ModLab news & patch notes

New major update is here! This version brings Volume from Alpha which let's you generate normal maps based on volume of objects. This feature might in many cases replace horizontal profiles, but there is nothing stopping you from combining both methods. Since picture is better than thousand words there are some examples: https://media.giphy.com/media/dBOMSIStNvbBfRoB77/giphy.gifhttps://media.giphy.com/media/dBOMSIStNvbBfRoB77/giphy.gif https://media.giphy.com/media/KxbGJ1I6HAl2UADNEk/giphy.gif https://media.giphy.com/media/UnxujdnxQURTHdxynu/giphy.gifhttps://media.giphy.com/media/UnxujdnxQURTHdxynu/giphy.gif Copyright/Attribution Notice: Irina Mir (irmirx) Big thanks to all Patrons. If you would like to support my projects then please visit https://www.patreon.com/RavensLab Patch notes:

• Volume from Alpha
• Improved behavior of curve editor
• Export for individual maps
• Serie of optimization
• New method for compressing bit range
• New preview modes
• Actuallized presets saving/loading

Well, yesterday release didn't went all that well. Not that are issues with ModLab, at least I am not aware of any, but both reddit posts got shutdown for whatever reason (I no longer even link Patreon nor Steam). Anyway, lets move on. We will do mix of Nebula and Normal maps options this time and see how it goes. 1) Nebula generator - Better asset control: For instance control over asset density, probability of placement of specific asset. Possibly using different noise function with preview for placement. 2) Nebula generator - Planets: I've implemented very advanced atmosphere rendering in past, together with semi procedural planet rendering (as you can see in Wayward Terran Frontier), throwing them into ModLab is viable project, altho list of sliders is going to be massive. 3) Nebula generator - 3D environment: Ability to place assets within pseudo infinite sphere and ability to export Spherical maps as well as Box maps. 4) Normal Maps - Generate volume: This is going to be funny one, there are two softwares that can generate volume from alpha map of sprite. One of them boosting it's only software it can do that. Anyway, I believe I can do better. It would work as alternative to Horizontal profile. It would most likely use distance field to check how deep is pixel within the sprite (further for edge, higher Z position it would have). Now the plot twist compared to competition: It would have curve editor, same as Horizontal profile and it would let you to reshape that volume, so you could make it to have spherical behavior, linear, flattened peak or whatever you desire. As always, big thanks to all Patrons, if you want to vote, please consider supporting this project on Patreon https://www.patreon.com/posts/27385010

This patch was focused around Nebula generator. This is first working interation of said feature that will let you export backgrounds into PNG. https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/49367b14fc88f49c928ab3cbf205fb4385fc9139.png Video tutorial will be up in few days, but in meantime there are some instructions:

• After starting ModLab click "Background" on top bar.
• Now you can start loading assets into Level 0.
• If you don't have assets or if you want to use default assets, you can click "Load Base assets" in bottom left menu to load 5 normal maps.
• From this point on, it should be straight forward: Poke sliders till you like the result.
• You can use left click to move omni lights, right click to open their content menu, espace to close that menu.
• Exported PNGs includes alpha channel, you can remove it in any 2D editor if you don't want to that channel. You are not losing anything because of existence of said data and when removed you will get the exact result you seen before export. You can however use it to create multiple layers and such.

Next Patron poll will including features extending background generator as well as normal map related feature. We will see what community will pick. As of the moment there might be some rough edges. If you encounter some fatal issue then please let me know. If you would like to support development of this app check Patreon https://www.patreon.com/user?u=7785848 Big thanks to all current Patrons.

https://www.youtube.com/watch?v=e_mjAM945ig Second tutorial describing how is ModLab used for processing ships for Wayward Terran Frontier: Zero Falls https://media.giphy.com/media/fLsoC1wIAEWwMG12LE/giphy.gif Patch notes:

• Switched Normal A and Normal B channels for combine
• Renamed Normal 2 map slot to Normal B
• Fixed issue with dominant light control drag taking priority over load buttons
• Added export png function for background (WIP)

