Testing and evaluating CatLabs "X Film 320 Pro (2022 version)

[Oren Grad]

I don't know much about how film base differs across different products. Perhaps someone can chime in and explain the differences.

Ilford 35mm film is coated on 0.125 mm / 5 mil acetate base, 120 film is coated on 0.110 mm / 4 mil acetate base, and sheet film is coated on 0.180 mm / 7 mil polyester base. Although there might be modest differences in development times to reach a given CI, I would not expect to see a material difference in fundamental sensitometric properties, including the curve shape. Indeed, although the curve provided in the FP4 Plus documentation is stated to have been based on roll (120) film, the same caption says "This curve is also representative of the 35mm and sheet film formats".

 

[markjwyatt]

Ilford 35mm film is coated on 0.125 mm / 5 mil acetate base, 120 film is coated on 0.110 mm / 4 mil acetate base, and sheet film is coated on 0.180 mm / 7 mil polyester base. Although there might be modest differences in development times to reach a given CI, I would not expect to see a material difference in fundamental sensitometric properties, including the curve shape. Indeed, although the curve provided in the FP4 Plus documentation is stated to have been based on roll (120) film, the same caption says "This curve is also representative of the 35mm and sheet film formats".

I think the biggest difference for sensitometry mis the b+f. This can vary a lot depending on the film, and the developer. Some bases are clear, some have inherent dyeing, pyro stains, etc.

 

[aparat]

I want to finish the discussion of film speed related to my test of Catlabs Pro 320.

Relative film speed

I think Phil Davis provided the most comprehensive manual on how to derive Effective Film Speed (EFS) values from a curve family. Unfortunately, the knowledge is spread over a number of somewhat obscure Beyond the Zone System (BTZS) print and electronic publications. The gist of the method is to first establish a Personal Speed Point (PSP) along the Relative Log Exposure (RLE) axis. The value of PSP needs to be determined empirically by doing a test on a film and developer combination that has been found to yield true box speed. In other words, the photographer needs to take the manufacture's box speed in good faith and use it to calibrate their process. I did this a while back (using Delta 100 and D76), and my PSP ended up being 2.36 RLE with the exposure of 3.24 Millilux Seconds. If you’d like to know the details of how I arrived at the number, please do not hesitate to ask. I have so far presented my Catlabs Pro 320 results based on that general approach.

Absolute film speed

In order to obtain a reliable measure of a film’s “absolute” speed, one needs to use a properly calibrated setup. Still, if such a setup is unavailable, I believe one can obtain approximately reasonable, “ballpark” estimates of film speed. One just needs to try to control the process as much as is reasonable in an amateur environment.

Theoretically, if one did one's calibrations correctly, the two types of analysis (relative and absolute) should produce similar EFS results, though it is probably unreasonable to expect them to be identical. The Catlabs Pro 320 EFS values (and all the other computed parameters) ended up being very similar across these two analyses. Below, I am showing the results in summary form. If you’d like a more detailed report, please let me know and I will make it available.

 

[Oren Grad]

I think the biggest difference for sensitometry mis the b+f. This can vary a lot depending on the film, and the developer. Some bases are clear, some have inherent dyeing, pyro stains, etc.

Yes, good point. The different base types and thicknesses could certainly affect the effective base density.

 

[aparat]

Ilford 35mm film is coated on 0.125 mm / 5 mil acetate base, 120 film is coated on 0.110 mm / 4 mil acetate base, and sheet film is coated on 0.180 mm / 7 mil polyester base. Although there might be modest differences in development times to reach a given CI, I would not expect to see a material difference in fundamental sensitometric properties, including the curve shape. Indeed, although the curve provided in the FP4 Plus documentation is stated to have been based on roll (120) film, the same caption says "This curve is also representative of the 35mm and sheet film formats".

This is impressive and very helpful info. Thank you!

 

[Oren Grad]

This is impressive and very helpful info. Thank you!

Many thanks to you for putting in all of the extra work to write up and share your findings here - really appreciated!

