To translate archival microfilm reel targets and frame indexes correctly, preserve the structure that allows a researcher to recover an original record: repository, microfilm publication, reel number, target card, frame or image position, original page or folio, date range and any filming notes. A translation of microfilm metadata is not simply a translation of old handwriting. It is an exact exercise in distinguishing different numbering systems, explaining how records were reproduced and keeping the evidentiary path intact.
The common failure is a perfectly natural sentence that points to the wrong item. A cataloguer may turn a roll number into a page number, a viewer image count into an original folio, a reduction ratio into a quantity, or an archival target into the first sentence of a historical document. Anyone translating archive finding aids, genealogy microfilm indexes, historical registers or digitized reels needs a source-first method that separates physical and digital identities. The same care is required whether the records describe people, land, courts, public administration or scientific observation.
Quick summary: capture the archive and collection first; type every identifier; distinguish publication, reel, target, exposure, folio and image-group numbers; preserve historical spelling and uncertainty; translate filming instructions separately from original content; carry technical ratios and date ranges accurately; and test the final citation by retrieving the exact same image.
Why microfilm translation is a record-navigation problem
Microfilm is a reproduction medium, not a synonym for the documents copied onto it. A reel may include a title target, introduction, technical test image, subdivisions marked by further targets, hundreds of historical pages, repeated exposures and a continuation notice. The archive’s own catalogue may have a publication series and roll-by-roll listing that leads to the right reel. The digitized version may add an entirely new numbering system that starts at the first image of the scan, including production pages. Each system helps a different step of retrieval; translating their labels as though they were interchangeable makes research unreliable.
For example, a source record might be described as a census page in a particular state and enumeration district, reproduced on a numbered NARA publication reel and shown at an image position in a modern viewer. The year and place help identify the historical event; the publication and reel locate the copy; the printed page or folio locates the entry inside the record; and the image number guides the user interface. A reader who is given only ‘page 142’ cannot tell which of those layers the number belongs to. Exact type labels are the difference between evidence and a guess.
The US National Archives’ microfilm preparation guidance explains that targets identify the records that follow, and that reduction ratios and image placements affect how material is filmed. Its census catalogue guidance demonstrates the importance of publication and roll numbers. FamilySearch separately distinguishes film numbers from modern Digital Image Group Numbers. These principles are useful internationally even though individual archives may have different codes, language conventions and filming arrangements. Always follow the source institution’s terminology where it defines one.
The twelve-stage method
1. Begin with the holding institution and source collection
Microfilm images are copies of records managed within a collecting institution’s finding and identification system. Different archives can use similar-looking roll numbers.
Write down the institution, collection or record group, publication identifier, reel or roll number and language of the original. Preserve the institution’s preferred citation form. A reel called 12 in one publication is not reel 12 in another. The prefix and archive provide identity, not decoration.
Before moving to the next reel, record a verification note: Compare all identifiers against the institutional catalogue instead of guessing from the photographed page. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
2. Distinguish a microfilm publication from the roll
NARA publications can comprise many individual rolls. Publication numbers, roll numbers and original archival series are three separate layers.
Make distinct ledger columns for publication ID, roll ID, series, unit title, inclusive dates and holding location. If a catalogue lists microfilm publication M653, roll 1007, do not turn 1007 into a document page or a new publication.
Before moving to the next reel, record a verification note: Test the translated reference by searching for the publication-and-roll combination. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
3. Recognize target cards as navigational text
Filming targets are inserted sheets that tell a reader what records follow or where one filmed group ends. They are not always the original document’s title page.
Label a photographed target as a filming target where its function is clear; translate the descriptor without treating it as authored historical testimony. A card saying END OF ROLL is a production notice, not a closing sentence from the original register.
Before moving to the next reel, record a verification note: Preserve both the location and the role of the target when transcribing the image sequence. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
4. Lock numbering and count systems
Roll, target, exposure, scanned image, folio, printed page, archival item and database result can each have a different number.
Store each number with a type label. Add a short note if the viewer numbers digital images rather than historical pages. A screenshot at image 438 of 1200 may reproduce printed page 387 or an unnumbered verso; none of those numbers should overwrite the others.
Before moving to the next reel, record a verification note: Return to the exact image using the cited identifier and capture which numbering layer resolved it. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
5. Handle filmed fronts and backs
A single leaf may create two distinct exposures, and historical books often use recto and verso references. A blank back may still be filmed.
