About reciprocity failure (BnW Film)

In my last YouTube video (found here), I mentioned something about reciprocity failure.
For photographers, reciprocity means that we generally need to add some time to our exposure when we exceed exposure times of 1s. This is only relevant for film photography. I won’t go into the technical details about what happens. That’s what Wikipedia is for: https://en.wikipedia.org/wiki/Reciprocity_(photography)

Instead, I want to touch on a few points that weren’t mentioned in the video and that I want to expand on. Also, note that this blog post refers to black-and-white negative films.

30s isn’t unreasonable

In the video, I made an exposure of 30s for a measured exposure of 4s for the nbr. 12 film by NoColorStudio. In the video, it might sound like 30s is unheard of, but it is not. It’s not far off what Fomapan 100 would require. At 4s measured exposure, Fomapan 100 calls for 21s. It was probably the chart I used in the field, even though I didn’t note this. I remember googling it, but I don’t remember what chart I used. It might also have been the chart on the Wikipedia article.

In the video, it does sound like I’m shocked by the 30s. This is down to my YouTube “persona”, not that I’m a different person when doing YouTube. I try to stay honest, but as you might have noticed, my presentation has changed from my earlier videos. This is partly due to me being more comfortable in front of the camera, but I’ve also started presenting more like I would if I were giving a public speech. I used to do many presentations in my previous job in different capacities. And I consider myself a decent public speaker, at least in my native tongue. I have started applying certain elements of that into my videos, which I think makes for better videos. One of the key points I use is to emphasise certain words and always believe in what you say. This might then come back at you, like in this instance. 30 seconds was a lot for that particular film and the point I wanted to make, but I realise it might sound like I’m shocked.

The reality is that it’s only 3 stops over, and it’s not unheard of for certain films. I might also have measured it a tad hot, as the sun was moving in and out of rather thick clouds. According to my notes, it was about 2 stops difference between sunny and cloudy. In other words I might have overexposed by 5 stops in total. I don’t think it would have made any difference if it was 3 or 5 stops for the nbr 12 film (by NoColorStudio); both would probably have been beyond saving.

It’s also worth noting that I’ve mostly used Ilford films for long exposures, which are the times I’m most used to. And why those were quoted in the video.

The following is to be found in Illfords document regarding reciprocity failure:

” …This uses the equation Tc = Tm^p
Tm is the metered (indicated) time, and Tc is the
corrected time. P is a factor calculated following a
range of exposure times.
The P-value differs between the different ilford films, but for instance HP uses 1,31. That would mean for a measured exposure of 4s, Tc=4^1,31=6,147…. seconds… “
(source, available 24-06-26 https://www.ilfordphoto.com/wp/wp-content/uploads/2024/05/Reciprocity-Failure-Compensation-v2.pdf )

Film matters

The film you use directly affects reciprocity failure and the compensation needed. Some films handle this better than others. Fuji Acros is famed for holding up well for long exposures, while Fomapan is famed for breaking down quickly.

It’s also worth noting it’s not a linear correlation.
Often referred to as the Schwarzchild effect.

“As the light level decreases out of the reciprocity range, the increase in duration, and hence of total exposure, required to produce an equivalent response becomes higher than the formula states; for instance, at half of the light required for a normal exposure, the duration must be more than doubled for the same result. Multipliers used to correct for this effect are called reciprocity factors” (source available 24-06-28 , https://en.wikipedia.org/wiki/Reciprocity_(photography))

Schwarzchild’s law is given as
$E=It^{p}\$
Where E is the effect of the exposure, I illuminance, and t is exposure duration.
It’s also what Ilford used above. This law has drawbacks and isn’t perfect, but it should be seen as a guideline.

As we all know, Flickr is a great resource for film photographers. I found a chart by Tobias Abel that illustrates the Schwarzschild effect compensation for different photographic materials.

Remember that the Schwarzchild effect is only a guideline and not an absolute. It’s also worth noting that the compensating effect isn’t linear, as seen by the formula. Also, note that the axles in Tobias Abel’s diagram are not linear (step-size differs).

“Schwarzschild effect compensation of different photographic films (ADOX, AGFA, FOMA, ILFORD, FUJI, KODAK) – V3.2” By Tobias Abel
https://www.flickr.com/photos/lennox_mcdough/8468252700/in/photostream/

According to his notes provided by Tobias Abel the graph was created by:

“The information of ILFORD and KODAK TriX films has been aquired by using vector graphic analysis of the graphs included in the film data sheets. The other graphs have been calculated by using the increment values supplied by the manufacturer (eg. for AGFA APX +1 f-stop for 1 to 10 seconds, +2 f-stops for 10 to 100 seconds, +3 f-stops for 100 to 1000 seconds) and a linearization method based on the logarithmic dependency of the correction factor.”

I believe Ilford uses the same graph on all its films, even though the reciprocity failures differ. So, it might be an average. But this gives a good overview of how different films are affected by reciprocity.

