ryobg

@stokerg thank you for this follow up. I think these descriptions are helpful, but not fully because as the reason I asked about these is I work with pixel precise information. And I want to know how the values are going to be transformed in more... mathematical sense. For example, it is common for for high-level work, Auto-leveling / normalising to cut a bit (few percentages) from the bottom or the top tonal range, so the shades in between stretch better. I'm looking for pure mathematical stretch in my case, but looking at these descriptions it is hard to infer what is going on actually. Thank you again.

Converting HDR to most other formats is likely a loss of information. The idea that you can convert a generic hdr, png, bmp, even jpg to most DDS formats and you won't loose information is misleading too.

Well, I might be minority and while I consider DDS opening / saving as a nice feature, I don't think is so critical or important. Usually, it's just one more step to convert from your exported raster to a DDS file using an external & specialised tool. At general case, it might be even just drag & drop / double click operation (if you have made a simple batch script for example). Also, it comes to mind, Unreal Engine for example, does not accept DDS files, it wants to compress them instead.

So, I have jumped in the hype train and went full ahead. Pros: 1. UI with potential, nice feel and uniformity. Workspaces! 2. Advance tools, layers and algorithms (hdr, panoramas, selection and etc.). 3. Fantastic performance! Cons: 1. Too basic UI customization: can't scale fonts or buttons, most panels can't be resized, toolbars are fixed, keyboard shortcuts are highly restricted 2. Weird workflow with channels. A simple decompose/recompose or layer blending mode (bound to the rgba whatever) would be much more convenient. 3. There is no way to see when a channel clips? I frequently normalize channels by using different algorithms, which leads to: 4. No scripting. The macros or that procedural texture layer are too basic and does not cover it. Waiting for how it will go in future!

Hello, I'm unable find any documentation about what are these actually doing. I'm working with a bit more precise pixel information and I am interested in the way how these are implemented. 1. At least I want to know according to what these are working (e.g. min/maxed axes from HSL/HSI/HSV/LAB or whatever). 2. Also, after I used the Black & White filter, none of the Auto-whatever actually increased the grey scale values? In my case, half of the dynamic range was exported only. 3. Side, but related question: is it possible to retrieve the min/max values of the raster during the definition of the procedural layer? Thank you and best regards. P.S. Scratch #2. Apparently, Auto-whatever needs to work on merged grayscale (my gray scale conversion was as layer).

Heya, I don't think that this solution can work for many people. I have already applied some OS scaling and I already use many applications which allow UI customisation (fonts, dpi scaling or else). If I'm to apply a system wide change it will turn everything upside down. The only long term solution is proper UI/UX implementation, let people adapt the tool to their needs. It does not need to be some fancy configuration interface where you specify what is what, even a simple file will enable most people. Of course, the bigger problem is that the software is mostly likely very rigid inside to allow such feature (albeit basic) to be implemented. Best regards.

I can't believe that the software internals are so bad that simple icon/button + font customisation is too hard?! Tons of portable image software allow such options. Hell, even 20+ year old Gimp you can make themes - custom icons, text scaling and whatnot and is much more portable. Sigh... anyway, hope in the next several years UI/UX customisation improves :)