Describe Image

Lesson 2.3: Describe Image (DI) — Structural Delivery Templates, Visual Taxonomy & Acoustic Fluency Optimization

Describe Image (DI) is the third speaking task type in Section 1 of the PTE Academic, appearing 3 to 4 times. Unlike Read Aloud and Repeat Sentence, which test integrated skills (Reading and Listening), Describe Image contributes points solely to your Speaking score.

Candidates are presented with a graphic prompt—such as a bar chart, line graph, pie chart, table, process diagram, map, or composite infographic. You are given precisely 25 seconds to study the image, after which a short tone sounds, and you have up to 40 seconds to deliver a structured oral description.

Because human examiners do not evaluate this task, success does not depend on deep economic analysis or exhaustive data interpretation. Success depends entirely on feeding the Pearson acoustic and NLP scoring engines what they expect: unbroken oral fluency, phonological clarity, key topical terminology, and high/low data comparisons structured within a repeatable delivery template.

1. Task Mechanics & Interface Flow

The interface for Describe Image follows an automated countdown sequence:

Phase	Duration	System State & Candidate Action
1. Preparation	25 seconds	“Current Status: Beginning in X seconds…” Identify title, axes, highest/lowest units.
2. Audio Alert	~0.5 second	A short BEEP sounds at second 25.
3. Microphone Live	40 seconds	Status switches to “Recording”. Deliver template-driven oral summary.

Critical Operational Rules

Wait for the Tone: Unlike Read Aloud and Repeat Sentence, you MUST wait for the audio beep before speaking. Words uttered before the tone sounds are cut off by the recording buffer, costing you initial content points.
The 3-Second Cut-Off Law: If you hesitate for 3 seconds after the beep, the microphone shuts off permanently and awards zero marks for that prompt.
The Optimal Sweet-Spot Finish (28–35 Seconds): You do not need to speak for all 40 seconds. Aim to complete your structured template between 28 and 35 seconds, then click [Next] immediately. Speaking right up to the 40-second boundary risks getting cut off mid-sentence, which disrupts pitch intonation and damages oral fluency.

2. Scoring Mechanics: Content, Fluency & Pronunciation

Describe Image is scored on a maximum of 15 points per item across three traits:

Scoring Trait	Points	Algorithmic Evaluation Focus
1. Content	0 to 5 pts	Mentions title, axes/units, key trends, max/min values, and a concluding overview.
2. Oral Fluency	0 to 5 pts	Rhythm, continuity, steady pace (120–140 wpm). ZERO pauses, stumbles, or self-corrections.
3. Pronunciation	0 to 5 pts	Vowel/consonant accuracy, syllable stress, absence of mumbling or breath pops.

The Content Gating Mechanism

The scoring engine evaluates whether your acoustic stream contains key vocabulary extracted from the image:

Image title and axis labels (e.g., “unemployment rate”, “percentage”, “years from 2010 to 2020”).
Comparative extremes (highest, lowest, peaked, declined).
Category names (e.g., “manufacturing”, “agriculture”, “services”).

Crucial Engine Rule: If your Content score registers as 0 (e.g., remaining silent or speaking on a completely irrelevant topic), the algorithm automatically zeroes out your Fluency and Pronunciation scores as well. However, once you mention 3 to 4 relevant keywords from the screen, you pass the content threshold and unlock full points for Fluency and Pronunciation.

3. The 25-Second Preparation Protocol: The “4-Point Scan”

Do not attempt to read every number or calculate complex percentages during your 25-second preparation window. Execute the 4-Point Scan:

Step	Time Window	Action Protocol
Step 1: Read the Title & Units	Seconds 01–08	Read the headline/title aloud in a whisper to prime your articulation. Note the unit of measurement (percentages, metric tons, degrees Celsius).
Step 2: Identify Extremes	Seconds 09–18	Locate the absolute highest bar, peak, or segment. Note its label.
Step 3: Pick One Category / Trend	Seconds 19–25	Locate the absolute lowest bar, trough, or segment. Take a steady breath, look directly at your opening line, and wait for the beep.