Big thanks to all my Patreons: Kageve Gonswic Zero Falls is Good Dmitri Klein Alex Yang Tiberiumkyle Kevin Zechner If you find this project useful, consider supporting it at Patreon: https://www.patreon.com/user?u=7785848

Don't worry, previous choices didn't go anywhere. We are just taking slightly different approach for this poll. Reason for this is simple, while overall user count is reasonably high (over 8 000), number of Patrons is fairly low. So, this time we will decide general direction. 1) Full 3D support - I will push ModLab in direction of properly supporting 3D. Recently I tested it with full unwrapped textures and I was really pleasantly surprised by results. There would be ability to load your 3D models to preview results in realtime, possibility of support for object space normals (I know it seems like outdated technology but information from object space normals can be incredibly useful) and so on. 2) Nebula generator and exporter - In this case I will focus on finishing background generator, add control over results and ability to export it as image, also as cube map and spherical map for 3D games. Later on things like adding randomly generated planet (using same technology as Wayward Terran Frontier, that means physically accurate atmosphere) and more. If you want to vote in this poll please consider becoming Patron and support development: https://www.patreon.com/posts/third-poll-22648462 Big thanks to all five existing Patrons.

Just removing alpha information from Roughness map export.

So, software was originally designed mainly for sprites but over time it grew into very powerful texture processing tool, used by almost 8 000 developers and modders. One of issue that came up from original design is interpolation around edges... NOT anymore. This verison of ModLab will let you chose UV wrapping sampler state. https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/c2803147a090885e45ba7d9b11472791fb02a978.png If you like the software and find it useful, then please consider supporting further development on Patreon: https://www.patreon.com/user?u=7785848 I you came across any critical issue please let me know, I will look into it as soon as possible.

v1.2.1 -DDS loading support -TGA loading support (without alpha) using .NET Targa Image Reader -Fixed crash on loading Height map before Dif -Automatic flat grey diffuse in case user wants to process just a normal map I had very little time to do some extensive testing altho this required some massive changes. If you run into any issue please let me know.

Second Patreon voting poll concluded some time ago and I bring goodies. Results shown that users want to improve roughness map, which was admittedly the weakest part of ModLab. To improve it I need some extra data, specifically curvature. This version of ModLab let you generate curvature from Normal map and use it to improve Roughness map. On top of it even if you provide your own Roughness map you can use new data to add some details to it. As seen here: https://media.giphy.com/media/4QFL474mkqMLFiXbWJ/giphy.gif Curvature itself is quite powerful information. For that reason if Curvature is generated it will get saved (unless exporting for WTF:ZF format) and can be used in any 2D editors. Red channel contains grey mid point information, green isolated edges and blue isolated cavities. https://media.giphy.com/media/mPIvPvbDuRtYJfKHCp/giphy.gif I made quick tutorial about using ModLab for sprite: https://www.youtube.com/watch?v=9lKMgj9lwKw Change log: -New Roughness processing! -Large amount of features relevant to roughness -Curvature from Normal map -Wrote bypass for PNG compression -Rare bug when part of saved image is missing should be fixed -Better GUI separators Big thanks to all Patrons. If you want to support ModLab, want to be able to vote regarding future development or simply want access to new version slightly sooner then please visit Patreon page: https://www.patreon.com/user?u=7785848 Also you can follow me @twitter https://twitter.com/JanOrszulik

So user had issue with windows open file dialog and asked for another method of loading files. Alright then, patch v1.1.1 is out and you can now load textures by drag and drop onto load file buttons. Also fixed one possible crash that I accidentaly discovered.

Patch it out! Since I've elaborated content of this patch in blog posts I will keep this short and just link them:

Main features:

Complete rework of medium and large elements, greatly increasing their quality and flexibility: https://steamcommunity.com/games/768970/announcements/detail/1671276544921786698 https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/b2fa1b5c310106e465ef1a948f85f0d87f6b99a2.png Optimization resulting in massive fps gain when processing large textures: https://steamcommunity.com/games/768970/announcements/detail/1669026014353287013 https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/01160601c1c84ff54ac31b79d076c49f3b5a17e1.png There are also some quality of life improvements as slider value normalization and left menu scrollbar mouse wheel response. I am going to monitor feedback for a week after which I will start Patreon poll for next set of features. Big thanks to all Patreon supporters. If you like this software and would like to support its further development please consider supporting on Patreon https://www.patreon.com/user?u=7785848

