วีดีโอ

@@cmluna90 Should be the same with some key replacements. If I remember Mac correctly, you will be replacing “control”with your “command” key and will be replacing “alt” with the “option” key (example: control shift c would be command shift c for you, or shift alt would be shift option) My “H” hotkey (to hide elements bounding boxes) is not a standard hotkey and would need to be added in your Live2D settings if you want to use it. Let me know if those key replacements work, if not I will check them tomorrow. I have a Mac its just for work so I dont use it often for Live2D-that being said I will check to make sure if needed 👍

Thank you so much, happy it helped. Really appreciate providing clarity on the issues you were having too, I'm sure it will help someone out!

thank you so much!

Hey Destova, I have been taking a break and haven't streamed in a while so forgive my knowledge being rusty, but you def shouldn't have to go as low as 40 on the shadow. I think I kept mine between 90-100% even. I would double check your two dynamic mask filters (toggle them both off in terms of visibility) and make sure both of those are working properly. If it seems like the gaming scene blur dynamic mask is the culprit, you can bump up your base values (try 0.15 or 0.2 instead of 0.1). I would also recommend even re-adding the dynamic mask filters and re-inputting values and mask sources. I recall that the StreamFX plugin would sometimes just...decide to not work, especially if you are changing things a lot. Unsure if that's the case with Exeldro's plugin, but it's worth a shot to re-add them in. There's a button in the bottom left of the filter for 'defaults' that should allow you to reset them, though deleting and adding from scratch might be a better option. In terms of the software you're capturing though, it shouldn't be any different if you are using transparency or spout2 capture. If you are using a chroma key to remove a background color however, you may need to first mirror the chroma keyed source to use as your initial vtuber source. I found it acts strange when you key it out on the same source that you add more masks on. There's also always the possibility that this could totally be obsolete at this point with new plugin & OBS updates. Hopefully that's not the case, wish I had more info for you!

@@flux_does_stuff Nah honestly that's amazing the amount of detail you got for being rusty! I'll take all of that into consideration the next time I feel like opening up my stuff to play with it. Thank you so much for replying!

Absolutely, Streamerbot has a separate tab for timers which you can link to actions. I don’t go over it in the video but can link documentation: docs.streamer.bot/guide/settings/timed-actions Since timed actions run an action every x period, and you want to have items go away after that time, you could have one action for the item toggle and one action to turn the items off. At the end of the initial action (to enable the toggle), have it enable the timer to toggle the item back off. This is how I did my colour change toggles! You can also keep it all within one action if you use some c# or some if statements.

@@flux_does_stuff Thank you so much for this! I'll give it a go here momentarily! Appreciate your help!

Got it the second time, I missclicked some values in dynamic mask and it end up being 20.00 instead of 2.00 😆

Hi Shahelion, Totally possible that OBS has changed, this is the 3rd version of this tutorial and stuff changes constantly so I wouldn’t be surprised if it is obsolete or needs another update. I will try out the LUTs when I am off of work or sometime this week and see if they are still working for me, then get back to you with a comment or comment edit. As long as the lighting scene you’re referencing contains only the game capture, it should not be mixing/adding color from sources that are not included.

Thank you! Absolutely, you would just make another parameter going from 0 to 1 and rig the eyes with opacity similar to to the mouth (and make sure to adjust the input settings in VTS). The eyes open/close mesh or related deformers should be inside of the squash stretch deformer in the hierarchy so they still move with the rest of the model. Because you would then have 4 pngs+ total it will get a little more complicated with overlap. I would recommend separating the eyes on their own layer in the .psd so they do not disrupt your speaking parameter (you could separate the mouth too, this tutorial was just very barebones with separation). If you don't have any way to separate them in the .psd, you can separate them by using the manual mesh tool (button to the left from the one used at 6:19, looks like a triangle with a pen) and mesh *just* the eyes in Live2D by clicking a mesh boundary around them instead of automesh. And area outside of that mesh boundary will not be visible, so you can still separate them in a pinch :) I will be doing an advanced version of this tutorial including a more in depth lip sync, hotkeys/toggles/animations, and autoblink at some point though!

