on the cusp of HP5+'s shoulder and leaves me no room for deviations/error in other process variables which is why I am inclined to dev to N-0.5 or N-1.

regards
Peter

I am nobody qualified to interject with any sensitometric evidence. Only intuition. And I'd tend to believe a lot of picture information and local contrast has been pushed up or over the shoulder. And the only hope would be undervelopment. Trading one bad thing for another in this case is the only possible hope. The first remedy that sprang to mind was pyro, but I have no experience with the stuff to back up my thought. However, "normal" development is something I'd discount immediately and close my mind to it.

Peter I think you are thinking too hard on this.

If these shots are really that important just shoot another roll the same way in a similar situation, cut it in half develop both ways and see which you prefer.

Mark, they REALLY are that important. It will take me another 2 days to travel out and back to get the shots again ! This exercise will provide all of us with valuable insights to apply next time we accidentally overexpose a roll of film by at least 3-4 stops. If my approach turns out not to work I suppose I could spend the time taking them again, but at least I want to give this roll the best opportunity for success.

I feel for you, many of us, me included, have already made these mistakes, we know what happens with our subject matter and our techniques and our tools. For me 4 over is an oh well got a little extra grain moment, 4 under though, now that gets me grumpy.

There are lots of variables here, the only real way to see what you're going to end up being acceptable given your subject matter, using your techniques and your tools, is to test.

I recently (and accidentally) overexposed a very important roll three stops and pulled lots in development to compensate (4min instead of 9 in DD-X). I had never done anything like this before, but the shots were all passable. It worked to my advantage in the end because the contrast of the shots would have been far too high otherwise. The grain wasn't bad either.

I recently shot some HP5+ in the Otway Ranges (dark) at 100asa. Up to 4 minute exposures to counter reciprocity.

I developed them at 5 minutes, down from my N of 10 minutes. Turned out fine.

If you're really that worried and/or they're that important, perhaps shoot some more HP5+ at the same exposure and run some test times?

LJH and heterolysis thanks for your feedback and ideas. And Mark yes I will test if I don't feel comfortable with my plan to dev for N-1.

Hi Peter,

My recommendation would be to not do a N-1 development but rather use a two-bath developer - such as Barry Thornton's. The two-bath developer will ensure that your highlights do not blow out and that the mid-tones will retain their normal contrast. You will probably need to print for longer but you will find that in 'printing down' down the negatives, the lower mid-tones and upper shadows will have more 'sparkle' or vibrancy of detail than you normally get. If you do a N-1 development you cannot be sure that this will bring the highlights down enough and will give you flatter negatives which you will then have to print on a harder grade of paper.

Also, several people have advised you that overexposure produces more grain. This a very long standing photographic old wives tale (been around since, at least, when Victor Blackman used to write a column in the UK's Amateur Photographer magazine), often repeated and passed on to subsequent generations without anyone really thinking it through logically. To explain:

a) On every negative that has ever made, you will have a range of tones from the shadows to the highlights. If it was a bright sunny day you will have at least 8 stops difference between the deep shadow to brightest highlight. This effectively means that the highlights have received 8 stops more exposure (i.e 'overexposed' by more than 8 stops) than the shadows. Are your highlights grainier than your mid-tone/shadows? If the concept that overexposure caused grain was correct, all of our photographs should have grainer highlights in comparison to the mid-tones/shadows - which simply is not the case. Where grain is the most noticeable in the print (especially in B&W) is in areas with a lot of brighter mid-tones (such as a blue sky, etc). This is simply because this tone is where we optically notice it more.

b) Those of us who test for our own personal Exposure Index (EI) all achieve differing results in accordance with our own preferences. In reality that means that, where I use an EI of 200 for Delta 400 processed in my choice of developer and a friend of mine, Johann, chooses to shoot Delta 400 at an EI of 400. Therefore, my films are 'overexposed' in comparison to his films and, according to folklore, my images should be grainier than his. In reality my images are far finer grained than his. The reason is that Johann likes noticeable grain. The real difference between our negatives is that he doesn't care about detail in the deep shadows and he uses HC110 dilution A with vigorous agitation..

The key thing with how grain appears in the final print is that it is affected by:

Type/speed of the film's emulsion
Type of developer used
Temperature of the developer
Time of development
Agitation method
Print developer
Grade of paper used

In conclusion, if you use a N-1 development you will not be sure of saving your highlights BUT you can be very sure that you will lower the film's overall contrast which will result in you needing to print on a harder grade of paper that will accentuate the appearance of grain (hence the cause of this misunderstanding that overexposure causes grain) and potentially make it harder to achieve a pleasing balance should you need to undertake dodging and burning (always trickier to get correct at higher grades).

As other have advised, the best would be to shoot a test roll with the same degree of overexposure with a scene that has similar contrast range to your important (as yet undeveloped) films and process it in a two-bath developer. I think that you will be very pleasantly surprised at the results.