Translate folio abbreviations consistently and distinguish page, leaf, recto and verso in the record. Folio 12v is not page 12, nor is it automatically the next numbered page in a digitized gallery.
Before moving to the next reel, record a verification note: Verify recto/verso against visible writing and original pagination before drawing conclusions. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
6. Understand reduction ratio as a technical measure
Reduction ratio compares original linear dimensions to the size of the film image. It describes capture scale and does not independently describe a document’s original paper size.
Keep ratios such as 12:1 as ratios; translate labels but do not convert them into percentages or confidence scores. A 12:1 reduction may be described as twelvefold; it is not an instruction to enlarge the digital file by exactly twelve without calibration.
Before moving to the next reel, record a verification note: Read the associated technical target and avoid unsupported dimensions. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
7. Separate frame layout and reading orientation
Single-page, double-page, portrait, landscape, cine and comic orientations affect how documents appear on film. Their notation may be defined by an institution or standard.
Transcribe the source layout code first, then provide a clear definition in the target language as needed. A two-page spread may occupy one exposure; translating one exposure as two images changes a retrieval count.
Before moving to the next reel, record a verification note: Compare layout symbols with the filming manual actually used. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
8. Preserve inclusive dates and uncertain dates
A target often lists earliest and latest dates of the material following it, while a particular entry may have another date or an uncertain year.
Keep range markers, circa indicators, incomplete dates and year conventions. Identify the calendar system if the archive specifies one. A roll labelled 1860–1865 is not evidence that every item belongs to all six years, nor that the subject lived throughout the period.
Before moving to the next reel, record a verification note: Distinguish roll-level dates from item-level dates in every citation. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
9. Retain original-language names before normalizing them
Person and place names in older documents can have variant spellings and scripts. Transliteration helps access but can also hide the original form.
Use transcription for what is visible, a separate transliteration if useful, and a clearly identified normalized search variant. A clerk’s ‘Johan’ should not silently become ‘John’ in the documentary transcript simply because an index uses that name.
Before moving to the next reel, record a verification note: Keep an image-level link so readers can evaluate the historical writing. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
10. Record defects without repairing the evidence
Microfilm can contain blurred exposures, skipped pages, repeated frames, density variation, camera shake or illegible text.
Record physical capture defects and uncertainty visibly. Refilming notes and missing targets should remain distinct from ordinary spelling uncertainty. If a line is unreadable, use an explicit illegibility note rather than inventing a plausible surname from adjacent entries.
Before moving to the next reel, record a verification note: A second reviewer should be able to distinguish unknown writing from unknown archive metadata. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
11. Distinguish a digitized film from native digital records
A digitized film roll can receive a digital image-group identifier and a viewer position unrelated to original frame counters.
Keep original film numbers alongside modern image group or DGS numbers where supplied. Do not call a DGS number a historical page number. FamilySearch’s help material distinguishes film numbers from Image Group Numbers; the fields should remain separate after translation.
Before moving to the next reel, record a verification note: Test the record in the current digital catalogue and specify the viewer or access route where numbers differ. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
12. Test citations by actual retrieval
A correct translation gives a researcher the ability to relocate the same image and verify the same statement.
Write a full reference with repository, collection or publication, roll, target or frame, item and stable image URL when available. If the translated citation points only to a general search homepage, the reader may never find the evidence again.
Before moving to the next reel, record a verification note: Give the citation to a second reader and see whether they reach the identical image without private instructions. This note should identify the observed image or catalogue record rather than a general impression that the translation feels correct. In archival work, reconstructability matters more than stylistic smoothness. If the original lacks a date, an identifier or a legible name, the translated description must not quietly manufacture one to make the index appear complete.
Twenty-four worked microfilm translation problems
All examples below are constructed to teach distinctions between archival layers. Publication and page values are illustrative unless a named source is cited, and they must never be used as real historical evidence without opening the relevant record. The translations below are English explanatory models that can be adapted to a specific target language after confirming that archive’s conventions.
Worked example 1: A target numbered 3
What appears: A photographed card reads ‘Target 3’ inside reel 27. The misleading interpretation: The translated index says ‘third original document’.
Reliable translation and explanation: Call it target number 3 on reel 27, and note that it identifies a filming subdivision. Target numbers refer to the film production’s navigation, not necessarily to original record numbers.
How to verify: Find the next filmed title or folder boundary to confirm its role. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 2: M and roll identifiers
What appears: A catalogue gives ‘M653, roll 1007’. The misleading interpretation: The translation reports page M653 and file 1007.