 

[aparat]

Tone reproduction

Tone reproduction analysis dates back to the 1950s and the work by Jones (and others). Unlike curve family analysis, it has not been as well documented in the literature since. I owe a lot to the Photrio (and Apug) community forum (esp. Stephen) where some of the otherwise obscure detail has been discussed in detail over the years. I still have a lot to learn, though. One of the goals of tone reproduction analysis is to simulate the entire journey of light from the scene (subject luminance), through the meter, the camera image (including flare), the negative, and the final the print. Below is a typical block diagram illustrating the idea.

In order to run this analysis, one needs to have sensitometric data of both film and paper. I did my paper analysis (Ilford MGIV FB Glossy) four years ago, so it probably is no longer relevant, as the emulsion seems to have changed (please, correct me if I am wrong on this), not to mention the fact that the choice of paper analysis type adds yet another layer of variation. I test paper according to the method described by Darkroom Automation, but it's just one of several paper testing methods in use today.

Tone reproduction analysis should be able to incorporate a few options available to the photographer, such as, whether to focus on shadow, mid-tone, or highlight information, how to use the exposure meter, how to choose one’s preferred print tonality, etc. I am showing just one set of such options here, mostly for the sake of brevity. Regardless, I think we can see one important detail, i.e., how one’s choice of film, exposure, and development, leads to a certain kind of tone distortion, esp. shadow and highlight compression.

Since I haven’t yet talked about camera flare, I thought I’d illustrate its effect in a series of three flare values (0, 0.5, and 1.3 stops). Flare, among other things, reduces overall subject luminance range to an effective in-camera luminance range (see Quadrant 1).


The last image shows how increased development time leads to an even more extreme tone distortion in the final print. It would be nice, I think, to do a similar type of analysis in a hybrid process. Something to think about.

 

[Romanko]

Since I haven’t yet talked about camera flare

I would like to learn more about camera flare, especially how it is modelled in sensitometric tests. Could you please point me to books and other resources on the subject? This information is not easy to find on the net.

It would be nice, I think, to do a similar type of analysis in a hybrid process.

This is a great idea! I am experimenting with reading and processing camera RAW files (programmatically using Python) and it turns out easier than I initially thought. Let me know if I could be of any help here.

 

[aparat]

I would like to learn more about camera flare, especially how it is modelled in sensitometric tests. Could you please point me to books and other resources on the subject? This information is not easy to find on the net.

I think that the section on flare in Phil Davis' Beyond the Zone System is quite good. It's fairly comprehensive and includes a simple method to get a rough measurement of flare for any given lens, without specialist instrumentation. I think flare is implemented in Win Plotter in the standard way, i.e., adding a fixed amount of non-image forming light to the model, though Davis talks about flare as a "flare density" rather than "flare factor," but it amounts to the same thing. Flare also affects print exposure, though in reverse, so to speak. I have read differing opinions regarding flare in testing photosensitive materials. Some say it's crucial to include it, others say that modern lenses are sufficiently immune to it. If you don't have the book, I'd be happy to scan the 3 pages on flare and send them to you.

This is a great idea! I am experimenting with reading and processing camera RAW files (programmatically using Python) and it turns out easier than I initially thought. Let me know if I could be of any help here.

That would be fantastic! I suppose most people who digitize film end up with RAW files, be it from digital cameras or scanners, so having a way to relate film sensitometry data to RAW file data would be great, esp. if it could help photographers produce negatives optimal for scanning. I remember struggling with a Nikon CoolScan 9000 ED years ago trying to produce the best negatives for scanning. It was a lot of work, but the results were very nice. I do regret selling the the scanner.

 

[Romanko]

Phil Davis' Beyond the Zone System is quite good

Thank you for the reference. I heard about this book but haven't read it yet. It should not be too hard to find even in Down Under.

having a way to relate film sensitometry data to RAW file data would be great

Most RAW processing software like RawTherapee and Darktable apply white balance, exposure and possibly other corrections when reading an image from RAW. I could not find a way do disable this processing so I started looking at libraries that would allow me more control. LibRaw seems to work for me. It is available in multiple programming languages including Python.