Sorry this post ended up being so long and best luck with rescuing your important films.

Bests,

David
www.dsallen.de

Michael R 1974 said:

Of course this all assumes your normal contrast is the same as Ilford's. As I said before, N-1 is a relatively mild contraction and should not cause too much shortening of the straight line, but at best you gain nothing.

The only thing you gain by reducing development is slightly reduced graininess which can slightly offset the inherent increase in graininess resulting from overexposure.

This is one of the most common mistakes people make with the Zone System. You cannot selectively move highlights down the normal curve by reducing development. There is no free lunch. Reducing development compresses contrast by altering the overall slope and/or the shape of the curve.

Click to expand...

This is also why I don't use reduced development. Extra exposure in the camera for me simply means extra exposure in the enlarger. Nothing more, nothing less. Mistakes in camera exposure simply can't be made "normal" by adjusting development, the damage is done when the shutter fires.

The math involved here is a hard task master.

For example those above who are suggesting a staining or two bath developer to "hold the highlights" are essentially suggesting modifying the shoulder to compress the highlight detail. Yes surely more detail may straight print inside the papers range but the top of the film curve has to flatten out to get you that effect. Conversely a "normal" non-compensating developer lets the film curve run out straight(er), the highlights may require burning to get them inside the papers print range, but greater local contrast is available.

Like Michael said, there's no free lunch.

I'd suggest developing the film normally in whatever developer you use that gives the least amount of speed enhancement.

Ilford Perceptol?

stand develop your film in caffenol C 25-30mins it will look completely normal

OK folks, I now have pictorially constructed what I was trying to explain above. I have used Kodak's published characteristic curves for TRI-X 400 as my example (Ilford doesn't publish the same data for HP5). I suggest that if one sufficiently overexposes an image then there is a greater compression of highlight tones if one develops for longer (see the following image). I hope this counters claims to the contrary.



Of course the overall image contrast will reduce but that is the trade-off which can be corrected for with a higher grade paper/MG filter.

Therefore if I go from N to N-1 development, any highlight compression should be reduced.

regards
Peter

First, I added 2 red lines meant only to compare the steepness of the shoulders. The upper line is steeper. The steeper line means better separation of tones, it really is that simple.

Not just the shoulder is flatter, the whole curve is flatter (compressed) with less development. Sure you then use harder paper but there is no gain in detail printed.

Second, is the captain obvious moment. While the theory can be compared using the Tri-X curve, the real shape of the HP5 curve is different. Its shouldering characteristics are different.



And the basic shape of the shoulder isn't absolute, it will change depending on your techniques. It can be manipulated. Below is a good article.

(there was a url link here which no longer exists)

develop in the slowest speed developer you have/can get. Rodinal, pyro
minimal agitation
add in restrainers
these in conjunction can lower your film speed by a couple stops

I wonder exactly how much speed you can lose before too much causes no development of image

Michael R 1974 said:

It doesn't.

You are comparing substantial overdevelopment (which also distorts the curve) to underdevelopment on the graph (development time for normal contrast is 7 3/4 min). A more appropriate comparison would be the bottom two curves. The lowest curve has lower contrast, and more compression of shadow detail too.

You also need to use a curve for the right film. If there is highlight compression on this curve with normal development it is because you exposed past the straight line, which is not what you said in your original post.

Click to expand...

Hi Michael, I was composing a reply to your 2nd last post when this one arrived so I thought I'd quickly chime in before a more thorough reply when I have time.

I don't think that I have left the straight line portion of the HP5 curve but I calculate it to be right on the edge and any variations in process variables could push it into the shoulder so my goal in slightly under developing would be to ensure I don't hit the shoulder. I know in the TRI-X case I compared the two extremes of over and under development, but that was to show the effect magnified which could/should still exist (if one has a sufficient amount of overexposure) as one transitions from a lower to a higher curve. Because I am unsure if I have encroached into the shoulder at N dev time, I wanted to reduce this chance of doing so by under developing.

Regarding your comments about HP5's shoulder being different, Ilford have not included any information in their HP5 characteristic curve so we cannot conclude anything but only assume. Even if its shoulder characteristics were different, I would be surprised if the general shape didn't generally follow that shown with TRI-X.

I will reply more about your red lines later tonight or tomorrow morning (OZ time of course !)

regards
Peter

markbarendt said:

First, I added 2 red lines meant only to compare the steepness of the shoulders. The upper line is steeper. The steeper line means better separation of tones, it really is that simple.

Click to expand...

