Revisiting Tri-X 400 After Three Decades

[chuckroast]

Exactly - the ISO standard reflects print quality considerations more than classic Zone System or other speed related determinations reflect print quality considerations.
A Zone System approach will generally give you more detail in the shadows, at the expense of highlight rendition.

I would suggest that's only true if you limit yourself exclusively to determining proper personal EI. But ZS in total asks you to figure out the development time that will hold highlights in bounds. For extreme cases where severe N- development would be incurred, there are techniques like divided development, compensating developers, SLIMT, EMA, semistand, etc. all of which can be put in service of ZS exposure management.

As I said upthread, this has a whole bunch to do with how each of us imaging "Zone III" should look and what you should see in "Zone VII".

And in most cases, people respond more to the highlight rendition in a print than they do to what can be found in the shadows.

Maybe. My impression is that people respond most strongly to the local contrast of the midtones assuming there is significant image real estate with midtone content.

Unless they are in the relative small subset of people who are Zone System devotees, who seem to gravitate more to the dark.
If you are unsure if someone is a Zone System devotee, it is often a good indicator if they have a beard and wear wide brimmed hats ......

If you occasionally wear wide brimmed hats, but never wear a beard, are you only half in the zone? Asking for a friend.

 

[Milpool]

Maybe. My impression is that people respond most strongly to the local contrast of the midtones assuming there is significant image real estate with midtone content.

Here's the official answer - just kidding - actually I don't remember exactly how I did this - I think it was based on some of Jack Holm's work.

If you occasionally wear wide brimmed hats, but never wear a beard, are you only half in the zone? Asking for a friend.

There's also the unsettled matter of how to express fractions of Zones in Roman Numerals

 

[chuckroast]

Here's the official answer - just kidding - actually I don't remember exactly how I did this - I think it was based on some of Jack Holm's work.

View attachment 371532



There's also the unsettled matter of how to express fractions of Zones in Roman Numerals

Oh, that's easy. Convert to Base 16 and use IEEE floating point arithmetic...

 

[Ian Grant]

The story goes that the earlier T-Max films did not reach film emulsion speed in this developer. It wasn’t an issue with D-76. Apparently D-76 was the release developer for the T-Max films.

The issue was Tmax films did not reach the box speeds in the ASA specified developer and methodology, or in D76 where results were not as good as the film's capabilities. Ron Mowrey (PE) said (on this Forum) that Kodak research used an unpublished Ascorbic based house developer and Tmax films were at their best in this.

So Kodak had a three-pronged approach, with no US competitor to object they had the ASA testing method changed, secondly they released Tmax developer which gave good shadow detail at box speed, and thirdly promised a new developer to give optimal results with these films.

The issue was a Swedish company's US Patent for Ascorbic film developers, which was quite broad, when that lapsed Kodak released Xtol which is their best developer for Tmax films. That's shown in the Kodak developer comparison chart.

But this thread is about Tri-X. In the video Andrew mentions improvements, a Kodak Tri-X datasheet dated 2004 claims significant improvements, many attribute this to 2007 but that's probably down to stock rotation. All Kodak films had significant hardening improvements just over 10 years ago, Tri-X was also improved again, many complaining it is now too like Tmax 400.

One point overlooked is the big price disparity between Tri-X & HP5 a box of 10 sheets of Tri-X is more than 25 sheets of HP5. A 50 sheet box of Tri-X is

I used HP3 (35mm), and it was OK, HP4 was awful, then HP5 which was way better than Tri-X (talking 1970s), but I mainly used HP5 for push processing photographing rock bands. I quickly switch to XP1 when that was release, then XP2, they push [process without the big contrast build up of conventional films.

It was around 30 years before I used HP5 again, it was the only good 400 ISO sheet film I could but in Turkey, and I needed the speed for hand held 5x4 work. Processed in Pyrocat HD the results are superb.

Ian

 

[snusmumriken]

These discussions about film speed puzzle me. I don't see how anyone can say whether a film does or doesn't live up to manufacturer's box speed in their own use, unless they also state how they typically meter the scene (as Andrew does). It also seems likely that the quality of your lens, and whether or not you use a lens hood, are going to affect the apparent emptiness of the shadows.

 

[chuckroast]

The difference between the ISO criteria and the Zone System is not the 0.1 above B+F. That density is part of the ISO criteria. The difference is where that exposure is relative to the metered exposure. The ISO measurement is essentially a special case of Delta-X which is a method for estimating the fractional gradient speed, and the fractional gradient was a relationship between the gradient at the shadow exposure and the overall gradient, with this relationship having been based on print judgement speeds. I

Do you happen to have a pointer to the standard? I'd like to poke about a bit in the actual math.