Reliable translation and explanation: Preserve publication M653 and roll 1007 as separate fields. A publication may be spread across many rolls, and the role of the prefix is part of retrieval.
How to verify: Attempt the lookup using both identifiers. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 3: Roll beginning marker
What appears: The first image is a title target followed by several unnumbered frames. The misleading interpretation: The translated description starts the content at image one.
Reliable translation and explanation: Describe the target as front matter and identify where the historical records begin. Digital image one is a capture position, not proof of original item one.
How to verify: Check image thumbnails and the next actual record. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 4: End-of-roll target
What appears: A card reads ‘End of Reel’. The misleading interpretation: The translation treats it as the last historical entry in the ledger.
Reliable translation and explanation: Label it a filming endpoint notice outside the historical transcript. It has a documentary function but does not represent testimony from the archive’s creator.
How to verify: Review its neighboring images and the catalogue’s roll listing. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 5: Continuation target
What appears: A volume continues on the next reel and the card states ‘Continued on Reel 14’. The misleading interpretation: The target summary says volume completed at reel 13.
Reliable translation and explanation: Preserve the continuation instruction and its exact destination. The reader may otherwise mistake a partial record series for a complete one.
How to verify: Retrieve the beginning of reel 14 to confirm the sequence. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 6: Original page 51
What appears: The page itself bears printed number 51 while the viewer says image 73. The misleading interpretation: The citation uses page 73.
Reliable translation and explanation: Use printed page 51 with a separate viewer image 73 reference. The two number systems serve different purposes and may change if images are inserted.
How to verify: Compare the visible original number with the interface counter. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 7: Folio 12 recto
What appears: An archival catalog lists ‘f. 12r’. The misleading interpretation: The translation calls it page 12.
Reliable translation and explanation: Expand as folio 12 recto where appropriate, keeping the original locator. A folio side differs from a numbered page and can be critical in manuscripts.
How to verify: Inspect whether the verso is a separate image. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 8: Folio 12 verso
What appears: An entry appears on the back of folio 12. The misleading interpretation: The translated index maps it to folio 13.
Reliable translation and explanation: Retain folio 12 verso or 12v, not the next leaf. This error moves the cited entry and may identify the wrong person.
How to verify: Check the adjacent recto and verso sequence. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 9: Double-page exposure
What appears: A camera frame shows two facing pages in one shot. The misleading interpretation: The description says there are two film frames.
Reliable translation and explanation: Explain one exposure containing a double-page spread. A frame is a capture unit, whereas the original document can contain multiple pages.
How to verify: Count exposures independently of visible pages. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 10: Blank page
What appears: The original volume contains a blank page deliberately filmed. The misleading interpretation: The translator deletes it and renumbers later images.
Reliable translation and explanation: Keep a brief indication of a blank image if the project preserves full sequence. Skipping an image may make all subsequent references appear off by one.
How to verify: Compare sequence counts before and after cleanup. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 11: Retake frame
What appears: The same sheet is filmed twice because the first exposure was blurred. The misleading interpretation: The translated catalogue treats it as two separate records.
Reliable translation and explanation: Mark the repeated exposure or retake without merging their metadata. Duplicate images do not necessarily mean duplicate archival entries.
How to verify: Compare handwritten content and original page marks. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 12: Illegible surname
What appears: A historical handwritten surname cannot be resolved confidently. The misleading interpretation: The translation chooses the most likely modern spelling without a note.
Reliable translation and explanation: Transcribe only readable characters and mark uncertainty or illegibility. An invented name can propagate into family trees and databases as false evidence.
How to verify: Retain a link to the image and invite independent review. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 13: Variant placename
What appears: A record gives an old village spelling different from today’s official name. The misleading interpretation: The translation substitutes the modern municipality and drops the original.
Reliable translation and explanation: Keep the original toponym and optionally add a modern equivalent clearly labeled as interpretation. A historical jurisdiction and a present administrative unit need not share boundaries.
How to verify: Check historical gazetteers before making equivalence claims. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 14: Date span
What appears: A reel target says ‘1891–1897’. The misleading interpretation: The translator labels every item as dated 1897.
Reliable translation and explanation: Describe the inclusive date range for the grouped material. The range indicates archival coverage, not an individual event date.
How to verify: Differentiate roll-level and document-level date fields. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 15: Circa date
What appears: A folder is marked ‘ca. 1905’. The misleading interpretation: The translation prints ‘1905’ as certain.