I don't have a densitometer so I experimented with estimating density of the negative from a DSLR capture. The measurements are relative, and you are only relying on the linearity of the camera sensor response. A histogram of the step wedge image should give an accurate way of calculating the characteristic curve.

I have some 70mm Technical Pan that I plan to use in my very old Voigtländer 116-type folder. Hence the interest in this thread and in particular in your testing techniques. My methods are very crude compared to yours, but they suffice for what I am doing.

 

[aparat]

Thank you for the reference. I heard about this book but haven't read it yet. It should not be too hard to find even in Down Under.



Most RAW processing software like RawTherapee and Darktable apply white balance, exposure and possibly other corrections when reading an image from RAW. I could not find a way do disable this processing so I started looking at libraries that would allow me more control. LibRaw seems to work for me. It is available in multiple programming languages including Python.

I don't have a densitometer so I experimented with estimating density of the negative from a DSLR capture. The measurements are relative, and you are only relying on the linearity of the camera sensor response. A histogram of the step wedge image should give an accurate way of calculating the characteristic curve.

I have some 70mm Technical Pan that I plan to use in my very old Voigtländer 116-type folder. Hence the interest in this thread and in particular in your testing techniques. My methods are very crude compared to yours, but they suffice for what I am doing.

I agree that you should get decently accurate densities from a linear scan of a step tablet. It would actually be a very useful technique for all of us, as densitometers are getting harder and harder to come by and service.

 

[aparat]

More on flare.

I was thinking about flare, prompted by @Romanko, and thought I'd include a simple flare demo based on the Catlabs Pro 320 data. There's an argument that flare should not be part of curve family analysis but should, instead, be included in tone reproduction analysis (Quadrant 1 in post #232) . I think it's a very good argument, and I don't mean to debate it. Still, I think it could be useful to simulate flare in a film curve plot, just to see how it would influence the negative. One can expect a slight lowering of CI in general, and a somewhat more significant lowering of shadow contrast, in particular. I have come across flare being modeled as a "flare factor" and as a "flare density," but the end result should be very similar. As my program is still not nearly ready for release, I thought I'd include a comparison with Win Plotter, if any of you want to try it out for yourselves. Win Plotter is, in my opinion, a very good program. It could be very useful even if one is not necessarily sold on the BTZS framework of which it is a companion app. I have an old Windows 10 laptop, and it runs very nicely on it.

I am attaching a PDF below. Pages 1-2 show the curve without flare, the 3-4 with flare of 0.02 density.

 

[Romanko]

As my program is still not nearly ready for release, I thought I'd include a comparison with Win Plotter

I am not familiar with Win Plotter. Which plots are generated by your software (green curves or black)? To me both look very good. I am not a big fan of putting a lot of stuff in the legend and prefer to use more of the plot real estate for the actual curve but these are just my personal preferences.

Pages 1-2 show the curve without flare, the 3-4 with flare of 0.02 density

Very interesting data. It really emphasises the importance of the whole system calibration including the lens/camera/film/developer/enlarger/paper/developer (or digitising/post-processing).

This film might give decent results with ancient uncoated triplets that give plenty of flare. I expect pre-flashing would work similarly.

 

[aparat]

This film might give decent results with ancient uncoated triplets that give plenty of flare. I expect pre-flashing would work similarly.

Yeah, it does look like that would tame some of the contrast. Win Plotter is pretty old, but works well. It's Windows only and cost around $30. Mine are the ones with blue and red dotted lines, pages 2 and 4. They do look kind of busy, with all the data crammed in. I'm still trying to find the most informative way to display some of the details.

 

[Oldwino]

I finally got around to running a roll of this stuff through Diafine. And it looks like the contrast is much more “normal“ as compared to my Rodinal results. Shot at EI200

Negs drying, will scan tomorrow for a closer look.

 

[CatLABS]

I finally got around to running a roll of this stuff through Diafine. And it looks like the contrast is much more “normal“ as compared to my Rodinal results. Shot at EI200

Negs drying, will scan tomorrow for a closer look.

cant wait!

