<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Alec Kunkel on Medium]]></title>
        <description><![CDATA[Stories by Alec Kunkel on Medium]]></description>
        <link>https://medium.com/@aleckunkel?source=rss-29d4aee17786------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*xt8rzDWRp--0lsgmQfvTZQ.png</url>
            <title>Stories by Alec Kunkel on Medium</title>
            <link>https://medium.com/@aleckunkel?source=rss-29d4aee17786------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 12:02:52 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@aleckunkel/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Four Resources to Generating Your Next NFT Collection]]></title>
            <link>https://medium.com/madebymckinney/four-resources-to-generating-your-next-nft-collection-24b7d4b5753f?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/24b7d4b5753f</guid>
            <category><![CDATA[nft]]></category>
            <category><![CDATA[code]]></category>
            <category><![CDATA[generative-art]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Thu, 03 Feb 2022 23:38:31 GMT</pubDate>
            <atom:updated>2022-02-03T23:38:31.491Z</atom:updated>
            <content:encoded><![CDATA[<p>Do you really think that the team behind Board Ape Yacht Club really hand-made all 10,000 variations? No chance! They automated the process using code. Generative art is a really fun concentration but it can be difficult to know where to start. Fortunately, there are some really great tools out there to get you started.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*CyWtUr24_J3764js" /></figure><p>Here are a few of my favorite tools to generate your next NFT collection.</p><h3>#1 Appsus NFT Generator — A Great Place To Start</h3><p>Coming in at number one is the Appsus NFT art generator. There are a number of NFT generators to choose from, but what makes this one special is that it comes with a user interface. With the tool open, you just need to select your asset directory, order each NFT layer, set rarity, and click generate. There is a really handy preview window that lets you see your NFTs come to life.</p><p>Check it out here — <a href="https://github.com/AppsusUK/NFT-Art-Generator">https://github.com/AppsusUK/NFT-Art-Generator</a></p><h3>#2 NotLuksus NFT Generator — Simplicity at Its Finest</h3><p>The NotLuksus NFT Generator is not far off from the first tool listed here. It does pretty much the same thing. Feed it your assets and you’ll be greeted with a collection of unique NFTs. The one thing I do love about this one is how bare-bones it is. As long as you’ve organized your folders correctly, it only requires one line of code to run.</p><p>View the code — <a href="https://github.com/NotLuksus/nft-art-generator">https://github.com/NotLuksus/nft-art-generator</a></p><h3>#3 Blend_My_NFTs — Create a 3D Collection</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*BPEVRSXCLnWEU3rZ.png" /></figure><p>This next tool is probably my favorite out of the bunch. This Cozy Studio is the genius behind the Blend_My_NFTs tool. This generative art tool is made specifically for <a href="https://www.blender.org/">Blender</a> and jam-packed with features.</p><p>The main gist of it is, to create your NFT collecting within a Blender file. Then run the included script file to automate the rendering of each NFT. There is a separate config file where you can specify all generative details.</p><p>I recommend checking out the documentation and Youtube channel to see everything it’s capable of.</p><p>Repo can be found here — <a href="https://github.com/torrinworx/Blend_My_NFTs">https://github.com/torrinworx/Blend_My_NFTs</a></p><h3>#4 Google Colab — A Different Approach</h3><p>If you’re not familiar, Google Colab is a tool used to run Python code in the cloud. What makes this a great tool for generating NFTs, is that you can use machine learning in a bunch of creative ways.</p><p>Using machine learning to create art has its own movement and there are many approaches to this method. If you’re interested in going down this rabbit hole, <a href="https://ml4a.net/">Machine Learning for Art</a> is a great resource to get you started.</p><p>One of my favorite approaches is using style transfer to generate art. I’ve included a gist below that you can open in Colab and run with your own dataset. <a href="https://gist.github.com/iamkunkel/033045f3feda2d009d93a081ed1ac1af">https://gist.github.com/iamkunkel/033045f3feda2d009d93a081ed1ac1af</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/0*YFnvSnolVDoEoq6b.png" /><figcaption><em>That notebook is what I used to create all of the apes on </em><a href="https://www.generateanape.com/"><em>https://www.generateanape.com/</em></a></figcaption></figure><p>These tools only scratch the surface of what’s out there! There is a lot out there you can find with a quick Google. And if you choose to use any of these in your next project, don’t forget to credit the creators :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=24b7d4b5753f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/four-resources-to-generating-your-next-nft-collection-24b7d4b5753f">Four Resources to Generating Your Next NFT Collection</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Scrape TikTok Data Using Python]]></title>
