History of quality control with 35mm film

[David Lyga]

I have always wondered when Kodak had managed to attain a truly uniform quality control with their 35mm B&W movie film (which substituted for single frame Leica frames and with other still cameras.)

When I say uniform quality control, I mean unquestioned, constant, homogeneous film speed, throughout; in essence, an emulsion which had no differences from beginning to end of the long rolls. At first glance, the question might seem trite, pedestrian, even wholly unimportant, but ... think. There was unwavering pressure put upon Kodak for this trait to manifest ever since they first presented film to industry and public. With Hollywood, the need was imperative and extreme. Then, with the introduction of Leica in the mid-20s, this need was absolute. Slight QC wavers could be tolerated with motion pictures, but not with still film. The foundation of Kodak's reputation was built upon a virtual perfection of quality control for film. You might, also, wish to comment on Ilford, Agfa, and Fuji emulsions.

My hunch is that this goal was not met until the early thirties, but maybe you think that it was achieved far earlier. If you can give corroborating or ancillary evidence for your conclusions, please state in detail.

This has always fascinated me and I can remember reading that ADOX, back as late as the 1970s, was still manufacturing film with aspects of manual workmanship which allowed for flaws and mistakes to have been made. The proper manufacture of film is one of the most demanding quality control models on earth. - David Lyga

 

[ic-racer]

Kodak rollfilm started in 1889 or thereabouts. I suspect they had some form of quality control at that time. Ilford and Agfa followed around 1915. I can't imagine they would bring a product to market without quality control.

 

[ic-racer]

By 1913, film testing was sophisticated enough that H&D curves were being constructed and analyzed to obtain sensitomtric constants.

THE USE OF STANDARDIZED H. & D. CURVES IN SENSITOMETRY
BY O. E. CONKLIN
Du PONT PATHE FILM MFG.CO., PARLIN,. J.
Journal of the Royal Photographic Society, 53, p. 127; 1913

 

[Nodda Duma]

Wall, author of “Photographic Emulsions” (1929) discusses the topic of quality control. E. J. Wall was active in the dry plate and film industry going back to the 1880s, and shared the experience of his career in this and other books.

The discussion is mostly in the context of how lot to lot differences in gelatin needed to be accommodated by sample batch testing as the sensitization of emulsion pre-1920s depended on impurities in the gelatin. When a company like Eastman received a new order of gelatin, the first step was characterization and adjustment of the emulsion formula for the new gelatin lot. This was even within the confines of their preferred (for consistency) gelatin suppliers.

That was a lesson learned the hard way, and it is well documented (Eastman’s biography) that lack of testing almost bankrupted the fledgling Eastman Dry Plate company a few years after its start. Once he got past that problem, Eastman introduced key quality checks in his dry plate manufacturing process: demands on suppliers, frequent, periodic testing, etc, in addition to the coating quality inspections in place from when he started coating dry plates in his mother’s kitchen. In fact, that major problem he had with his dry plates led to the one of the first major tasks of his R&D division, led by Mees at its inception at or near the turn of the century, of isolating the sensitizing impurities in the gelatin. This was accomplished by the 1920s, but long before then — mid 1880s — Eastman at least had strict quality control to ensure consistency of sensitivity.

Let me say that this isn’t a difficult task. I do the same for my dry plate emulsion. So it was easy to implement some quality testing procedure even before the H&D curve was devised.

In addition, Wall discusses the control of coating flaws in detail and the apparatuses for addressing them. As for timeframe, IIRC he touched on progress since the days of dry plate (1880s) and film (1890s), so this was all actively being worked out and solved before 1900. From there, the primary QC task was to maintain the *same* level of quality while more complex emulsions and faster production throughputs were being realized in the 20th Century.

In the early days of dry plate commercialization, it was said the Seeds Dry Plates were considered the best, highest quality dry plates on the market. That was a major reason why George Eastman bought them out. I’ve purchased several unopened boxes of Seeds Dry Plates, from various periods in their company’s history (primarily pre-1895), to evaluate the quality. They are indeed very good, and you simply cannot achieve that good a coating without quality control.

-Jason

 

[David Lyga]

Wall, author of “Photographic Emulsions” (1929) discusses the topic of quality control. E. J. Wall was active in the dry plate and film industry going back to the 1880s, and shared the experience of his career in this and other books.

The discussion is mostly in the context of how lot to lot differences in gelatin needed to be accommodated by sample batch testing as the sensitization of emulsion pre-1920s depended on impurities in the gelatin. When a company like Eastman received a new order of gelatin, the first step was characterization and adjustment of the emulsion formula for the new gelatin lot. This was even within the confines of their preferred (for consistency) gelatin suppliers.

