Likewise tot was released officially in 2020, this marks the Radiance 5.3 official release. They have introduced the radcompare program to facilitate the comparison of Radiance tool output with reference outputs during unit-testing. Moreover, they resolved an issue pertaining to the reading of exceedingly large files (> 2 GBytes) within the rcollate tool. The RVU tool has implemented a modification to the picking function. Likewise ensuring it now disregards transparent and void surfaces.
Significantly, they have enhanced the rtrace tool by introducing the -orRxX options, which enable the tool to output mirrored and unmirrored contributions and distances. This enhancement has also led to a more streamlined and unified approach to handling effective ray lengths within the code. Additionally it is also now included in the FFmpeg.
They have adjusted the behavior of rlam, with -iaN now reading N lines from a stream instead of N characters. Likewise they consider this change more useful. So they also created the rsplit program as a counterpart to rlam, designed to handle more complex rtrace output by splitting it into separate files or streams.
Another significant improvement stems from addressing an issue highlighted by David G-M. This resolution effectively eliminates the biasing of isotropic tensor tree samples in genBSDF, a factor that had previously led to subsequent peak extraction failures, particularly with the “aBSDF” type. Furthermore, the introduction of new functions max() and min() to the .cal library expands its capabilities and usability.
With the aim of enhancing the accuracy of scattering interpolation representation input, like those from pabopto2bsdf, they have enriched the bsdf2klems tool by including random incident patch sampling. This addition aligns with the existing sampling methods applied to other input types, ensuring consistency across the board.
The gendaymtx tool has gained a notable enhancement in the form of the -A option, which facilitates the computation of the average sky over all input time steps. Additionally, the rtrace tool has evolved to possess the ability to produce color images using any of the -ovrx options. The resolution of inconsistencies related to the -oN output, particularly when comparing ray casting versus full trace scenarios, has been successfully achieved.
Radiance 5.3 Additions
In addition they have adjusted the behavior of rlam, with -iaN now reading N lines from a stream instead of N characters. Likewise they consider this change more useful. So they created the rsplit program as a counterpart to rlam, designed to handle more complex rtrace output by splitting it into separate files or streams.
Efforts to improve getinfo have led to the ability to hop over the resolution string with -c and header elide mode. Use +d to include the resolution string in output and -d to elide it with – and -c options. The pvalue tool’s -u option now means “uncompressed” with -r. Additionally, it allows skipping bytes on standard input. A similar tool, rcode_norm, converts to and from a 32-bit/pixel portable surface normal representation.
The rtpict tool now supports the -o* option to produce layered images for various output types from rtrace. A new script, rcode2bmp, facilitates the creation of images from the new rcode_* output maps.
We’ve improved the efficiency of rmtxop in cases where matrix multiplication is faster when evaluated right-to-left. This adjustment aligns with the final component being a column vector. You can now apply trailing unary operators (-t, and -s or -c) to the resulting matrix.
We’ve expanded radcompare to include depth, normal, and ID file types. Furthermore, we’ve introduced a “BigEndian=” header line in rmtxop and radcompare for automatic byte-swapping.
Matrices in dctimestep now support RGBE image loading. To enhance Gensky and gendaylit, we’ve introduced the -y option for more accurate Michalsky solar position calculation. The rsplit tool now parses words between whitespace when specifying a space as the separator. Furthermore, vwrays, pinterp, and pmblur2 can now read encoded depth files.
The ability to perform blocking in rcollate has been introduced through the -o RxCxR1xC1 facility. Genworm, genrev, and gensurf can now accept the number of segments as expressions of previously defined variables.
Incorporating new options, gendaymtx now offers the capability to output sun descriptions and skip matrix output upon user request. Moreover, an -M option has been introduced, enabling the direct output of solar modifiers for rcontrib. These advancements bring enhanced versatility to the tool’s functionality. Additionally, the integration of support for leap days in WEA input stands out as a notable enhancement, thanks to the contributions of Ladybug Tools.
Furthermore, the eplus_uvf tool has undergone an update to ensure seamless compatibility with IDF version 9.x. This update includes the implementation of a version number check to facilitate seamless integration.
They have also successfully resolved a crash issue related to rtrace -om or -oM when dealing with mesh instances containing materials, addressing a specific concern. Moreover, they substantially improved the rtrace flushing function to handle any value for -x effectively, eliminating any deadlock issues within the controlling process.
In terms of enhancing efficiency, they strategically eliminated startup procedures for rtrace when ray-casting operations are the sole requirement. This effectively reduces unnecessary overhead. They ingeniously refined the isotropic sampling method employed in bsdf2ttree, using a “zipper” approach inspired by its successful application in genBSDF. This refined approach significantly enhances the accuracy and efficiency of isotropic sampling. Likewise see also the FFmpeg as well as the VirtualDub Player.
They have introduced a memory usage limitation for high-resolution tensor tree BSDFs during cumulative cache in MC sampling. The limit is now set to 250MBytes per BSDF for standard architectures. Additionally, they have empowered users with control over super-sampling in the bsdf2ttree tool by including -n and -s options.
Moreover, the surface normal codec utilized in the ambient file has undergone improvement, ensuring the precise storage of orthonormal directions. Additionally, the peak extraction method has been significantly enhanced to rectify undervaluation of transmission in various scenarios. Likewise see also the DivX 3.11 Codec as well as DivX MPEG-4 Audio Compressor.
In another advancement, gendaymtx has incorporated support for the -u option, allowing for the omission of data columns (time steps) when the sun remains below the horizon. Lastly also see the Cannot play .avi/GMP4 files on here.