FP4+ ISO 4000

Got curious about something, I bulk loaded some Expired FP4+, which normally has a dMin of 0.4 in Xtol Replenished for 7 minutes at 24 degrees celsius.

Bracketed it shot from 125 to 4000 in 1 stop increments, loaded it in E-6 (standard time), stopped it at reversal bath and then fixed it.


The dMin/fog level is 1.26, quite high, however, even the 4000 shot had reasonable density.

Living in Grain City


This would be interesting to try with some fresh film.. say Tri-X or Delta 400

Perfecting Adox CMS 20 Part I



Decided to revisit Adox CMS 20, and get the best out of it I possibly can, also completely out of Adotech II developer, so comparisons will come to that whenever I can get more.


First up, I’ll start up straight with the third attempt as it’s been the most satisfactory so far, as in quite useable.


The third attempt was developed in a split-bath I formulated as follows:


Bath A (5 minutes)

The asorbic acid and sodium carbonate were mixed together before adding, to neutralise the acidity.

Ingredient Amount
Sodium Sulphite 30g/L
Sodium Metabisulphite 10g/L
Metol 12g/L
Asorbic Acid 3g/L
Sodium Carbonate 1.87g/L
Potassium Bromide 50mg/L
Potassium Iodide 2.5mg/L

Interactive Whiteboards by PolyVision

Bath B (4 minutes)

Ingredient Amount
Borax 30g/L
Potassium Iodide 2.5mg/L

dMin was 0.06, dMax was 2.37

Rated speed was ISO 20.

Zone V is middle grey camera metering (Canon AE-1), and measured 1.17

Here is the response curve, I think I screwed up between Zone X and XI and there should be another 2 stops there.


Correcting for that, it shows a range of 12 stops each with decent separation, there’s another 2 stops at the end of the chart but little separation.


and ISO 20 looks like a perfectly good exposure, the catch is however, exposure above Zone V (btw the numbering is messed up, “1” is Zone 0, so 6 is “Zone V”, I’ll have to redo the chart later) starts to flatten out and has low (in this context) contrast after Zone VII, while exposure below Zone V has massive contrast. Think of it as a contrast mis match, it becomes very difficult to display both good midtones, shadows and highlights at the same time.

I think 5 minutes is too long in Bath A, I want to reduce the Asorbic Acid and go back to 3 minutes.


We need to reduce the contrast of the steep part, make it less upswept, and/or increase highlight contrast, basically get a straighter line. In reality I would like to get a relatively smooth straight line as possible for macro/global contrast of large areas, but have massive local contrast in smaller areas.



Scan made with Flextight 949 (8000 dpi)



Konica Centuria Chrome 100 and E-6 modification

Got a few rolls of expired Konica Centuria Chrome 100, the base is fairly fogged, with the dMax reading approx 1.29, using a custom first developer, the dMax improved to approx 1.58 (about 1 stop of more density) while retaining clear highlights, I would like to improve the dMax to about 2.4+

I think I could do that with increased Potassium Bromide, and increasing the vitamin C to phenidone ratio to increase the contrast of the developer without having to adjust the time, otherwise, more potassium bromide, a little less solvent (potassium thiocyanate) and longer time.

Unmodified raw scan from Flextight 949 (8000 dpi crop), shot at ISO 32, think it was f/2.8


Some adjustments with levels


The formula is as follows:

Process at 38 degrees celsius for 6 minutes, followed by acidic stop bath, then process from there as normal E-6 (start in reversal bath or light fog and start in colour developer).

10g/L of Sodium Sulphite

14g/L of Potassium Carbonate

6/gL of Sodium Bicarbonate

0.3g/L of Phenidone

15g/L of Ascorbic Acid (pH neutralised with Sodium Bicarbonate in a separate solution)

2.2g/L of Potassium Bromide

0.8g/L of Potassium Thiocyanate

4/gL of Potassium Hydroxide

5mg/L of Potassium Iodide (added from a solution)

C-41 modifications, sharpness and playing around with Lucky Color 200

I decided to play around with some Lucky Color 200 for a few reasons.

I wanted something cheap and somewhat grainy to test C-41 developer modifications with, in order to see how much one can influence characteristics through development changes.

The first roll I shot normally, I achieved about 50 lp/mm from 80:1 to 4:1 contrast, though that may need re-testing to confirm it is that low normally at the high contrast side. The other thing that cropped up was that it appeared faster than it’s 200 rated speed. Probably a 400 film.