            <link>https://medium.com/madebymckinney/scraping-tiktok-data-using-python-a2fde787f160?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/a2fde787f160</guid>
            <category><![CDATA[code]]></category>
            <category><![CDATA[tiktok-app]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Fri, 28 Jan 2022 18:44:29 GMT</pubDate>
            <atom:updated>2022-01-29T12:16:46.398Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*pRQRZd4XulRsSLkN.png" /></figure><p>Today we’re going to chat about something slightly different than 3d development and design! This week we’re going to break down how to scrape data from Tiktok using Python. Specifically, we’re going to scrape all the video data on a user’s timeline. To give you an idea of where we are headed, I’ve broken down the project into 3 steps:</p><ol><li>Set up a virtual environment</li><li>Install the required dependencies</li><li>Scrape and cleaning data</li></ol><p>Just a heads up, this tutorial is going to assume you’re already set up with python and at least a little familiar with writing code.</p><h4>Step 1: Setting up your virtual environment</h4><p>Before we jump into the code, we’re going to kick off by creating a virtual environment. Open up your terminal and navigate to a place where you want to keep your files organized.</p><p><em>Pro Tip: Use “mkdir” to create folders and “touch FILENAME” to create files directly from your terminal</em></p><p>Start with the following line to create your virtual environment.</p><pre>python3 -m venv NAME</pre><p>After you created your venv, activate it using the following command.</p><pre>source NAME/bin/activate</pre><p>This will give you an environment isolated from your base python installation. Setting up a virtual environment is a great way to test new code without the risk of breaking anything globally. If you have any additional questions/issues, I recommend checking out the documentation <a href="https://docs.python.org/3/library/venv.html#creating-virtual-environments">here</a>.</p><h4>Step 2: Installing the required Dependencies</h4><p>The next bit of setup we’re going to tackle is installing the dependencies to our venv. To scrape all of the data, we’re going to use this <a href="https://github.com/davidteather/TikTok-Api">unofficial API</a>. While we’re here, let’s also go ahead and install Pandas.</p><pre>pip install TikTokApi Pandas</pre><p><em>Important Note: At the time of writing this, I was running into issues with this exact package. If you run into a captcha bug, uninstall that package and use the one below.</em></p><pre>pip install git+<a href="https://github.com/Daan-Grashoff/TikTok-Api">https://github.com/Daan-Grashoff/TikTok-Api</a></pre><p><em>If you cloned the repo, you can install the dependencies using “</em>pip install -r requirements.txt&quot;</p><h4>Step 3: Writing Our Code</h4><p>Finally, with everything installed, we’re ready to start writing some code! Create a new file, “Main.py”. This will serve as our entry point to using the API. Open up your file in your preferred text editor. I’m a huge fan of VS Code but use what works for you.</p><p>The first thing is to import our dependencies and instantiate the TikTokAPI. The API variable will be used throughout our code.</p><pre>from TikTokApi import TikTokApi<br>import pandas as pd</pre><pre>api = TikTokApi.get_instance()</pre><p>Within our file, I’ve broken down the code into three functions: one that asks for a username, one that scrapes the data, and finally a function to clean the data.</p><h4>3a. Asking for user input</h4><p>The first function, “inputUserID”, kicks off by asking for a username. Once we’ve received the username, we get to use the TikTok API for the first time. Specifically, we pull and return the “userID” and “secUID”. These values are required when pulling profile feed data.</p><pre>def inputUserID():<br>  userName = input(&quot;Enter Username: &quot;)<br>  userInfo = api.get_user(userName)<br>  userID = userInfo[&#39;id&#39;]<br>  secUID = userInfo[&#39;secUid&#39;]</pre><pre>  return userID, secUID</pre><h4>3b. Scrape the data</h4><p>Let’s start by defining a few variables we will use throughout the function. Don’t worry about these values just yet. As we walk through the remainder of the function I’ll explain the purpose of each.