Currently working on implementing subtitles in the module.
You will be able to custom code them within a table, even using translated content.
There is a small difficulty because I’m actually working on two modules: this video module and another module for music. The subtitle part was coded on the music module, and it will be a little more time-consuming for me to embed it in the video module.
I translated the showcase song into English and Russian. Take a look and listen to both versions:
Russian audiotrack and Russian subtitles.
Japanese audiotrack and English subtitles.
The subtitles use their own code and are not dependent on the audio. You can change them exactly how you want. In the future, I might also add custom properties to the code so you can change more stuff, including the color, size, etc.
This feature is not yet finished. Expect updates and improvements soon.
This is the only downside to this method of playback. You need to upload each frame individually. The only way to make it easier is to skip every second frame, which will not make a very noticeable difference, except if there is a lot of moving objects. As I stated you can use the asset manager and bulk import the images. You can do 200 at a time. You will get rate limited though. So you will need to either wait 24 hours or get someone else to bulk import them. I was able to upload all 2.1k frames for the showcase video in about 3 days (In those 3 days I was also coding the video player, you can do it in like 3 hours.) Also while uploading your frames, make sure there is nothing that the auto moderation will think is inappropriate. Any text needs to be blurred even if it does not violate anything, a simple 2 minute Photoshop will save you weeks of waiting for Roblox support to reply. That’s all I really have to say for now until I find a possible solution to upload them faster. As for everyone else reading this, if you know how you can upload decal images with http requests please leave a reply. I checked the API myself and couldn’t find anything related to that.
it’s impossible to get the ids quickly, and you can use alt accounts to bulk upload them, Maybe you can try the Open Cloud Assets API to upload decals?
I have searched wide before this, but I see this for the first time. This might actually be the solution. I’m going to try it out. Thanks for telling me. Time to remember my 5th grade Python knowledge
Can confirm that this does work. I couldn’t remember my Python knowledge from 5th grade so I had to remember my Javascript experience. I was able to write a script that will post an image and return you the assetID.
For anyone interested the code is below. I will make a guide on how to use it once I’m done with the “subtitles” feature. I will write a script that will automatically upload all frames from a folder and store their IDs to the table needed for the module to work. If the rate limitation is actually 60 uploads per 60 seconds then that means you will be able to upload you bad apple video in like 30 minutes, fully automatically, compared to the painful day of clicking, switching alt accounts, getting moderated, and copy pasting numbers to a table.
This module is pretty cool, it’s pretty tedious and all but you can get cool stuff done with it, awesome job! Heres something that we could do with the help of @bloodbonniekingnoob1 and your module
Good job with this!
I have made a python script to upload assets. It’s very simple to use and automatically puts everything together as it shoud. This is only a test version that is not guaranteed to work. Currently known issues: when you get your account moderated it will error. The only way to save your IDs is to copy them from the terminal output and paste them somewhere to store for further usage. You won’t be able to know how much exactly you have done, but if you want use a tool to check the amount of lines, that will be the amount of frames, remove the old ones and start uploading again. Beware, it’s easy to get banned. If you ask who exactly will be banned, it will be the user that created the API key, regardless if it’s a group or a user one.
How to make an API key? Go to your creator dashboard, under the credidentials tab.
Now create a new API key, name it whatever, select API system as “Assets API” and make sure to add both read and write permissions. Add 0.0.0.0/0 as the IP address and create your key. Make sure to copy it and to never share it. Now store it somewhere in a safe place.
How to use the python script? First of all, make sure you have Python installed. Open up a terminal and type python --version. If it works then you have Python. Make a new .py file, paste the script from below. Open it and paste your API key in the given position (look the comments). Now depending on who owns the API key, a group or a user, enter the group ID or your User ID of your account, make sure to change .Group() to .User() and vice versa. After that, use the video method in the links section to extract your frame. You need to put your .py file inside an empty folder, and then make a folder called “files” inside. Don’t put the .py file inside, it must stay outside the files folder. Extract your frames into that files folder. After you are done, open up a terminal, navigate to the directory using cd in terminal.
Now you need to install the libraries needed. Type pip install rblx-open-cloud. If pip doesn’t work for you find a guide online on how to fix your Python installation. And now simply python main.py.
Be patient, your connection might be slow. It takes quite a lot of time to get those images posted. After it’s done, it will stop printing new assetIDs, it will make a text file called output.txt. Open that, this is now your frames table ready to be pasted in Roblox Studio.
If you get banned while you are still uploading it will error, in case it was only 1 asset that was banned, copy the IDs from the terminal output and store them for later. Count the lines to know at which frame it stopped and continue later.
This is not the final script, I just want to share it while I’m at it for people to test and possibly find an improvement that I missed.
The Python script:
import rblxopencloud
import os
creator = rblxopencloud.Group( # Make sure to change this to .User( if you are using a userid, and to .Group( if you are using a group.