 

[Oldwino]

Here's a few samples shot at ~200, during a break in my jury duty Tuesday, developed in Diafine. Leica IIIa & Voigtlander 28f3.5. Shadow detail is pretty good, yet the film retains a unique, "edgy" look to it. Negatives were straightforward to scan, and they look like they'd print well, too.

 

[MattKing]

Interesting.
I wonder if this is a film that enjoys a speed boost in Diafine.
In my experience, many current films don't get much of a speed boost from Diafine, even if their older incarnations did (e.g. Tri-X).
Diafine is, of course, a fairly low contrast developer. So it may be perfectly suited to this film.

 

[aparat]

Here's a few samples shot at ~200, during a break in my jury duty Tuesday, developed in Diafine. Leica IIIa & Voigtlander 28f3.5. Shadow detail is pretty good, yet the film retains a unique, "edgy" look to it. Negatives were straightforward to scan, and they look like they'd print well, too.

These are fantastic! Thank you for sharing. I am particularly intrigued by your choice of developer. How do you think it differs from a more conventional developer, such as D76, particularly in terms of shadow contrast? Also, is the Voigtlander a vintage or modern lens? I curious as to its flare performance. Sorry, one last question. In picture #2, do you happen to remember if the sky was clear-blue or was it a bit "hazy" or partly cloudy?

 

[Huss]

Here's a few samples shot at ~200, during a break in my jury duty Tuesday, developed in Diafine. Leica IIIa & Voigtlander 28f3.5. Shadow detail is pretty good, yet the film retains a unique, "edgy" look to it. Negatives were straightforward to scan, and they look like they'd print well, too.

View attachment 320145 View attachment 320146 View attachment 320147 View attachment 320148

Those look great!

 

[Oldwino]

These are fantastic! Thank you for sharing. I am particularly intrigued by your choice of developer. How do you think it differs from a more conventional developer, such as D76, particularly in terms of shadow contrast? Also, is the Voigtlander a vintage or modern lens? I curious as to its flare performance. Sorry, one last question. In picture #2, do you happen to remember if the sky was clear-blue or was it a bit "hazy" or partly cloudy?

Diafine is a compensating developer, which I think helps tame the highlights of this film. It also tends to be lower in contrast, which also is a plus for the CatLabs. It still has a higher contrast look to it, which I like, but I think the shadow detail is pretty good in these.
The Voigtlander is the Color-Skopar 28mm f3.5 LTM version, so it is "modern", but almost vintage at this point. It's a nice sharp lens. It is pretty flare resistant.
Regarding the sky - it seems like the CatLabs likes to make darker blue skies. It was a tad hazy that day, with some sparse high clouds. The film seems to render blues fairly darkly. I didn't use any filters on any of these.

I should also add that I shot these at 160-200 (I didn't meter for every shot), and I think the film got the Diafine speed boost effect, particularly in light of your tests which seem to indicate a ISO rating of around 80 (is that right?).

 

[Huss]

Ach, roll #8 in still fresh DF96 shows severe bromide drag w CatLabs 320 Pro... No issues with other film at this point. Using my same agitation techniques as with other films that do not show bromide drag.
Not a dig on CatLabs, I guess DF96 is not a decent match for it. So now that is three films for me that don't work well with DF96 - Catlabs 320, Silberra Orta, Adox CMS 20.

 

[MattKing]

Ach, roll #8 in still fresh DF96 shows severe bromide drag w CatLabs 320 Pro... No issues with other film at this point. Using my same agitation techniques as with other films that do not show bromide drag.
Not a dig on CatLabs, I guess DF96 is not a decent match for it. So now that is three films for me that don't work well with DF96 - Catlabs 320, Silberra Orta, Adox CMS 20.

Is that in 35mm, or 120?

 

[Huss]

Is that in 35mm, or 120?

35mm. I'm about to develop a roll of Acros II 35mm in the same DF96 batch - will run a test strip first - to check to make sure that it is not the DF96 that is at fault. But in my extensive use with it, it has lasted at least 16 rolls.

 

[Dwight Anderson]

Just a teaser for tomorrow, Catlabs 120 shot at 100 and developed for 20% less time. Not the best composition, but I'm quite happy with the results.