CDIP common mistakes
CDIP · Understand the traps and how to avoid them.
Treating resolution (dpi/ppi) as the sole determinant of scan quality, so the candidate recommends 600 dpi for every document type and ignores bit depth, colour mode, and compression.
Resolution is the most visible scanner setting and the easiest to memorise, so candidates anchor on it and forget that a 600 dpi 1-bit scan of a colour photograph loses more information than a 300 dpi 24-bit scan. The exam deliberately pairs resolution with bit depth and colour mode to test whether the candidate understands the whole capture chain.
What to do instead
For any capture scenario, state all four parameters together: sampling rate, bit depth, colour mode, and compression. Learn typical targets: 300 dpi bi-tonal for typed text with OCR, 300-400 dpi 24-bit for photographs and mixed content, 600 dpi bi-tonal for small print or poor originals, and 200-300 dpi for access copies. When a question changes only one parameter, ask what the other three are doing.
Choosing a compression or file format by habit (always PDF, always JPEG) instead of matching format to content type and preservation intent.
Candidates use PDF and JPEG daily and assume they are universal. The exam tests whether the candidate knows that JPEG is lossy and unsuitable for text or master images, that PDF/A is the archival profile, that TIFF is a common master format, and that OCR text belongs in a searchable layer rather than being baked into the image.
What to do instead
Build a small decision table: master images (TIFF or JPEG 2000, lossless), access and distribution (JPEG, PDF), archival documents (PDF/A), text with search (PDF with hidden text layer). For each, note one reason and one risk. When a question mentions long-term preservation or legal retention, check for PDF/A before selecting plain PDF.
Assuming metadata is captured automatically and completely by the scanner, so index fields, audit trails, and provenance are treated as optional extras.
Scanner software often does populate basic metadata, which creates the impression that metadata is free. The exam tests whether the candidate knows that descriptive metadata, capture device settings, operator identity, and date/time usually need deliberate configuration and validation.
What to do instead
Treat metadata as a capture requirement, not an afterthought: identify which fields are mandatory, which are auto-generated, and which need validation at ingest. Know the difference between descriptive, structural, administrative, and technical metadata, and be ready to say where each is created.
Designing workflows that omit exception handling, so the described process works only for clean, single-page, correctly oriented documents.
Textbook workflows are drawn as clean linear flows, and candidates reproduce that shape. Real capture includes misfeeds, skewed pages, duplicates, separator sheet errors, and rejected batches, and the exam probes what happens when a step fails.
What to do instead
For every workflow you describe, add the exception path: detection method, quarantine or hold queue, notification, correction step, and re-entry point. Practise describing a batch workflow end to end including what the operator sees when a document fails quality control.
Confusing scanning throughput with end-to-end process throughput, and quoting scanner speed as the answer to capacity questions.
Vendors quote pages per minute, so candidates equate scanner speed with system capacity. The exam tests whether the candidate accounts for preparation, indexing, quality control, rescanning, and system handoffs, which usually dominate total time.
What to do instead
When sizing a workflow, estimate each stage separately and identify the bottleneck. Use realistic assumptions: preparation and indexing often take longer per page than scanning, and rework rates of a few percent can consume a large share of capacity. State your assumptions explicitly in the answer.
Treating image quality control as a visual glance rather than a measurable check against defined criteria.
Quality control feels subjective, and candidates describe it as 'check the images look right'. The exam expects knowledge of specific defect classes and how they are detected and documented.
What to do instead
Learn the common defect categories (skew, folded corners, streaks, speckle, missing pages, wrong orientation, cut-off margins) and the idea of sampling plans and acceptance thresholds. Be able to say what a QC operator records and what triggers a batch to be reworked rather than accepted.
Mixing up output resolution, physical page size, and scaling, so a question about printing a scanned page at a given size produces a dpi figure that is off by the scale factor.
The relationship between pixel dimensions, physical dimensions, and effective dpi is arithmetic that candidates skip under time pressure. Scaling a page up by two in each direction halves the effective dpi, and that step is frequently forgotten.
What to do instead
Work the arithmetic every time: effective dpi equals pixel dimension divided by output size in inches. Practise a few conversions in both directions, and check whether the question asks for the file's stored dpi or the effective dpi at output size.
Answering distribution questions without considering the recipient's constraints, such as accessibility, file size limits, viewer compatibility, or the need for a fixed, non-editable rendition.
Candidates focus on producing a correct file and stop there. The exam frames distribution as delivering usable content to a specific audience, which brings in accessibility tagging, downsampling for web delivery, and format choice for print versus screen.
What to do instead
For each distribution scenario, name the audience, the delivery channel, and the constraint that drives format choice. Know that accessible PDF requires tagged structure, alternative text, and logical reading order, and that web delivery often means downsampled images and optimised file size.
Citing a standard by name or number without knowing what it actually constrains, for example assuming PDF/A governs scanning resolution or that any TIFF is archival.
Standards are memorised as labels because the numbers are distinctive. The exam tests the scope of each standard, and a label-only answer collapses when the question asks what a conformance level permits or requires.
What to do instead
For each standard you rely on, write one sentence on what it covers and one on what it does not. Distinguish archival format standards from imaging guidelines and from records management frameworks, and be careful with conformance levels and profiles.
Treating security and compliance as a single checkbox, so retention, access control, encryption, and audit logging are collapsed into one vague answer about 'keeping data safe'.
Security questions feel intuitive, and candidates give a general assurance rather than naming the control. The exam distinguishes controls by purpose: confidentiality, integrity, availability, and evidential accountability.
What to do instead
Separate the controls explicitly: authentication and authorisation for access, encryption in transit and at rest for confidentiality, hash or audit trail for integrity, retention and disposition rules for lifecycle, and logging for accountability. Match the control to the risk named in the question rather than listing all of them.