 

[Milpool]

The issue was Tmax films did not reach the box speeds in the ASA specified developer and methodology, or in D76 where results were not as good as the film's capabilities. Ron Mowrey (PE) said (on this Forum) that Kodak research used an unpublished Ascorbic based house developer and Tmax films were at their best in this.

So Kodak had a three-pronged approach, with no US competitor to object they had the ASA testing method changed, secondly they released Tmax developer which gave good shadow detail at box speed, and thirdly promised a new developer to give optimal results with these films.

The issue was a Swedish company's US Patent for Ascorbic film developers, which was quite broad, when that lapsed Kodak released Xtol which is their best developer for Tmax films. That's shown in the Kodak developer comparison chart.

But this thread is about Tri-X. In the video Andrew mentions improvements, a Kodak Tri-X datasheet dated 2004 claims significant improvements, many attribute this to 2007 but that's probably down to stock rotation. All Kodak films had significant hardening improvements just over 10 years ago, Tri-X was also improved again, many complaining it is now too like Tmax 400.

One point overlooked is the big price disparity between Tri-X & HP5 a box of 10 sheets of Tri-X is more than 25 sheets of HP5. A 50 sheet box of Tri-X is

I used HP3 (35mm), and it was OK, HP4 was awful, then HP5 which was way better than Tri-X (talking 1970s), but I mainly used HP5 for push processing photographing rock bands. I quickly switch to XP1 when that was release, then XP2, they push [process without the big contrast build up of conventional films.

It was around 30 years before I used HP5 again, it was the only good 400 ISO sheet film I could but in Turkey, and I needed the speed for hand held 5x4 work. Processed in Pyrocat HD the results are superb.

Ian

Yes the infamous Swedish patent.

As I recall it was specifically the ASA developer the early T-Max films didn't like, not D-76. According to John Sexton who was consulting for Kodak and testing the T-Max films at that time, D-76 was the release developer and the films worked excellently in it. Based on discussions I had with Ron on another forum years ago the internal Kodak EAA developers (at least one of which Ron disclosed) were non-solvent "surface" developers for test purposes. It is unlikely emulsions such as T-Max would have ever performed their best in a no-sulfite developer.

 

[Milpool]

Do you happen to have a pointer to the standard? I'd like to poke about a bit in the actual math.

I'm not sure if I'm allowed to link it but you can find it by googling ISO 6 1993. Wikipedia probably also has an article showing the ISO "triangle" if you look up film speed or emulsion speed or something. However the standard itself won't help you a whole lot with the math, Delta-X, fractional gradient etc. history and the relationship to Zone System tests.

There are some publications you can read which will help. One is the original paper by Nelson/Simonds which led to Delta-X. The paper was called Simple Methods for Approximating the Fractional Gradient Speeds of Photographic Materials. Another one worth reading is the 1960 Nelson paper Safety Factors in Camera Exposures which led to the doubling of speeds.

You could also seek out summaries written by Stephen Benskin - who at one time was a member here and on LFPF. Maybe he's still here. He has written extensively about exposure theory and tone reproduction, the history of it from the original L. A. Jones "first excellent print" studies that led to fractional gradient speeds, then Delta-X, ISO etc.

 

[Sirius Glass]

I have been using Tri-X at box speed since 1960. I never left it. I started using HP5+ around 2010 when I started shooting 4"x5" again and there was not Tri-X then.
I still use Tri-X and use HP5+ when using 4"x5" or on occasion in other formats. My developer preference is replenished XTOL or PyroCat.

 

[chuckroast]

I'm not sure if I'm allowed to link it but you can find it by googling ISO 6 1993. Wikipedia probably also has an article showing the ISO "triangle" if you look up film speed or emulsion speed or something. However the standard itself won't help you a whole lot with the math, Delta-X, fractional gradient etc. history and the relationship to Zone System tests.

There are some publications you can read which will help. One is the original paper by Nelson/Simonds which led to Delta-X. The paper was called Simple Methods for Approximating the Fractional Gradient Speeds of Photographic Materials.

Found here:

http://b-wtechnik.pp7.uk/downloads/articles/Simple%20Methods%20for%20Approximating%20the%20Fractional%20Gradient%20Speeds.pdf

Another one worth reading is the 1960 Nelson paper Safety Factors in Camera Exposures which led to the doubling of speeds.\

Found here:

https://www.photrio.com/forum/attachments/safety-factors-in-camera-exposure-pdf.40145/

You could also seek out summaries written by Stephen Benskin - who at one time was a member here and on LFPF. Maybe he's still here. He has written extensively about exposure theory and tone reproduction, the history of it from the original L. A. Jones "first excellent print" studies that led to fractional gradient speeds, then Delta-X, ISO etc.

