Reconstructing E-E KOZLOV's photo archive from the 1980s • Chapter 6. Svema and Tasma filmstrip edge markings • © Hannelore Fobo, 2021

Because 35 mm films were originally developed for film cameras, those sold for cameras were not provided with sequential numbers either, at least not initially. In other words, a specific frame had no reference to a specific number, which made it difficult to select a particular picture for printing. The films Valentin Kozlov used around 1960 were all unnumbered and had no imprint of the company name, either. I couldn’t find out when exactly Soviet factories introduced sequential numbering. We still find some of those “unidentified” films in Evgenij Kozlov’s archive, but the large majority are Svema and Tasma films numbered along the film edges, below the sprocket holes – the perforation allowing transporting the film to the next frame. Accordingly, the company name is above the top edge sprocket holes. Additionally, most Svema and some Tasma imprints display production dates. Neither printed the film type or film speed.



Example of a film without numbers nor imprint. Boris Grebenshikov, Timur Novikov, Georgy Guryanov and Konstantin Kinchev. Spring 1985. (See also Chapter 5 more >>)		Example of a Svema film with film numbers and imprint 6 Svema M6 85. (E-E) Evgenij Kozlov at Andrey Medvedev's place.

A typical Svema imprint would be 11 CBEMA M1 86, where “11” most probably refers to the exact location of the production (the production plant). “CBEMA” is Cyrillic for Svema, “M1” is the first month of the year and “86” the year 1986. Thus, the film was produced by Svema in January 1986. Two such films are in Kozlov’s archive. One of them documents a Pop Mekhanika performance from 20 October 1986 (see Chapter 9 more >>). The other one was used for an art project. Evgenij Kozlov photographed two pregnant women and created one of his “USA-CCCP” series with these pictures. Since the series itself is from 1987, it is possible that he took the pictures also in 1987. Be that as it may, If an event cannot be dated otherwise, the film production date helps us to indicate the earliest possible date: “January 1986 or later”. In fact, considering that the films had to be delivered to the shops and that shops might sell any time after, January 1986 is not very probable.

We find a similar system in some of the Tasma films, for instance 8-TASMA-4-84-Б corresponding to April 1984 (I cannot say what the Cyrillic B at the end stands for). Unfortunately, most Tasma films and some Svema films provide only some limited information, such as “27 TACMA” or “6 CBEMA”, giving no clue to the year.



8-TASMA-4-84-Б Unidentified person and Oleg Kotelnikov, 1984		11 Svema Vladislav Mamychev Monroe at Kozlov's studio "The Russian Field", Leningrad, 1990. See vintage contact print more >>

Sequential numbering follows various systems. In date-specific Tasma and Svema films, the numbers appear with two consecutive numbers for each frame, for instance 5 and 6 or 7 and 8. This type of numbering is useful when shooting with a half-frame camera instead of a full-frame camera, because a half frame, as the name suggests, is only half the format of a normal frame. Half frame is the typical format of 35 mm movie films. This is why a “camera” film, when shot with half frames, can be projected to the screen with a 35 mm movie projector, and if you connect several of those films, you get some kind of animated film consisting of single shots. In terms of sprocket holes, a half frame has four holes instead of eight, and in this case, each frame gets its own number, 5 or 6, 7 or 8. In western films, “A” numbers are used instead of consecutive numbers, but they follow the same principle of marking half frames: 1, 1A, 2, 2A etc.

Fragment of Boris Kazakov's half-frame 35 mm film from 1998. Each frame has a width of four sprocket holes and is provided with its own number.
Kazakov, a Petersburg artist and film-maker, shot the film at (E-E) Evgenij Kozlov's studio "The Russian Field 2" when he came to Berlin to present one of his films at the Berlinale. In the foreground: E-E Kozlov, in the background artist and film-maker Evgeny Kondratev. On the very right Hannelore Fobo shooting with Kozlov's Nikon camera.
The Kozlov & Fobo Collection, Berlin.

Yet for some reason that I couldn’t figure out, the distribution of numbers is not completely uniform on those dated Svema and Tasma films. I haven’t checked them all, but, for instance, 4 CBEMA M10 84 (inventory name BB in my nomenclature; see Chapter 9, Numbering the films more >>) and the above-mentioned 8 TACMA-4-84-Б (invent. name DN) print them with alternating distances of three and five sprocket holes (3-5-3-5-3-5 etc). Consequently, with half frames, you may get two numbers on one frame and none on the following frame. This occurs when the first number is near the first sprocket hole and the second number is near the fourth sprocket hole, then the third number, “skipping” sprocket holes 5 to 8, is at the ninth sprocket hole and so forth.

