Invited Paper
Development of a photographic package printer: a case study R. Barry Johnson∗, SPIE Fellow and Life Member Consultant, 1527 Chandler Rd., Huntsville, AL 35801, USA and Department of Physics, Alabama A&M Univ., 4900 Meridian St., Normal, AL, 35762, USA ABSTRACT A specialty area in the commercial photographic industry involves simultaneously producing a plurality of high-quality photographs of varying size and shape from a single photographic negative or digital image. The images are formed on a large piece of photographic paper by a set of lenses having specific magnifications and appropriately located between the negative and paper. Package printers are typically reconfigurable to allow different sets of images to be created; however, such reconfiguration is time consuming. Most often, a package printer is configured and then devoted to a specific format. The case study presented in this paper covers the system requirements, design and fabrication of the various lenses, exposure and color balancing of the lenses, alignment and tolerancing. An interesting aspect of this package printer project was that the client literally built everything in-house including the mechanical housing, film and paper transports, lamp houses, lenses, and coatings. A critical element of the design, fabrication, and assembly of these package printers was tolerancing. Since a large number of these package printers was to be manufactured for their inhouse, management needed assurance that the unit could be reasonably manufactured and would be reliable in the several plants around the world. The emphasis of this case study is on the challenge of producibility which required close attention to the capabilities of the various fabrication groups, assemblers, and technicians employed by the client. The project was successful and untold billions of photographs have been made by these package printers. Keywords: photographic printing, package printers, lens design, tolerancing, color balancing, lens fabrication.
1. INTRODUCTION A specialty area in the commercial photographic industry involves simultaneously producing a plurality of high-quality photographs of varying size and shape on a single sheet of photographic paper from a single photographic negative or digital image. The most common application of such printers is in the production of packages of photographs commonly purchased at schools, churches, community groups, and individuals. A variety of firms specialize in both taking the photographs and their production. The images are formed on a large piece of photographic paper by a set of lenses having specific magnifications and appropriately located between the negative and paper. Package printers are typically reconfigurable to allow different sets of images to be created; however, such reconfiguration is time consuming. Most often, a package printer is configured and then devoted to a specific format since the printer is used for high-volume production typically for multiple work shifts and five to six days per week (preventive maintenance is generally performed on the down day). Figure 1 illustrates a generic package printer configuration that utilizes a pair of identical lenses, denoted as A and B and mounted on a baffle plate. Each lens forms an image of the negative upon the photographic paper since the lenses are used off their respective optical axes. Unlike typical enlargers, package printers have a fixed negative-to-paper distance so the image size is determined by the focal length of the lens. To print simultaneously say three images of different sizes, then each of the three lenses would have to have the correct focal length to achieve the desired magnifications and the lenses would be located at different distances between the negative and paper. The spatial location of the images on the paper is of course determined by the transverse positions of the lenses. Although not illustrated in Figure1, care must ∗
[email protected] Optical System Alignment, Tolerancing, and Verification IV, edited by José Sasián, Richard N. Youngworth, Proc. of SPIE Vol. 7793, 779303 · © 2010 SPIE CCC code: 0277-786X/10/$18 · doi: 10.1117/12.863833 Proc. of SPIE Vol. 7793 779303-1 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
be take in laying out the format of the images when different magnifications are used so that none of the lenses mounted on the baffle structure vignette the light entering or leaving any other lens.
Figure 1. Generic package printer configuration with the light house/source not shown (no scale).
The case study presented in this paper discusses the author’s consultation with a client company some years ago in an enhancement program for their package printers. The client produced vast quantities of photographic packages that were sold primarily in the USA, but also in a variety of other countries. The company was highly vertically organized, designed and built most of their own equipment, and had their own marketing, sales, and photographers. In particular, they built their own customized cameras utilizing off-the-shelf lenses and their own package printers. The number of cameras and package printers they produced and maintained was quite large. The client maintained a significant inhouse machine shop and optical shop capability for their own use, i.e., no external work was performed. It became evident to the client that to remain competitive and a leader in their marketplace, enhancement of the optics for their package printers was necessary for particularly the medium-to-large format photographs (5” by 7”, 8” by 10”, and 11” by 14”). Their internal management and engineering quandary regarded whether to purchase off-the-shelf lenses or design and build their own lenses. Their investigation to locate off-the-shelf lenses that would meet their requirements was basically unsuccessful as was attempting to have manufacturers customize existing lenses. What is unusual about package printers is that the magnification and F-number of the optics need to be tweaked somewhat depending upon the application. For some years the client had been manufacturing their own lenses for the smaller format prints such as wallet and 3” by 5”. These lenses utilized the classic Dialyte configuration illustrated in Figure 2. For their purposes, this lens form produced acceptable results. The client heuristically learned to tweak their basic design by adjusting the air spaces as a means of controlling the magnification and the image quality. One of the first consultation tasks was to analyze these lenses and formally establish the useful range of magnification and Fnumber, and the corresponding lens spacings for magnification adjustment. It was also determined that the existing lenses could meet their small-format and quality requirements and, consequently, no redesign of these lenses would be necessary.
