Attention
Sound-Induced Flash Illusion
Also called: SIFI, double-flash illusion, illusory flash effect, fission and fusion illusions, flash-beep illusion
Measure how sound changes what people see: how often one flash with two beeps looks like two flashes, and two flashes with one beep like one.
6-8 minTimed: desktop or laptop with a mouse or trackpad
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 101
- Evidence grade
- D
- Registered clinical studies
- 1
What it measures
A measure of how sound changes what people see. A white disk flashes once or twice below a cross with zero, one or two short beeps, and participants press 1 or 2 for the number of flashes they saw. One flash with two beeps is often seen as two flashes (fission), and two flashes with one beep as one (fusion). The score is the share of one-flash, two-beep trials seen as two flashes.
Constructs
- Audiovisual integration
- Multisensory perception
- Temporal binding of sight and sound
- Visual perception
Research uses
- Multisensory integration research
- Aging and falls research
- Autism and schizophrenia research
- Within-study group comparison
How the paradigm works
The sound-induced flash illusion shows how sound changes what people see. A white disk flashes once or twice below a cross while zero, one or two short beeps play, and participants press 1 or 2 for the number of flashes they saw, ignoring the beeps. Shams, Kamitani and Shimojo (2000) found that a single flash with two beeps is often seen as two flashes, an illusion called fission. Andersen, Tiippana and Sams (2004) also found the reverse, fusion: two flashes with one beep seen as one. The illusion holds while the beeps fall within about 100 ms of the flash (Shams and colleagues 2002), and it varies a great deal from person to person, which makes it a common measure of audiovisual integration in studies of aging, autism and schizophrenia (Keil 2020).
More about the paradigm
Each trial shows one of six conditions: one or two flashes with no beep (the visual controls), one flash with one beep and two flashes with two (the matched controls), one flash with two beeps (fission) and two flashes with one beep (fusion). A flash lasts one frame of a 60 Hz screen and two flashes start four frames apart, about 67 ms, shown as whole frames at the screen's measured rate; the beeps are 10 ms tones of 3.5 kHz, as McGovern and colleagues (2014) used over headphones, timed so the first is heard with the first flash. The standard session has 10 rounds of 8 trials, each control once and each illusion twice, 80 in all, after 8 practice trials on the controls with feedback, in about 7 minutes with the sound setup.
The test needs headphones and normal hearing. Before it, a loudness step sets a comfortable volume and the headphone check of Woods and colleagues (2017) checks that headphones are worn; a failed check stops the test unless the study chooses to continue. The browser reports how late its sound output plays, and the beeps are scheduled to allow for it; most wireless headphones add a delay it cannot see, so wired headphones are best.
Scores
- Fission Rate%
- Share of one-flash, two-beep trials answered two flashes.
- Higher means a stronger illusion
- Fusion Rate%
- Share of two-flash, one-beep trials answered one flash.
- Higher means a stronger illusion
- Fission Effect%
- Fission rate less the share of single flashes with no sound answered two.
- The illusion beyond errors without sound
- Fusion Effect%
- Fusion rate less the share of double flashes with no sound answered one.
- The illusion beyond errors without sound
- Fission d-primed'
- Telling one flash from two with two beeps: hits are two-flash, two-beep trials answered two; false alarms one-flash, two-beep trials answered two.
- Lower means the beeps blurred one flash and two more
- Fission Criterionz
- Response criterion with two beeps; negative leans toward answering two.
- Informational
- Fusion d-primed'
- Telling one flash from two with one beep: hits are one-flash, one-beep trials answered one; false alarms two-flash, one-beep trials answered one.
- Lower means the beep blurred one flash and two more
- Fusion Criterionz
- Response criterion with one beep; positive leans toward answering one.
- Informational
- Accuracy Without Sound%
- Share right on one or two flashes with no beep.
- Below 0.75 is flagged for review
- Accuracy With Matching Beeps%
- Share right when the beeps matched the flashes.
- Higher is better
- Median Answer Timems
- Median time from the question to the answer.
- Informational
- Displays on Time%
- Share of answered trials whose flashes showed for their planned frames with none dropped.