Alright, so other part of this update: Optimization First of all I need a reasonable goal. Processing a 4k x 4k texture on a 4 years old mid tier GPU at 60 fps seems like a reasonable goal, so let's aim for that. To make it fun let's include the latest changes with the most performance hungry settings (using largest possible kernels for small and medium elements). So initial test: Load texture, Generate From Diffuse set, set the highest radius for both elements + using most complicated kernel aaaaand we are at 17 fps. Alright, so I need to make this aprox 3 times faster. In terms of rendering this is kinda scary especially considering that I was trying to make this efficient in first place and I am not going to compromise quality in any way. First of all let's do some profiling to see what's the most expensive, I have pretty good idea but it is good to confirm and see how exactly expensive it is. What we have learned at this point is that VS graphic debugging sucks. The moment I am under 60 fps (thus the point you want to profile) I am unable to capture a frame and the whole debugging process glitches. Well it's not like I am mad, I am just furious. Ok then, let's profile smaller texture and just use the ratios between individual processing steps. Okay, so I can now capture a frame, I can get quick analysis but I can't get extended "Error have occured" - Thx for info, I almost didn't notice. *flips table* Alright , let's roll with Nsight. Luckily for me Nsight works like a charm. Profiling confirmed what I expected. While at it I've noticed minor Color Surfaces mismatch and here we have another fps. In other words 53 ms per frame, I would need to get to aprox 16.7 ms for frame to be at 60 fps. Currently worst offenders are processing added recently and for those there is no possible optimization as they are simply blurs with large kernel (remember when I said that I wont compromise quality, so no, I am not going to undersample these, blurs shaders used here already make advantage of linear interpolation). The most reasonable way at this point is to process these only when required, thus on value change, it is obvious approach, really which I was avoiding for few minor reasons but I guess it's time. TL DR version of what's going to happen: There are several parallel processing branches running at same time, some are merging at some points, all are contributing to final result. Goal here is to update minimum portion of the "tree" system when their input changes. After these changes we are suddenly at 46-47 fps with very minor fps drop (3-4fps) when moving sliders, everything still responds in realtime, 13fps/5ms to go. So let's give same treatment to billateral processing. https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/65cb0842f7d782ae2680dcd26b18763c0e9bc7fc.png Changes above resulted in improvement from 17 to 56 fps with worst case scenario drop to 49 fps for split second (realistically unnoticeable). I've tested also some 8k x 2.5k textures and performance was still very reasonable. This concludes optimization pass for patch 1.1. Main feature still remain changes regarding large and medium elements as mentioned in previous post: https://steamcommunity.com/games/768970/announcements/detail/1671276544921786698 So when is v1.1 going to be released? Probably during next week, no promises there, I want to clean up few things, finish slider value normalization and add few other QoL improvements (scroll wheel to scroll left side of GUI incoming). As always I would like to thanks all Patrons, you are the reason why I am working on this patch. Patreon: https://www.patreon.com/user?u=7785848

First Patreon poll concluded some time ago and since then I am have been working on the two most voted options. Since talking about optimization would be boring I will talk about other one, which is nothing else than quality of medium and large generated elements.

The issue:

As I stated in the poll I knew about the issue. I generally discouraged people from using medium and large elements and stick with pixel perfect small ones which would do for most of the cases but not always and it was limiting factor for Modlab's flexibility. The issue was caused by my approach for medium and large elements which was to simply undersample input data thus processing smaller images and then upscaling them back. https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/87d0487bd881cac1b7db79f4047c2644ffd1b664.png As you can see on the images above that caused some artifacts and that wont do.

