Summary: We recognize smells almost instantly—often before we consciously notice them. New research shows that the decisive steps for odor identification occur within the first fraction of a second inside the olfactory bulb, not primarily in the cerebral cortex as previously assumed.
Using a process the authors call “temporal filtering,” the brain relies on the earliest nerve signals produced during the initial 50 milliseconds of a sniff to identify an odor while suppressing later, background signals.
Key Facts
- Sniff timing: Mice take fast sniffs (about 250–500 ms), while human sniffs are longer (1–3 seconds). Regardless of species, the critical recognition signal appears at the very start of each sniff.
- Precision optogenetics: Researchers used a custom circuit-mapping microscope and optogenetic stimulation to pinpoint and control neural activity with millisecond precision, revealing exactly when specific signals fire.
- AI potential: Temporal filtering—prioritizing the first, most informative signals and ignoring later noise—could inspire more efficient artificial intelligence approaches for processing large sensory datasets.
- Parallel in vision: The result echoes recent vision research showing that the retina performs substantial preprocessing before signals reach the visual cortex.
Source: NYU Langone
Overview: Mice use rapid interactions among nerve cells in the olfactory bulb to distinguish odors within a fraction of a second. The study demonstrates that the earliest-activated signals in the olfactory bulb provide a concentration-invariant code for odor identity, while later signals can be filtered out.
Led by scientists at NYU Langone Health and published online in Nature Neuroscience on April 14, the work shows that a small subset of olfactory signals—those that fire within the first milliseconds of inhalation—determine which smell is perceived. In mice, a full sniff cycle lasts roughly 250 to 500 milliseconds; human sniff cycles are slower, typically between one and three seconds (1,000 milliseconds = 1 second).
The experiments focused on how millions of olfactory sensory neurons in the nose connect to clusters called glomeruli in the olfactory bulb, which then link to groups of mitral and tufted cells (MTCs). The team discovered that glomeruli and their associated MTCs that activate within the first 50 milliseconds robustly represent odor identity across a range of concentrations. In contrast, MTCs connected to later-activated glomeruli showed responses that depended on odor concentration.
The researchers describe the mechanism as temporal filtering: an early burst of excitatory input at the start of a sniff creates a brief window of excitability in the olfactory bulb, followed by prolonged inhibition. That early window transmits the defining odor signal while later inputs are suppressed, effectively blocking background or overlapping odors from altering the initial identification.
Importantly, the same pattern of early glomeruli–MTC activation recurred for a given odor regardless of how concentrated it was. Once that early pattern was established, subsequent activation by background odors could not override or pass through; the system prioritized the first reliable signal and decorrelated later activity.
“These results challenge a fundamental assumption that most sensory computations happen downstream in the cortex,” said co-senior investigator Dmitry Rinberg, PhD, professor of neuroscience at NYU Grossman School of Medicine. “We show how the olfactory bulb itself implements a rapid temporal filter that identifies odors.”
Co-senior investigator Shy Shoham, PhD, director of the Tech4Health Institute at NYU Langone Health, noted the broader implications: the same temporal filtering principle may shape other sensory systems and could be applied to improve computational networks that face large streams of sensory data.
The study’s methods relied on advanced optogenetics and a new circuit-mapping microscope developed by lead investigator Mursel Karadas, PhD, which allowed the team to stimulate and monitor individual nerve signals in the thin, outer layers of the olfactory bulb with high spatial and temporal resolution.
Next steps include exploring how temporal filtering helps distinguish closely related odors—such as lemon versus orange or different berry scents—and how these early temporal patterns support fine discrimination among similar smells.
Funding for the research came from National Institutes of Health grants U19NS107464, U19NS112953, and R01DC022320. Other NYU Langone contributors include co-investigators Jonathan Gill and Sebastian Ceballo.
Key Questions Answered:
A: The olfactory bulb can encode “coffee” within about 50 ms, but it takes additional time—hundreds of milliseconds—for that information to reach the cortex and enter conscious awareness. Encoding is fast; conscious recognition follows slightly later.
A: Not necessarily in all cases. The system is tuned to capture the initial burst of input. A sharp, quick inhalation gives a clear early window of information; repeated rapid sniffs can refresh that window and help the brain filter out background odors more effectively.
A: Yes. Many AI systems suffer from sensory overload when attempting to process all incoming data. Implementing temporal filtering—prioritizing the first, most informative moments and down-weighting later noisy inputs—could yield faster, more accurate artificial sensors.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The original journal paper was reviewed in full.
- Additional context was provided by the editorial staff.
About this olfaction and neuroscience research news
Author: David March
Source: NYU Langone
Contact: David March – NYU Langone
Image: The image is credited to Neuroscience News
Original Research: Open access. “Rapid temporal processing in the olfactory bulb underlies concentration-invariant odor identification and signal decorrelation” by Mursel Karadas, Jonathan V. Gill, Sebastian Ceballo, Shy Shoham & Dmitry Rinberg. Nature Neuroscience
DOI: 10.1038/s41593-026-02250-y
Abstract
Rapid temporal processing in the olfactory bulb underlies concentration-invariant odor identification and signal decorrelation
Sensory systems must extract stable, relevant signals from dynamic, noisy environments. Animals that depend on smell must identify odors reliably despite wide fluctuations in concentration; yet odor receptor activation varies with concentration. Using an all-optical approach in awake mice, the study mapped connectivity between odor receptor channels (glomeruli) and mitral and tufted cells (MTCs) while monitoring their responses. The earliest-activated glomeruli and MTCs robustly encoded odor identity across concentrations, whereas later-activated elements were concentration dependent. The olfactory bulb exhibits a short temporal window of excitability at sniff onset followed by sustained inhibition, implementing a rapid temporal filter that stabilizes identity signals and decorrelates responses between odors.