Lucky 200 @ 1600 (no push)

Lucky 200 @ 1600

1600 (no push)


3200 (no push)


The first developer modification didn’t turn out great, the second one was a lot better, it involved adding 7.2g/L of Potassium Bromide to normal C-41 developer and processing at a time of 4 minutes 15 seconds, which is normally a 2 stop push.

In this case, I want to lose speed while increasing processing time. Less development with more exposure to improve grain, but with a longer time with restrainer instead of a simple pull process in order to increase contrast rather than decrease it, hopefully in order to pull a bit more detail than normal. The is idea is to suppress the larger grains of the emulsion from contributing to the image.

In this case I got 70 lp/mm @ 80:1 contrast, 62 lp/mm @ 4:1 contrast, and 45 lp/mm @ 1.6:1 contrast at an exposed ISO of 50 and 100, my chart went down to about 15 lp/mm at the distance I used it at, and could not resolve any of the 1.23:1 contrast patterns.

Now, I also got about 90-100 lp/mm @ 80:1 and 4:1 contrast with this same lens at the same aperture (f/5.6) on Agfa Precisa CT 100 (the stuff produced for Agfa by Fuji), I’ve seen it elsewhere stated to get 120-135 lp/mm at 4:1 contrast.

At f/5.6, I managed about 150 lp/mm at 80:1 out of Adox CMS 20. Excessively sharp on CMS 20, but my lens (Canon FD 28mm f/2.8) isn’t to the same standard as used for those tests that got much higher detail on Precisa.

So my results have greater loss then theirs does. Where as due to CMS’s extreme resolution and being very thin, there’ll be minimal loss.

Now considering, Precisa is allegedly Sensia, or at least close to it, Fuji rates Sensia at 55 lp/mm @ 1.6:1 and 135 lp/mm @ 1000:1, although it’s stated these are often conservative, as excellent lenses show what they can put on these films at low contrast.

So let’s assume the peak is 135 lp/mm for Precisa (Even though their lens should also have loss on it), and 150 lp/mm peak at f/5.6 for my lens, applying the resolution formula, where lens is the lens resolution, film is the film resolution, and r is the resolution of what gets put on the film:

1/lens^2 + 1/film^2 = 1/r^2

Then, using this, 1/150^2 + 1/140^2 = ~9.9314×10^-5

Or r = 100.3 lp/mm, which matches my results.

Now if I use this to work out the peak film resolution in the developer modification, what I -might- get with a much more excellent lens

1/r^2 – 1/lens^2 = 1/film^2

1/70^2 – 1/150^2 = 1.5964×10^-4

Or about 80 lp/mm, which compared to 50 lp/mm (which only goes up to 53 lp/mm with this formula) is a significant improvement. It is a 51% increase from 53 lp/mm.

Showing this technique does have some merit.

Now I suspect this film is actually a 400 speed film repackaged as 200. So if I were to use this on Ektar, I would probably rate it at 25. Though first roll.. bracket exposure tests to find an ideal exposure.

I think this can be further improved even more by tweaking, greater amount of Potassium Bromide for example. A small amount of Potassium Thiocyanate to give finer grain (Though not needed on Ektar, but would be good to experiment).

The application of this would to get essentially, a much slower speed Ektar (or other 100 speed colour negative) without having to overexpose a few stops and compromise contrast, saturation and resolution, but actually give a resolution benefit.

This same source holds that Ektar gets a 105 lp/mm peak, this would put it at almost 160 lp/mm if that 51% improvement figure would hold (and that 50 lp/mm was correct for normal processing of Lucky 200). With my lens that would allow me to hit that figure they achieved. O something like 150 lp/mm with something like the Biogon 25.

Now if we still had Ektar 25, which was rated at 200 lp/mm, that would be an excellent landscape colour negative film at the right speed without resorting to funny processing.

I should mention, that these figures are barely-resolved, they’re there, they are just barely visible, soft, but there. The higher you can push this resolution, the better the image will be at a lower resolution, ie: being able to resolve 120 lp/mm will give superior spatial frequencies at 80 lp/mm then something only just hitting 80 lp/mm @ peak. They will be cleaner, clearer, and more well defined. And 80 lp/mm is excellent definition when it is very well defined, rather than only just barely visible.