I think it's hard to avoid due to the nature of additive layers. The only real way to mitigate it is via LUT or color grading, haven't found much of a way around it :(

LOL thank you, I tend to cut it off cause I have a really hearty loud laugh lol, try my best to limit myself being obnoxious in tutorials haha

Thanks for watching! Actually CSP supports layer comps now as well (ver 3.0 only) but still may be useful for easy exporting. I wish you the best of luck with your own picrews!

Totally doable with a gif! Most .gif files can be imported directly to a .psd and will import each frame as a layer. If I'm understanding correct then you probably have 2 sets of the same floating frames--one set with mouth open and another with mouth closed. You can use the same method in this video, but around 7:15, you'll have an extra step of grouping all floating mouth open frames in 1 deformer and all mouth closed frames in another deformer (hold shift or control to collect them and make a new deformer). Then continue on and rig those two deformers as if they were your mouth open & mouth closed. Deformers can use opacity and multiply so it should still work the same, you'll just have one deformer holding multiple components for open/closed instead of one artmesh for each. As for the floating animation, there's two options for you, but its a little more complicated because both will require you to use the animation menu. It boils down to 'rig the floating in live2d' as you suggested, OR 'use the frames in frame by frame animation.' If its a unique animation I would do option 2 so you can keep it! If it's simple then you can probably rig it easier in Live2D. I'll try my best to explain both but maybe can do a video on this at some point! You will have to make a new parameter for the floating which you will rig depending on which route you go. L2D Deformer Method Hold ctrl and select your mouth open/mouth closed deformers and make a new deformer 'floating' for both of them. Then you can make a new parameter 'floating' or something that will go from 0 to 1. Make your 'floating' deformer active on this new parameter using your two keyform button. At 0 it stays the same, and at 1 you can hold shift and drag the deformer up. Frame by Frame Import all your frames for the levitation, add a parameter for the levitation, and make the bounds equal to the number of frames (ex. if it's an 8 frame loop, you'd go from 0 to 7). Then rig visibility for each frame--ex. frame 1 would be visible on param=0 with all other frames invisible, frame 2 visible on param=1 with all other frames invisible, and so on. Once you have that rigged, you can make a new animation file (set to 1920x1080p, and make sure your framerate matches what you use in VtubeStudio, ex if you use 60fps then the animation needs to be 60fps). Drag your model into the screen and resize it so you can see it ok (the placement doesnt matter too much since we arent exporting to video or anything). You'll want to open the dropdown 'live2d parameters' so you can view your parameters, and set keyframes for your floating parameter. If you do the deformer way, you really only need 3 keyframes (bottom keyframe, top keyframe, then back to bottom to loop). You can open the dopesheet and set the graph to be ease in so it looks smooth. If you do it the frame by frame way, you will have 1 keyframe for each frame, and you'll have to set your dopesheet to be a stepladder so it doesnt flash/blend your frames. You can export the animation file and add it to your model folder. Exporting the actual model is the same but you may need a larger texture atlas to fit all your frames. The last thing is that in VtubeStudio you'll have to set your idle animation to your floating animation! I know that's a lot of info and its hard to translate text to working on something, but if anything I hope it can guide you a bit or know what to look up for assistance. If you have questions feel free to comment back or dm me on twitter or something!

Oh and sorry, forgot to say, if the ball is attached to your pngs, you can detatch it in your psd file by selecting it and erasing or using ctrl+x and ctrl+v. Then when you import, it will be in its own layers / artmesh.

do spout2's .dll file show up in your obs plugins folder at all? it will be in maindrive:/Program Files/obs-studio/obs-plugins/64bit if it's not there, the install was unsuccessful and i would recommend manually dragging the .dll file to your plugins folder or reinstalling it

You mean in OBS? My recording settings are: Format: MPEG-4 (.mp4) Encoder: NVIDIA NVENC H.264 Audio Encoder: FFmpeg AAC However, if you mean recording transparent video--to my knowledge, there's no way to record OBS videos with an alpha channel. Recording in .mov will fill void area with black, and does not record transparency. If you want to record a video with 'transparency' you will need to fill the background with greenscreen (color source) and remove it in post.

