Identify the kind of reference

Write whether the item is inspiration, a physical cutting, an existing article or a labelled development direction. State what it can answer and what remains unknown. A photograph can communicate rhythm and surface but cannot establish every article property. A website direction can explain a preference while having no material or measured dimensions. An existing article still needs its current identity and applicable information retained with the project rather than assumed from an old label.

Give the physical sample a stable identity

Retain the supplied ID or assign a project reference that your team can use consistently. Keep the confirmed article code in a separate field when it becomes available. Use the physical ID on photographs, measurements and trial records. If you compare two cuttings, do not merge their most attractive features into one supposed article. Keep both identities and ask the textile team which actual candidate can support the combination of preferences.

Keep measurement evidence with dimensions

A ruler, flat physical measurement or documented method is needed when a repeat value becomes important. Record units, direction and the endpoints used for a complete repeating feature. A displayed photograph size is not the sample size. If you have no reliable measurement basis, leave the dimension open and state that the reference is for appearance. Ask the team for the actual article dimensions before using them in a pattern layout or required-metreage estimate.

Label face, back and orientation

Take or request clearly identified views of the intended garment face and any back view needed for the question. Record length direction and label a reversed or rotated comparison. The same physical sample should remain identifiable across close and larger views. Do not turn an arrow overlay into a proven nap property or alter the photograph to make its pattern or colour look like a different article. Keep the original view available for context.

Confirm sharing permission

Use images and designs that you can share for the intended discussion. Record the source and any limits on reuse. The local observation tool does not upload your image, and the website does not automatically attach it to a factory enquiry. You choose what to include in your own message. If rights or the reference's origin are unclear, explain the appearance in words and request an authorised sample rather than treating the image as freely reusable product photography.

Translate the reference into a garment question

State which feature you want to retain and where it matters in the garment. A close texture, larger-panel rhythm or pocket direction should lead to a sample task. Keep composition constraints and target dimensions in separate fields so the team can propose suitable articles without assuming unverified properties. Request the next physical sample and written information needed to resolve the question. Retain the response with the original reference and identify any change of article or stage.

Reconstruct which views actually belong together

A reference folder may contain a close texture, a worn coat and an old label without evidence that they describe the same cloth. Start by mapping those relationships rather than assembling them into one specification. Identify which view visibly belongs to which object and which connection comes only from someone's recollection. A matching background or filename can suggest a link, but it does not establish a common article. Keep an uncertain close crop separate from the garment view until its connection is confirmed. Ask the person who supplied the material for the original sequence, an uncropped view or an explanation of what they intended each image to show. If that information is unavailable, preserve the useful appearance observations without joining the identities. The close image might communicate surface detail while the coat communicates the desired scale of a finished panel. Those can be two design inputs, provided neither is presented as evidence of the other's composition, dimensions or supply. The next sample request should name the relationship still missing and say whether recovering it would change the choice. This prevents time being spent reconstructing an irrelevant filename while a genuinely important link, such as the identity of the proposed face, remains unresolved.

Reconcile conflicting photographs without inventing an average cloth

Suppose one image suggests a crisp, light pattern and another a darker, softer field, both described as the reference. Compare what the files actually contain before deciding that the cloth changed. One may show a folded sleeve under directional light, while the other is a broad flat cutting. Look for cropping, orientation, backing and garment curvature that could explain different presentations, but retain these as possible explanations when the capture history is unknown. Request a paired view of the same identifiable object if it can still be photographed. Do not recolour the files until they agree or choose the lighter photograph for colour and the sharper one for texture without stating the resulting combined brief. When the source cannot clarify the conflict, ask the design owner which observed feature should guide the next physical comparison. The answer can be a restrained overall impression from one view and a separate texture ambition from another. It cannot establish that an existing article possesses both. Keep the rejected interpretation alongside the chosen direction so a later reviewer does not restore it from an attractive old image. A reconstructed reference is useful when it makes those choices explicit, even if the original material remains unidentified.