Reliable translation and explanation: Retain approximately or circa in the target-language date. Eliminating the uncertainty can create an invented chronological assertion.
How to verify: Check whether a second source verifies the exact year. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 16: Partial volume
What appears: A target says ‘Vol. 4, pages 200–350 only’. The misleading interpretation: The translation labels the reel as the whole of volume four.
Reliable translation and explanation: Keep the limited page range in the title and citation. A researcher searching for page 150 will otherwise be misdirected.
How to verify: Inspect reel continuation notes and catalogue coverage. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 17: Reduction 12:1
What appears: A technical target records reduction ’12:1′. The misleading interpretation: The translated description treats 12 as a frame or an original page count.
Reliable translation and explanation: Keep a reduction ratio and explain that it is a linear scale measure. Image size is not a direct record count or identifier.
How to verify: Read the camera or project technical notes. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 18: A quality-control chart
What appears: A test chart appears between title targets and filmed documents. The misleading interpretation: The translated item list indexes it as a historical photograph.
Reliable translation and explanation: Identify it as a filming or image-quality test target. It describes the reproduction process, not the archival collection subject.
How to verify: Check the target caption and chart structure. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 19: A folded map
What appears: An oversize map is photographed in several sections. The misleading interpretation: The translation says the archive holds several different maps.
Reliable translation and explanation: Explain that the exposures are sectional views of one original map, where supported. The overlap and order are necessary to reconstruct the object.
How to verify: Compare map headings, edges and filming guidance. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 20: Mixed scripts
What appears: A register uses two scripts for the same place name. The misleading interpretation: The translation retains only a romanized reading.
Reliable translation and explanation: Preserve the original forms when legible and label transliterations separately. Different scripts may encode distinctions useful for later historical research.
How to verify: Document the transliteration convention and retain image evidence. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 21: Reel versus box
What appears: A collection guide identifies box 9 and film roll 9. The misleading interpretation: The translated citation collapses both into ‘9’.
Reliable translation and explanation: Use typed references: original box 9; reproduction roll 9. Matching numerals do not imply the same physical or intellectual unit.
How to verify: Check the collection’s arrangement and filming correspondence. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 22: Digital image group
What appears: A FamilySearch resource has a film number and a DGS number. The misleading interpretation: The translation calls the DGS number a reel number.
Reliable translation and explanation: Label it as the digital image group identifier and keep the film number distinct. Digital grouping can reflect the scanning platform rather than the original film arrangement.
How to verify: Follow the current resource to see which identifier retrieves which group. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 23: Changing viewer positions
What appears: A platform adds a front target, shifting image display counters. The misleading interpretation: The translator silently changes the original folio references.
Reliable translation and explanation: Preserve original locators and supplement with the new viewer position when useful. A platform sequence can change while the original archive record remains the same.
How to verify: Keep a stable item reference and document the date of access. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Worked example 24: A citation without evidence
What appears: A translated article names a person but links only to a general archive home page. The misleading interpretation: The researcher assumes the statement has an easily verifiable source.
Reliable translation and explanation: Supply repository, publication, roll, target or image, and record locator when available. Without retrieval details, even an accurate transcription becomes difficult to check.
How to verify: Ask a colleague to independently recover the same photographed page. A successful translation lets another researcher repeat the retrieval and see the same recorded evidence. When the two readings point to different originals, keep the institution’s number, date or original-language term in its own field, and make any editorial explanation explicitly separate from the transcription. The goal is a usable document trail, not a story that sounds plausible.
Common failure modes, with repairs
Failure 1: Collapsing all numbers into page
Retain separate columns for roll, target, frame, original page, folio and digital image counter; never leave an unlabeled numeral in a citation. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 2: Ignoring the film publication prefix
Keep the archive’s publication identifier together with the reel number; otherwise unrelated reels can appear to be the same item. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 3: Confusing targets with original titles
Describe target cards as technical navigation when appropriate and do not attribute their wording to the creator of the historical register. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 4: Renumbering after removal of blanks
Do not delete images from a transcription sequence without documenting how the displayed sequence differs from the original copy. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 5: Inventing dates from range labels
A roll-level span is not an item-level event date; copy qualifiers and uncertainties into the translated record. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 6: Modernizing names without disclosure
Provide original script, a transliteration and a normalized variant in separate fields, where relevant and supported. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 7: Misreading reduced-scale notation
Preserve the numerical ratio and source’s unit explanation; do not infer sizes without the required calibration. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 8: Merging distinct photographs of one object
A double exposure, a retake and a multi-section foldout have different documentary meanings; preserve the capture relationship. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 9: Treating a digital image group as a physical reel
Record platform-specific image identifiers alongside, not instead of, the microfilm and source collection. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Failure 10: Publishing untestable citations
Use a citation someone else can act on: repository, series, publication, roll and page/folio or image reference. The repair is a two-reader reconstruction test. One reader should see the original and fill an identity ledger; the other receives only the translation and attempts to locate the same image. If they do not arrive at the same entry, compare the layers from the outer collection down to the frame, not from the person’s name outward. This isolates the point where the translation changed the retrieval path.