- Below 0.9 is flagged for review
- Beep to Flash Offsetms
- Median time from the first flash reaching the screen to the first beep being heard, by the browser's clocks; 0 is planned.
- Near 0 is on time
- Reported Output Delayms
- Delay of the sound output that the browser reported, null when it reported none.
- Informational
- Headphone Check Passed0/1
- 1 when the headphone check passed.
- 0 is flagged
Use and evidence
101 papers from 2021 to 2025 name this task in their title or abstract, up 20% compared with 2016 to 2020. That ranks it 186 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 23 |
| 2024 | 10 |
| 2023 | 22 |
| 2022 | 18 |
| 2021 | 28 |
| 2020 | 24 |
| 2019 | 18 |
| 2018 | 17 |
| 2017 | 9 |
| 2016 | 16 |
| 2015 | 18 |
| 2014 | 17 |
| 2013 | 8 |
| 2012 | 11 |
| 2011 | 10 |
| 2010 | 12 |
| 2009 | 2 |
| 2008 | 4 |
| 2007 | 4 |
| 2006 | 1 |
| 2005 | 5 |
| 2004 | 2 |
| 2003 | 0 |
| 2002 | 1 |
| 2001 | 0 |
| 2000 | 0 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade D
Little published measurement evidence found.
- Papers on test-retest reliability
- 0
- Papers on reliability or validity
- 26
- Registered clinical studies using it as an outcome
- 1
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Display
- A white disk 80 pixels wide, its center 200 pixels below a gray cross on black.
- Flashes
- One or two flashes of 1 frame of a 60 Hz screen (about 17 ms), 4 frames apart (about 67 ms), shown as whole frames at the measured refresh rate.
- Beeps
- None, one or two 10 ms tones of 3.5 kHz at the level set in the loudness step. The first beep is heard with the first flash and the second as far after it as the second flash.
- Conditions
- One or two flashes alone, one flash with one beep and two with two, one flash with two beeps (fission) and two flashes with one beep (fusion).
- Fixation
- 600 to 1,000 ms, drawn at random for each trial.
- Answer
- Press 1 or 2 for the number of flashes seen, with no time limit, after the flashes and beeps.
- Length
- 10 rounds of 8 trials (6 to 14), each control once and each illusion twice in a random order: 80 trials.
- Practice
- 8 trials with feedback, each control twice. No illusion trial is practised.
- Sound setup
- A loudness step, then the headphone check of Woods and colleagues (2017). A failed check stops the test unless the study chooses to continue.
- Scoring
- The share of one-flash, two-beep trials answered two (fission) and of two-flash, one-beep trials answered one (fusion), each also against the same flashes with no sound, with d-prime and criterion against the matched controls.
- Timing
- Trials whose flashes did not show for their planned frames are left out. The time from the first flash to the first beep, by the browser's clocks, is saved with every answer.
The test needs headphones and normal hearing. Shams and colleagues played the beeps from loudspeakers; this version uses headphones, as McGovern and colleagues (2014) did. The browser reports how late its sound output plays and the beeps are scheduled to allow for it, but most wireless headphones add a delay the browser cannot see, so wired headphones are best. Sizes are not calibrated to the viewing distance. No percentile: susceptibility to the illusion is not a better or worse score. The scores support research and do not diagnose.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: sound_induced_flash_test from the scored run, one or two flashes answered with no time limit. Only the four control conditions (flashes alone, or with as many beeps as flashes) have a right answer (chance 50%); the illusion trials are the measure and are not judged. Below chance on the controls invalidates and not above chance is flagged; one key on 90% of answers invalidates when the controls are not above chance. A headphone check that stopped the test invalidates, and one that failed while the test ran on is flagged. Under 75% right on the flashes with no sound, no fission rate, a run short of its planned trials, or fewer than 90% of trials with their flashes on time is flagged for review. Answer times are not checked: the answer comes after the flashes and beeps, when it may already be decided.