</p><pre>def getUserVideos(<em>userDefID</em>, <em>secDefUID</em>):  <br>  # Sets up a few variables that will be used throughout the code<br>  cursorValue = 0<br>  hasMore = True<br>  df = pd.DataFrame()</pre><p>Out of the box, the API is limited to pulling 30 posts. To get around that, we will use a couple of the variables we set up at the beginning of the function. Specifically cursorValue and hasMore.</p><p><em>The cursorValue lets the API know where on the page those 30 videos ended</em></p><p><em>The hasMore value lets the API if there are more videos on the user’s timeline</em></p><p>Using the power of a while loop, as long as hasMore is true, we’ll continue to pull more data based on where the cursor is.</p><pre>  while hasMore:<br>  <br>  else:<br>    print(&quot;No more data&quot;)</pre><p>Within this while loop, we start by making a call to the api. There are a couple of values being passed through:</p><ul><li>userID/secUID: These came from our initial function</li><li>cursorValue: The current cursor place on the page</li></ul><p>After each iteration, we update those variables and pass them back.</p><pre># Function to pull video data until all have been pulled<br>  while hasMore:<br>    # Makes initial call to the API<br>    TikTokList = api.user_page(<em>userID</em>=<em>userDefID</em>,<br>    <em>secUID</em>=<em>secDefUID</em>,<br>    <em>cursor</em>=cursorValue)</pre><pre>    # Cleans the data after it&#39;s scraped<br>    data = cleanData(TikTokList[&#39;itemList&#39;])<br>    df = df.append(data)</pre><pre>    # updates our variables based on the latest scrape<br>    cursorValue = int(TikTokList[&#39;cursor&#39;])<br>    hasMore = TikTokList[&#39;hasMore&#39;]</pre><p>So, that code will run until “hasMore” returns false. Once we’ve got all of our data, the last step is to output it into a csv.</p><pre>print(&quot;No more data&quot;)<br>    df.to_csv(&#39;UserVideos.csv&#39;)</pre><p>And violà! Here is our final function to scrape user data.</p><pre>def getUserVideos(<em>userDefID</em>, <em>secDefUID</em>):  <br>  # Sets up a few variables that will be used throughout the code<br>  cursorValue = 0<br>  hasMore = True<br>  df = pd.DataFrame()</pre><pre>  # Function to pull video data until all have been pulled<br>  while hasMore:<br>    # Makes initial call to the API<br>    TikTokList = api.user_page(<em>userID</em>=<em>userDefID</em>,<br>    <em>secUID</em>=<em>secDefUID</em>,<br>    <em>cursor</em>=cursorValue)</pre><pre>    # Function used to clean<br>    data = cleanData(TikTokList[&#39;itemList&#39;])<br>    df = df.append(data)</pre><pre>    # updates our variables based on the latest scrape<br>    cursorValue = int(TikTokList[&#39;cursor&#39;])<br>    hasMore = TikTokList[&#39;hasMore&#39;]</pre><pre>  else:<br>    print(&quot;No more data&quot;)<br>    df.to_csv(&#39;UserVideos.csv&#39;)</pre><p>At this point, we’ve implemented an input to ask for a username and a function that pulls all of the profile data, but if you were to look at the result, you’d see a nested object not fit for analysis. Let’s convert that object into a nicely formatted CSV.</p><h4>3c. Processing our data</h4><p>You probably saw that in our while loop above, we had a “cleanData” function being called, that’s what we are going to be creating here.</p><p>Within our API response, we have a whole bunch of data. And a lot of that data is nested data. Here is an example of what that data looks like:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/972/1*SQnnhEpMfrgLOlPIoZimag.png" /></figure><p>It starts with our itemList and within that, we have information for each video. Each video also has its own nested elements (those are the items with a plus next to them). What we need to do is create a function that uncouples all of these elements and writes their children to a new “row”.</p><p>When our data is ran through the function below, we process each data pull into a flattened dictionary that then gets converted to a Pandas dataframe. While converting our data, we also look for any value that is stored in our nested_values array. If the criteria is met, we expand it into it’s own element.</p><pre>def cleanData(<em>data</em>):</pre><pre>  nested_values = [&#39;video&#39;, &#39;author&#39;, &#39;music&#39;, &#39;stats&#39;, &#39;authorStats&#39;]</pre><pre>  # Creates a dictionary for our df to be stored in<br>  flattened_data = {}</pre><pre>  for id, value in enumerate(<em>data</em>):<br>    flattened_data[id] = {}<br>    # Loop through each element<br>      for prop_id, prop_value in value.items():<br>      # Check if nested<br>      if prop_id in nested_values:<br>        for nested_id, nested_value in prop_value.items():<br>        flattened_data[id][prop_id + &#39;_&#39; + nested_id] = nested_value<br>      # If it&#39;s not nested, add it back to the flattened dictionary<br>      else:<br>        flattened_data[id][prop_id] = prop_value</pre><pre>  return pd.DataFrame.from_dict(flattened_data, <em>orient</em>=&#39;index&#39;)</pre><p>Now every time we want to clean our data, we just run it through this function. What gets returned is a pandas dataframe that we can then write to a csv.