CR vs Digital Detector: Compare the Complete NDT Workflow

Define the terms before comparing quotations

Computed radiography, usually abbreviated CR, uses an imaging plate and a reader to produce a digital image. A digital detector array system acquires the image through an electronic detector without that separate plate-reading step. Suppliers sometimes use digital radiography as an umbrella term and sometimes use DR specifically for the detector-array approach. Write the intended meaning into the comparison so a proposal described as digital does not hide which acquisition workflow is being offered.

Carestream's CR white paper describes imaging plates, scanner, workstation and software, and display as system components. Its HPX-DR detector page describes a detector-array product. These sources establish the distinction; they do not determine which system meets a particular contract. This guide addresses purchasing and records. It does not specify exposure, radiation protection arrangements, technique approval or image acceptance criteria.

Start with the deliverable the customer will accept

Ask the technical owner to identify the required image record, interpretation environment, associated metadata and final report. Record the applicable procedure and customer requirements without substituting a product brochure for them. The purchase question is whether the proposed workflow can produce and preserve that deliverable under the intended conditions. A system can create an attractive image while leaving an unsupported gap in the required record or review process.

List the outputs separately: acquired image, retained original data, approved processing information where required, image identity, interpretation record and customer package. Assign an owner to decide each requirement. Do not assume that a flattened picture or PDF is an acceptable replacement for the required digital image dataset. If several customers require different packages, include representative examples in the trial. Otherwise, a successful demonstration for the easiest customer may conceal additional software or manual work needed for the others.

Map the CR plate and reader sequence

Create a process map that follows one imaging plate from its identity and intended assignment through the authorized acquisition process, reading, image association, review and return to its next permitted use. Keep technical handling instructions in the manufacturer's documentation and approved procedure. The procurement map identifies where information changes hands. It should show who can establish that the image read from a particular plate belongs to the correct asset and examination location.

Ask how the proposed workflow handles several plates arriving together, an unreadable identifier or a plate that must be held for investigation. The HPX-1 Plus page lists supported plate and cassette arrangements; those are configuration details to verify for the offered reader. Do not assume every size, plate type or handling accessory is included. Record the actual plate families, carriers, cassettes and supporting items in the quotation.

Map the detector and acquisition sequence

For the detector workflow, identify the detector, acquisition interface, workstation or tablet, connection method, power arrangements and software. Follow the record from the assigned asset identity to the captured image and reviewer handover. Removing the separate plate-reading stage does not remove the need to confirm identity, completeness or correct file storage. The trial should demonstrate those steps with the exact proposed equipment and access conditions.

Ask the supplier to distinguish detector functions from functions performed by optional software or interfaces. If a demonstration uses a different computer, connection or accessory from the quotation, record the difference and resolve it. A portable panel may depend on a particular controller or workstation arrangement. The buying decision should evaluate that complete arrangement. Do not treat a detector-only price as directly comparable with a CR package that already includes a workstation, reader software and review access.

Assess physical fit without relying on category stereotypes

Compare the actual offered imaging medium and detector arrangement against the representative task. Avoid assuming that all CR plates have the same flexibility or that every digital detector is rigid. The market contains different designs, and the relevant evidence is the manufacturer's documented scope for the quoted model. The technical owner should evaluate access and image requirements without improvising handling or bending beyond the product instructions.

Provide a representative geometry description and ask which parts of the requirement the supplier can demonstrate. Record any holders, protective items or fixtures required for that arrangement. A medium may fit physically while the complete supported configuration does not. Conversely, a packaging difference may make an option feasible that would be excluded by a broad category assumption. Preserve the demonstrated configuration and limitations so later purchasers do not substitute a different plate or detector based only on a familiar technology label.

Compare the complete elapsed workflow

Measure the time between a defined starting point and a reviewable, correctly identified record using a controlled demonstration. Include the preparation and handover steps relevant to the intended work. For CR, this may include moving plates to the reader and associating images with the job. For the detector package, it may include connection, transfer and workstation availability. The authorized technical team determines the demonstration process; the buyer records its scope and results.

Do not multiply a brochure's fastest image rate by annual workload and call the result productivity. Identify queues, shared equipment and review delays. If one reader serves several acquisition teams, examine how competing work is scheduled. If one detector workstation is needed both for acquisition and interpretation, examine that dependency too. Compare completed and accepted records per representative work period, with assumptions visible. This provides a defensible business case without presenting a controlled trial as a guarantee for every site.

Establish identity before testing image exchange

Use fictional jobs with deliberately similar identifiers during the software trial. For example, create asset HX-240 and asset HX-241, each with a similarly named examination location. Ask the supplier to demonstrate how images are assigned, reviewed and corrected if a user selects the wrong job. Keep this a nonproduction exercise. The purpose is to see whether the record remains understandable and whether an authorized correction can be distinguished from the original assignment.

For CR, include the association between the physical plate or cassette workflow and the resulting file. For a detector, include the acquisition session and detector identity where required. Neither technology makes labeling errors impossible. Record the information that survives export and the information retained only inside the acquisition database. If the final customer package omits a field the organization needs, identify a supported delivery method before acceptance rather than relying on a handwritten note to bridge the gap indefinitely.

