Information Storage Methods


Any large project requires that it be broken down into smaller parts. The methods I employed are outlined here and were continuously developed to assist with storage and distribution of the material acquired. They evolved over 25 years and are presented in hindsight as clear cut decisions. In reality, options were tried and discarded until methods proved worthwhile.


Research can be broken down into three areas:

Perhaps the most important points to remember are:

  1. Establish goal(s) and stick to the plan

    Provides focus and clear targets

  2. Reappraise goal(s) and progress annually

    The challenge of the activity can significantly enhance the ego of the researcher. This can have severe lifestyle consequences, e.g. 'obsession'

For every part in the research activity there must be a simple strategy. This is to ensure that all material and knowledge gained can be dealt with quickly. No back tracking or 'stuffing around'. The long term strategy I devised for this research was:


The basis for the photography was the N scale modeling interest. My modeling method was to apply scratch build bodies onto modified commercial underframes. After careful thought and watching others use randomly recorded detail photographs I evolved a photographic method that ensured sufficient photographs could be recorded with the minimum amount of time and effort. To establish this method two systems needed to be created:

The system for side and end identification needed to be simple enough to avoid crawling under a vehicle. Whilst many side and end identifiers were common, the application for photography seemed unique. I'm sure many people may feel this is an unnecessary step. Quite true. However there are many types of rolling stock that have no identifiers and there is value in a strategy that records the methods used.

For each type of different identification method a small 3"x5" card was filled out. The card contained

Cards are then used as reference. Vehicle groups associated with identification are clearly shown as 'Type xxx" where this would be 'Type PL1' from the example above.

As this is purely an arbitrary system, at some point the references to the descriptions must be given. This will assist in the conversion from your arbitrary system to other systems or the official naming convention used.

The number of photographs determined to be the absolute minimum were ten photos:

  1. [s1p] - perpendicular side 1
  2. [s1q] - three quarter side 1
  3. [s2p] - perpendicular side 2
  4. [s2q] - three quarter side 2
  5. [e1p] - perpendicular end 1
  6. [e1q] - three quarter end 1
  7. [e2p] - perpendicular end 2
  8. [e2q] - three quarter end 2
  9. Roof | Top | Interior
  10. Traffic photo, generally loaded and three quarter view

This method can be further evolved into an exact description of the subject in a photo. For example

By recording the perpendicular of side or end, then moving 3-4 paces left or right and taking another view, a 'flip' stereo pair is recorded. 'Flip' stereo is my term for two photos within 3-5 paces of each other of the same subject. Quick eye scans of both images can reveal detail not obtainable from either photo, e.g. detail depth.

The photographic system was further developed into three more sections:

The side and end identification system uses arbitrary features of different rolling stock and was generally determined thus

This is obviously the minimum and was admirably suited for 'N scale' where I tended to stop the detail photos when I had no intention of modeling small detail. Of course for modelers of larger scales, more specific photographs would be required. The establishment of a set photo pattern insures that given minimum time limits and perhaps the excitement of a new find, the basic details would not be missed.

As awareness and knowledge increased the number of additional 'value' detail photographs were recorded for overall interest:

From all these photographs it became easier to view rolling stock as having three 'layers'

  1. Main structure
  2. Components
  3. Lettering

With this in mind it was possible to photograph vehicles in service, yet model them for another era by knowing the component and lettering changes. Obviously original prototype photos were superior but in the absence of these or diagrams, at least a start can be made.

Random Selection

 To build up an information base for vehicles required more than detail photographs. At 1973 on VR, there were about 300 groups ( called 'classes') for about 23000 vehicles covering passenger, suburban, guards vans, freight and maintenance stock.

With increased awareness, more structural types were found and these were recorded as soon as possible. Within a short time it became apparent that random sampling by percentage might find more differences. A target 10% by class was set, plus any vehicles deemed immediately unusual.

Within nine months the target had been reappraised to capture every Victorian Railways vehicle in service. This was introduced to maximise the effort. In most cases I had to walk past vehicles to photograph others. Why not record the lot then? At this time some 6,000 four wheel vehicles were destined to the scrap heap within five years. A golden opportunity to record vehicles which had an average age of 60 years. This photography achieved the following:

To ensure that as many vehicles as possible were photographed it was decided to embark on field trips. These were intense forays to metropolitan and country rail yards on a regular basis. By persistently returning to the same locations on regular intervals, the cross section of available rolling stock to photograph was increased. Not only were some locations more viable than others but trip routing was selected on the basis of best lighting conditions for each location. Generally trips in the Northern areas were clockwise and trips to the west and south west were anti clockwise. This took advantage of sunlight and yard orientation during the day. Examples of locations and frequency are:

Trips were based on curiosity and spaced sufficiently far apart to ensure a fresh input of stock to photograph. Unusual activity and happenings (tests, loads) took precedence.