Yes, expressions are totally possible to rig for pngtubers using the same method standard .moc3 live2d models would use! Each expression will need to be tied to its own parameter in Live2D (example: excited expression with star eyes as a parameter going from 0 to 1, with 1 being 'on/visible' and 0 being 'off/not visible') and then you can toggle this using a hotkey in VtubeStudio. For pngtubers, rigging something like the example above using a simple opacity change would be very easy to implement. More complicated Live2D stuff would let you change the physics intensity or even toggle/add an animation. And if you separate your layers more (ex. each 'speaking' and 'not-speaking' png might consist of 3-4 pngs, one for hair, one for eyes, etc.) you can use VtubeStudio to change the colors and stuff like that. Lots of options for sure! I plan on doing an advanced version of this tutorial when I have some time and it will cover animations/autoblinking, expressions, and a more advanced lipsync. No guarantee on when that'll be out though, so if you want to do some research atm, i'd recommend watching a video on 'rigging live2d expressions' and then 'setting up hotkeys in vtubestudio.' Hope that helps, happy rigging :)

@@flux_does_stuff I appreciate you taking the time to respond! I'll look more into the expressions in other videos, and I look forward to your future videos! ♥

Are you still having this issue? I don't believe it should happen but you can also try installing exeldro's source mirror and mirror the source to other scenes if it's still happening.

Hi, Unfortunately I will not be redoing this video unless there is a major update for OBS/shaderfilter or a new way to execute it, especially since this is already the 3rd video on this topic. This is an intentionally fast paced video as I disclosed in the intro. It's intended to be an update for anyone that was using the previous versions of this method and would like to change to the new dynamic mask shaderfilter from streamFX. The other two videos, which are at a slower pace and include informative blurbs in regards to the method & the usage of source mirrors, are both linked in the description. I recommend watching those for supplementary help, they all have timestamps if you are looking for something in particular. Additionally the original method by Zephyr is extremely similar to this but with the old plugins--that video is also linked in the description if you would like to use that for reference. If you have questions you are free to comment for any help.

hi there wordoftruthministriesocean4325 you need steam to download vtubestudio. so unless you use a different program to run your model, you will have to at least use steam to download it first before you can run it without: github.com/DenchiSoft/VTubeStudio/wiki/Starting-without-Steam

Okay? That's crazy?

Hiya! Do you mean prgmitchells blur? There is a note in the description about its removal with some alternatives linked. I recommend composite blur by FiniteSingularity: github.com/FiniteSingularity/obs-composite-blur It functions very similar, but if you run into issues, there's also a new version of this tutorial available that is more current: th-cam.com/video/BsbFCIovf0w/w-d-xo.htmlsi=S3QkkbUfFfgiFHCw

Thanks for watching! I understand your point but I will continue to use the zoom effect because personally it helps me learn (i picked it up from another streamer for that reason). I also know it can be helpful for people hard of sight or people that do not have access to better video quality and would have a harder time following along without being able to zoom fully/see menus. This video in particular was definitely faster paced (it was meant to be a cut and dry version of the older two versions of the tutorial, so anyone returning wouldn't have to listen to me go over old topics) and I can see why the zoom speed combined with the pace of the video could cause confusion. I will definitely take that to heart and take more care/time when zooming for any faster paced videos I make in the future. Thanks!

@@flux_does_stuffno worries at all! The tutorial is great I just set it up for my own stream and it looked gorgeous especially during my painting streams. Thank you so much for posting this I always wondered how this was done, much love <3 !

I agree with you :)

thank you so much 🧡

I think i made a mistake, i put the model file into picrew data and nothing poped up into vtube studio pls help :(

@@executrixofficial Hiya! You mean you moved your model to the folder and it won't show in your list of models? First open and close the menu a few times, or close and re-open VTS. VTS does not instantly update the model list when you add a new model in, so it may take a moment for it to show up. Did you make sure that your model data was stored into its own folder before adding to VTS Live2D Models foldeR? VtubeStudio does not read model files alone--the moc3 file, json files, and texture folder (with texture pngs) will need to all be together within their own folder. If that doesn't help just explain in a bit more detail and Ill help as best I can!