Two frames from the BB film (4 Svema M10 84) displaying the film numbers in the typical Svema sequence of three + five sprocket holes. More frames from the BB film below and in Chapter 9 more >>
Kozlov painted the negative with nos 13 / 14 with a felt pen. We later printed this and some other painted negatives on a colour paper. See Chapter 11 more >>.

In this respect, the undated Svema and Tasma films are more regular: their numbers are printed at exact 4/4 intervals. The undated Tasma films conform to the system of double numbers adopted for Western films: 1, 1A, 2, 2A etc. To complicate the situation further, there are also dated Tasma films with double numbers: 3-TASMA-6-80-Б (film DV) applies a system of 1, 1, 2, 2, while keeping to the 3-5 distribution.



Performance with members of the Kino band: Alexander Titov, Yury Kasparan, Georgy Guryanov, and Viktor Tsoy, 1984. The undated Tasma films conforms to the system of double numbers also adopted for Western films: 1, 1A, 2, 2A etc, with a regular distribution of numbers. The numbers are mirror-inverted. More pictures from the same film (AB) are availabe in Chapter 9 more >>.		(E-E) Evgenij Kozlov at his Peterhof flat and studio "Galaxy Gallery" more >> with one of his paintings from 1980 more >>. He shot the picture with the help of the shutter release cable see Chapter 1, as shown by the position of his left shoulder. The 3-TASMA-6-80-Б (film DV) applies a system of double numbers – 1, 1, 2, 2, etc. – while keeping to the 3-5 distribution. The numbers are mirror-inverted. The frame has the double number 10.

There is another irregularity in the numbering system: film numbers and film frames do not always align. This is the case when even film numbers – and, likewise, uneven film numbers – do no stand at exact intervals of 8 sprocket holes.

They do so with the EE film (CB M10 1987), but they don’t in the case of DO: the DO film numbering gives us 40 frames, but is actually has 39 frames. Here, frames and numbers do not coincide because the numbers stand too close to each other, and the same goes for the BB film. Thus, in the BB film, number 1 “sits” below the third sprocket hole of the frame, while after 19 frames, instead of number 39 being below the third number, there is a no. 41 below the fifth sprocket hole.

By contrast, they stand a little bit too far from each other in some other films, like the AM film, so that the number of frames exceeds the number “pairs” by one frame. In both cases, the numbers “move” along the frames – left, when they stand too close, and right, when they stand to far from each other.

If this sounds confusing, this is only the beginning, because with Svema and Tasma films, some other features must also be taken into consideration.

Although the films were sold with 36 frames, Kozlov shot up up to 39 frames. The films were longer than their western counterparts: 165 instead of 160 cm. The width of a negative, including the space between frames, is circa 3.8 cm; accordingly 39 x 3,8 is 148.2 cm, and the remaining end pieces of the film on either side were used for the tongue and cusp. Film EE (CB M10 1987) starts at number 1/2 and goes up to 78, which corresponds to 39 frames.

Yet such a “normal” counting is an exception. The film numbers actually go in a circular way, and frame numbers may start anywhere in the numbering scale between 1 and 88 or 90 (Svema and Tasma single numbers), or between 1 and 44 or 45 (Tasma double numbers). Having reached the highest number, they continue with 1.

In the case of DO (85 in 4 CB M6 85), you get the following sequence for the first four frames: 85/85, 87/88, 89/90/, 1/2. Accordingly, the last of DO’s 38 film frames is number 73/74. Put differently, no frames exist for no. 75-84. Most dated films follow this principle. Other examples are 13 CBEMA M7 84 (film AM, going from 72 to 88 and continuing 1 to 50) and 14 CBEMA 10M 84 (film AY, going from 47 to 88 and continuing 1 to 34).

Marianna Tsoy making up Alexander Titov for Kozlov's photo session with the Kino band. Kozlov shot three films shot to choose a group portrait for his cover of Kino's album “Nachalnik Kamchatki” more>>
The picture is from film AM (13 CB M7 84). In this frame, 88, the highest number, is followed by 1.

Two frames with Katya Selitskaya from Kozlov's photo session "Fashion Show", 1984. See Chapter 7 more>>
Film AY (14 CB M10 84). The picture with frame numbers 87 / 88 comes before the picture with frame nubmers 1 / 2

So what happened to the missing numbers? Although we perceive film numbers only on exposed frames and not on the unexposed end pieces of a film, it appears that numbers continued along the entire film roll from which the 165 cm pieces were cut. That is to say, in the case of 88 numbers, they went from 1 to 88 + 1 to 88 + 1 to 88 and so forth, without leaving any unmarked gaps for the end pieces – the cusp and leader.