That was a lesson learned the hard way, and it is well documented (Eastman’s biography) that lack of testing almost bankrupted the fledgling Eastman Dry Plate company a few years after its start. Once he got past that problem, Eastman introduced key quality checks in his dry plate manufacturing process: demands on suppliers, frequent, periodic testing, etc, in addition to the coating quality inspections in place from when he started coating dry plates in his mother’s kitchen. In fact, that major problem he had with his dry plates led to the one of the first major tasks of his R&D division, led by Mees at its inception at or near the turn of the century, of isolating the sensitizing impurities in the gelatin. This was accomplished by the 1920s, but long before then — mid 1880s — Eastman at least had strict quality control to ensure consistency of sensitivity.

Let me say that this isn’t a difficult task. I do the same for my dry plate emulsion. So it was easy to implement some quality testing procedure even before the H&D curve was devised.

In addition, Wall discusses the control of coating flaws in detail and the apparatuses for addressing them. As for timeframe, IIRC he touched on progress since the days of dry plate (1880s) and film (1890s), so this was all actively being worked out and solved before 1900. From there, the primary QC task was to maintain the *same* level of quality while more complex emulsions and faster production throughputs were being realized in the 20th Century.

In the early days of dry plate commercialization, it was said the Seeds Dry Plates were considered the best, highest quality dry plates on the market. That was a major reason why George Eastman bought them out. I’ve purchased several unopened boxes of Seeds Dry Plates, from various periods in their company’s history (primarily pre-1895), to evaluate the quality. They are indeed very good, and you simply cannot achieve that good a coating without quality control.

-Jason

Thank you for a truly in-depth explanation which covers all areas. I always knew that QC existed from the earliest days (film: i.e., 1889) but QC is not always QC I am certain that you can understand. One can have rather stringent requirements but, still, there are troubling deviations. I was after the ultimate QC, if I dare phrase it that way. Perhaps, that 'event' happened even before the thirties.

What piqued my interest with this was watching old silent films where the consistency did not always seem perfect. Perhaps, there is more to this than with the film's QC. Indeed, DEVELOPMENT of that film plays a major part here, and getting the replenishment factors completely ironed out was a formidable task, for certain. There are many factors which finally allowed films by the late thirties to be fully complete with regard to a consistency which matches that which we take for granted today. - David Lyga

 

[Nodda Duma]

Hi David,

yep, I think you hit the nail on the head there. A couple additional thoughts to add context:

Quality control in the past was driven by customer / manufacturer expectations: What is “good enough”? In the 1950s era United States, tied to the post-WW2-era trend of standardization (to streamline the ramp-up of industry in case of war), the expectation of quality was codified in standards and specifications. Those standards have been consistently updated — ISO standards today, particularly ISO 9001 in a generalized industry sense — so you can actually go out and find what aspects of quality control the manufacturers adhere to, both from a general manufacturing sense as well as specifically applied to the film industry.

As an engineer, in the more general sense, I tend to shudder at the massive juggernaut of QC standards... not from the “necessary evil” perspective, but more from a “this is codified basic common sense and thus shouldn’t be necessary” perspective. But... I’ve been around long enough to understand that quality control as codified in specs and standards is a mutually beneficial “shared experience” of industry, and helps the progress (and safety) of technology in a good way by displaying the art of the possible. I like to think the scientific and technical community shares that sentiment. As for standards and specifications (of which QC is one part), it allows for universally understandable technical communication. “35mm film” actually means something, and isn’t arbitrarily defined from manufacturer to manufacturer.

 

[David Lyga]

Hi David,

yep, I think you hit the nail on the head there. A couple additional thoughts to add context:

Quality control in the past was driven by customer / manufacturer expectations: What is “good enough”? In the 1950s era United States, tied to the post-WW2-era trend of standardization (to streamline the ramp-up of industry in case of war), the expectation of quality was codified in standards and specifications. Those standards have been consistently updated — ISO standards today, particularly ISO 9001 in a generalized industry sense — so you can actually go out and find what aspects of quality control the manufacturers adhere to, both from a general manufacturing sense as well as specifically applied to the film industry.

As an engineer, in the more general sense, I tend to shudder at the massive juggernaut of QC standards... not from the “necessary evil” perspective, but more from a “this is codified basic common sense and thus shouldn’t be necessary” perspective. But... I’ve been around long enough to understand that quality control as codified in specs and standards is a mutually beneficial “shared experience” of industry, and helps the progress (and safety) of technology in a good way by displaying the art of the possible. I like to think the scientific and technical community shares that sentiment. As for standards and specifications (of which QC is one part), it allows for universally understandable technical communication. “35mm film” actually means something, and isn’t arbitrarily defined from manufacturer to manufacturer.