Repair the scale evidence before it reaches a layout decision

An incomplete ruler crop often looks more authoritative than it is. The scale may have been photographed with another cutting, its endpoints may fall outside the crop, or the image may have been resized separately from an accompanying overview. Check whether the available files show the ruler and the relevant patterned surface together and whether their connection to the identified object is preserved. This is a check of evidence continuity, not a method for deriving dimensions from pixels. If that continuity is broken, withdraw the image from the dimensional part of the brief while retaining its appearance role. Ask for a physical cutting or a fresh measurement record that can be linked to the correct reference. Explain the decision requiring the information: a full front, a pocket relationship or an initial layout discussion may need different representative coverage. Do not borrow a scale from a neighbouring photograph or use an object of assumed standard size to bridge the gap. Until a reliable basis arrives, the sample maker can discuss visual priorities, but any dependent placement or quantity remains provisional. A later confirmed measurement belongs to its actual sample; it does not retrospectively prove the scale of every older reference image.

Turn a composite preference into alternatives that can be answered

When the only surviving evidence points in several directions, divide the request by the garment decisions it should support. Identify the feature that must be retained, the feature that can change and the feature for which the team has no reliable evidence. For example, the broad garment view may establish the designer's preferred visual rhythm while the close crop raises a question about a softer surface. Ask for identifiable candidates that explore that relationship, and state if a change of material or construction is acceptable. Do not require an exact copy of an unknown cloth while also requesting properties taken from an unrelated sample. If the candidates will answer different parts of the brief, keep those purposes distinct. A candidate selected for its panel appearance may still require a separate surface decision. Where no feasible proposal preserves every preference, return the conflict to the design owner rather than silently dropping the least convenient one. This is particularly important when a small reference looks attractive but does not represent the area of the proposed coat front. The resulting request should let the textile team explain a usable route or a remaining constraint, without having to guess which unverified feature the buyer regards as essential.

Establish a replacement benchmark when the original cannot be recovered

If the original cutting has been lost, decide whether the project is seeking that article or developing a new candidate guided by the surviving evidence. Those are different tasks. For new development, the design owner can select an authorised physical candidate as the benchmark for the next decision and describe exactly what it replaces. Keep the original image set as the history of the request, with its missing information still marked. The new cutting becomes the physical reference for its own observed properties, not a recovered identity for the lost material. Before using it in a larger trial, check that its available area and state represent the intended comparison; a small remnant may not show the required panel relationship. Tell the sample maker which decisions can now use the replacement and which still depend on article information or construction work. If the original later reappears, compare it as newly available evidence rather than overwriting the candidate's record. The project may keep the new direction, revisit a specific feature or discover that its earlier interpretation was wrong. Preserving both references makes that decision reviewable and prevents an apparent recovery from silently changing work already based on the replacement.

Read a constructed reference as an object with a history

An existing garment can be a valuable reference while giving a misleading impression of the cloth originally supplied. Before describing its fabric, identify the state you can actually inspect: a worn coat, an unfinished sample, a display garment or an object with an uncertain history. Look at the feature that prompted the enquiry in relation to its surroundings. A softly rolling collar, a flat supported front and an exposed sleeve may present the same nominal cloth differently. Do not combine those local effects into a claim about one unworked fabric finish.

Ask what is known about construction and subsequent handling. Relevant information could include an altered collar, replacement lining, local pressing or an earlier repair, but an unexplained appearance does not prove any of those events. Record an account as an account, with its source, rather than turning recollection into a confirmed processing history. A visible seam or attachment can identify where to ask the next question without establishing what lies underneath. Keep the reference intact unless its owner and the responsible maker have agreed an appropriate examination.

Compare accessible areas that can inform the question. A visible panel beside the collar may show the broader rhythm, while a turning shows the effect the buyer actually likes. Neither automatically represents untouched cloth. If a concealed allowance can be viewed without disturbing the object, identify that location and its limits; it may have different exposure and construction from the face. Do not declare it an original control merely because it was hidden. Request suitable physical material when the required observation cannot be made on the garment.