</p><p>And with our last function wr, we’re just left to call them at the end.</p><pre>userInfo = inputUserID()</pre><pre>getUserVideos(userInfo[0], userInfo[1])</pre><p>Give it a try using python Main.py. You’ll be promoted for a username and then a csv with all of your data!</p><h3>Final Thoughts</h3><p>At this point, you should now have a working scraper that pulls data from TikTok. It’s a great start, but it can absolutely be taken further. Maybe you want to deploy your code and automate it to run. Or connect your script to the Google Sheets API so that your data can be sent directly to a sheet. This code can be used as a building block to your own reporting dashboard!</p><p>If you’re interested in just getting the code, you can find that in the github repo here: <a href="https://github.com/iamkunkel/TikTok-Performance-Scraper">https://github.com/iamkunkel/TikTok-Performance-Scraper</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a2fde787f160" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/scraping-tiktok-data-using-python-a2fde787f160">Scrape TikTok Data Using Python</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[One Week With Nvidia Omniverse]]></title>
            <link>https://medium.com/madebymckinney/one-week-with-nvidia-omniverse-d52684580e35?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/d52684580e35</guid>
            <category><![CDATA[3d]]></category>
            <category><![CDATA[nvidia]]></category>
            <category><![CDATA[omniverse]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Thu, 20 Jan 2022 13:19:58 GMT</pubDate>
            <atom:updated>2022-01-31T21:44:26.162Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/560/0*y-cgACmzH1lIIo-r.png" /></figure><p>It’s been about two weeks since Nvidia Omniverse was officially released to the general public. Since then I’ve just begun to explore what it’s capable of. In this post, I’m going to break down what Omniverse is and a few of my favorite features so far.</p><h3>WTF is Nvidia Omniverse?</h3><p>After spending a week with Omniverse, it’s still hard for me to define it. And I think that was intentional. On the surface, Omniverse is a collection of tools to help streamline your 3d workflow but underneath it&#39;s the foundation to redefine the entire process.</p><p>Dramatic? Maybe a bit. But if you think about it, this has been a long time coming. As long as I’ve been around we’ve hacked together these 3d “pipelines”. Usually, you&#39;re stuck having to create the same asset multiple times across each platform. With Omniverse, you’re given an entirely new way to work between applications and teams.</p><p>But with so much functionality, where do you start? Here are a few of my favorite features so far!</p><h3>1. Seamless Blender Bakes</h3><p>First on the list is the Blender beta application. A very minor feature compared to the rest of the platform but for Blender creators, potentially, a huge timesaver. This new tool lets you bake light data for an entire Blend file into one image texture with just a few clicks.</p><p>After you’ve downloaded this beta, you get access to the following panel. Within it, you get access to a few handy USD import and export options and the ability to bake your scene.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YKacwrXKI3Ud3tLLl8euCw.png" /></figure><h3>2. Audio2Face</h3><p>Next up is Audio2Face. This ML-powered tool generates mocap data based on an audio clip you give it. It really is as simple as uploading your own WAV file and then assigning it within the parameters panel. Right away the placeholder model will come alive. From here you can take the motion data and remap it to your own custom models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0PsQ0PDI-llRHuQfjZsyYw.png" /></figure><p>This is potentially a massive time saver when you compare it to setting up shape keys for all facial positions and then streaming mocap data from your mobile device.</p><h3>3. Asset Browser</h3><p>Keeping an organized asset library isn’t always easy, but Omniverse has some great features to find your files. It starts with ML object detection! Anytime you add a new asset you’ll be able to search for it without having to label it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DM_pAhGGOosaYmBkcExS2A.png" /></figure><p>It’s also worth noting that there are connectors built for Sketchfab and some of the other cloud asset sites. That way you can peruse those models alongside your local ones.</p><h3>4. Collaboration Features</h3><p>Probably one of the most notable features on launch is the ability to spin up a local server and work collaboratively. Once you’ve completed the setup, you can have a team working on the same scene with live updates in Omniverse.