It would be interesting to delve into the ongoing REASONS for the determined level of QC. ALL will admit that perfection is a level which can, theoretically, never be fully achieved. So, it comes down to the interplay of 1) common sense, 2) dollars, and a 3) theoretical underpinning which holds perfection as a 'godly' religion for its own sake.

In its latter years, Kodak almost achieved a synergism of the three. But, WHY? I think that the reputation of Kodak was such that to be less than truly obsessive would have been economically, socially, and industrially counterproductive. Few who are not of my age, or close to it, can begin to understand the level of status that that corporation was tethered to during the mid and latter 20th Century. - David Lyga

 

[MattKing]

Consistent high quality minimizes waste and therefore maximizes profit when production volumes are high.
And at its zenith, Eastman Kodak produced very high volumes.
It is important to understand as well that Eastman Kodak made a lot of money selling chemicals, paper and (to a lesser extent) equipment to labs.
If the film you make for photographers is inconsistent, your lab customers will end up with unhappy customers of their own.

 

[btaylor]

In terms of motion picture film I would argue the opposite- that QC was much more critical from roll to roll. When editing requires film from different rolls to be abutted to one another it is critical that they match. There is an awful lot more flexibility when printing still frames. Back when movies were still shot on film a production would buy all the film from a single emulsion batch before shooting started to minimize variance.
I would also say that in terms of the early motion pictures you have seen there are a myriad of factors that influenced what actually you got to see. My first introduction to silents was probably the 16mm prints I saw at a pizza parlor when I was a teenager. Poor worn dupes of old prints I suspect, and they were run at the wrong speed. Quite awful from a IQ standpoint but wonderful films. A few years ago I saw a gorgeous restoration of FW Murnau’s “Sunrise” in a high def format. The cinematography and IQ was stunning- we sometimes get a false impression of what films looked like in the early days. Sorry to get a bit OT.

 

[btaylor]

Consistent high quality minimizes waste and therefore maximizes profit when production volumes are high.
And at its zenith, Eastman Kodak produced very high volumes.
It is important to understand as well that Eastman Kodak made a lot of money selling chemicals, paper and (to a lesser extent) equipment to labs.
If the film you make for photographers is inconsistent, your lab customers will end up with unhappy customers of their own.

Yea, this. The “Yellow God”! Their quality and range of products was amazing. When I was growing up if you wanted the best you bought the yellow box. I used all sorts of off brand, “expired” materials from Freestyle, but when it counted I always used Kodak.

 

[guangong]

In terms of motion picture film I would argue the opposite- that QC was much more critical from roll to roll. When editing requires film from different rolls to be abutted to one another it is critical that they match. There is an awful lot more flexibility when printing still frames. Back when movies were still shot on film a production would buy all the film from a single emulsion batch before shooting started to minimize variance.
I would also say that in terms of the early motion pictures you have seen there are a myriad of factors that influenced what actually you got to see. My first introduction to silents was probably the 16mm prints I saw at a pizza parlor when I was a teenager. Poor worn dupes of old prints I suspect, and they were run at the wrong speed. Quite awful from a IQ standpoint but wonderful films. A few years ago I saw a gorgeous restoration of FW Murnau’s “Sunrise” in a high def format. The cinematography and IQ was stunning- we sometimes get a false impression of what films looked like in the early days. Sorry to get a bit OT.

You said exactly what I wanted to say, but much better. I used to order 16mm 10 rolls at a time from Kodak with emulsion numbers on box, but since I shot a variety of subjects outdoor it didn't really make much difference, but for a major studio production small variations would be noticed by audience.
But silent films during silent era were shot on crystal clear nitrate base. Long ago, maybe at MOMA in ealrly 1950s, I saw a couple films projected on nitrate and the effect was stunning. A noticeable difference between results from nitrate and safety film.

 

[Kino]

I would say with motion picture film, they had that nailed by the early 1910's for black and white (of course) film.

Adjusted for inflation, films have always pretty much cost the same amount (as far as media, processing and so on) since the very early silent era. It was, and is, still an astronomically expensive medium to produce and bad quality film was not to be tolerated...