4. The Universal High-Scoring Response Template

To maintain unbroken fluency at 120–140 words per minute across any visual without freezing, apply the Patel PTE Universal Image Template:

Template Section	Target Timing	Standard Delivery Phrasing
1. Introduction	0:00 – 0:08	“The given [chart type] provides detailed information about [Title / Topic], measured in [Units].”
2. Highest Value	0:08 – 0:16	“Looking at the significant trends, it is evident that the highest figure is observed in [Category A], reaching approximately [Value A].”
3. Lowest Value / Contrast	0:16 – 0:24	“Conversely, the lowest value is seen in [Category B], which stands at roughly [Value B], followed by moderate numbers in [Category C].”
4. Conclusion / Overview	0:24 – 0:32	“In conclusion, it can be clearly seen that [Category A] accounts for the dominant share throughout the period.”

5. Adapting the Template to Complex Visual Types

Type A: Standard Statistical Visuals (Bar Charts, Line Graphs, Pie Charts, Tables)

Strategy: Use the standard high/low template directly.
Demonstration (Bar Chart – Renewable Energy Production in 2024):
“The given bar chart provides detailed information about renewable energy production across various countries, measured in gigawatt hours. Looking at the significant trends, it is evident that the highest figure is observed in solar energy, reaching approximately 450 gigawatt hours. Conversely, the lowest value is seen in geothermal power, which stands at roughly 50 gigawatt hours, followed by moderate numbers in wind and hydroelectric power. In conclusion, it can be clearly seen that solar energy accounts for the dominant share throughout the recorded timeframe.”
(Time: ~30 seconds | Words: ~78 | Fluency: 5/5 | Content: 5/5)

Type B: Process Diagrams & Life Cycles

Adjustment: Replace “highest/lowest” with “initial step/final stage”.
Adapted Process Template:
“The given process diagram provides detailed information regarding the sequential stages of [Process Name / Topic]. In the initial stage, the process begins with [Step 1 / Input], which undergoes several transformative phases including [Step 2 / Step 3]. Following this, the material is processed through [Step 4] before reaching its terminal phase. In conclusion, it can be clearly seen that the entire operation culminates in the production of [Final Product / Output] through a cyclical framework.”

Type C: Geographical Maps (Past vs. Present or Comparison)

Adjustment: Replace “highest/lowest” with “directional expansions/modifications”.
Adapted Map Template:
“The given visual illustration provides a comparative overview of [Location Name] between [Year A] and [Year B]. Looking at the northern quadrant, it is evident that substantial infrastructural development occurred, featuring the addition of [Feature 1]. Conversely, the southern and eastern sectors witnessed the replacement of [Old Feature] with modern [New Feature / Green Space]. In conclusion, it can be clearly seen that the entire area underwent a comprehensive urban transformation over the specified duration.”

6. High-Frequency Traps in Describe Image

Trap Category	Candidate Mistake	What the AI Engine Expects
Speaking Before Tone	Talking during the preparation stage; the first sentence is never recorded.	Waiting for the electronic beep, then starting immediately.
Number Overload	Attempting to read 10 different exact statistics (“34.5%, 12.8%, 91.2%”), causing stumbles.	Stating 1 or 2 approximate figures (“approximately 40%”); prioritizing smooth speech over raw data density.
Self-Correction	Correcting a misread label (“in France, sorry, in Germany”).	Continuing smoothly without pausing. The AI penalizes hesitation far more than an inaccurate data read.
Speaking to 40.0s	Getting cut off mid-sentence while delivering the conclusion.	Completing the conclusion smoothly by second 30–33 and clicking [Next].

7. Targeted Action Drills for Lesson 2.3

The “Template Muscle-Memory” Sprint:
Recite the 4-part Universal Template aloud 5 times in a row without looking at any image until the transition phrases (“The given… Looking at the significant trends… Conversely… In conclusion…”) flow without hesitation.
The 25-Second “4-Point Scan” Drill:
Open 10 random graphs (bar, line, pie, map, process). Spend strictly 25 seconds on each image writing only 4 items on your scratchpad: Title | Units | Peak / Start | Trough / End.
The 30-Second Hard-Stop Recording Drill:
Record 5 complete Describe Image responses using a countdown timer. Train yourself to complete the concluding sentence and hit the stop button between the 28th and 34th second.

PTE Academic

Curriculum