</p><p>This feature isn’t totally new, but what’s cool is that your team doesn’t have to be working in the same app. You can have someone in Maya and someone in Unreal Engine and see all of those updates live!</p><h3>5. Omniverse Kit</h3><p>Last on the list is a bit different than the others. Instead of this being a specific feature, it’s the ability to build on top of the Omniverse platform. <a href="https://docs.omniverse.nvidia.com/prod_kit/prod_kit/developer_api.html#kit-long-python-api">Omniverse Kit</a> is the Python-based API that comes with the infrastructure to build extensions.</p><p>Alongside the API, they’ve also released the application Omniverse Code to help assist in building your own features.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/918/1*NWKIH-JqEufm1mX7SAHnOg.png" /></figure><p>These five things only really scratch the surface. As you get started with Omniverse, I recommend checking out the <a href="https://www.youtube.com/watch?v=dvdB-ndYJBM&amp;list=PL4w6jm6S2lztLazLC7P0I4SnX3gxdL1Ad">Nvidia YouTube channel</a>. They’ve put together a great resource for getting started with the platform. Even if you’re a team of one, there is a very good chance you’ll be able to use it!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d52684580e35" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/one-week-with-nvidia-omniverse-d52684580e35">One Week With Nvidia Omniverse</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building the Ultimate Cloud-Based Blender Asset Library]]></title>
            <link>https://medium.com/madebymckinney/building-the-ultimate-cloud-based-blender-asset-library-df8c7f881d35?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/df8c7f881d35</guid>
            <category><![CDATA[asset-management]]></category>
            <category><![CDATA[blender3d]]></category>
            <category><![CDATA[blender]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Mon, 10 Jan 2022 13:03:09 GMT</pubDate>
            <atom:updated>2022-01-10T13:03:09.030Z</atom:updated>
            <content:encoded><![CDATA[<p>I know this might not be the sexiest writeups on Blender, but it’s one that needs to be written. We’re not going to be using geometry nodes to create a seamless satisfying animation. No. Instead, we’re going to be talking about organizing your Blender asset library.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-WsXsc9F4KjD3_gm5e2pJQ.png" /></figure><p>Just before we jump into my setup, I figured it’d be worth kicking it off with a quick breakdown of the asset library and how it works.</p><h3><strong>How the Asset Library Works</strong></h3><p>Blender 3.0 included the long-awaited asset library. A new system designed to make accessing your models, textures, and character poses ultra-convenient. You can find your library by toggling any window to the “asset library”.</p><p>This new tool works in two ways.</p><ol><li>Accessing assets within the current Blend File.</li><li>Accessing asset libraries defined through your Blender preferences menu.</li></ol><p>The trick to using assets across Blend files is marking a model as an asset and then storing that blend file in a folder you’ve marked as an asset library. To mark an asset, just right click on your model through the layer panel and click “mark as asset”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*72IVC21BzW7YhSD9KoTF1Q.gif" /></figure><p><em>A quick tip before marking your models as an asset, apply all your transformations using CTRL + A. This will give you a much better preview image and ensure everything is orientated correctly when you want to use it.</em></p><p>After you’ve marked something as an asset you can continue to organize things even more granularly through catalogs. And you can nest, and nest, and nest these catalogs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*CpFhi_ZZMSBnmE_484p54A.gif" /></figure><p>Before you start using your new asset, I highly recommend also filling in where the model came from and a description about it through the “Asset Details” window. That way if you ever have any questions, you can quickly reference that information.</p><p>It’s a pretty slick new feature, but to use it effectively, we need to make sure our assets are organized.</p><h3>Organizing Your Library</h3><p>I’m a bit of a weirdo, in the fact that I’m constantly jumping between my two computers. So when considering how to set up the library I opted for a cloud solution. Specifically <a href="https://www.google.com/drive/">Google Drive</a>.</p><p>To get this to work correctly, you need to use the <a href="https://www.google.com/drive/download/">desktop application</a>. The desktop app gives you access to your drive from your folder explorer.