Practice and transfer exercises
Start with a deliberately artificial four-roll collection. Give each roll a publication label, a front target, an original page range, and several viewer image positions that include blank and technical images. Ask the learner to translate five index entries into another language. An answer is correct only if the colleague can retrieve the original record, not merely if the target sentence reads naturally. Score errors by layer: repository, publication, reel, target, original locator, digital position and content transcription. Keep an error log and repeat the exercise with a different arrangement.
Next use an example with two number systems: a scanned reel with a digital image-group ID and a source catalogue with a physical microfilm number. Explain which identifier belongs to the repository, which belongs to the digitization platform, and which tells the viewer where the image is displayed. Introduce a target card announcing the next subdivision, a retaken frame, a blank verso and an uncertain handwritten date. This set tests almost every error a naive translation workflow is likely to conceal.
For advanced transfer, compare two institutions with different conventions. Translate a roll target from one archive and a genealogical index record from another, then explain why their apparently identical labels cannot be standardized without a source-specific glossary. A competent translator should know when to use an international explanatory phrase and when to preserve an institutional code exactly. Their final citation should identify the image, describe any uncertainties and remain honest about material that cannot be read.
Frequently asked questions
What is a microfilm target?
A target is a filming sheet inserted to identify material, subdivisions or technical information. It should not automatically be transcribed as part of the historical document. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Are a reel and a microfilm publication the same thing?
No. One publication can contain multiple rolls or reels, and the publication prefix is part of the identifying reference. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Is a film frame always one page?
No. It can represent a single page, two facing pages, a foldout section, a target or another exposure arrangement. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Can I translate roll numbers into my own filing scheme?
Keep the source identifiers unchanged. A local crosswalk can supplement them, but should not replace the original locator. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
What does a reduction ratio mean?
It describes the relation between dimensions on the original object and its image on film. Do not read it as a page count or proof of resolution. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
What if the microfilm is blurred?
Mark the affected words as uncertain or illegible. Consult other copies where available, and do not invent missing names or figures. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Is image 50 always page 50?
No. Viewer positions, targets, blank images, retakes and original numbering can create different sequences. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
What is a DGS number?
FamilySearch uses an Image Group Number, also called a DGS number, to identify a set of digital images. It differs from a traditional physical film number. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Should I modernize old spelling?
Keep documentary transcription separate from normalized search forms. Readers should be able to see what the historical page actually records. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
How do I cite a digitized film?
Include the relevant institution or repository, collection/publication and reel identifiers, item locator and a stable direct image reference if the service provides one. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
What if the catalogue and target disagree?
Record the discrepancy rather than choosing the version that looks tidier. Ask the archive or compare its original arrangement guidance. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Can OCR replace manual transcription?
OCR can help locate obvious strings, but handwriting, faded images, old fonts, target cards and mixed languages require verification against images. When uncertain, keep the original identifier and explicitly mark the editorial explanation rather than replacing the original text. A useful quality test is always practical: can a second reader reach the same photographed page, in the same series, without being told a private shortcut? If not, the translation needs a stronger locator or a clearer explanation of its limits.
Sources and next learning routes
The US National Archives’ microfilm preparation guidance covers targets, orientation, exposures and reduction ratios. The NARA census microfilm catalogue instructions explain publication-and-roll retrieval. FamilySearch’s film-number explanation and DGS image group guide clarify physical and digital identifiers. These sources support the method but do not resolve an unreadable image on the translator’s behalf.
Review the original translation-method owners for meaning rather than isolated words, context and register, cultural and non-literal language, and glossaries and quality checks. Develop source vocabulary with the Vocabulary Learning Hub and practise precise English explanation via How English Works.