41
J
Figure 2. Basic air-spaced Dialyte lens1 configuration used for the wallet and 3” by 5” format lenses.
Proc. of SPIE Vol. 7793 779303-2 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
With regards to the medium and large format lenses, reasons for the client to choose to design and build their own lenses were as follows: • • • • • • •
• • •
Increases corporate control over printer manufacture Maintain optical enhancements as trade secrets Lenses and assemblies manufactured in client’s optics and machine shops Increased utilization of corporate capital investment Long-term cost effectiveness Increased contributed value in printer manufacture since a greater fraction of manufacturing costs are expended internally Realize maximum optical performance − Resolution − Color/density balance − Magnification Improved print quality Maximum number of common lens elements between different lenses In-depth understanding of the functional character of each designed lens assembly.
The case study presented in this paper covers the system requirements, design and fabrication of the various lenses, exposure and color balancing of the lenses, alignment, and tolerancing. The lenses described in this paper are the result of an intensive design effort that afforded the client with an improved cadre of enlargement lenses that can be manufactured in the client’s optics shop. Significant contributions to this effort were made by client’s staff through numerous technical interchange discussions. Close coordination of the technical effort allowed realistic trades to be made between the numerous optical and mechanical parameters throughout the design effort. The final result of this study was a set of six lenses that were designed and optimized specifically to meet the client’s unique printer needs.
2. REQUIREMENTS A number of requirements and objectives for the design of the client’s enlargement lenses were established. These items were as follows: • • • • • • • • • • • •
Specific magnification for each lens (values given in Table 2-1) Fixed negative-to-paper distance Minimum color difference between lenses Minimum number of elements in each lens Minimum number of different element curvatures for each lens Maximum number of common inter- and intra-lens element curvatures Excellent resolution of the entire image Less than 1% distortion over the image field (goal) Match lenses to spectral response of paper (rather than to the eye response) Maximize optical speed of each lens consistent with cluster lenses formats and mechanical considerations Determine manufacturing tolerances/sensitivities for each lens. Keep requirements within the capabilities of the client’s machine and optical shops.
The negative-to-paper distance was agreed upon with the client to be established at a nominal 847 mm, thereby allowing a reasonable degree of compatibility with their existing 5x7 and 8x10 Printers. It was determined that six lenses needed to be designed to meet the needs of the three medium-to large formats and two negative sizes used. Table 1 states the desired magnifications and focal lengths. In the design process, the exact focal length was not set as a hard objective to achieve but rather the design goal was based on achieving the specified magnification. The reason for this was to produce prints of precise sizes. The client’s cameras used 120 and 70 mm film and included reticle markings to indicate the size of various formats. The nominal 120 negative mask was 1.68” by 2.425” and the 70 mm negative mask was 1.95” by 2.823”. The masks for the small formats were 1.5” by 2.131” and 1.5” by 2.190”. Being that their print
Proc. of SPIE Vol. 7793 779303-3 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
formats were already market-proven, it was important to consider the 3D spatial positions of the various lenses forming the lens clusters to produce the desired image positions, image quality, color balance, and to have no mechanical interference or vignetting of one lens by another. Table 1 Magnifications and focal lengths for medium-to-large format printer lenses. NEGATIVE 70 mm 120 70 mm 120 70 mm 120
PRINT SIZE 5x7 5x7 8x10 8x10 11x14 11x14
MAGNIFICATION 2.41 2.80 3.47 4.04 4.91 5.71
FOCAL LENGTH 175.6 mm 164.3 mm 147.1 mm 134.8 mm 119.1 mm 107.5 mm
A question of importance pertains to how “good” does each lens need to be to achieve the client’s goals. The penalty of requiring a lens to be better than needed can be high. Such factors as manufacturing costs, lens complexity, and maintainability will be seen, in general, to increase. Therefore, the specific use of each lens should be considered. The driving factor is that the print will be viewed by a human observer at a closest viewing distance of 250 mm. How good can the human eye see? The answer to this question is very complex since numerous parameters are involved. However, at the time of this investigation, a rational solution could be taken from the work of Campbell and Green2 in which they present contrast sensitivity as a function of spatial frequency. From their work, one can determine that the limit for contrast sensitivity is about 8 line pairs per mm on the print with 6 lp/mm being more typical under normal viewing conditions. The optical aperture or entrance pupil radius for each lens is a function of its use in a specific lens-cluster format. This fact prohibits the specification of a unique optical aperture; hence, a formula was given on each element drawing for every lens to provide a guide to the selection of the proper element diameters as a function of entrance pupil radius and field angle. In other words, the same basic lens for say the 5x7 format may have several realizations, i.e., different barrel diameters and aperture sizes depending upon where the lens is used in a given lens cluster. On each lens assembly drawing, a formula was given that relates the entrance pupil radius and aperture stop diameter.
