Digital Restoration

[Senaqua Hildreth]

AconDigital Restoration Suite is a collection of four plug-ins for professional grade audio restoration. Restoration Suite is available for PC and Macintosh and supports the VST, VST3, AAX and AU plug-in formats. Both 32 bit and 64 bit versions are available.

Restoration Suite 2 is a suite of four cutting edge plug-ins for audio restoration and noise reduction. The plug-ins are available as VST, VST3 or AAX for both Windows and Mac or as AU plug-ins for Mac. There are 32 and 64 bit versions for Windows. The Mac version is 64 bit and runs natively on both ARM (such as Apple M1) and Intel processors.

DeNoise is a plug-in designed to reduce noise such as hiss, wind noise, buzz and camera noise. The noise can be reduced automatically in the adaptive mode or after measuring the characteristics of the noise in the noise profile mode. New in version 2 are dynamic noise profiles that capture the dynamic properties of the noise so that noise that fluctuates over time, such as wind noise, can be effectively reduced. The new algorithm has also been greatly improved and is now even less prone to typical de-noising artifacts.

Version 2 introduces the novel dynamic noise profiles that help reducing noise that varies randomly over time, such as wind noise or rustle from lavalier microphones. Where the earlier versions merely captured a static noise print with time-constant noise levels, the dynamic noise profiles capture statistics from the noise to be reduced. The noise suppression algorithm then estimates the most suitable noise threshold curve for the noisy input signal using the measured statistics.

Version 2 introduced support for Mid/Side (M/S) processing, which can reduce unwanted fluctuations in the stereo image. A new automatic fine-tune button triggers automatic estimation of the hum noise frequency.

Digital photograph restoration is the practice of restoring the appearance of a digital copy of a physical photograph that has been damaged by natural, man-made, or environmental causes, or affected by age or neglect.

Digital photograph restoration uses image editing techniques to remove undesired visible features, such as dirt, scratches, or signs of aging. People use raster graphics editors to repair digital images, or to add or replace torn or missing pieces of the physical photograph. Unwanted color casts are removed and the image's contrast or sharpening may be altered to restore the contrast range or detail believed to have been in the original physical image. Digital image processing techniques included in image enhancement and image restoration software are also applied to digital photograph restoration.

Photographic material is susceptible to physical, chemical and biological damage caused by physical forces, thieves and vandals, fire, water, pests, pollutants, light, incorrect temperature, incorrect relative humidity, and dissociation (custodial neglect).[citation needed] Traditionally, preservation efforts focused on physical photographs, but preservation of a photograph's digital surrogates has become of equal importance.[1][unreliable source?]

Fragile or valuable originals are protected when digital surrogates replace them, and severely damaged photographs that cannot be repaired physically are revitalized when a digital copy is made.[2] Creation of digital surrogates allows originals to be preserved.[3] However, the digitization process itself contributes to the object's wear and tear.[4] It is considered important to ensure the original photograph is minimally damaged by environmental changes or careless handling.[5]

Images that are digitally reproduced and restored often reflect the intentions of the photographer of the original photograph.[citation needed] It is not recommended[according to whom?] that conservators change or add additional information based on personal or institutional bias or opinion.[citation needed] Even without copyright permission, museums can digitally copy and restore images for conservation or informational purposes.[citation needed]

At one point, in order to better illustrate what is wrong with the physical restoration of the ostensible Leonardo da Vinci painting of the Salvator Mundi, I got the bright idea of digitally correcting the most egregious errors as a visual aid to hammer home my argument. I only intended to work on the face, but then got the idea to do a full-fledged recreation as a more convincing visual argument, and as a potential portfolio piece.

I started with the cleaned state of the painting, before the restorer began her transformation of it into the unfortunate spectacle we see now. First I restored it as best I could, and then I tried to recreated what it might have looked more like.