Thanks!

 

[Ian Grant]

Yes the infamous Swedish patent.

As I recall it was specifically the ASA developer the early T-Max films didn't like, not D-76. According to John Sexton who was consulting for Kodak and testing the T-Max films at that time, D-76 was the release developer and the films worked excellently in it. Based on discussions I had with Ron on another forum years ago the internal Kodak EAA developers (at least one of which Ron disclosed) were non-solvent "surface" developers for test purposes. It is unlikely emulsions such as T-Max would have ever performed their best in a no-sulfite developer.

I have John Sexton's article pre Tmax release, it was Kodak themselves who stated D76 was not the best developer for the films. It does not unlock the optimal sharpness and gradation, and film speed.

Ian

 

[Keith Tapscott.]

I know Ilford quote ID11 as their standard developer for film speeds but what exactly is the ASA (ISO) standard black and white film developer?
Does anyone know the formula for it?

 

[mshchem]

I enjoy the videos. I haven't shot Tri-X in roll films for a very long time. TMY in XTOL

 

[Milpool]

Keith, here is the formula

Metol 0.5g
Sodium sulfite (anh.) 40g
HQ 1g
Sodium carbonate (anh.) 1.5g
Sodium bicarbonate 1.0g
KBr 0.2g
pH 9.4 +/-0.2

In 1993 this (as well as the specified fixer formula) was removed from the standard.

I know Ilford quote ID11 as their standard developer for film speeds but what exactly is the ASA (ISO) standard black and white film developer?
Does anyone know the formula for it?

 

[Keith Tapscott.]

In 1993 this (as well as the specified fixer formula) was removed from the standard.

Does that mean that the film manufacturer can choose which ever developer they like for their ISO ratings?

 

[Anon Ymous]

Does that mean that the film manufacturer can choose which ever developer they like for their ISO ratings?

Yes, they can.

 

[Milpool]

Does that mean that the film manufacturer can choose which ever developer they like for their ISO ratings?

Yes, as well as the rest of the processing conditions (agitation etc.). However the standard does stipulate the manufacturer needs to make the processing details available, which seems to mean upon request - ie it doesn’t have to be disclosed unless asked.

 

[chuckroast]

Yes, as well as the rest of the processing conditions (agitation etc.). However the standard does stipulate the manufacturer needs to make the processing details available, which seems to mean upon request - ie it doesn’t have to be disclosed unless asked.

So, that would suggest that a straight up comparison of two film rated at the same ISO isn't meaningful unless you know the developer that was used. At best, we could say they are approximately the same speed.

In my first, very cursory and very brief look at the papers you suggested, it looks to me like a simple density measurement doesn't really tell the whole story (which you did mention upthread). The standard appears to take into account the slope (1st derivative) of the H/D curve in the shadows and again in the highlights. Do I have this right? (Again, I only glanced at these papers, so I could be way wrong.)

 

[Milpool]

So, that would suggest that a straight up comparison of two film rated at the same ISO isn't meaningful unless you know the developer that was used. At best, we could say they are approximately the same speed.

In my first, very cursory and very brief look at the papers you suggested, it looks to me like a simple density measurement doesn't really tell the whole story (which you did mention upthread). The standard appears to take into account the slope (1st derivative) of the H/D curve in the shadows and again in the highlights. Do I have this right? (Again, I only glanced at these papers, so I could be way wrong.)

Yes basically the fractional gradient ("0.3G") speed (which links back to the first-excellent-print studies) reflects that ultimately tone reproduction is a matter of contrast and that the shadow contrast relative to midtone/average contrast is important in determining the speed of an emulsion. The Delta-X method essentially combines a fixed density criterion (the familiar 0.1 above fb+fog) with the fractional gradient method within the ISO method.

When the ISO triangle is satisfied, Delta-X is basically 1 stop - ie the shadow exposure where the gradient is approximately 0.3G is 1 stop below the exposure where net density is 0.1 above fb+fog. The exposure where net density is 0.1 above fb+fog lies 3 1/3 stops below the metered exposure and the shadow exposure (0.3G speed) is 4 1/3 stops below the metered exposure. So, although on the surface the ISO method might appear to just be based on a simple fixed density speed criterion, the contrast parameters in the standard effectively "build in" the fractional gradient method by way of Delta-X.