In fact, if we consider that 88 frame numbers correspond to 44 full frames, we can easily see that they exceed a 165 cm piece film, because 44 frames make up 167.2 cm (and 45 frames 171 cm). If “exceeding” numbers went into the next filmstrip, this would explain not only the circular system of numbers, but also the arbitrariness of starting numbers as part of a rotating system. The lack of congruence between numbers and frames (when there are less numbers than frames) would enforce the arbitrariness.

There is still another, even stranger aspect. It concerns the way the numbers appear on the filmstrips: on Svema, they may stand at the bottom edge, or they may stand upside down at the top edge (with the date upside down at the bottom). Basically, it is the filmstrip rotated to 180 degrees. Accordingly, while we see the numbers ascending from frame to frame when they are in the correct position, we see them descending when the stand up-side down. Examples for descending numbers are 9 CB M5 84, film AH (from 45 to 1 and 88 to 57) and 5 CB M7 84, film BA (from 22 to 1 and 88 to 51).

But why were so many films flipped? It actually seems that there are more “flipped” than “unflipped” films.

The simplest hypothesis is that all flipped films were shot with the LOMO 135 SV. Because it has the take-up spool on the “wrong” side” of the camera, the cartridge is placed into the camera upside down. In this case, it is normal to have the film flipped and the numbers standing upside down on the upper edge.

The LOMO 135 SV has the take-up spool on the “wrong” side” of the camera – accordingly, the cartridge is placed into the camera upside down.
Photo: Hannelore Fobo

But this hypothesis didn’t convince me, because many of the “flipped” films with descending numbers actually display the typical FED-2 effect that sees images close to or even overlapping with the lower perforation. It therefore simply makes no sense to relate all flipped films to the LOMO 153 VS and all unflipped films to FED-2.

The second, more daring hypothesis is that “flipped” films were rolled up in the factory from the “wrong” end, and this assumption needs some explanation. It is based on how the numbers are printed onto the film edge. I hereby rely on the information alan_marcus | 2 published on photo.net in 2010: “During the film manufacturing process, the edge printing is applied. This happens before spooling onto rolls.” External link

It means that how the numbers appear on a film is determined by the way those 165 cm film pieces are is rolled up. Obviously, if we start at one end, we get one result, and if we start at the other end, we get the opposite result.

Remember that Soviet films rolls were sold in two forms, with and without a spool. In both cases, they have to be loaded into a cartridge, and then the cartridge is placed into the camera. We will consider these cases separately, starting with a film X rolled on a spool.

For the sake of simplicity, we will assume that the last number at the free end of film X, the leader of tongue of the film, is 1; accordingly, the last number at the “inner” end of the film attached to the spool, the cusp, is 88.

This is what we expect them to be when the film goes directly into the cartridge, because once the cartridge is loaded into the camera and we start shooting pictures, the film is wound up on the take-up spool from the tongue to the cusp, and thus with numbers ascending: 1-2-3-4-5… 85, 86, 87, 88. Again, we are talking about images in a camera; if we could look through the back of the camera, we would see the images rotated by 180 degrees, and therefore the numbers would actually be at the upper edge of the film, standing upside down, and we would perceive them descending.

Example of a film with a correct numbering sequence starting at 1/2 with the first frame. Using a photo editing programme, this principle is demonstrated with a black and white negative strip that was placed rotated by 180 degrees, the way images are projected onto the film (how we would see them looking through the back of the camera). For the sake of clarity, the frame numbers were highlighted.
The black end – the leader (right), of which a large piece was cut off when processing the film – goes into the take-up spool.
Evgenij Kozlov took these portraits of artist Igor Smirnov (Sten) at his Leningrad studio “The Russian Field” in 1990.The LOMO 135 SV has the take-up spool on the “wrong” side” of the camera – accordingly, the cartridge is placed into the camera upside down.
Photo: Hannelore Fobo

But everything changes when the same piece of film X is delivered without a spool and wound up on a spool by the costumer, as this inverts the position of the end pieces and, accordingly, the direction of frame numbers. Consequently, loading the cartridge into the camera, the end piece with the leader that goes into the take-up spool is now 88 instead of 1, and shooting the film causes a descending order of numbers: 88–87-86-85. But that isn’t all. Winding film X around a spool also means that, in comparison with film X sold on a spool, it will be rotated by 180 degrees when placed into the cartridge. The left-right inversion is coupled with a bottom-top inversion. If we looked at the film through the back of the camera, with the images standing upside down, we would see the numbers standing upright at the lower film edge. Otherwise said, the numbers no longer stand in the same direction as the images (as in a film sold on a spool), but in the opposite direction.