3. DESIGN Once the design requirements were established, several generic lens types were examined as a solution, each being more complex. Since the client was intending to make the new lenses internally, a visit was made to the client’s machine and optics shops to see first-hand what their capabilities were, both human and equipment. This allowed me to gain confidence in understanding their experience working various glasses, coating lenses, holding tolerances, etc. One item of equipment was identified for purchase that would enhance and speed up their test and alignment activities. The selection of proper glasses for the client’s lenses goes hand-in-hand with the selection of the generic lens form. The generic type that appeared to meet the requirements, shown in Figure 3, was found to be of the Double Gauss3 or Biotar4 class. After reviewing candidate optical materials that the client had both working and coating experience, four glasses were found to be compatible. It should be noted that although a number of glasses had the desired refractive properties, many excessively absorbed blue light which would aggravate the desired color balance. The glasses selected were LaK N9, SF 1, SF 4, and SK 16, all manufactured by Schott. The generic Double-Gauss lens shown in Figure 3 basically comprises a positive-negative meniscus lens pair, a stop, and a negative-positive meniscus lens pair. There are a number of variations of this lens form, primarily where one or both of the negative lenses are made into either air-spaced or cemented doublets or triplets which can significantly improve the optical performance of the lens and increase the cost to build. The client had determined that the F-number of the lenses needed to be in the range of F/8 to F/22, and typically close to F/8 in order to maximize print production throughput. Also, the client and I determined the field-of-views of the lenses when placed into the numerous cluster formats they used. From the data developed, the maximum field-of-view for each lens was established. The
Proc. of SPIE Vol. 7793 779303-4 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
magnifications for these lenses, shown in Table 1, are modest and nicely suited for the Double-Gauss lens due to its quasi-symmetry. No vignetting was to be allowed; however, because the client was building the optics in-house and the variety of lenses was reasonably small, an acceptable design consideration was to change the diameter of the same type lens to avoid mechanical interference or mitigate vignetting.
Figure 3. Generic Double-Gauss3 or Biotar4 configuration considered for the medium-to-large format lenses.
The Double-Gauss lens has been found by numerous lens designers to be the solution to their design problems often by increasing the complexity of the configuration. In this case, I found that the minimum number of lens elements needed to achieve the desired requirements was five with two of the elements forming a cemented doublet. This doublet replaced the right-hand side negative element shown in Figure 3. Often one finds that a doublet is needed to be used for the left-hand side negative lens, but fortunately for the client this wasn’t the case. The lenses were design using a lens design program I wrote that used both multiplicative damped least squares and orthonormal optimization. The aberrations used in the correction process were based upon real rays and are described in Chapter 4 of Ref. 1. The 11x14 – 120 lens was designed first since it was deemed to likely be the most difficult to design. The remaining lenses were expected to be a variation of this “base” lens which was found to be true. As is typical of the design of this lens type, the choice of glass is very important. Using the aforementioned glasses resulted in achieving excellent color correction with a practical minimum of secondary chromatic aberration that would meet the program objectives. The final configurations for the six lenses are shown in Figures 4–9. In an effort to minimize the manufacturing costs, a significant effort was made to maximize the number of surfaces that have the same curvature as another surface within this set of six lenses. Of the ten surfaces required for each lens, only six different master test plates were needed. Also, it should be noted that in several cases the same surface curvatures were used in several lenses which further reduced costs. The client had learned through experience that customers would complain if the several photographic prints on a sheet appeared to have been exposed differently. The exposure tolerance was stated to not exceed 3% between prints on the same sheet. Consequently, it was necessary to determine the transmittance of each lens in each of its planned positions. To accomplish this required adjusting the aperture stop diameter of each lens at its various cluster positions in order to modify their numerical apertures such that an exposure balance was obtained. The transmittance computation accounted for the thickness of the glasses, spectral transmittance of the glasses, obliquity of the beam entering the lens, and the coatings. A similar analysis was made for the smaller format lenses. The assembly drawing for each lens had equations for each element’s clear aperture, the stop diameter, and barrel dimensions as a function of the lens position in a given cluster. Another critical parameter that the client had learned through experience is that customers are often sensitive to distortion in the prints. The client had made a number of field tests before embarking upon this enhancement effort and found that the distortion must be less than ±1%. As can be seen from Table 2, all designs meet this requirement. The field angles shown for each lens are normalized fractional values of 0.4, 0.7, and 1.0. The full field angle is reference to the radius of the circumscribed circle containing all of the print formats potentially formed by a lens cluster upon a photograph sheet of paper. Examination of Table 2 reveals that the distortion contains high-order contributions. This was intentionally done to achieve the appearance of low distortion over the portion used of the total field-of-view of the 5” by 7” and the 8” by 10” lenses. The 11” by 14” (120 negative) lens has interesting distortion in that the distortion
Proc. of SPIE Vol. 7793 779303-5 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
becomes slightly negative (≈ 0.5%) by the fractional field angle of 0.4 and remains at this value over the rest of the fieldof-view.