</p><p>Before you start organizing assets, let&#39;s set up a general 3d asset folder. Within that, I’d recommend creating subfolders for each type of asset. In the screenshot below I’ve broken mine into “Materials”, “Models”, and “HDRIs”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*-URskVzsACamqw1UDbbd_w.png" /></figure><p>I also recommend making your asset folder “available for offline access”. It’s going to make grabbing your assets much quicker from within Blender.</p><p>If you’ve used the asset library before, you’ll know that when opening a library you’ll often get a delay. To fix this issue, I’ve opted to divide up my entire asset library into sublibraries. That way if I’m looking for a 3d model of a brain, I don’t have to sift through all the different furniture assets I have.</p><p>At the time of writing this here is a list of my sublibraries:</p><ul><li>Anatomy</li><li>Architectural</li><li>Buildings</li><li>Clothing</li><li>Electronics</li><li>Food &amp; Drink</li><li>Furniture</li><li>People</li><li>Plants</li><li>Transportation</li></ul><p>Each of which is also divided up into subfolders. So for example, if you open up the transportation folder you’ll see folders for “Cars”, “Planes”, etc. These subfolders won’t have any effect on the way our assets are displayed, but they do help me stay sane trying to keep all my Blender files organized.</p><p>Once, you’ve got your folders set up, you can begin adding each as a library within Blender. Below is what my setup looks like:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/639/1*sTMr9CMUZmfl7hSWT4HkFg.png" /></figure><p>So now that our libraries are loaded, how do we organize things in Blender? That’s where catalogs come back in! And, unfortunately, there isn’t a one size fits all setup that I’ve found effective. Each library has its own organization and to go through each one would be a waste of time! Instead, I’ve gone ahead and included a repo that has each of the libraries broken out for you. Feel free to check it out here: <a href="https://github.com/iamkunkel/blenderassetlibrary">https://github.com/iamkunkel/blenderassetlibrary</a></p><p>After you clone it to your local (or cloud storage), just go into Blender and mark each folder as an asset library. At this point, you’ll be set to start building your collection. Fair warning it gets wildly addicting.</p><p>If you take a look at my repo, you’ll notice that each of my asset catalogs matches pretty closely to the folder setup. I’ve found that keeping these consistent makes finding and logging assets much easier.</p><p>Without going through each library, that’s the primary setup! You can absolutely add to these folders, or if you download additional libraries just add them through the preferences menu.</p><h3>Creating Assets</h3><p>If you have been following along, by now you should have a pretty killer asset library setup, but how do we add to it? It’s honestly really simple:</p><ol><li>Create a new Blender file.</li><li>Save within the corrisponding folder.</li><li>Create your model and mark it as an asset.</li><li>Within that Blend file, toggle to the corresponding catalog and organize as necessary.</li></ol><p>After adding and organizing, don’t forget to add asset details as well!</p><p><em>Another quick tip for you: If you need to edit an asset, just right-click on the thumbnail and you’ll be able to open up the file from the asset library. This will save you from having to navigating through all of the folders within your setup.</em></p><p>So now you’re familiar with how the asset library works, how to organize a library, and how to add to it! And as you start to work on your setup, feel free to experiment and change it up. Find something that makes sense to you. And please comment with any questions or tips that have helped you organize things!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df8c7f881d35" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/building-the-ultimate-cloud-based-blender-asset-library-df8c7f881d35">Building the Ultimate Cloud-Based Blender Asset Library</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What’s in Store for 2022]]></title>
            <link>https://aleckunkel.medium.com/whats-in-store-for-2022-9e0fe6f761c?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/9e0fe6f761c</guid>
            <category><![CDATA[2022]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Tue, 04 Jan 2022 22:28:20 GMT</pubDate>
            <atom:updated>2022-01-04T22:28:20.308Z</atom:updated>