UserID/GroupID, api_key="API_KEY"
) # Change this to your API key you created and copied. Make sure not to share it.
assetIDsArray = []
def uploadAsset(currentFile):
with open(currentFile, "rb") as file:
asset = creator.upload_asset(
file, rblxopencloud.AssetType.Decal, "Example Asset", "Decal Description"
) # Feel free to change the name and description, it doesn't matter.
if isinstance(asset, rblxopencloud.Asset):
print(asset.id)
assetIDsArray.append(asset.id)
else:
while True:
status = asset.fetch_status()
if status:
print(status.id)
assetIDsArray.append(status.id)
break
fileDirectory = './files'
for fileName in os.scandir(fileDirectory):
if fileName.is_file():
uploadAsset(fileName.path)
content = "local FrameIDs = {\n"
for i in range(len(assetIDsArray)):
content += " " + str(assetIDsArray[i]) + ",\n"
content += "}"
with open('./output.txt', 'w') as file:
file.write(content)
file.close()
This script uses a third-party Python library called rblx-open-cloud. More information about it here:
Please share feedback from this script if you used it. I’m still not sure if it fully works. I was able to upload the first 300 frames of the bad apple video, but I’m currently experiencing some internet issues and can’t really be for sure if it did work. My frames appear as “ghost frames” and I don’t know what caused it, probably some privacy issue from permissions. If you experience the same bug, please tell me.
Use the script at your own risk. I do not take responsibility on how you use it. Don’t upload too much with it, since it’s not still confirmed that it works due to the “asset privacy” issue I have. First to try upload a few frames and confirm they do work and appear.
I figured out why the images appear gray. I made an ooppise while reading the documentaion of the Open Cloud API and thought that would upload an Image asset, but in reality it’s a decal. A decal and an Image are different. An Image type uses rbxassetid://id, while a decal can’t use that. You are still able to convert it though! I’m going to stop making new replies until I finish v1.1 with the subtitles and finished python script to upload them + a new guide on how to use it. Thanks for all compliments, @Haxonium your work is pretty cool! Would love to see how your game will look like in the future, seems interesting.
For more information how to convert a decalID to an imageID read here (before I document it in the post):
Also sorry for any grammar mistakes in my replies, I don’t have enough time at the moment. I write most stuff from my phone.
unfortunate that we have to write our own modules and screenshot a bunch of frames because roblox doesn’t want to let us use their already existing video category.
One thought I have is why you can’t “buffer” the videos, where instead of loading the entire video, instead you load only, say, the first 15 seconds, and then while those 15 seconds play, you load the next 15 seconds, and so on and so forth. I think this would help loading times a ton.
I think you got confused while reading the script. That’s literally what it does. It loads the first 400 frames of the video at once, then it starts playing them and once the frames start to become low, for example there are only 100 more frames loaded, the script will load the next frames 200, because it uses :Destroy(), the old frames are no longer rendered on the screen, so it doesn’t create more lag. Also I’m sorry for the big delay, I’m currently experiencing issues with my internet provider and there hasn’t been internet for the past few days. I will get the update done once I solve that issue. If you know Python then you won’t have an issue completing the script yourself, the hardest part was to make the new loading function work since you need to do it on the server.
I’m going to provide a small tutorial for this. I’ve got a website for getting the frames and a script that adds them to a dictionary automatically (may work on it more to skip frames that have the same id in case you accidentally inserted 2 of the same frame)
Update: turned the script into a small module. Things you can do: choose the path of the folder/model/container or whatever
Choose if you’d like to use getchildren or getdescendants.
Oh also you have access to the table, just do Module.FrameTable or something like that.
im trying to add more features since there’s barely anything i can add, but I’m thinking of adding an optional boolean that adds the table to an “universal” module that holds all of the dictionaries.
Also it no longer works because I’m the best at coding, especially modules
edit: module is fully done. I will provide a community resource post soon
Here’s my module for making your life a little easier: Asset dictionary
also here’s the website: https://www.onlineconverter.com/
(By the way, my module works on any type of downloadable asset, aka animation, decal, texture, mesh, image label, image button, shirt (not crap keyboard) and pants)
I am glad to hear that you have made this module. I hope more people will find it useful.
News on Update v1.1!
News on Update v1.1!
I’m almost done finishing up some stuff. Expect the update pretty soon (in a few hours or tomorrow).
What changed?
A full guide with a video on how to use the module and import the frames (with the new Python script).
Improved loading times and performance on the low end. Automated some calculations so you don’t have to go around the module yourself.
Added subtitles, which you will be able to code yourself. They depend on their own format, just like real subtitles. You will need to sync them with the music manually; of course, if you don’t like them or don’t need them, you don’t have to use them. Once again, they are customizable; you can pick the container to play them in. More features to the formatting will come soon. I’m planning to add some features like being able to pick the color or even add a gradient. I also might make it word-for-word with animations in the future, but I still don’t know how to realize that, so I don’t guarantee it.
Many more things are subject to change; if you have suggestions, don’t hesitate to leave a reply.
The update is not out; it will be in a few hours or tomorrow as of writing this.
Oh, I thought it was loading the whole video. When I went to the demo place and opened the Shift+F3 menu, it appears to be trying to pre-load about ~1300 frames, which I assumed was the whole video.
Implemented a solution for uploading a large number of assets. Use the script here.
Changes
Changed the resample mode of each frame to Pixelated (configure it in the module) also changed the scale type to fit, both of these changes are optional, you can remove them if the scale adds gray lines and cuts on the screen or if the resample mode makes it too pixelated or low quality on your screen
Image.ResampleMode = Enum.ResamplerMode.Pixelated --// Improves quality on larger screen resolutions. (rarely makes it worse looking)
Image.ScaleType = Enum.ScaleType.Fit --// Adds gray lines if the video does't fit t
Imrpoved documentation.
Better performance.
Added a FPS argument to the function to easily change it.
Removed some useless stuff that caused lag and rearranged some parts to improve readability and performance.
Merged the “Usages” and “Showcase” parts together.
I removed most details from the post so it’s easier to read.