The photographic guidelines were:


Field book

When the photography escalated so did the logistical problems:

To assist this endeavour a 'Field book' was established. To quickly set up numbers a special rubber stamp was made. By stamping number blocks onto A5 pages, all number groups could be easily accommodated without worrying about gaps. The number stamp had the following layout:


00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99



00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

To represent 199 numbers in a class, the first block shows 01 - 99, the next block is prefixed by '1' to represent 100 - 199, etc. With this system, vehicles could be found within about 10 seconds.

The 'Field book' required restamping every two years due to wear and tear and rolling stock changes. With the start of computer operations in 1984 to assist the hobby, a 16K handheld computer was programmed to store 12,600 vehicles numbers. This was the fleet size at that time. Any vehicle could be accessed in 8 seconds.

Negative Indexing

My start in 35 mm photography began in 1965. Up until 1975, no indexing method was organised. This is fairly typical for those with small numbers of photos. Most items can be found quickly by a perusal of all the photos on file. My method to this time was to store the developed films as negatives and only print photos as required.

Negatives were stored as film strips in chronological order using a system created in 1965. Slides went through three numbering systems until 1980 when a fourth was implemented. The numbering methods are:

Note: Any numbering system requires three functions

  1. Ability to find the object based on the number
  2. Ability to replace the object accurately, to find its position within the order
  3. Ability to find stuff reasonably quickly without an index


With the sudden ramping of the rolling stock photography two index systems were implemented:

  1. Index for rolling stock (locomotives, vehicles)

    Pages were stamped with number blocks. Film numbers were written under the vehicle number to record the location of the negative. A quick view of the film would find the exact negative. This combined speed, convenience and the fact that few photos would actually be 'recalled' for use. As there was insufficient room for more than one photo, a supplementary system was devised. An 'X' placed within the vehicle/film number indicated direction to another "Duplicate Photo Index". This index comprised of 1" x 3" cards specially chopped. By 1985, some 17,000 vehicles had been photographed with some 5,000 duplicate cards. Access to any of these photos took about 60 seconds.

    Note: To date, all non railway photographs have not been indexed, say 25% of the collection

  2. Detail Photo Index

    To track the detail photographs and to establish a knowledge base, a Detail Photo Index was created. Quite simply this was a card file system that had a card with a vehicle photo glued to it. One side was the photo and the other contained the stamped columns that described the sides and ends. Negative numbers were added to the columns as films were indexed. There was room on the card for other interesting details that were recorded.

    As every film was developed and indexed, each vehicle was compared against the current photo cards on file. Any vehicle photographed that varied significantly by lettering or structural difference caused a new card to be created.

    This system ran for a brief five years from 1975 and some 1500 detail cards were created. The work was so labour intensive that by 1980 it was stopped. By this time attention was starting to focus on data research.


Data Collection

Very rarely does data collection just happen. It must be pursued. The type of commitment will ultimately provide the key to the amount of data. Time is the other factor. The more time available, the more chances of getting the right stuff.

To assist in the endeavours of data acquisition, the following points are worth remembering:

Data Storage

If research was simple all the information would come bundled in the correct order. Such is not the case. The variables are

So how is all this taken on board and stored? Great difficulty. The information storage path evolved from 1978:


Distribution of the data forms an important part of the research. The biggest hang-up to published distribution is the perceived lack of information. Some are reluctant to finalise and publish until all data is at hand. Ultimately, the reason for collecting is to dispense it. The various methods are:

Guidelines to behavior

Ever wondered why companies and organisations don't like photographers and info junkies? Here's a few

So the following "rules" apply if continued access is required


Digital Photography

In the ensuing years digital photography has gradually made inroads into the 35 mm scene. What has been outlined above still applies but must be adapted to suit the digital environment. 'Film' means digital storage and battery life.

Whilst there is some control over the size of the image ( i.e. pixel width and height ) the best option is to go to maximum resolution unless the close up details only warrant a lower setting.

Perhaps the biggest down side to digital is the inability to enlarge portions of the image sufficient to determine some detail. With 35 mm, it would be a fair comment to say that for most uses 35 mm photographs were recorded on a medium which recorded in a greater resolution than required. In some cases, years after the photograph was taken a small interesting portion of the image (not of interest at the time of the photo) could be enlarged to obtain some information.

With digital resolutions up to 2560 pixels wide/5 Mb chips this is not possible at the moment. So we effectively lose that ability. That is why I stress taking photos at the greatest resolution possible. If the main subject is known, fully framed and is in close proximity then a lower resolution is appropriate.


That's about it. I hope the outlines provided here can assist those make a start with their efforts. I have tried to include all the influences I have discovered as well as all the mistakes I've made over the years.

Good luck


Peter J. Vincent, March 2003