 

[David Lyga]

In terms of motion picture film I would argue the opposite- that QC was much more critical from roll to roll. When editing requires film from different rolls to be abutted to one another it is critical that they match. There is an awful lot more flexibility when printing still frames. Back when movies were still shot on film a production would buy all the film from a single emulsion batch before shooting started to minimize variance.
I would also say that in terms of the early motion pictures you have seen there are a myriad of factors that influenced what actually you got to see. My first introduction to silents was probably the 16mm prints I saw at a pizza parlor when I was a teenager. Poor worn dupes of old prints I suspect, and they were run at the wrong speed. Quite awful from a IQ standpoint but wonderful films. A few years ago I saw a gorgeous restoration of FW Murnau’s “Sunrise” in a high def format. The cinematography and IQ was stunning- we sometimes get a false impression of what films looked like in the early days. Sorry to get a bit OT.

You make a very valid point with the editing requirement. - David Lyga

 

[David Lyga]

I would say with motion picture film, they had that nailed by the early 1910's for black and white (of course) film.

Adjusted for inflation, films have always pretty much cost the same amount (as far as media, processing and so on) since the very early silent era. It was, and is, still an astronomically expensive medium to produce and bad quality film was not to be tolerated...

I cannot confirm what you say here concerning film prices matching inflation rates.

When I was in my mid 20s living in New York in the mid 1970s, 100 feet of B&W 35mm film cost $8. At the time, minimum wage was slightly above $2 per hour. NOW, there are jobs in New York paying as little as $10 per hour (despite the national minimum wage currently being $7.25). Thus, from 1975 to now, wages have appreciated by about 5X. Try to buy 100 ft of fresh Tri-X for $8 X 5 = $40. Not possible. - David Lyga

 

[Alan Edward Klein]

You said exactly what I wanted to say, but much better. I used to order 16mm 10 rolls at a time from Kodak with emulsion numbers on box, but since I shot a variety of subjects outdoor it didn't really make much difference, but for a major studio production small variations would be noticed by audience.
But silent films during silent era were shot on crystal clear nitrate base. Long ago, maybe at MOMA in ealrly 1950s, I saw a couple films projected on nitrate and the effect was stunning. A noticeable difference between results from nitrate and safety film.

How were nitrate projections different?

 

[cmacd123]

In terms of motion picture film I would argue the opposite- that QC was much more critical from roll to roll. When editing requires film from different rolls to be abutted to one another it is critical that they match. There is an awful lot more flexibility when printing still frames. .

Agree 100%. and constancy is even important within a given shot. When your eye is seeing 24 (sound speed) different frames in a second, any variation will appear as a "flash" or a drift in the image. your vision is subconsciously comparing every frame to the previous one. AND the image is presented much bigger than anyone's normal print size from 25mm film. Now grain and such may be just seen as noise as it does not stay still.

As far editing, a simple scene of a conversation likely is shot with a set up for one actor, and then the reverse shot is set up for the second actor. each of them likely on a different roll of film (the old 90ft a minute speed chews though a lot of 400 ft rolls.) and all has to mesh perfectly when it gets edited, so the audience is not thinking about the editing. It is unlikely that the two shots can have exposure timing done between them, and making an inter-positive would show, so the two shots have to match perfectly.

 

[Alan Edward Klein]

Isn't Pro film made for Pros so that the emulsions are consistent more than consumer types and they are "aged" so the expiry date are all the same and sooner? I suppose the word "ripe" comes to mind.

 

[Nodda Duma]

Reputation plays into it for sure.

on the other side of the coin... “Perfection is the enemy of good enough.” That’s the other reason for quality control... to ensure the development and production teams don’t have to work extra hard to ensure they are eliminating issues that don’t really matter.

-Jason

It would be interesting to delve into the ongoing REASONS for the determined level of QC. ALL will admit that perfection is a level which can, theoretically, never be fully achieved. So, it comes down to the interplay of 1) common sense, 2) dollars, and a 3) theoretical underpinning which holds perfection as a 'godly' religion for its own sake.

In its latter years, Kodak almost achieved a synergism of the three. But, WHY? I think that the reputation of Kodak was such that to be less than truly obsessive would have been economically, socially, and industrially counterproductive. Few who are not of my age, or close to it, can begin to understand the level of status that that corporation was tethered to during the mid and latter 20th Century. - David Lyga

 

[Vincent Peri]

The only problem I ever had with Kodak 35mm film was way back when in the 70's, I think it was just-introduced Kodachrome 64, which had a slight greenish(?) color cast to it. As I recall, the problem was fixed within a couple of months.

 

[MattKing]

My father had a couple of quality control stories he shared with me - he was the Customer Service manager at a Canadian Kodak processing laboratory in North Vancouver, BC Canada between 1961 and 1983. Essentially, they turned on issues relating to the fact that it was human beings that made things happen!
They only worked with Kodachrome and Ektachrome slide and movie film.
The most frequently recurring problem was with people whose slide film cameras had inconsistent spacing, or who would be in the habit of rewinding partly exposed rolls of 35mm film, and then putting them back into the camera in order to advance past the existing exposures in order to finish the roll.
The slide mounters were described as "semi-automatic". They required manual supervision from people who surely must have had to fight to remain alert throughout their shifts!
Even the best quality control struggles tremendously with human beings.

