Research ArticleTheory/New Concepts, Cognition and Behavior

Visual Stimulation Under 4 Hz, Not at 10 Hz, Generates the Highest-Amplitude Frequency-Tagged Responses of the Human Brain: Understanding the Effect of Stimulation Frequency

Talia L. Retter and Christine Schiltz
eNeuro 17 June 2025, 12 (6) ENEURO.0426-24.2025; https://doi.org/10.1523/ENEURO.0426-24.2025

Article Figures & Data

Figures

  • Figure 1.
    Figure 1.

    EEG responses at F as a function of visual stimulation frequency. Data derived from Retter et al. (2021a) (electrode Oz; N = 16 people; stimuli = natural face images; Butterworth 4th-order filter from 0.05–140 Hz). The amplitude of only the first harmonic (F, i.e., 1F) is plotted (black line), as in many analyses in the literature. The maximal F response occurs to visual stimulation at 10 Hz here (arrow). Note that the amplitude pattern of responses to stimulation does not follow the trend of baseline noise activity (light gray line; from the 40 Hz stimulation condition), which is highest close to 0 Hz and falls steeply until near 0.5–1 Hz, and then tends to decrease gradually, often with a smaller local peak around the alpha band (8–12 Hz) range.

  • Figure 2.
    Figure 2.

    EEG responses to periodic visual stimulation in the time domain. Data derived from Retter et al. (2021a) (electrode Oz; N = 16 people; stimuli = natural face images). The response amplitude range, decreasing as stimulation frequency increases, is highlighted in color and given in each panel. a, b, Some peak response deflections to stimulation at 3 and 5 Hz are fit approximately with segments of a 10 Hz sine wave (red highlights). c, At 10 Hz stimulation, the response becomes cyclical and is fit well with a 10 Hz sine wave (2 cycles highlighted). d, At 15 Hz stimulation, the cyclical response is condensed to 15 Hz (2 15 Hz cycles highlighted in yellow).

  • Figure 3.
    Figure 3.

    EEG responses as a function of visual stimulation frequency (Oz). Data derived from Retter et al. (2021a) (electrode Oz; N = 16 people; stimuli = natural face images). a, As in Figure 1, the amplitude of only the first harmonic (1F, i.e., F) is plotted in black, leading to an apparent signal peak ∼10 Hz over the occipital cortex (marked here with an arrow). Additionally, the missing information of the higher harmonic frequency response amplitudes within each stimulation frequency condition is included here in colored lines (see the key); the first four harmonics are labeled illustratively for the 3 Hz stimulation frequency (1F = 3 Hz; 2F = 6 Hz; 3F = 9 Hz; 4F = 12 Hz). b, In a recommended analysis, the response amplitude is summed across harmonic frequencies (Fs; F's below 50 Hz for each stimulation condition; see the colored line plots in panel a), leading to the largest response amplitude at the lowest stimulation frequency tested, 3 Hz (marked with an arrow; error bars indicate ±1 SEM across participants; noise level (with which to correct the signal measurement) given by a light gray line).

  • Figure 4.
    Figure 4.

    EEG responses as a function of low visual stimulation frequency. Data derived from Retter and Rossion (2016) (electrode PO10; N = 16 people; stimuli = natural face vs object images; Fs summed up to 19 Hz; as in Fig. 3 of that study; and its Fig. 2 for frequency-domain harmonic distributions and Fig. 4 for time-domain amplitudes). a, The baseline-subtracted amplitude of Fs is stable across the lowest stimulation frequencies, from 1.14 to 2.50 Hz, and is decreased only at the higher stimulation frequency of 4.16 Hz (red arrow). b, The time-domain deflections’ amplitude is stable across the lowest stimulation frequencies, from 1.14 to 2.50 Hz, and is decreased only at the highest stimulation frequency of 4.16 Hz (red arrow; full-contrast time period represents one full cycle at 4.16 Hz = 240 ms), in correspondence with the Fs frequency-domain amplitudes.

  • Figure 5.
    Figure 5.

    EEG responses as a function of visual stimulation frequency (PO10). Data derived from Retter et al. (2021a) (electrode PO10; N = 16 people; stimuli = natural face images). All notations match Figure 3. a, Unlike at Oz, the amplitude of only the first harmonic (F, i.e., 1F) over the occipitotemporal cortex leads to a signal peak at 3 Hz. b, The response summed across harmonics for each stimulation condition also leads to a signal peak at 3 Hz at this recording location.

Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Visual Stimulation Under 4 Hz, Not at 10 Hz, Generates the Highest-Amplitude Frequency-Tagged Responses of the Human Brain: Understanding the Effect of Stimulation Frequency
Talia L. Retter, Christine Schiltz
eNeuro 17 June 2025, 12 (6) ENEURO.0426-24.2025; DOI: 10.1523/ENEURO.0426-24.2025
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

  • Published on: (29 August 2025)
    Page navigation anchor for 4 Hz? It depends!
    4 Hz? It depends!
    • Sven P. Heinrich, Physicist, Neuroscientist, Eye Center, Faculty of Medicine, Medical Center – University of Freiburg, Freiburg, Germany

    This article addresses an important practical aspect of event-related potential research: which stimulation frequency to choose to obtain the “largest” responses in studies that use frequency-tagging. The authors raise many important points, and I applaud them for having reviewed this highly relevant issue. Their conclusion, as conveyed by the title of the article, is that visual stimulation below 4 Hz generates the highest-amplitude responses.

    This is a straightforward conclusion, but unfortunately, the 4 Hz threshold is not universally valid. The article seems to target readers who may not be fully familiar with the mathematical underpinnings of the issue and who may thus be particularly prone to being misled. A more guarded message would be: If the cycle duration is long enough to avoid the superposition of responses to successive individual stimuli, then one can capture the full underlying single-stimulus response through frequency-domain analysis. Otherwise, it depends. Importantly, the “it depends” regimen may extend well below 4 Hz when slow response components are involved.

    Because of potentially complex superposition effects that arise when the cycle length is shorter than the total duration of a single-stimulus response, a stimulus frequency that yields a high overall amplitude is not necessarily one that shows a large (absolute or relative) effect as the result of an experimental manipulation or pathophysiological process. Further, if the goal is t...

    Show More
    Competing Interests: None declared.

Subjects