The solution:

To greatly improve quality I changed how processing of mentioned elements works. Instead of simply undersampling, input data are processed by filtering and since all processing in ModLab is so far realtime it provides decent options. Not only that artifacts were removed and quality greatly increased, but new options are now available. You can now chose radius of medium and large elements as well as Z profile of processing filter giving you even more control over results. https://media.giphy.com/media/1ykANmMsAANHW5unZE/giphy.gif https://steamcdn-a.akamaihd.net/steamcommunity/public/images/clans/31777225/b2fa1b5c310106e465ef1a948f85f0d87f6b99a2.png

What's next:

I will now spend some more time working on other most voted option, which is optimizations. I have some plans on that matter but don't expect miracles. Generating and processing normal maps in realtime simply aren't cheapest operations. Once everything is ready I will release v1.1 patch. There is chance I will release some tutorial videos or at very least I will be streaming how I use upcoming version of ModLab to process some sprite assets stream will be later on available on Youtube and linked here.

Special thanks:

I would like to thank Patrons, for your financial support and your help in deciding direction of ModLab development. Patreon: https://www.patreon.com/user?u=7785848

All Patrons (in this round this includes suporter tier) are able to vote in current poll which will run till Monday 12th March after which I will start working on implementation of winning option. 1) Image export for Background generator - In this iteration I would be aiming to create export into one layer 2D image. Resolution would be settable between 512x512 to 4096x4096. In future this would open path for further improvement of space background generator and very possibly export of spherical 360 3d map. 2) Ability to chose shader model - Schlick's approximation is currently used for rendering, on top of that it is slightly modified version. I would like to offer option to pick between several shader models. First additional model would be very basic phong model (which is much more primitive than currently used Schlick). Main motivation behind this would be creating framework itself for realtime shader model swap. In future this would lead to ability to preview textures in specific rendering environment being identical or close to Unity/Unreal Engine. 3) Optimization - realtime processing is done on GPU but it can be demanding, especially on older GPUs.. or on any GPU if used for processing large textures. There are several optimizations that can be done. Largest one would be to process texture only on value change and that would be first step. In future this will open way for some additional functionality that would use gained performance. 4) Increased quality of Medium and Large elements - I am not very happy with with medium and large elements quality when generating normal map from Height map or Diffuse map. Luckily I know exactly what is the cause and what to do with it. For cost of some performance (and of course some work) I can greatly improve quality of these elements. Patreon link: https://www.patreon.com/user?u=7785848

All Patrons (in this round this includes suporter tier) are able to vote in current poll which will run till Monday 12th March after which I will start working on implementation of winning option. 1) Image export for Background generator - In this iteration I would be aiming to create export into one layer 2D image. Resolution would be settable between 512x512 to 4096x4096. In future this would open path for further improvement of space background generator and very possibly export of spherical 360 3d map. 2) Ability to chose shader model - Schlick's approximation is currently used for rendering, on top of that it is slightly modified version. I would like to offer option to pick between several shader models. First additional model would be very basic phong model (which is much more primitive than currently used Schlick). Main motivation behind this would be creating framework itself for realtime shader model swap. In future this would lead to ability to preview textures in specific rendering environment being identical or close to Unity/Unreal Engine. 3) Optimization - realtime processing is done on GPU but it can be demanding, especially on older GPUs.. or on any GPU if used for processing large textures. There are several optimizations that can be done. Largest one would be to process texture only on value change and that would be first step. In future this will open way for some additional functionality that would use gained performance. 4) Increased quality of Medium and Large elements - I am not very happy with with medium and large elements quality when generating normal map from Height map or Diffuse map. Luckily I know exactly what is the cause and what to do with it. For cost of some performance (and of course some work) I can greatly improve quality of these elements. Patreon link: https://www.patreon.com/user?u=7785848

First minor patch, generally quality of life changes.

v1.0.1

• Fixed window maximalization and snapping
• Ctrl + Scroll for zoom to position
• Smoother scroll
• Removed camera autocentering
• Inverted zoom
• Removed some old license note*
• Patchlog

*To be clear, you can use ModLab for both Commercial and Noncommercial projects.

First minor patch, generally quality of life changes.

v1.0.1

• Fixed window maximalization and snapping
• Ctrl + Scroll for zoom to position
• Smoother scroll
• Removed camera autocentering
• Inverted zoom
• Removed some old license note*
• Patchlog

*To be clear, you can use ModLab for both Commercial and Noncommercial projects.