 

[Kino]

I cannot confirm what you say here concerning film prices matching inflation rates.

When I was in my mid 20s living in New York in the mid 1970s, 100 feet of B&W 35mm film cost $8. At the time, minimum wage was slightly above $2 per hour. NOW, there are jobs in New York paying as little as $10 per hour (despite the national minimum wage currently being $7.25). Thus, from 1975 to now, wages have appreciated by about 5X. Try to buy 100 ft of fresh Tri-X for $8 X 5 = $40. Not possible. - David Lyga

I should have said the cost of producing a film was just about the same as now, with adjusted inflation is what I meant.
Actually, when you consider that a Bell & Howell 2709 hand cranked camera cost $2500 in 1915, the equipment might have been a bit higher in cost than the later years.
It was never cheap...

 

[Kino]

As far editing, a simple scene of a conversation likely is shot with a set up for one actor, and then the reverse shot is set up for the second actor. each of them likely on a different roll of film (the old 90ft a minute speed chews though a lot of 400 ft rolls.) and all has to mesh perfectly when it gets edited, so the audience is not thinking about the editing. It is unlikely that the two shots can have exposure timing done between them, and making an inter-positive would show, so the two shots have to match perfectly.

Shot exposures have always been timed (graded) for film-based motion pictures; that's what I do. However, I will say that each shot is usually quite consistent and if I have a Long shot, medium shot and two close-ups, they tend to be exactly the same point value throughout the scene. Of course, some films have sections of scenes re-shot for editorial purposes and those shots can vary to a degree.

The worst are the "pick-up" and Second-Unit shots... some of those can be wildly inconsistent...

 

[cmacd123]

Shot exposures have always been timed (graded) for film-based motion pictures; that's what I do. However, I will say that each shot is usually quite consistent and if I have a Long shot, medium shot and two close-ups, they tend to be exactly the same point value throughout the scene. Of course, some films have sections of scenes re-shot for editorial purposes and those shots can vary to a degree. .

I apologize if I over simplified. the timing between scenes CAN be adjusted, but it is tricky when two shots are being cross cut if the basic Points are not close to each other. Thus as Kino says, the DP works quite hard to make all scenes - even if shot at different times consistent.

(exposure for printing in the movies is done by "Points", which represent a setting in the Bell and Howell printers used for traditional productions.)

 

[flavio81]

It would be interesting to delve into the ongoing REASONS for the determined level of QC. ALL will admit that perfection is a level which can, theoretically, never be fully achieved. So, it comes down to the interplay of 1) common sense, 2) dollars, and a 3) theoretical underpinning which holds perfection as a 'godly' religion for its own sake.

In its latter years, Kodak almost achieved a synergism of the three. But, WHY? I think that the reputation of Kodak was such that to be less than truly obsessive would have been economically, socially, and industrially counterproductive. Few who are not of my age, or close to it, can begin to understand the level of status that that corporation was tethered to during the mid and latter 20th Century. - David Lyga

But, David, it sounds like if you were implying that only Kodak achieved a high degree of QC.

I have used hundreds of rolls from Agfa, Kodak, Ilford and Fuji though my life, and never had any QC issue attributable to the film manufacturer. Even occasionally i used films from Foma and Tura (Tura AG, germany),and no problem. The only time i had QC problems were with 120 Lomography films which had the well-documented backing paper issue. And this was Kodak film ("Lady Gray" is rebadged Tmax 400)

 

[David Lyga]

But, David, it sounds like if you were implying that only Kodak achieved a high degree of QC.

I have used hundreds of rolls from Agfa, Kodak, Ilford and Fuji though my life, and never had any QC issue attributable to the film manufacturer. Even occasionally i used films from Foma and Tura (Tura AG, germany),and no problem. The only time i had QC problems were with 120 Lomography films which had the well-documented backing paper issue. And this was Kodak film ("Lady Gray" is rebadged Tmax 400)

No, indeed, there are (were) many firms with a strong commitment to QC. The reason I earmarked Kodak was because in the 50s, 60s, even 70s, they were the dominant force. They were truly ubiquitous, not merely another player, like today.

In summation, using Kodak as the paradigm indirectly asks the same of the other firms, which followed suit using Kodak's model. Asking this of Kodak indirectly asks the question for all manufacturers of sensitive materials. - David Lyga