This paper presents a new method for rendering high dynamic range (HDR) images on conventional displays. The method is simple, efficient, robust, and easy to use. It manipulates the gradient field of the luminance image by attenuating the magnitudes of large gradients. A new, low dynamic range (LDR) image is then obtained by solving a Poisson equation on the modified gradient field. The method effectively compresses the dynamic range of HDR images while preserving fine details and avoiding common artifacts such as halos, gradient reversals, or loss of local contrast. It also enhances ordinary images by revealing details in dark regions. HDR images are becoming increasingly common in computer graphics. They offer advantages over standard low dynamic range images and are useful in various applications. However, displaying HDR images on low dynamic range (LDR) displays is challenging. This paper addresses this problem by proposing a new technique for HDR compression that enables HDR images to be displayed on LDR devices. Previous work on HDR compression includes global (spatially invariant) mappings and spatially variant operators. Global mappings, such as tone reproduction curves (TRCs), are simple but may not preserve local contrasts well. Spatially variant operators are more flexible but can be complex. The proposed method is a gradient domain HDR compression technique that manipulates the gradient field of the luminance image by attenuating large gradients. This approach is effective, computationally efficient, and avoids halo artifacts. The method works by first computing the logarithm of the luminances, then identifying and attenuating large gradients. The gradients are then used to solve a Poisson equation to reconstruct a low dynamic range image. The method is applied to various types of images, including HDR radiance maps, panoramic video mosaics, and medical images, with good results. The method is based on the assumption that the human visual system is not very sensitive to absolute luminances but responds to local intensity ratio changes. The algorithm manipulates the gradient field of the luminance image by attenuating the magnitudes of large gradients. This approach preserves fine details and avoids common artifacts. The method is implemented using a multi-resolution approach, where gradient attenuation factors are computed at different levels and propagated to the full resolution. The gradient attenuation function is computed in a top-down fashion, and the final gradient attenuation function is used to solve a Poisson equation to reconstruct the low dynamic range image. The method is tested on various HDR images, including those of a streetlight on a foggy night and a Stanford Memorial church. The results show that the method effectively compresses the dynamic range while preserving fine details and avoiding halo artifacts. The method is also used for LDR image enhancement, where it reveals details in dark regions and maintains good contrast. The paper concludes that the proposed method is a simple, efficient, and robust technique for HDR compression. Future work will explore the application of this method in various areas, including denoising, edge manipulation, and non-photorealThis paper presents a new method for rendering high dynamic range (HDR) images on conventional displays. The method is simple, efficient, robust, and easy to use. It manipulates the gradient field of the luminance image by attenuating the magnitudes of large gradients. A new, low dynamic range (LDR) image is then obtained by solving a Poisson equation on the modified gradient field. The method effectively compresses the dynamic range of HDR images while preserving fine details and avoiding common artifacts such as halos, gradient reversals, or loss of local contrast. It also enhances ordinary images by revealing details in dark regions. HDR images are becoming increasingly common in computer graphics. They offer advantages over standard low dynamic range images and are useful in various applications. However, displaying HDR images on low dynamic range (LDR) displays is challenging. This paper addresses this problem by proposing a new technique for HDR compression that enables HDR images to be displayed on LDR devices. Previous work on HDR compression includes global (spatially invariant) mappings and spatially variant operators. Global mappings, such as tone reproduction curves (TRCs), are simple but may not preserve local contrasts well. Spatially variant operators are more flexible but can be complex. The proposed method is a gradient domain HDR compression technique that manipulates the gradient field of the luminance image by attenuating large gradients. This approach is effective, computationally efficient, and avoids halo artifacts. The method works by first computing the logarithm of the luminances, then identifying and attenuating large gradients. The gradients are then used to solve a Poisson equation to reconstruct a low dynamic range image. The method is applied to various types of images, including HDR radiance maps, panoramic video mosaics, and medical images, with good results. The method is based on the assumption that the human visual system is not very sensitive to absolute luminances but responds to local intensity ratio changes. The algorithm manipulates the gradient field of the luminance image by attenuating the magnitudes of large gradients. This approach preserves fine details and avoids common artifacts. The method is implemented using a multi-resolution approach, where gradient attenuation factors are computed at different levels and propagated to the full resolution. The gradient attenuation function is computed in a top-down fashion, and the final gradient attenuation function is used to solve a Poisson equation to reconstruct the low dynamic range image. The method is tested on various HDR images, including those of a streetlight on a foggy night and a Stanford Memorial church. The results show that the method effectively compresses the dynamic range while preserving fine details and avoiding halo artifacts. The method is also used for LDR image enhancement, where it reveals details in dark regions and maintains good contrast. The paper concludes that the proposed method is a simple, efficient, and robust technique for HDR compression. Future work will explore the application of this method in various areas, including denoising, edge manipulation, and non-photoreal