​@@flux_does_stuffI followed your advice and it worked this time thank you for making this tutorial. I just have 2 question about 1.how am I supposed to record / stream with vtubestudio, Is it already built into the app or do I need a different program 2. Now that I've uploaded my model will it automatically save save in Vtubestudio?

@@executrixofficial 1-- To stream you will need to download & install a streaming software like OBS and link your streaming platform to it. There are tons of tutorials on how to get it setup! After installing and getting used to the app, you will need to add VtubeStudio as a source in OBS. The current best way to do this is using a plugin called Spout2. I have a tutorial on my channel for that here: th-cam.com/video/zA7v_DNVxOI/w-d-xo.htmlsi=Nm7EPIYhRni4zOUE 2-- Yes, your model will now always exist in VtubeStudio so long as you don't edit your Live2D Models folder. Any hotkeys or model adjustments made to your model will save to the model's JSON file. I do recommend backing it up from time to time (as with any files it's good practice). But yes, it will stay there and be accessible every time you open the app now :)

Aa okay, good to know, I remembered that wrong--thank you for the correction. That's actually great because people don't have to use their trial up for this quick rig!

i always happen to drop an f bomb early so I try to take it out in case youtube nerfs me HAHA

Hiya! Within Vseeface general settings, there should be a section dedicated to Spout2. Make sure that the box is checked by “Use Spout2.” You can also name the sender, which you should see in your list of senders under the OBS spout source. Let me know if that works!

thank you ill try it right now!@@flux_does_stuff

heya, OBS is open-source and SLOBS is unfortunately a version of OBS which basically has a ‘wrapper’ over it. it changes some things for QoL, but runs a bit worse and unfortunately will not support all OBS plugins since it is not *exactly* OBS. If the plugin isn’t working right in SLOBS, my first thought is that it’s probably just because SLOBS doesn’t support it or hasn’t made a version of it they can support. I do not use SLOBS personally so I cannot speak much for troubleshooting, sorry! I would highly recommend swapping to OBS if you want easier compatibility with plugins. Plus, as I said, it’s effectively the same as SLOBS (since they just took the source code), and it runs better! Here’s a tweet regarding the situation: twitter.com/obsproject/status/1460782968633499651?s=46&t=7Iuf4udafN1FCXB_54SJKQ

Hiya Mable, Did you set your background to transparent in VTS? You'll have to change your background--it will be the option close to the bottom labeled color picker (it will have a 'special' label next to it). You can select it and toggle on 'transparent in capture' and it should work.

@@flux_does_stuff It worked! thank you so much

atm I don't believe tiktok has a public api so if they are using SB with tiktok interactions its probably something very complicated using c# code! however if they are just using it for effects, scene transitions etc, then they are probably just using SB with OBS and streaming to tiktok from their PC. if you want to connect your SB to OBS, you will need to download and install OBS's websocket plugin and connect the address in the server/client list in SB. then you'll be able to use SB to control scenes, sources, and audio. there are lots of helpful tutorials on getting into SB, I would recommend Nutty's channel, his tutorials go over the websocket install: th-cam.com/video/CcXAs-qZ0Ys/w-d-xo.htmlsi=DNskIQZDcBATNsXC Andilippi and GaelLevel are also great people to learn from!

@@flux_does_stuff Thanks for the reply! Yeah I'm realizing it's something really confusing involving code and I honestly have up 🥲 I really appreciate it though!

@@flux_does_stuff I found out they use something called Lumia Stream with it no code necessary;0 Although I'm not that smart but hopefully I can figure it out, this video was very informative regardless! And Ty for the response (sorry if this is a double reply my other reply isn't showing up on my screen so 😭)

thank you so much!

its totally worth it i promise hahaha

Playing with the values would be the best bet, yes! You can pull down the input values in the 2nd one from 5 to something lower, or adjust the base values. If you want, there's an older version of this tutorial (also without using StreamFX) which is based on blending methods. If you do it this way, the effect is naturally less strong and a bit darker. The plus side is you can just adjust it with a color correction filter and change the opacity. Link here: th-cam.com/video/pPARYmVWITw/w-d-xo.htmlsi=a2vb6-k5wU3Mz6Qk