Separate image processing from preservation

Ask what the software retains as acquired data and what is stored as a processed view, presentation state or exported image. Have the technical reviewer explain which forms are required for the intended record. Do not assume that a convenient screenshot contains the information needed for later interpretation. The demonstration should show the documented relationship between the original record, authorized processing and the report presented to the customer.

Request a test export that another reviewer opens on the intended receiving system. Check identity, orientation, annotations and the information required by the contract. An image that appears on screen is only the first step. The recipient should be able to explain what was received and how it relates to the accepted record. The separate DICONDE acceptance guide provides a structured exchange trial; a format claim alone does not prove every function works between the chosen software versions.

Plan around the different shared resources

Build a dependency list for each proposal. A CR reader may be a shared resource for many plates, while a detector package has its own acquisition and connection dependencies. Ask what happens operationally when each critical component is unavailable. This is a continuity planning question, not a prediction that one technology is inherently more reliable. Obtain current service arrangements, supported substitutes and the evidence needed before a replacement configuration can be used.

Compare the cost of maintaining the required service level. Include the relevant plate inventory, reader access, detector availability, review licenses and replacement arrangements. Avoid calling reusable imaging plates free consumables or assuming a detector has no maintenance and replacement costs. Ask for the supplier's documented care requirements and commercial terms for the actual products. The coordinator should preserve those terms separately from technical acceptance, because a favorable warranty does not establish that a substitute configuration meets the approved task.

Review the workstation and archive as part of both offers

Ask whether the quoted computer supports acquisition only or also the required interpretation and reporting functions. Identify display requirements through the applicable technical process. Check the proposed storage, backup and restore workflow with representative image files. Both CR and detector systems can create digital records that become inaccessible if software access, database recovery or archive ownership is left undefined.

Create a small archive recovery trial using approved test data. Restore the selected records to the intended review environment and confirm that their identity and required supporting information remain available. Record the versions used and any manual steps. If recovery depends on a vendor service, document that dependency and its commercial scope. Do not assume that copying a folder is equivalent to preserving a database-backed inspection record. The accepted archive should support the future review obligations established by the organization and its customers.

Use a fictional comparison to expose the real constraint

A hypothetical U.S. inspection company compares a CR package with a detector package for a mixed workload. The detector demonstration produces an image quickly, but the offered license permits only one workstation role at a time. The CR proposal includes separate review access, although the planned shared reader creates a queue when two teams return together. Neither observation establishes a universal winner. They identify different constraints that need to be priced and evaluated for this company's workload.

The team asks for a revised detector offer with the necessary review access and tests a revised CR scheduling assumption. It compares the complete cost and the time to an accepted customer package. The technical owner separately decides which tasks each configuration can support. The resulting purchase may include one technology, both or a narrower service scope. The useful outcome is an explicit decision based on actual dependencies, rather than choosing the equipment with the most impressive isolated acquisition demonstration.

Write a comparison worksheet that preserves evidence

Use one row per requirement with columns for intended task, required output, proposed configuration, supporting source, trial evidence, unresolved issue, responsible owner and disposition. Give CR and detector proposals separate response columns. Mark a feature as demonstrated only when the agreed evidence exists. A catalog statement, supplier promise and completed local trial are different evidence types; distinguish them rather than converting all three into the same check mark.

Include commercial assumptions alongside the relevant technical dependency. For example, place the review-license cost beside the requirement for simultaneous review, and the replacement arrangement beside continuity of service. This keeps hidden additions visible. Record exclusions plainly. A supplier may offer an excellent package for part of the workload without supporting every planned application. A qualified partial response is more useful than a broad yes that nobody can connect to a defined configuration or deliverable.

Close the purchase with an end-to-end acceptance record

Acceptance should follow a representative fictional job through the approved trial, image association, review, report, export and archive retrieval. Identify the exact equipment and software used. Assign any remaining technical or commercial conditions before the package is committed to production. The record should explain what the organization purchased, what was demonstrated and what remains outside the accepted scope. Retain the supplier documentation that supports those conclusions.

NDT Connect's equipment tracker can hold the reader, detector and associated assets by name, type, serial number, manufacturer, model, due date, status and notes. It does not acquire radiographs, read CR plates, interpret images or provide an image archive. Keep the imaging records and technical evidence in the selected controlled systems. Use the tracker for asset coordination and concise references to those records. The completed buying decision should leave a supported imaging workflow with a clear path to the required customer deliverable.

Distinguish replacement of film from expansion of capability

A company moving from film may need to preserve existing customer access while adding a digital workflow. Identify which customers accept digital originals, which require a particular exchange package and which require additional approval before the transition. Do not describe the purchase itself as customer authorization. The coordinator should keep a customer-deliverable matrix so the same acquisition choice is not assumed to satisfy every contract. This is especially useful when a company serves both established maintenance accounts and new manufacturing programs.

Keep historical film records and new digital records associated with the correct asset without implying that scanning an old film recreates the original acquisition dataset. If historical comparison is part of the intended service, demonstrate how the reviewer finds both records and understands their different origins. Ask what metadata must be entered manually and who verifies it. The equipment decision may need a separate historical-record viewing or digitization arrangement; that cost should not be hidden inside an assumed benefit of buying a CR reader or detector.

Record the transition boundary in the acceptance file: which jobs use the new workflow, which retain an existing process, and who approves later expansion. That boundary prevents an administrative rollout from silently becoming a technical authorization for an untested application.