Why these rules
Accuracy significantly below chance (one-sided exact test, p < .05) cannot come from guessing; it means the response mapping was reversed or the task was not followed (Slick, Sherman and Iverson 1999 treat below-chance forced-choice performance as definite response bias). Chance-level accuracy can be a real result in impaired participants, so it is only flagged. One option on at least 90% of answered trials, when the correct options are balanced and accuracy is not above chance, means the answers did not depend on the items (the task counterpart of long-string responding in surveys; Curran 2016). Without conflicting sound, one and two flashes are easy to tell apart (Shams, Kamitani and Shimojo 2002, Figure 4), so the controls show whether the task was followed; on the illusion trials a wrong count is the illusion itself. The fission and fusion rates mean little when the flashes alone could not be counted, or when the sounds may not have reached both ears through headphones.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Rounds | 10 | 6 to 14 |
| When the headphone check fails | Stop the test | Stop the test, Continue the test |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Timed test: participants need a desktop or laptop with a mouse or trackpad. Needs headphones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Sound-Induced Flash Illusion task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Shams L, Kamitani Y, Shimojo S, 2000, 2002), with the standard settings (rounds: 10; when the headphone check fails: Stop the test). The main score was fission rate (share of one-flash, two-beep trials answered two flashes).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Shams, L., Kamitani, Y., & Shimojo, S. (2000). What you see is what you hear. Nature, 408(6814), 788. https://doi.org/10.1038/35048669
- Shams, L., Kamitani, Y., & Shimojo, S. (2002). Visual illusion induced by sound. Cognitive Brain Research, 14(1), 147-152. https://doi.org/10.1016/S0926-6410(02)00069-1The flash, the beep and the window of about 100 ms in which the beeps act.
- Andersen, T. S., Tiippana, K., & Sams, M. (2004). Factors influencing audiovisual fission and fusion illusions. Cognitive Brain Research, 21(3), 301-308. https://doi.org/10.1016/j.cogbrainres.2004.06.004The fusion illusion: two flashes with one beep seen as one.
- McGovern, D. P., Roudaia, E., Stapleton, J., McGinnity, T. M., & Newell, F. N. (2014). The sound-induced flash illusion reveals dissociable age-related effects in multisensory integration. Frontiers in Aging Neuroscience, 6, 250. https://doi.org/10.3389/fnagi.2014.00250The six conditions, the beeps over headphones, and d-prime and criterion for fission and fusion.
- Keil, J. (2020). Double flash illusions: Current findings and future directions. Frontiers in Neuroscience, 14, 298. https://doi.org/10.3389/fnins.2020.00298Control conditions in laboratory and online studies.
- Woods, K. J. P., Siegel, M. H., Traer, J., & McDermott, J. H. (2017). Headphone screening to facilitate web-based auditory experiments. Attention, Perception, & Psychophysics, 79(7), 2064-2072. https://doi.org/10.3758/s13414-017-1361-2The headphone check.
Source papers
- Shams L, Kamitani Y, Shimojo S (2000). What you see is what you hear. Nature, 408(6814):788.N 8, Adults
- Shams L, Kamitani Y, Shimojo S (2002). Visual illusion induced by sound. Cognitive Brain Research, 14(1):147-152.N 16, Adults (19 to 41)
- McGovern DP, Roudaia E, Stapleton J, McGinnity TM, Newell FN (2014). The sound-induced flash illusion reveals dissociable age-related effects in multisensory integration. Frontiers in Aging Neuroscience, 6:250.N 50, Adults 18 to 30 and 65 to 88
Questions
How long does the Sound-Induced Flash Illusion take?
About 6 to 8 minutes, as listed in the test library.
What device does a participant need?
Timed test: participants need a desktop or laptop with a mouse or trackpad. Needs headphones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Fission Rate, Fusion Rate, Fission Effect, Fusion Effect, Fission d-prime, Fission Criterion, Fusion d-prime, Fusion Criterion, Accuracy Without Sound, Accuracy With Matching Beeps, Median Answer Time, Displays on Time, Beep to Flash Offset, Reported Output Delay and Headphone Check Passed. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Sound-Induced Flash Illusion used in research?
101 papers from 2021 to 2025 name it in their title or abstract, up 20% compared with 2016 to 2020. That ranks it 186 of 395 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.14.0, counted 2 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Sound-Induced Flash Illusion have?
In the published research, 0 papers report its test-retest reliability and 1 registered clinical study uses it as an outcome (evidence grade D in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Rounds (6 to 14) and when the headphone check fails (Stop the test, Continue the test). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.