Periodic Reporting for period 4 - ULTRA-FAST (VIDEOGRAPHY OF ULTRAFAST PHENOMENA USING THE FRAME CONCEPT)
Reporting period: 2023-10-01 to 2025-03-31
Dynamic ultrafast phenomena—single-occurrence events taking place on the picosecond timescale (1 ps = 0.000000000001 s) or faster—are notoriously difficult to study due to the extreme temporal resolution required. While high-speed photography has enabled us to capture isolated moments of such events—such as plasma formation—since the early 2000s, it offers only a fleeting glimpse, insufficient for understanding the complete evolution of the process. To unravel the physical, chemical, or biological mechanisms at play, a single snapshot is not enough; these events must be filmed in real time. The ULTRAFAST project focuses on developing and applying video technology capable of capturing such ultrafast processes—millions of times faster than what current technologies can resolve or where they fail due to insufficient sensitivity.
Filming on femtosecond timescales requires extremely intense and brief illumination—essentially, a high-speed version of flash photography. In this project, we use a laser source that emits light pulses lasting just tens of femtoseconds to illuminate the scene. Each pulse provides a glimpse of the object, and by firing a rapid succession of such pulses, we construct a complete ultrafast film. The challenge lies in the fact that all pulses occur within such a short timeframe that the camera records them all at once in a single image. To separate the frames, each pulse is uniquely encoded with a recognizable modulation pattern. Sophisticated algorithms then decode this single image, extracting and correctly ordering the individual frames to reconstruct the ultrafast video.
This approach to ultrafast videography is unique in its compatibility with spectroscopic measurements—it enables interactions with atoms and molecules via light. A central goal of the ULTRAFAST project is to apply this spectroscopic sensitivity to investigate ultrafast phenomena that would otherwise remain inaccessible. Another major objective has been to enhance detection sensitivity, allowing visualization of objects or signals that are typically too faint to observe. This has been achieved by increasing the sophistication of the computational analysis, enabling the detection of subtle traces left by the sample within the encoded data.
Filming on femtosecond timescales requires extremely intense and brief illumination—essentially, a high-speed version of flash photography. In this project, we use a laser source that emits light pulses lasting just tens of femtoseconds to illuminate the scene. Each pulse provides a glimpse of the object, and by firing a rapid succession of such pulses, we construct a complete ultrafast film. The challenge lies in the fact that all pulses occur within such a short timeframe that the camera records them all at once in a single image. To separate the frames, each pulse is uniquely encoded with a recognizable modulation pattern. Sophisticated algorithms then decode this single image, extracting and correctly ordering the individual frames to reconstruct the ultrafast video.
This approach to ultrafast videography is unique in its compatibility with spectroscopic measurements—it enables interactions with atoms and molecules via light. A central goal of the ULTRAFAST project is to apply this spectroscopic sensitivity to investigate ultrafast phenomena that would otherwise remain inaccessible. Another major objective has been to enhance detection sensitivity, allowing visualization of objects or signals that are typically too faint to observe. This has been achieved by increasing the sophistication of the computational analysis, enabling the detection of subtle traces left by the sample within the encoded data.
The ULTRAFAST project is based on the FRAME technique that, in short, is an imaging technique that allows the user to store several images in a single photograph (i.e. a film in a picture) using an image-coding approach. Exactly how many images a single photograph can hold depends on a variety of factors, such as the resolution of the sensor, the object under study, how the images are being coded. In the project we have investigated how these factors influence the obtained result and demonstrated the potential of achieving +1000 images within a single photograph. The number of images a sensor may hold simultaneously is an important factor since the amount of images the photograph holds sets the length of the extracted film, which needs to be long enough to cover the entire event - the more images the film contains, the more information and conclusions can be drawn from the film.
As described, ultrafast videography by means of FRAME requires ultrafast laser technology. However, FRAME videography can be performed using less complex light sources as well, which can facilitate the spread and usage of the technology. We have investigate the flexibility of FRAME and demonstrated how it can be combined with modern, state-of-the-art, yet inexpensive LED illumination technology to breach MHz frame rate videography, thus offering a cost-effective alternative to high-speed cameras.
The unique image-coding methodology with FRAME can be used for more than videography and we have also demonstrated the ability to store (1) spectral information into a single photograph for highly sensitive snapshot hyperspectral imaging and (2) three-dimensional data, demonstrated by, for the first time, visualising a plasma channel in 3D. Hyperspectral imaging is routinely used within both industry and science and we believe the versatility of image-coding using FRAME will open new technological avenues. Coded light can make inherently “colorblind” cameras, such as intensified cameras, spectrally sensitive, which also has been demonstrated within the ULTRAFAST project.
