Astronomers did not continuously film 3C 345’s jet for 27 years. They observed it 116 times between 1995 and 2022, then used a neural representation to infer a changing radio image consistent with those irregular measurements. The work appears in the peer-reviewed Nature article Video reconstruction of variable VLBI observations with neural fields, led by Marianna Foschi. Their kine algorithm produced a time-continuous polarimetric reconstruction and a two-dimensional map of the apparent motion through the jet. This is an article applied to an exceptionally well monitored blazar. The reconstruction is limited by real radio interferometry, but the frames between observation dates are modeled estimates, not missing photographs retrieved from an archive. The source points almost towards us 3C 345 is a blazar with a redshift of 0.593. Its central supermassive black hole drives a relativistic jet inclined only 3 to 6.8 degrees from our line of sight. That close alignment makes the jet glow and creates the geometry behind its apparent faster-than-light motion. The 116 epochs came from the MOJAVE monitoring program at 15 gigahertz. They were collected with the Very Long Baseline Array, ten radio antennas spread across the United States and operated together as a virtual telescope thousands of miles wide. That method is related to the long baseline technique behind the Event Horizon Telescope. Space Daily previously described how widely separated observatories can act as an Earth-sized instrument. The gaps between antennas still leave incomplete spatial sampling, so creating an image requires careful reconstruction. Kine unites space and time. Traditional processing typically reconstructs each observation independently. Small differences in telescope coverage and noise can cause the resulting sequence to flicker, complicating attempts to distinguish physical motion from image-to-image artifacts. Tracking broad Gaussian components also reduces the changing flow to a few selected entities. Kine, on the other hand, represents brightness and polarization as a function of right ascension, declination, and time. A multilayer neural network predicts those quantities, a forward model converts them into radius measurements that the array should have recorded, and optimization reduces the difference with the real data. The team has released code, observations, and supporting products. The method is a physics-constrained reconstruction. It is not a text to video generator. The network adapts directly to interferometric visibilities and related observables, while learning correlations in space and time. Sharpness Comes with Interpolation Validation testing gave the simultaneous reconstruction an average effective resolution of about 113 microarcseconds, about 4.2 times finer than the nominal 475 microarcsecond beam. Its total dynamic range was about 500,000, about 140 times the conventional CLEAN result for this particular data set. These factors are not universal promises for every radio observation. The authors state that profits depend on the quantity, quality and cadence of the data. Much of the improvement in resolution already appeared when cinema processed individual epochs, while combining all 116 epochs greatly contributed to the dynamic range. For optical flow analysis, the continuum model was sampled at regular times. The most distant interpolated frame was 5.7 months from a real observation. Synthetic tests designed to resemble the 3C 345 indicated that motion-preserving interpolation remained reliable for approximately six months. Still, measured epochs and inferred frames are not the same. Nothing traveled faster than light. The bright components appeared to move between 10 and 13 times the speed of light, while the surrounding mass flow averaged between 9 and 12 times the speed of light in the same region. These are projected apparent velocities, not local velocities through space, and do not violate relativity. A plane traveling close to the speed of light and pointing almost towards the Earth almost reaches its own previous light. The material travels a real distance between two emissions, but the second signal has a shorter path to us. The arrival time interval is compressed, making movement across the plane of the sky appear superluminal. No plasma or information surpasses light locally. The paper’s synthetic tests found that kinematics combined with optical flow could recover apparent speeds of up to at least 23 times the speed of light in data of this quality, above the fastest motion measured in 3C 345. That test addresses whether the algorithm itself would suppress fast apparent motion. Bright knots may not be strong shocks. A long-standing interpretation treats the compact bright components in relativistic jets as traveling shocks that compress the magnetized plasma. Kine allowed the team to compare those pattern speeds to the local flow around them, rather than just tracking the movement of selected nodes. The knots and surrounding plasma had comparable apparent velocities. The reconstructed polarization also lacked the localized increase that the authors expected from strong shocks. Together, those findings led them to favor excessively dense or unusually emissive regions over strongly impacted plasma. This interpretation remains conditional. Some shock conditions can produce similar flow patterns and velocities, and the velocity map assumes that changes in radio emissions track the physical motion of the plasma. The result reduces one explanation rather than eliminating shocks from relativistic jets in general. The work complements previous high-resolution aircraft studies, including Space Daily’s coverage of Centaurus A, but adds a dimension of continuous time that a collection of static images cannot provide. A method as important as the particular jet The immediate result is a more detailed description of how 3C 345 curves, ejects bright structures and transports plasma. The broader proposal is that the same approach could process other long-duration VLBI monitoring programs and short, rapidly changing observations from the Event Horizon Telescope. The algorithm does not create additional observations. It uses patterns shared at nearby epochs to estimate a continuous source compatible with the measurements and then validates that process with synthetic data whose true evolution is known. That distinction is essential to read the result. Kine has converted 116 scattered radio data sets into the highest-resolution continuous reconstruction yet of this plane, not direct images of each day over 27 years. With that limit visible, reconstruction becomes a new way of measuring movement that isolated snapshots could not reveal. About this articleThis article is for general information and reflection. It is not professional advice. For your specific situation, consult a qualified professional. A personal measure of deep time. Look what has changed in the universe since you were born. Start with a date. Get a personal, detailed journey through everything that kept moving after you arrived. Six measurements obtained. No predictions. No astrology.