'
\
/S
('ft (k'(
L\\\ t
1
--
i\
I
i71
i1I
/k
'\'
..-.
5.00±.05 __-.. 22.9±1 ___.4
'/7{
-.1
12,25±,05
2.51±.05
Figure 4. Lens for printing 5” x 7” images from 70 mm negatives.
I 933*
-.
5.67±03
-
8.1±1
I.-
I..-
.05
-1k--
.20±.05
Figure 5. Lens for printing 5” x 7” image from 120 negatives.
\:f\
(Y
.
r \\
'
'
\ \
iI
U
L
.Q±.nl
12.0
Figure 6. Lens for printing 8” x 10” images from 70 mm negatives.
Figure 7. Lens for printing 8” x 10” images from 120 negatives.
STOP9
/ I \'
k
'
" '1
(
"\-
18.5±.i
\:....
I
'
\\ '
1;.
I1iNY
\± .6o±.02
kk\
.55±.02 -1.- L.48 ±.10
4.50 -
I
15.1t.1
18 . 66
Figure 8. Lens for printing 11” x 14” images from 70 mm negatives.
05
11.81±05
Figure 9. Lens for printing 11” x 14” images from 120 negatives.
Although it will not be discussed in detail, an important subsystem in the package printers is the light house. The client had spent considerable effort in developing their custom light house that included both monitoring and control of the output flux and the color temperature. The uniformity of illumination of the entrance pupils of the various lenses was quite good. A part of their routine maintenance of the printers was to measure the output spectrum of the light house. If the lamp aged too much, its spectrum would shift out of acceptable boundaries, then the lamp would be replaced and the light house realigned. As with other parameters, the client had learned through experience of customers’ sensitivity to not only having excellent image resolution, but good perceived realism of color in the prints. The subject of color balance is very important in cluster lenses configurations and multiple format printers. Color balance between lenses can
Proc. of SPIE Vol. 7793 779303-6 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
be maintained to an adequate level if the same glasses (also same relative thicknesses) are used and if the same coatings are applied to the elements. This was successfully demonstrated in client’s small-format lenses. Analysis of the largerformat lenses indicated that as long as the same coating is applied to each like element of the six new lens designs, adequate color balance should be realized between these lenses. However, an additional analysis indicated that if the large-format lenses are used in conjunction with the small-format lenses in a cluster lens format, inadequate color balance will likely result unless color compensation filters are utilized. The nominal color contributions of the lenses were computed in a manner compatible with the American National Standard ANSI PH3.37-1969. There are two approaches to achieve the desired color balance, viz., provide the smaller-format lenses with color compensation filters or provide the larger-format lenses with color compensation filters. In general, it would preferable to put filters upon the larger-format lenses since there are fewer of them in a given cluster. However, further analysis indicated that a superior color balance could be realized if the filters were placed upon the smaller-format lenses. Placing the filters on the largerformat lenses would result in a slight residual color imbalance that a skilled observer could detect. The exact color compensation filters were determined experimentally, but were very close to the analytically computed filter values. Table 2 Distortion of medium-to-large format lenses. LENS 5x7-70
5x7-120
8x10-70
8x10-120
11x14-70
11x14-120
FIELD ANGLE 0.4 0.7 1.0 0.4 0.7 1.0 0.4 0.7 1.0 0.4 0.7 1.0 0.4 0.7 1.0 0.4 0.7 1.0
DISTORTION 0.41% -0.15 -0.70 0.84 0.44 0.10 0.91 0.55 0.00 -0.63 0.25 -0.29 0.07 0.06 -0.27 -0.48 -0.45 -0.53
Manufacturing tolerances are always a concern to the designer, particularly with a minimalist approach is being followed, i.e., in this case attempting to minimize (1) the number of lens elements needed, (2) the number of surface curvatures and test plates, and (3) number of cemented surfaces. The client had a rather well-equipped machine shop with skilled machinists. The mechanical spacing tolerance of the elements determined by the optical design software was compatible with the machine shop’s capability. In general, the shop could hold better than 0.003 mm. Examination of Figures 4–9 shows that the tightest inter-element spacing tolerance is 0.01 mm. Accounting for tolerance buildup in the element spacers, the shop was quite capable of making the parts which they indeed did. The capability of the optics shop was also found to be adequate. The radius tolerance of most of the surfaces was about 0.3–0.5% and 0.25–0.5 fringe irregularity. The most difficult lens to fabricate was L5 shown in Figure 6. The radius tolerances