Thanks for persevering with me Mark. Unfortunately I don't see why the absolute value of the steepness of the curve in the highlight/shoulder section is the thing to compare. I say this because I will be scaling the ENTIRE curve up or down based on the paper grade I select. For the sake of discussion, consider this scaling process to be proportional across the entire curves by assuming that the paper's transfer function is linear (I know it isn't but just assume it is for this point). So imagine I scale the upper and lower TRI-X curves to a position mid-way between those two curves. The shoulder sections will NOT lie on top of each other, and there will still be less highlight separation in the scaled top curve. I can tell this by observing that the ratio of D2/D6 (=21.3) is less than D1/D7 (=10.9). If they would lie on top of each other those ratios would match. Ignore D3 and D4 as I extrapolated those parts of the curves.


As you mentioned I should be comparing the middle curve to one below it, but the effect is too small to observe with the limited data present in those curves. If there was more data in the curves showing shoulders for each dev time then it would be easy to compare two middle curves, but in lieu of that missing data it is not unreasonable to assume the effect we see at the longest dev time exists in the middle curves at their shoulders if sufficient over exposure occurs.

markbarendt said:

Second, is the captain obvious moment. While the theory can be compared using the Tri-X curve, the real shape of the HP5 curve is different. Its shouldering characteristics are different.

Click to expand...

As I noted above, Ilford have not included any should information in their HP5 characteristic curve so we cannot conclude anything only assume. Even if its shoulder characteristics were different, I would be surprised if the general shape didn't generally follow that shown with TRI-X.

markbarendt said:

And the basic shape of the shoulder isn't absolute, it will change depending on your techniques. It can be manipulated. Below is a good article.

(there was a url link here which no longer exists)

Click to expand...

Very informative thank you. At least I know not to reduce my agitation lest I further compress my highlights !

Time now to walk the dogs and get to bed.

regards
Peter

If you change how you process your film, and you shot a normal contrast scene, your final negative will come out with lower contrast, and the only way you can counter that is to use a higher contrast paper.
If you develop 'normal' with an overexposed negative, you will retain the full negative contrast, which enables you to print on a normal contrast paper. The only thing you have to do is to open up your enlarger lens, use a stronger light source, or simply just expose the paper longer.

This gives you MORE flexibility at printing time, because you can still default to a higher grade paper if you need it for artistic reasons. But if you have already used that higher contrast to compensate for a low contrast negative, well you're out of options. Some of it can be remedied by using a high contrast print developer, like Dektol, of course.

Anyway, if it were me, I would shoot another roll/sheet of HP5+ in similar conditions with the same gross overexposure, and then test my way to the best likely outcome before​ I processed the very important ones.

PeterB said:

I don't see why the absolute value of the steepness of the curve in the highlight/shoulder section is the thing to compare. I say this because I will be scaling the ENTIRE curve up or down based on the paper grade I select.

Click to expand...

A bigger real difference is easier to get more contrast from, especially if you need to burn in or are fighting against the paper's curve to get that detail printed.

PeterB said:

As you mentioned I should be comparing the middle curve to one below it, but the effect is too small to observe with the limited data present in those curves.

Click to expand...

The two ideas above are related and the effect is too small to see in real life too. The point is that there is no gain to be had. It's the equivalent of moving your wallet from your left pocket to the right pocket, it doesn't change how much money is in your wallet.

One reason that I almost never adjust film development/change the contrast rate is that my normal film development (regardless of my camera exposure placement) sets me up to get the best prints from my paper.

I only move away from my normal paper grade as a last resort. Moving away from my normal print grade compromises the print because the the shape of the paper curve starts changing too. That doesn't mean I won't change paper grade, just that I use it sparingly.

PeterB said:

As I noted above, Ilford have not included any should information in their HP5 characteristic curve so we cannot conclude anything only assume. Even if its shoulder characteristics were different, I would be surprised if the general shape didn't generally follow that shown with TRI-X.

Click to expand...

You are right that with the info at hand we are guessing a bit, but just because Tri-X 400 shoulders a certain way doesn't mean HP-5 will follow suit. Flip a couple pages from where you got your graph and look at TXP/Tri-X 320 it doesn't shoulder like Tri-X 400 and it is even made by Kodak.

PeterB said:

Very informative thank you. At least I know not to reduce my agitation lest I further compress my highlights !

Time now to walk the dogs and get to bed.

regards
Peter

Click to expand...

Exactly, that is why in this particular situation the compensating development schemes and developers aren't the right answer.

A little extra agitation would mitigate a good bit of the process variables that you are worried about.

Sleep well.

Agree that between N and N-1 will serve you well.

The shoulder is a response of the film to light, it's not necessarily controlled by development. But some developers can create shoulders. You don't want a shoulder.

You are going to have a dense negative. A dense flat negative may be harder to work with.

This example shows HP5P 135 developed to a CI of 0.56 in Xtol. The graph on the left shows normal exposure with a resulting negative density range of 1.06. The graph on the right shows a three stop "over" exposure with a resulting negative density range of 1.02.