            <content:encoded><![CDATA[<p>The new year! A chance to start fresh and make goals that we inevitability drop halfway through February. While traditional goals are usually to drink less or make it to the gym, I’ve set my sights on growing my technologist toolkit. Aka, exploring technology that I see potential in. Here is where I’m planning to spend my time in 2022.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ExXM5vhDhVeps-CNTV-loQ.png" /></figure><h3>Low Level: Rust</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dQzZxWXOs6rH1cTD.jpg" /></figure><p>First on the list, Rust.</p><p>Maybe it’s just the circles I run in, but for the past month, I’ve seen Rust come up enough that it finally caught my eye. To the point that during the holiday I spent some time reading into what it actually is and the fundamentals. So far, I’ve loved everything I’ve read. Extremely fast, made for safe code, and on top of that a healthy community.</p><p>I know there are a ton of different facets to Rust, so I’ve decided to spend my time mainly around the crypto and WASM side of things.</p><h3>Time to Make a Switch: React</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*A3K546qpnoNWO2ai" /></figure><p>Okay, a switch is pretty dramatic, more a baseline understanding of the React framework. I’ve been using Vue for a long time and am a huge fan; I, however, also believe in looking over the walled garden at other options. And this year, I’ve decided to enter the world of React.</p><p>React has been the most popular framework for a long time, and the ecosystem is thriving! I feel like I owe it to myself to look into it further and understand at least the basics.</p><h3>Enter the Meta: Unreal Engine 5</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*M1e6VVS56OfBZpph.jpg" /><figcaption>Demo of UE5 capabilities</figcaption></figure><p>Throughout the past year, I have grown to love Unreal Engine. And while I’ve spent hundreds of hours within the platform, I feel like I’ve just scratched the surface. Unreal Engine 5 is full of amazing new features that I’ve been dying to try. Things that are going to make creating photorealistic scenes more accessible than ever. In 2022 I want to level up and started creating with the new engine.</p><h3>A New Way to Model: Photogrammetry</h3><p>At this point, I’ve had a bit of experience going photo to model but I still want to play more. Most of my models and experience has come from <a href="https://poly.cam/">Polycam</a>. They’ve made photogrammetry incredibly easy. To the point where you can get really decent models all from your iPhone, but it has its limits.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*3DOhSW-5BiPehLvrEkRTKA.gif" /><figcaption>A Polycam Demo</figcaption></figure><p>This year, I’m going to get serious with <a href="https://www.capturingreality.com/">Reality Capture</a> and start converting everything to the meta space.</p><h3>Procedural Everything</h3><p>Last on the list, is less a peice of software, but more a concept. The concept of doing everything procedural and non destructive. Last year Blender announced the Everything Nodes project and released the first version of geometery nodes. Having just a little bit of experience with these systems is enough for me to be a believer.</p><p>I truly think this is the beginning of an entirely new way to create. One in which making updates to your creation is changing a value rather than starting from scratch.</p><h3>Let’s Get Started!</h3><p>The realist in me knows that this is a lot. People make entire careers out of each of these things so the likelihood that I will end the year as an expert in even one of these is slim to none. I hope this post mainly serves as something I’ll be able to reflect on at the end of 2022. Oh and one more thing:</p><h4>Typing up a Storm: Medium</h4><p>This year I want to document my progress across Medium. If you are interested in the mind of a creative technologist throw a follow my way!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9e0fe6f761c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Eyes in the Back of My Head: An AI Alternative to Looking Over Your Shoulder]]></title>
            <link>https://medium.com/madebymckinney/eyes-in-the-back-of-my-head-an-ai-alternative-to-looking-over-your-shoulder-c43045d3210?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/c43045d3210</guid>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Tue, 26 Oct 2021 17:03:32 GMT</pubDate>
            <atom:updated>2021-10-26T17:03:32.636Z</atom:updated>