This is why the ISO speed is a meaningful thing - as opposed to what some people would have us believe (ie that it is either an egghead laboratory test having nothing to do with photography or that it is some sort of conspiracy, either case requiring us photographers to find the "real" speed on our own).

As for the comparison of films rated at the same ISO without knowing what developer the manufacturer used etc. not being meaningful, I think that way overstates the case. In reality unless one is using extreme processes and/or special purpose/poorly formulated developers, the speeds are comparable to within a small fraction of a stop. Whether you use TMax, XTOL, D-76, DD-X, Microphen, Rodinal etc. there speed differences will generally be very small. While Kodak doesn't tell us exactly what they did, they indicate the ISO speed will be reached in most developers. For what it's worth I have found that to be the case for the Kodak, Ilford and Fuji films I've used/tested.

 

[chuckroast]

Yes basically the fractional gradient ("0.3G") speed (which links back to the first-excellent-print studies) reflects that ultimately tone reproduction is a matter of contrast and that the shadow contrast relative to midtone/average contrast is important in determining the speed of an emulsion. The Delta-X method essentially combines a fixed density criterion (the familiar 0.1 above fb+fog) with the fractional gradient method within the ISO method.

When the ISO triangle is satisfied, Delta-X is basically 1 stop - ie the shadow exposure where the gradient is approximately 0.3G is 1 stop below the exposure where net density is 0.1 above fb+fog. The exposure where net density is 0.1 above fb+fog lies 3 1/3 stops below the metered exposure and the shadow exposure (0.3G speed) is 4 1/3 stops below the metered exposure. So, although on the surface the ISO method might appear to just be based on a simple fixed density speed criterion, the contrast parameters in the standard effectively "build in" the fractional gradient method by way of Delta-X.

This is why the ISO speed is a meaningful thing - as opposed to what some people would have us believe (ie that it is either an egghead laboratory test having nothing to do with photography or that it is some sort of conspiracy, either case requiring us photographers to find the "real" speed on our own).

As for the comparison of films rated at the same ISO without knowing what developer the manufacturer used etc. not being meaningful, I think that way overstates the case. In reality unless one is using extreme processes and/or special purpose/poorly formulated developers, the speeds are comparable to within a small fraction of a stop. Whether you use TMax, XTOL, D-76, DD-X, Microphen, Rodinal etc. there speed differences will generally be very small. While Kodak doesn't tell us exactly what they did, they indicate the ISO speed will be reached in most developers. For what it's worth I have found that to be the case for the Kodak, Ilford and Fuji films I've used/tested.

Thanks - that's very helpful.

 

[Augustus Caesar]

Yes basically the fractional gradient ("0.3G") speed (which links back to the first-excellent-print studies) reflects that ultimately tone reproduction is a matter of contrast and that the shadow contrast relative to midtone/average contrast is important in determining the speed of an emulsion. The Delta-X method essentially combines a fixed density criterion (the familiar 0.1 above fb+fog) with the fractional gradient method within the ISO method.

When the ISO triangle is satisfied, Delta-X is basically 1 stop - ie the shadow exposure where the gradient is approximately 0.3G is 1 stop below the exposure where net density is 0.1 above fb+fog. The exposure where net density is 0.1 above fb+fog lies 3 1/3 stops below the metered exposure and the shadow exposure (0.3G speed) is 4 1/3 stops below the metered exposure. So, although on the surface the ISO method might appear to just be based on a simple fixed density speed criterion, the contrast parameters in the standard effectively "build in" the fractional gradient method by way of Delta-X.

This is why the ISO speed is a meaningful thing - as opposed to what some people would have us believe (ie that it is either an egghead laboratory test having nothing to do with photography or that it is some sort of conspiracy, either case requiring us photographers to find the "real" speed on our own).

As for the comparison of films rated at the same ISO without knowing what developer the manufacturer used etc. not being meaningful, I think that way overstates the case. In reality unless one is using extreme processes and/or special purpose/poorly formulated developers, the speeds are comparable to within a small fraction of a stop. Whether you use TMax, XTOL, D-76, DD-X, Microphen, Rodinal etc. there speed differences will generally be very small. While Kodak doesn't tell us exactly what they did, they indicate the ISO speed will be reached in most developers. For what it's worth I have found that to be the case for the Kodak, Ilford and Fuji films I've used/tested.