were 0.2% for the outer surface and 0.3% for the inner surface with 0.25 fringe irregularity. This 38 mm diameter lens has a thickness of 3 mm with a tolerance of 0.02 mm, which is quite precise. The client’s shop had to take significant care in fabricating this element. The other elements typically had a thickness tolerance of 0.03–0.05 mm which is still in the precision range, but quite achievable. The reason for the rather precise thickness values was to maintain the aberration correction. Had the negative meniscus lens been also made a doublet, many of the tolerances could have been relaxed somewhat. However, the client elected to minimize the number of elements as previously mentioned and suffer the tighter tolerances based upon their confidence in their optical shop’s skill and the anticipation of overall lower unit costs.
Proc. of SPIE Vol. 7793 779303-7 Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
4. SUMMARY The client was very pleased with the outcome of the project. Perhaps the most important aspects of this being a successful project were obtaining a clear and concise statement of the needs of the client, having close coordination with the client’s engineering and shops’ staffs, and having a good understanding of the machine and optical shops’ capabilities. Two questions often asked are (1) how much did the client pay for the consultation and (2) how long did it take to do the job? The answer to the first question is more than they wanted and less than what the services were truly worth to them. The answer to the second question is that the project was done in two parts. I would not take the full engagement until a preliminary feasibility investigation was done. This activity took about 50 hours spread over about two months and included a trip to the client’s facility. The reason for taking a couple of months was the client began to realize that they didn’t fully know exactly what enhancement requirements were needed to meet the corporate objectives. Once these requirements were “cast in stone,” the feasibility of meeting the requirements was investigated. Once I was confident of a high probability of success, a proposal was prepared and submitted for 400 hours of consultation time spread over about 10 months. The lens design and drawing preparation were performed over a period of about a month and was relatively simple to estimate, but the time over which interactions (including another site visit) with the client would occur was generally unpredictable. For example, acquiring the necessary glasses took a few months longer than originally expected. Fabrication of the mechanical and optical parts went well. It should be mentioned that a lot of effort was expended by the client’s staff in redesigning the several package printers to accept the new clutters of lenses. The client built hundreds of these printers and was able to maintain and extend their edge over their competition. After the immediate need for lenses was met, the client continued to have the optics shop fabricate additional sets of lenses of all format sizes. After about two years, their management determined that they had enough lenses in storage to last them for the next 20 years. Since management determined there was no need for the optics shop any longer, they shut it down and dismissed the staff. The client’s management had an unexpected issue arise as the enhanced printers were brought into service in their various production facilities. This issue was that the quality of the prints made using the new printers was demonstrably better than those made by the older printers. Fortunately the production in each facility was organized by geographic regions so that each region had specific printers dedicated to it. Marketing staffs from regions using the older printers complained, but little could be done to placate them since it took many months to upgrade the printers. Over the years, these package printers have produced billions of photographs and made millions of people very happy.
REFERENCES [1] Kingslake, Rudolf and Johnson, R. Barry, [Lens Design Fundamentals, Second Edition], Academic Press, Burlington, MA, and SPIE Press, Bellingham, WA, 355–363 (2010). [2] Campbell, F. W. and Green, D. G., “Optical and retinal factors affecting visual resolution,” J. Physiol., 181, 576–593, http://jp.physoc.org/content/181/3/576.full.pdf, (1965) [3] Ref. 1, 363–377. [4] Smith, Warren J., [Modern Lens Design], McGraw-Hill, New York, 319–353 (2005).
Proc. of SPIE Vol. 7793 779303-8 View publication stats
Downloaded from SPIE Digital Library on 21 Jul 2011 to 76.29.230.54. Terms of Use: http://spiedl.org/terms