I ended up making 3 videos about the Salvator Mundi. The first two are parts 1 &2, but the third is not part 3. Rather, it is a comprehensive, stand-alone, documentary-length analysis, in which my recreation of the painting is one of many custom graphics and animations I created for the video so one always has something visually engaging to look at.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Digital restoration is a rapidly growing methodology in cultural heritage whereby images of art objects are computationally manipulated to visualize their original appearance or reveal hidden information without actual physical intervention1,2,3,4. Digital restoration is increasingly playing a role in interpreting and displaying an artwork when it is severely damaged5,6 or when it has been stripped of historically significant information7.

As has been recognized by UNESCO since 1980, moving images are a fundamental part of the world's Cultural Heritage8. Throughout the twentieth century, films were coloured with light and heat-sensitive dyes incorporated into the emulsion layers. Today, these films often exhibit colour degradation, fading, colour loss, bleaching, and colour change8, thus necessitating their digital restoration9,10,11. For motion pictures, the film is commonly restored by scanning using an RGB scanner and manually processed with dedicated software, such as Photoworks Photo Editor 202112, DaVinci Resolve 17 by Black Magic13, and Paintshop Pro by Corel14, to re-balance the colour and adjust the colour saturation and contrast10. Conventional digital restoration is laborious, with the resulting appearance reliant upon the restorers' skills and judgments about what looks appropriate.

This study proposes a machine learning algorithm that avoids subjective choices in restoring differentially faded film. As described in more detail below, a vector quantization algorithm is proposed that exploits a sparse representation of spectral reflectance data obtained from degraded and non-degraded films. After registration of representative degraded and non-degraded frames, a joint dictionary is learned from these data sets, which calculates a restored representation for the entire film. Spectral data were first processed using a simple codebook approach and further improved by a multi-codebook method capable of restoring frames with different degradation effects. The method proposed here provides more accurate results than those obtained with the currently available restoration software.

In response to these subjective approaches, several algorithms have been developed to automatically restore digitalized films with minimal intervention9,15. Several of these techniques successfully detect scratches or lacunae, and this missing content is in-painted using standard techniques4,16,17,18. However, for faded colour, most existing models assume homogeneous reduction in colour and hue across the image frame. Only one deep learning algorithm, based on latent space translation, trains with paired synthetic data19 to compensate for uniform fading. For more severe and inhomogeneous colour loss, the algorithm is prone to failure. Other CNN algorithms focusing on colorization of black and white films20,21,22,23 rely on synthetic training data sets that have the same limitation when it comes to uneven fading. In historical films, the degradation of colour usually varies across and within each frame, so restoration models trained using many homogeneous synthetic images may impose inaccuracies or even colour distortions. Another approach, based on what is known as the Automatic Colour Equalization Model11, imitates the mechanics of the human visual system, optimizing colour contrast, saturation, and balance according to human perception and aesthetics rather than restoring the film to its original appearance. Such methods are generally ineffective when attempting to restore artworks to a historically accurate state, as is the central requirement in the cultural heritage sector.

Practical restoration of differentially degraded colour film thus remains an unsettled problem. Here we propose advanced tools, such as spectral imaging, to face the challenges imposed by the complexity of colour degradation in historical films. Hyperspectral imaging has been increasingly applied to the analysis and conservation of important artefacts24,25,26. The fine spectral resolution afforded by optical reflection spectroscopy, down to nanometre resolution, enables the capture of degradation phenomena of film at high spatial and spectral resolution, which is otherwise hard to identify with the conventional RGB captures. By combining spectral imaging with advanced machine learning algorithms, the limitations of using synthetic data alone is overcome, given the large amount of spectral data that may serve as the input to the algorithm. In addition, machine learning also handles the challenge of processing large amounts of data which is often a major concern in cultural heritage applications. Such methodology has already been reported in the study of illuminated manuscripts where hyperspectral imaging and a deep neural network were combined to perform the spectral unmixing and quantitative estimation of pigment concentrations27. Another important work on the degraded medieval manuscript28 proposed a codebook algorithm to fuse the hyperspectral data and XRF data that successfully revealed the hidden content through the correlated spectral mapping. Although no application of this approach has been reported on film restoration, those research projects open the door for a novel solution to the colour degradation problem in damaged historical films.

3a8082e126