The example from above presented with an “incorrect” numbering sequence of descending numbers written on the lower edge; note that the small original numbers can still be seen on the upper edge.
The numbers no longer stand in the same direction as the images – they are now in the opposite direction.
Photo: Hannelore Fobo

Finally, after processing film X, we rotate it by 180 degrees to look at the images in their correct position. Obviously, with film X sold on a spool, everything is fine, but with film X sold without a spool, the numbers are now standing upside-down at the upper film edge, where we read them backwards, that is, descending: 88-87-86-85.

The same two versions of a numbering sequence with the images in an upright position. The correct version starting at 1/2 delivers an ascending order with numbers standing at the lower film edge, while the incorrect version starting at 88 delivers a descending order with numbers standing upside-down at the upper film edge.

Basically, with a film sold on a spool, we get only one 180 degrees rotation of images – the moment we look at the processed film – and this flips the film numbers into the right place. With a film sold without a spool (as Evgenij Kozlov would buy them), we get two such rotations, so that the second rotation undoes the effect of the first rotation, flipping the numbers back to the wrong place (2 x 180° = 360° = 0°). Therefore, my assumption is that Svema and Tasma often sold film rolls without a spool that should have gone on a spool – without bothering to adapt the edge printing sequence during the manufacturing process to correct the inversion of the film direction. I say “often” because there are indeed “correct” films with numbers ascending at the bottom, for instance film CH (6 CB M6 85) or film BB (4 CB M10 84).

If my assumption is correct, then we can say that with Svema, this lack of accuracy was rather the rule than an exception, and the same goes for dated Tasma films.

We may further develop these reflections: what if all Svema films sold without a spool were actually “not correct”? In this case, those with the numbers flipped, like AH and BA, would have been shot with the FED-2, and those with the numbers in the correct position, like BB or CH, would have been shot with the LOMO. The LOMO has the cartridge on the “wrong” side, and this undoes the effect of flipped numbers originating from winding the film around a spool.

It is, however, is difficult to maintain this idea. CH is an example of a symmetric film exposure, and it might have been shot with a LOMO, but BB displays the asymmetric film exposure of the FED-2, and BB cannot have been shot with a Lomo. Or can it? Here, the situation seems to get out of hands. We would have to produce proof of the contrary: that Kozlov’s FED-2 always exposes frames closer to the lower perforation, and not just most of the time, and we would have to prove that his LOMO 135 SV never shoots films in this way. Plus we would have to prove this for the past. But even if the Lomo shoots like the FED-2, this would just be enough to speak of the possibility that all unspooled films were sold with flipped numbers, of something that cannot be ruled out altogether.

Though the question is not essential for reassembling Kozlov’s archive, I still find it intriguing, because it shows that coming to conclusions involves a large variety of parameters that we may be able to define without being able to prove that they actually apply.

There is yet another aspect to this enigma. Undated Tasma films also have inverted numbers, but they stand inverted at the bottom edge — and they are ascending. Actually, as demonstrated by the AC film, a Tasma-6 film, we not only see them standing upside down (rotated to 180 degrees, like in the previous example), but mirror-inverted (flipped left to right), too.

Two frames fom the AC film, a Tasma-6 film, with members of the Kino band – Yury Kasparan, Georgy Guryanov, Alexander Titov, and Viktor Tsoy, 1984.
This films shows the frame numbers at the lower edge in ascending order, but rotated by 180 degrees and mirror inverted.

If we turn the whole film upside down – rotate it by 180 degrees, the way the images are projected onto the film in the camera – the numbers stand upright in descending order. While this corresponds to the first "correct" example of a filmstrip placed into the camera, they nevertheless remain mirror-inverted: we read numbers and letters from left to right.

The problem is solved if we rotate a filmstrip around its horizontal axis, so that we now look at the side with the photo emulsion – the “wrong” side of the film, so to speak.

Note that we haven’t discussed that side of the negative film previously. We always looked at the shiny or “right” side of a dated Svema or Tasma film. All we did was rotating the film by 180 degrees in a two-dimensional plane, but we haven’t treated it as a three-dimensional object with a shiny front side and a reverse emulsion side so far.

Et voilà, now that we look at the reverse side and turn the film upside down, the numbers stand true sided, right way up, at the top edge. Unfortunately, the image is now “wrong” – flipped up/down, left/right.

We actually perceive the image in just the same way as if we were looking onto the film emulsion through the camera’s lens (placing the camera in front of us), because a lens projects an upside down and mirror-inverted image of a non-transparent object onto a screen. The screen, in this case, is the side of the film with the film emulsion.

Therefore, in this game of rotating and inversing images, it is very important to remember each time not only where we stand in relation to the film, but also which side of a film we are looking at. And if it hasn't been clear to us from the beginning, we have now seen that if on a film, images and numbers stand in opposite directions – up-down or left-right, or both – no rotating around any of the axes will ever get this right.