I agree that the ISO speeds are consistent, but I and almost everyone else who has "tested" exposures (in my case by bracketing at 1/2 stop intervals) and has reported his results here and elsewhere, finds that 50% to 75% ISO yields better results. This has nothing to do with the Zone System. I discovered this 50 years ago! One way to deliver consistent shadow detail is to measure the open shadow area, using 125% ISO (e.g., EI 160 for FP4). It amounts to the same thing.

 

[Andrew O'Neill]

I agree that the ISO speeds are consistent, but I and almost everyone else who has "tested" exposures (in my case by bracketing at 1/2 stop intervals) and has reported his results here and elsewhere, finds that 50% to 75% ISO yields better results. This has nothing to do with the Zone System. I discovered this 50 years ago! One way to deliver consistent shadow detail is to measure the open shadow area, using 125% ISO (e.g., EI 160 for FP4). It amounts to the same thing.

You mean you realised that 50 years ago. You certainly didn't discover it.
EI's are personal. Everyone who works seriously with B/W films knows that.

 

[Milpool]

What I’d say is a preferred EI should be based on a series of results. For example if one finds he is consistently struggling to get the desired shadow reproduction, decrease the EI. If that EI differs from the ISO speed I can’t argue against it. Whatever works, works. The only point I’m making is that if rather than basing the EI on results one instead embarks at the outset on some sort of “speed test”, it is worthwhile asking some questions: What am I testing? What are the targets/criteria based on? What is the relationship (if any) between the test criteria and print/scan/output quality? What does the test result tell me (if anything)?

It can be a testy topic (pardon the pun), but you must admit in many cases people seem to simply be parroting and/or doing all this “personal EI test” business by rote. You often read that doing a personal EI test is the first thing a serious craftsman should do. But why? That part is usually gobbledygook. And it is no surprise when people run these tests they usually end up somewhere around 2/3 stop slower than the ISO speed. There is no mystery there, but I doubt most testers know why it works out that way.

I agree that the ISO speeds are consistent, but I and almost everyone else who has "tested" exposures (in my case by bracketing at 1/2 stop intervals) and has reported his results here and elsewhere, finds that 50% to 75% ISO yields better results. This has nothing to do with the Zone System. I discovered this 50 years ago! One way to deliver consistent shadow detail is to measure the open shadow area, using 125% ISO (e.g., EI 160 for FP4). It amounts to the same thing.

 

[Augustus Caesar]

You mean you realised that 50 years ago. You certainly didn't discover it.
EI's are personal. Everyone who works seriously with B/W films knows that.

Well, when I started doing this, my negatives printed better. I had been using a variety of cameras (moving from screw-mount Yashica, to Nikkormat, to Leicaflex).

 

[Augustus Caesar]

What I’d say is a preferred EI should be based on a series of results. For example if one finds he is consistently struggling to get the desired shadow reproduction, decrease the EI. If that EI differs from the ISO speed I can’t argue against it. Whatever works, works. The only point I’m making is that if rather than basing the EI on results one instead embarks at the outset on some sort of “speed test”, it is worthwhile asking some questions: What am I testing? What are the targets/criteria based on? What is the relationship (if any) between the test criteria and print/scan/output quality? What does the test result tell me (if anything)?

It can be a testy topic (pardon the pun), but you must admit in many cases people seem to simply be parroting and/or doing all this “personal EI test” business by rote. You often read that doing a personal EI test is the first thing a serious craftsman should do. But why? That part is usually gobbledygook. And it is no surprise when people run these tests they usually end up somewhere around 2/3 stop slower than the ISO speed. There is no mystery there, but I doubt most testers know why it works out that way.

You'll note that ASA speeds were determined on the basis of print results. The goal was to find the lowest exposure that would provide excellent results. This worked, and I agree with this methodology. Then, after WWII, faster films became available. The cameras of the time (mostly leaf shutters) were designed for slower films, and this forced users into using the smallest apertures, because maximum shutter speeds were not very high (1/300 was the usual top speed for most cameras of the period). When this is done, the exposure is proportionally greater than it would be with larger apertures. So, 1/50 @ f/16 is not equivalent to 1/400 @ f/5.6 with a leaf shutter. This is because the entire aperture is uncovered for a proportionally longer time when the aperture is small. Nonetheless, the powers that be decided to double the film speeds in 1960. This would mean that an exposure of 1/100 @ f/16 under the old speed would become 1/200 @f/16 with the new higher speed. But that is the wrong approach. The correct thing to do would be to inform photographers who have leaf shutter cameras to use higher shutter speeds when they are using small apertures.

This whole thing was ironic, because right then, cheap Japanese SLRs with focal plane shutters started flooding the market, which made the change unnecessary. We are now stuck with this stupid ISO system.