One of the sharpest colour negatives I have personally used has been Konica Centuria Super 100 in 120, although I shot it at 50 and pulled it 1 stop. It was ridiculously good for nature images, not so good for portraits (just looked odd).

25/06/2013 Update

I managed to get 77 lp/mm out of Lucky 200 @ 50 and 100 this time round.

Using the same process time of 4m 15s, but with 10g/L of KBr added and 300mg/L of Potassium Thiocyanate added.

This pushes the film resolution to 90 lp/mm assuming 150 lp/mm is the peak resolution of my lens.

Even with the large potassium bromide increase, 50 and 100 still looked like very good speeds, this would be attributable to the speed increasing effect of potassium thiocyanate that would help uncover more developing sites as it’s a powerful solvent.

Perhaps there is again a little more wiggle room with more KBr and a longer time again.

But this now represents a +80% increase. If this works just as well on Ektar, that’d shift it up to 190 lp/mm.

I am guessing at this point, with this amount of additives and processing time, a good exposure index for Ektar would be 25.

Kodak D-76 vs Homebrew D-76 differences

Quick test results, both were processed at the same time (5 minutes), at the same dilution (stock), at the same temperature (22.5 degrees celsius – the ambient temperature to make consistency in comparison easier)

The film is Kodak T-Max 100, several exposures were made, the film was cut in half, one half processed in Kodak D-76, and the other half processed in the follow D-76 Recipe

Metol 2g/L
Sodium Sulphite 100g/L
Sodium Tetraborate 2g/L
Hydroquinone 5/gL

Here is a crop and then digital blow-up (to help spot the difference) between the two. Please click to enlarge them to view the full size.

The results are that, with this recipe, the grain is a little softer, and image detail is a little less sharp. But the negatives appear identical in terms of both overall contrast and tonality and density. They were treated identically, no sharpening applied, these are output from a raw scan off an Imacon 949, crop and levels (identical in both) applied.

One could dilute this developer for sharper grain and detail.

Kodak D-76
Kodak D76

D-76 Formula
D76 Formula

Adox CMS 20 Experiments

First Attempt

So I was curious about exactly what I could force-fit into Adox CMS 20, as it seems to be a particularly difficult film to deal with.

One of the problems is that any overexposure, or highlights are lost. What I notice is Adotech II has a strong solvent built into it. So I would strive to minimise all solvents in any custom developer for CMS 20, and maximise available dMax for more highlight room.

A quick test of the film before processing, shows it is already very thin to start with compared to others, the density pre-processing measures 0.9, 1.22, 1.54. Developing a cut piece in some Agfa Neutol print developer in room light, washing, drying, I get a reading of 2.30, 2.25, 2.44.

So theoretically, you can still get a nice high dMax to fit a lot of tonal range in. So if we aim to have a limit of 2.0 controlled by available developer in the split-bath, it just might work.

So my idea came basically down to a split-bath developer optimised for CMS 20. I settled on either Metol or Phenidone as sole developing agents. As on their own they are both very low contrast and soft working. With phenidone being supercompensating, as it prevents too much density build up by it’s secondary products inhibiting development, but also produces some fog, so getting reasonable dMax and dMin separation can be tricky. That might just be the ticket though.

Though there may be no reason at all we cant use some hyrdoquinone or ascorbic acid as well.

My first recipe consisted of:

Bath A
10g/L Sodium Sulphite
3.5g/L Sodium Metabisulphite
0.24g/L Phenidone
pH ~6.7-6.8
3 minutes development

Bath B
30g/L Sodium Carbonate
2g/L Potassium Bromide
pH ~9.4
6 minutes development

My results from ISO 100 to 3, all had very low dMax, so this is not suitable. The highest dMax I can find in the images on the film is about 0.1, the images are visible (because the base is essentially clear).

My next test, I would significantly increase the phenidone amount, and half the potassium bromide in Bath B.

I would like to get to a normal looking working bath, then try potassium hydroxide in Bath B, for a high pH, and see if there is an increase in speed and/or local contrast over carbonates and bicarbonates and meta or tetraborate etc.

Second Attempt

For the second attempt, I will try approximately the following:

Bath A
10g/L Sodium Sulphite
3.5g/L Sodium Metabisulphite
1.0g/L Phenidone

Bath B
30g/L Sodium Carbonate
1g/L Potassium Bromide

New Split Bath Colour Negative Developer Part 1

First time playing around with colour split bath in a long time, since people are talking about it on APUG again.

