I keep raw files. As new process versions of Lr/ACR come out, you can upgrade the processing with a click of a mouse -- even for entire directories -- it you want. I never store JPEGs for reuse; just PSDs, PSBs, and TIFFs. Every time you modify and save a JPEG file, you rerun the lossy discrete cosine transform algorithm all over again; that's IQ loss you don't need.
So when you modify a TIFF file (I don't use the other ones you describe), you aren't losing any quality do to rerunning the algorithm?
No. As I think I've told you before, lossless compression is indeed lossless.
Is TIFF lossless compression?
TIFF files support several compression methods, two of the most common being ZIP and LZW. Both are lossless compression schemes, meaning they reduce file size without discarding any image data.
LZW, which stands for Lempel–Ziv–Welch, is a dictionary-based compression algorithm developed in the 1980s. In the context of TIFF files, LZW is identified by a compression tag value of 5. The algorithm works by identifying repeated sequences of data, storing them in a dictionary, and then replacing repeated sequences with references to this dictionary. This method is fast and widely supported. It works well with images that have large areas of uniform color, such as line drawings or indexed-color graphics. LZW compression was once encumbered by patents, which discouraged its use in some open-source software for a time, but those patents have since expired. While LZW is simple and relatively efficient, it does not include entropy coding techniques like Huffman or arithmetic coding, which limits how much compression it can achieve on more complex images.
ZIP compression in TIFF files uses the same DEFLATE algorithm found in ZIP file archives and PNG images. It is identified in TIFF files with a compression tag value of 8. DEFLATE combines sliding window compression, similar to LZ77, with Huffman coding to further reduce redundancy. This method tends to achieve better compression ratios than LZW, especially on photographic content where image complexity and noise reduce the effectiveness of dictionary-based approaches. Although ZIP compression is generally slower than LZW, particularly during decompression, it is well supported in most modern imaging software, including Adobe Photoshop, and Lightroom. However, compatibility may be an issue with older software that has not been updated to recognize or decode ZIP-compressed TIFFs.
The choice between ZIP and LZW often comes down to the type of image and the software in use. LZW may be preferred for compatibility and for simpler images with large uniform areas. ZIP is usually better for compressing photographs and continuous-tone images, especially when file size is a priority and the software ecosystem supports it.