Schwarzchild’s law limitations

Many use the previously mentioned Schwarzchild’s law to calculate the reciprocity failure. For example, Ilford. But this law is a bit limiting, and originally, Schwarzchild suggested the E=I x t^0,86. Which proved to be of limited usefulness. And therefore, we have the more modern version E=I x t^p (source, avalivle 24-06-28, https://en.wikipedia.org/wiki/Reciprocity_(photography))

Many other theories and experiments have been done, and I’m not going into more details about them at this point. However, I found this very promising calculation by a user called Jano on Flikr. He states in a comment when asked about the source of the equations, that “…The particular equation was developed using 3 equations from data points made with the chart. So, I developed the equation.”

Source: https://www.flickr.com/photos/janokelly/6804638225/
Avalible 24-06-28 by user ”Jano”

Reciprocity isn’t that important.

As you might know, I’m a big fan of Fomapan and shoot it extensively. Yet, it’s one of the worst-performing films when considering reciprocity. This is a trade-off: Fomapan is a great, cheap, general-purpose film. The price value for Foma films is unmatched. It’s one of the cheapest films in the EU, readily available in 35mm, 120, 4×5, 9×12, and 8×10.
Foma also makes paper, chemicals, enlarger filters, etc. I consider them one of the most important manufacturers keeping analogue photography alive alongside Ilford and Adox*. They are the main companies providing all products needed for an analogue workflow at affordable prices, at least in the EU.
That’s one reason I favour Foma and Fomapan, as I want to support the companies that keep our hobby alive.

It’s also worth noting that reciprocity failure only kicks in at 1 second. Honestly, I rarely get into reciprocity failure territory. During the summertime, this rarely happens outside, even for 100-speed films. During winter, a 200—or 400-film will often be enough to keep it over 1 second.

*Kodak also provides this, but often at least double the price (in the EU). For example 50 sheets (4×5) of Tri-X costs 267,99 €, Ilford HP5 costs 129,98 €, Adox CHS 100 II costs 90€, and Fomapan 400 costs 43,50 €. (prices incl. VAT from Retrocamera.be as of 24-06-25). Bergger is another manufacturer that at least provides film and paper on the market, but I haven’t tried any of their products yet.

Fomapan 100 reciprocity failure

Foma’s datasheet doesn’t provide much information regarding reciprocity failure and the Schwarzchild -effect. All that is provided is the following information (source, available 24-06-25: https://www.foma.cz/en/fomapan-100)

Exposure (seconds)	1/1000–1/2	1	10	100
Lengthening of exposure	1x	2x	8x	16x
Correction of aperture number	0	-1	-3	-4

Schwarzchild effect for Fomapan 100, according to Foma’s datasheet

Instead, I’ve used this table that originated from a Reddit post by the user “vaughanbromfield” post-link https://www.reddit.com/r/AnalogCommunity/comments/ttem9y/fomapan_100_exposure_reciprocity_correction/ (available 24-06-25).

Worth noting is that in that same thread, another user, “FlyThink7908”, had problems when trying out this chart and had negatives coming out dense (overexposed). I, myself, haven’t done any extensive testing, but I can’t recall my long-exposures coming out overexposed.

Sidenote: This chart is calculated using the formulas mentioned in the section “Schwarzchild’s law limitations”.

Measured	Measured
1 s	2 s
–	3 s
–	5 s
2 s	7 s
–	10 s
–	15 s
4 s	21 s
–	30 s
–	42 s
8 s	59 s
–	1 min 21 s
–	1 min 51 s
15 s	2 min 32 s
–	3 min 26s
–	4 min 39 s
30 s	6 min 17 s

Fomapan 100, Reciprocity failure

NoColorStudio nbr. 12 and paper film

In the video, I made one photograph on the nbr. 12 film by NoColorStudio (source, available 24-06-25 https://www.nocolorstudio.com/shop/p/no12-baryta-sheet-film).
This is a slightly different film that uses Baryta paper as its base, which gives it a unique look. It’s also a slow film, at an ISO of 6.

In my video, I mentioned that I suspected you might not need to add any time for reciprocity failure. My reasoning for that is as follows.

I couldn’t find any information about reciprocity failure on the NoColorStudios webpage. However, there’s another manufacturer that makes paper film, and that’s Washi Film. The Washi W-film is a film with a Kozo Paper base. And in its datasheet, this is mentioned:
“the emulsion used for this film is similar with black
& white paper emulsion, there is no problem to use it for
long exposure in pinhole camera. It can also be used as
printing paper (grade 3).”
(Source, available 24-06-25 https://filmwashi.com/datasheet/W25_en.pdf)

This, to me, hints that you shouldn’t add any time for reciprocity failure.
It also states that the Washi W-film is similar to a black-and-white paper emulsion, which led me to believe it might be the same for NoColorStudio nbr. 12.
In the back of my mind, I have an old note that paper doesn’t suffer reciprocity failure to any notable degree. I haven’t found a source that backs me up, though. However, in the darkroom, it’s not uncommon to have exposure times exceeding 10s; to my recollection, paper always performs linearly.