The resulting brief can preserve the finished effect without assigning it entirely to the fabric. For example, describe a collar appearance to investigate in a new assembly and separately request a cloth candidate for the broad field. If the buyer requires the exact old article, state that the constructed reference alone has not identified it. If the buyer instead wants a similar garment result, let the material and pattern teams propose an assessable combination. This distinction prevents an attractive worn detail from becoming an unsupported instruction to reproduce an unknown finishing process.

Check whether different areas really belong to one cloth reference

One garment does not necessarily give every visible area the same evidentiary role. A buyer may photograph a pocket, front and undercollar and assume that their differences reveal one fabric's face, back and direction. Before making that connection, trace the actual component boundaries. Establish which photographed areas are continuous material and which are separate attached pieces. A matching colour or shared garment label cannot prove that a replacement pocket, contrasting underside or repaired panel has the same article identity.

Use an overview to locate each close observation on the physical object. Include the surrounding join or boundary where it can be shown without damage. Ask whether the feature continues across the area being discussed or stops at a construction boundary. If an isolated patch has another surface, keep several explanations open: a different material, a different exposed side, a local working history or an incomplete view. The aim is to identify what additional evidence would separate them. Do not diagnose a repair or mislabelling solely because two fragments look different.

Where the owner knows a component was replaced, retain that fact beside the particular area rather than applying it to the entire garment. The original body and replacement can both contribute useful design observations, but the request should identify which contributes each feature. A new article that resembles the pocket is not thereby a recovered identity for the body. If the relationship is unknown, give the areas separate reference labels within the object until clarification is obtained.

A useful outcome can be narrower than complete identification. The team might confirm that the broad front is the intended appearance target while leaving the underside irrelevant to the proposed design. Alternatively, the underside may be essential to an intended reversible arrangement, making the missing material relationship consequential. Ask that question before spending effort on a hidden component that the new garment will never expose. When the original object cannot settle it, request an appropriate candidate and construction discussion. Keep the distinction between a reference assembled from several observed features and evidence that one actual cloth supplies them all.

Test whether a disputed feature belongs to the material or its representation

A small herringbone feature may appear different in a messaging preview, a presentation export and the original file. Treat that as a question about the representations before interpreting it as a textile change. Start with one identified source photograph and compare the versions that were actually circulated. Keep their crop boundaries and any known editing steps visible. If a coloured edge or extra line appears only in one version, do not trace it into the fabric specification as though the physical reference had acquired another yarn.

Ask the sender for the original available file and an overview of the same object. Compare whether the disputed feature can be located at the same place, rather than trying to make every displayed image the same size. Moving between a preview and the source can help identify an inconsistency, but screen zoom still supplies no physical dimension. A feature that changes with display or export is a reason to seek a clearer material view. It is not enough to identify a particular camera fault, prove image manipulation or establish the underlying construction.

For a hypothetical example, a close crop might show a thin bright boundary that the broad view does not resolve. A fresh linked photograph and inspection of the actual cutting could show whether that boundary is present on the cloth. If physical access is unavailable, describe it as uncertain and avoid making it the defining requirement. The buyer can still choose the overall impression visible in the reliable view, or ask for a candidate with a deliberately described feature. Those are new design choices, not confirmation of what the disputed file originally depicted.

Preserve an unannotated version when adding explanatory marks. A drawn junction, contrast enhancement or sharpened edge can help someone find the question but should remain distinguishable from the source observation. If the marked version has already circulated without that distinction, issue the original and the explanation together. Simply replacing the picture leaves recipients unsure whether the target changed. The next sample request should identify the material feature to investigate, not demand reproduction of an artifact that nobody has confirmed on the object.

Use a loaned reference before access disappears

A reference may remain available only for a meeting or a temporary loan. Before returning it, identify which consequential questions require direct access and which can later be discussed from photographs. Surface feel, the relationship of a rolled detail and the identity of an obscured component may not be recoverable from an attractive overview. Ask the relevant reviewer to name the observation they need. This is not a reason to dismantle the object or extend permission beyond what its owner allowed.

Prepare a short route through the object around those questions. Locate the important face, component boundaries and accessible labels, then connect the close views to the overview. If the object cannot be laid flat, state that limit instead of arranging it forcefully for a measurement photograph. If a component cannot be exposed without alteration, record the unanswered question and ask whether a separate reference or explanation is available. A blank dimension with a clear reason is more useful than an apparently precise value taken from a curved or inaccessible area.