The world's fastest detector is called a streak camera, commonly used to study molecular dynamics occurring on a picosecond timescale. The technical solutions that make the streak camera fast do, however, yield certain issues such as image blur and loss of temporal contrast. In the ULTRAFAST project we have investigated whether these artifacts can be suppressed by simply applying a similar image-coding strategy as used in FRAME and have, thus far, been able to demonstrate experimentally an improved temporal contrast when combining the streak camera with FRAME-based image-coding.
Besides the above mentioned work, the two core research directions in the ULTRAFAST project have been the development and application of (i) interferometric ultrafast videography, and (ii) a novel diagnostic concept we refer to as coherence lifetime imaging (CLI), designed for visualizing ultrafast chemical processes. Interferometry, while highly sensitive, has previously been incompatible with ultrafast videography. By redefining FRAME’s analysis algorithms, we succeeded in extracting hidden interferometric signals, achieving high-fidelity imaging of nature’s fastest events.
Inspired by fluorescence lifetime imaging (FLI), CLI overcomes key FLI limitations. While FLI is restricted to fluorescent targets and nanosecond timescales, CLI probes spectral features from molecular rotation and vibration—about 100 times faster—thus accessing a new class of ultrafast chemical phenomena.
As described, ultrafast videography by means of FRAME requires ultrafast laser technology. However, FRAME videography can be performed using less complex light sources as well, which can facilitate the spread and usage of the technology. We have investigate the flexibility of FRAME and demonstrated how it can be combined with modern, state-of-the-art, yet inexpensive LED illumination technology to breach MHz frame rate videography, thus offering a cost-effective alternative to high-speed cameras.
The unique image-coding methodology with FRAME can be used for more than videography and we have also demonstrated the ability to store (1) spectral information into a single photograph for highly sensitive snapshot hyperspectral imaging and (2) three-dimensional data, demonstrated by, for the first time, visualising a plasma channel in 3D. Hyperspectral imaging is routinely used within both industry and science and we believe the versatility of image-coding using FRAME will open new technological avenues. Coded light can make inherently “colorblind” cameras, such as intensified cameras, spectrally sensitive, which also has been demonstrated within the ULTRAFAST project.
The world's fastest detector is called a streak camera, commonly used to study molecular dynamics occurring on a picosecond timescale. The technical solutions that make the streak camera fast do, however, yield certain issues such as image blur and loss of temporal contrast. In the ULTRAFAST project we have investigated whether these artifacts can be suppressed by simply applying a similar image-coding strategy as used in FRAME and have, thus far, been able to demonstrate experimentally an improved temporal contrast when combining the streak camera with FRAME-based image-coding.
Besides the above mentioned work, the two core research directions in the ULTRAFAST project have been the development and application of (i) interferometric ultrafast videography, and (ii) a novel diagnostic concept we refer to as coherence lifetime imaging (CLI), designed for visualizing ultrafast chemical processes. Interferometry, while highly sensitive, has previously been incompatible with ultrafast videography. By redefining FRAME’s analysis algorithms, we succeeded in extracting hidden interferometric signals, achieving high-fidelity imaging of nature’s fastest events.
Inspired by fluorescence lifetime imaging (FLI), CLI overcomes key FLI limitations. While FLI is restricted to fluorescent targets and nanosecond timescales, CLI probes spectral features from molecular rotation and vibration—about 100 times faster—thus accessing a new class of ultrafast chemical phenomena.
Filming fast is not the same as being able to observe fast events. At ultrashort timescales, even the brightest phenomena appear dim, often producing signals too weak to detect. Within the ULTRAFAST project, we have made significant advances toward high-fidelity imaging under such challenging conditions, opening new avenues for studying transient, ultrafast phenomena.
A further breakthrough lies in our ability to detect a wide range of spectral signals simultaneously. This capability could lead to a new generation of high-speed sensing technologies, with promising applications in fields such as biomedical diagnostics.
We have also shown that our image-coding approach—originally developed to embed multiple frames in a single photograph—enables a higher level of multiplexing than conventional techniques. In such multiplexed data, the two spatial dimensions are provided by the sensor, while the third dimension is typically time (in video) or wavelength (in hyperspectral imaging). However, no previous technique has, to our knowledge, demonstrated the addition of a fourth dimension. In the ULTRAFAST project, we have experimentally verified that FRAME image coding can achieve this, enabling the creation of true multidimensional images.
A further breakthrough lies in our ability to detect a wide range of spectral signals simultaneously. This capability could lead to a new generation of high-speed sensing technologies, with promising applications in fields such as biomedical diagnostics.
We have also shown that our image-coding approach—originally developed to embed multiple frames in a single photograph—enables a higher level of multiplexing than conventional techniques. In such multiplexed data, the two spatial dimensions are provided by the sensor, while the third dimension is typically time (in video) or wavelength (in hyperspectral imaging). However, no previous technique has, to our knowledge, demonstrated the addition of a fourth dimension. In the ULTRAFAST project, we have experimentally verified that FRAME image coding can achieve this, enabling the creation of true multidimensional images.