As I couldn’t find any information about reciprocity failure on the NoColorStudio webpage, I e-mailed them and asked about it. And got the following response:
“…You are absolutely right about reciprocity failure, it is non-existent for most exposures. I had found that for exposure 10minutes or longer, perhaps a 1/2 stop can be increased for good measure…”
(Source: e-mail conversation with Arnas Spaka of NoColorStudio as of 24-06-23).

This puts it to rest for me. I won’t be adding any time to my long exposures of NoColorStudio nbr. 12 in the future, something I should probably look into before using the film. I don’t know if this holds true for all other paper films. I suspect it’s the same for at least Washi W and Ilford Direct Paper, but I haven’t looked further into it.

Developing

Another thing to consider is what developer you use and how you develop your film.
Personally, I’ve developed all my successful long exposures normally and without adjusting developing times. All have been developed using normal inversion agitation (four at a start and then two every half-minute) and normal developing times. It’s worth mentioning that I rarely do long exposures and try to avoid going over 5-10 seconds. I think the longest I’ve done is 30 seconds, and a few have been around the 15-second mark.

However, when researching this, I came across two interesting things. In a thread on Flickr, user “Spotted bit” stated the following.

“When compensating for reciprocity failure the highlights are pushed more than they need to. For example, if your meter reads 1 second, then zone V is already in the failure range and need compensation, but zones Vii or Viii need no compensation, and they will be overexposed. Decreasing the developing time controls those highlights. You can calculate how much they are “pushed” and decrease time accordingly.”
(Source, available 24-06-25, https://www.flickr.com/groups/67377471@N00/discuss/72157633721255138/)

User “Joosttermeer” states in the same thread the following
“I recall that I read somewhere that you have to lower the development time bij 10-20%, but I’m not sure.”

This statement is also backed up by DigitialTruth Photo (famous for their Massive dev Chart), which gives the following guidelines for B&W
“1 sec: expose for 2 secs, underdevelop by 10%
10 secs: expose for 1 min, underdevelop by 20%
100 secs: expose for 20 mins, underdevelop by 30%”
(source, available 25-06-25 https://www.digitaltruth.com/data/reciprocity.php)

It makes sense to me that you should underdeveloped to some extent for optimal results.
However, if I were to start doing more long exposures, I would take the exposure guidelines with a grain of salt and try to do some tests with my own equipment. I can also speculate that you probably don’t need to compensate if you use a compensating developer. For instance, I believe semi-stand development using Rodinal should achieve the same effect.

I also found a blog post by S.J. Godfrey, which states that the Dynamic Range increases as a side effect of reciprocity failure. I haven’t dug deeper into this, and he only quotes Night Photography by Lance Keimig as a source in the post. (source, available 24-05-26 https://sjgodfrey.com/blog/2011/6/9/long-exposures-film-reciprocity-failure.html)

This sounds possible to my ears and in line with the thoughts about highlights becoming overexposed and shadows still being underexposed. In other words, it would extend the dynamic field.

According to a blog post on “shootitwithfilm” by James Baturin (he’s described as an exposure film photographer), he states the following:

“While reciprocity failure always requires you to compensate by lengthening exposure time, I’ve found that compensation in development is only really necessary in images with a really wide range of tones.”
(source, available 24-06-25 https://shootitwithfilm.com/understanding-reciprocity-failure/ )

This idea would also be supported if a scene has a narrow dynamic range. Then, everything in the scene should logically experience the same degree of reciprocity failure.

End thoughts

This is where my research ends for this time. As I’m not a long-exposure fanatic, it only becomes academically interesting. In practice, I will only compensate for reciprocity using manufacturers’ guidelines as a starting point (except for the Fomapan, as mentioned previously in this blog post).

I will avoid exposure times exceeding 10s, as they need very little compensation in developing. If necessary, I’ll try semi-stand developing in Rodinal for those few occasions I exceed 10s.

It’s also worth remembering that “normal” negative films have some latitude for exposure. It’s not uncommon to read that you don’t need to compensate for development if you over or underexpose with one or two stops (depending on the film). So even if you mistimed your exposure, forgot to account for reciprocity failure, or didn’t compensate in development, the negatives might still be perfectly usable. Of course, perfect exposure is always preferred, but I won’t lose sleep if I miss exposure with a stop or two.

Methods & dissclaimer

I have very little experience with long exposures exceeding 5-10s. Therefore, I did a lot of googling. I’ve tried to give sources and proper credit. I wouldn’t consider this academically sound research; most sources are blog- or forum posts. These may or may not be reliable sources; I haven’t cross-checked the information to any greater degree than mentioned in the post. If it sounds reasonable to me, I’ve included it. But as always, don’t take my word for it; do your research, and perhaps more importantly, do your experiments with your own equipment.
See this blog post as a guideline and a starting point for further research.

Sidenote: The date format used in my writing is YY-MM-DD
Sidenote: I consider this blog post a living document. When and if I find some new and relevant information I will update this post