Before the meeting ends, have participants point to the feature behind their descriptive words. Soft, close and clean may otherwise refer to different things in the later enquiry. Confirm whether they selected an overall panel impression, a local surface or a constructed relationship. Record an alternative interpretation if it would materially change the next request. This resolves a reference-language problem while the object is available, rather than asking the supplier to infer the intended feature from several incompatible notes.

After return, retain the object's role without claiming continued physical control of it. List which observations were made directly and which questions still need a new sample. If a later candidate is preferred, identify it as that candidate; do not relabel it as a retained part of the loan. A future comparison with the original may need renewed access. Where that is impractical, the design owner can make a deliberate next decision on the available evidence. The missing access should remain visible, particularly if the next person is being asked to judge feel or construction from images alone.

Your decision and record

Herringbone Reference Diagnosis: Old Garments, Repairs and Image Artifacts
CheckDecideKeep with the reference
RoleDistinguish inspiration, direction and article evidence.Reference type and intended use.
IdentityUse one stable physical ID across records.Sample label and confirmed code if available.
DimensionsKeep real values connected to a method.Ruler, units, direction and endpoints.
ViewsLabel face, back and any orientation change.Close and larger files with the same ID.
RightsShare material within its permitted use.Source, permission and limits.

Questions for this step

Can an image be the only article specification?

It can describe appearance, while composition, dimensions and supply need identified article information. Keep the image's limits explicit.

Should I assign a code to an inspiration?

A project reference can help organise it, but it should not be presented as a real factory article code. Keep those identities distinct.

What if I cannot share the photograph?

Describe the visual priorities in words and ask for an authorised reference or sample. Do not invent permission or publish it as product photography.

Do two crops with the same filename identify one physical sample?

Not by themselves. Files may have been duplicated, renamed or exported from different source images. Look for an original view or a reliable explanation linking the crops to the same object. If the connection cannot be established, keep separate project references and describe the appearance each contributes. A shared filename can remain a retrieval clue, but should not merge article identities or allow one crop's scale, face or colour evidence to be assigned to the other.

Can a button or hanger supply the missing scale?

Do not assume its physical size or that it shares the relevant image plane. Even a familiar-looking object may differ from the size imagined, and garment curvature can complicate its relationship to the cloth. Retain the photograph for visual discussion and request a linked physical measurement where dimensions affect the decision. The missing scale should remain visible as a limitation, rather than becoming a precise-looking repeat value that the next person cannot verify.

What if the missing original arrives after a candidate has been selected?

Give the recovered item its own confirmed identity and compare it with the selected candidate for the feature that matters. Preserve the existing decision and identify any proposal to change it. Recovery does not automatically cancel useful development work or establish that the candidate was an exact match. Ask the relevant design and technical owners whether the new evidence changes the next trial, an accepted appearance or only the historical explanation of the brief.

Can a hidden seam allowance establish the old fabric's original finish?

Not automatically. Identify the exact area and what is known about its construction, exposure and any previous work. A hidden allowance may help explain a difference, but it is not necessarily an untouched counterpart to the visible face. Avoid altering the only reference to create a comparison. If the original supplied state matters, ask for an identified retained cutting or suitable article information, and keep the allowance observation limited to the part actually inspected.

Does a pattern changing between image previews prove the photograph was falsified?

No. Establish which files and representations were compared, retain the available original and seek a linked view of the actual material. A display-dependent difference does not identify its cause or prove what the cloth's construction must be. Keep the disputed detail out of confirmed specifications until it can be resolved. The image may remain useful for a broader appearance preference that does not depend on that uncertain feature.

Can several parts of a borrowed garment share one reference code?

A project code can identify the object, while separate area labels preserve the role of its pocket, body or underside. Sharing that object code must not imply that all components are the same cloth or have the same history. Link each observation to its location and state which relationships are confirmed. If a later request combines features from different components, call that a proposed development goal rather than an established specification of one article.