I’ve played around with them before, hit and miss results. This isn’t the Dignan NCF-41 formula either, this is something I’ve arrived at after a lot of pondering over previous results.

Bath A preparation
Pinch of Sodium Sulphite
CD-2 – 9g/L
Sodium Sulphite – 9.6g/L
Sodium Bisulphite – 0.54g/L

Measured pH 7.1

Bath B preparation
Potassium Carbonate – 50g/L
Potassium Iodide – 1.4mg/L
Sodium Thiocyanate – 220mg/L (I’m sure my thiocyanate needs drying out too btw).

Measured pH 11.1

Developing time:
A – 3m 15s
B – 6m

Test roll was Agfaphoto Vista Plus 400

It came out rather well.

Here are the first samples, this shot is at ISO 400.

Scans were made on the Imacon 949 at 8000 dpi and down sized to 4000 dpi.

(click images for full view)

Full Frame

Detail Crop (4000 dpi)

Out of Focus Grain Quality Crop (4000 dpi)

The thing that grabs me is that the detail and grain are both very good for ISO 400 colour neg. It is very much improved over a roll from the same batch in standard C-41.

What I really like is how the grain looks – the grain is there, but it is smooth and continuous. Not sparse and empty as I normally see, and it’s not colour speckles either. I really like this sort of grain.

More to come.

More samples from this roll

I like the way the exposure of the sky is balanced with the foreground exposure, rather than separated, the shaded area where the car is is very very grainy. This was shot at Sunny 16 for ISO 400.


Testing Colour Quality of Film Scans


My program for looking at colour fidelity in film scans (mainly flextight scans) finally is starting to work, all functions in 16-bit precision seem working so far with LibTiff (The VS Studio .NET built in way of handling 16-bit images only returns 13-bits! Bad!).

However it can open any 16-bit TIFF file (though only up to a certain size, larger ones like 500mb/80mp scans get out of memory errors, will have to work arround, though probably don’t need to analyse that high anyway).

What I call “unique colour factor” is the percentage out of the maximum number of possible unique colours allowable in that image.

IE: A 12mp image cannot have more than 12 million unique colours as there is only 12 million pixels.

(Hue is above 16-bits as I decided to use a scale of 360*360)

Please note the below results aren’t the same image on each camera, just something to play with at the time being, however I do not think that a dSLR will be able to check all the boxes for colour fidelity at the same time in the same image.

Results remain to be seen.

Random Nikon dSLR image analysis:

Random Velvia 35mm on Flextight analysis:

Quick Scanner Comparison

Threw this together, it’s some Velvia 50 (lightly expired, shot at 25 and pulled 1 stop).

Epson 4990 vs Plustek 8200i vs Flextight 949, and for kicks the Fujifilm SP 3000 minilab scanner (Free Export BMP, 12×18).

The Plustek 8200i is rated at 3250dpi by filmscanner.info, and the 8100 (And 7400) is rated at 3800dpi, but doesn’t have IR dust removal (for those that want that feature). Looking forward to the Plustek 120 since a Plustek rep said it’ll be better than any of their 35mm scanners (allegedly, so 35mm will scan better on the 120), and supposed to surpass the CoolScan 9000 in terms of IQ.

5219 500T Split Development #1

Fiddling around with split development.

For this recipe I used:

Bath A:

100mL/L of Fuji N1-CR (or simply double the amount of part C as you would use per litre of C-41 developer, or essentially you can also use about 11g/L CD-4)

pH was about 2.5 at this point, add sodium sulphite until pH rises to 6.5, then added potassium carbonate till it was a touch over 7.

Then added white vinegar until pH dropped to 6.8 (wasn’t much)

Added 120mg/L of phenidone

Water up to 1L

Development Time: 4.5 minutes

Bath B:

14g/L Potassium Carbonate

15.3g/L Sodium Tetraborate

590mg/L Potassium Bromide

490mg/L Sodium Thiocyanate (50%)

10mL/L Hydrogen Peroxide 3% (Added this after initial 30s of agitation with bath B as I forgot to put it in beforehand)

Development Time: 6 minutes

Rinse, stop.

Bleach, fix, wash, dry, etc.

Below are some flextight 949 scans.

The dMax is way too low on these results (including edge code), I should have raised the pH more initially with the carbonate to say 8 then dropped it back down to under 7 (the CD-4 has H2SO4 attached which needs neutralisation).

ISO 3200 (no filter)