            <content:encoded><![CDATA[<p>On November 2nd, we’ll be headed back to the office. After almost two years of the safety of working from home, I realized that soon I’m going to face a challenge I’d almost forgotten: People walking behind me as sit at my desk. How am I going to get away with watching Netflix at work now?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/475/1*3frATmizw8UHbfPbgv9TOg.gif" /></figure><p>But that’s all good! Because I think I have a solution! I’m going to build a tool that gives me eyes in the back of my head. Aka a live detection tool that alerts me when a person is approaching me from behind. That way I’ll never be caught off guard again. And thanks to machine learning this should be a snap!</p><p>In terms of the tech stack, I’m going to keep things fairly simple for this little experiment. The front end will be built with Vue and then on the back end, I’ll have TensorflowJS doing all of the image processing.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/785/1*OpNd7eyTyxDKNCHmGf8OhQ.png" /></figure><h3>The Front End</h3><p>Because this app is hardly going to be seen, I’m going to keep everything super minimal. At its core, it’s just a camera stream, but to make it a little more fun, I’m going to set that video inside an eye. So it really feels alive!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Dm20CF5AJFiMJDhYgY_PtA.png" /></figure><p>I then went ahead and slapped a fancy logo at the top and had the first half of the project done.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0KTFL1OnhOI2dvl2jRqFLg.png" /></figure><h3>The Back End</h3><p>This is where things actually get interesting and where we get to start playing with machine learning. There are a few basic components to the back end: the camera feed, loading the object detection model, and alerting the user to an incoming person.</p><h4>Setup the Webcam</h4><p>Getting going with object detection is a pretty simple process, but the most important thing we will need is the feed. To do that I’ll create a video element that mounts the webcam when the page loads.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9c57cc44dc5771de36472f30f23283bf/href">https://medium.com/media/9c57cc44dc5771de36472f30f23283bf/href</a></iframe><h4>Loading the CNN Model</h4><p>Of course, I could spend the time collecting images of people, segmenting and labeling each, and finally training the model but that takes a long time. Not only that, but why would I when there is a perfectly good model bundled with Tensorflow JS.<a href="https://github.com/tensorflow/tfjs-models/tree/master/coco-ssd"> Coco-ssd</a> is a pre-trained model that out of the box can recognize 80 different objects. One of those being people!</p><p>Loading the model is really simple and only takes a couple lines of code.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e8e2012795db89203a26efb6e7b2f863/href">https://medium.com/media/e8e2012795db89203a26efb6e7b2f863/href</a></iframe><h4>Send Signal</h4><p>With the camera and model loaded, we’re ready to move on to actually interpreting the model output. Everything is stored in a pretty simple object. Within it you get the item recognized, confidence level, and then the position of the object. Knowing that, I threw together this little function that looks for two people in frame. Whenever two people are detected it will notify me!</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/52a08c6c42d48d75c5d9e11261e17c32/href">https://medium.com/media/52a08c6c42d48d75c5d9e11261e17c32/href</a></iframe><p>To actually send the notification, I ended up just using an NPM package that I found online. It was called<a href="https://www.npmjs.com/package/simple-web-notification"> Simple Web Notifications</a> and for this demo it worked great!</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/87613506b78757599f896ca508f1df7f/href">https://medium.com/media/87613506b78757599f896ca508f1df7f/href</a></iframe><h3>The Final</h3><p>With just a couple of hours of coding, I’ve managed to give myself eyes in the back of my head. I’ll never have to worry about someone catching me while I’m streaming the latest season of You. All of this meaning, that I’m officially ready to head back to the office!</p><p>Machine learning can sometimes seem like a black hole but the tools out there have made it really easy to pick up. Even for a novice coder like myself, I can throw together a working app super quickly. And while this demo is just basic object detection, there are a ton of other really cool features built into Tensorflow JS that you can check out<a href="https://www.tensorflow.org/js/models"> here</a>. Want to give yourself eyes in the back of your head? Check out the live demo here: <a href="https://mleyes.netlify.app/">https://mleyes.netlify.app/</a></p><p>And full code can be viewed here: <a href="https://github.com/iamkunkel/ml-eyes">https://github.com/iamkunkel/ml-eyes</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c43045d3210" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/eyes-in-the-back-of-my-head-an-ai-alternative-to-looking-over-your-shoulder-c43045d3210">Eyes in the Back of My Head: An AI Alternative to Looking Over Your Shoulder</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Happy Little AI Trees: Bob Ross Meets Nvidia Canvas]]></title>
            <link>https://medium.com/madebymckinney/happy-little-ai-trees-bob-ross-meets-nvidia-canvas-d8a2117eb71b?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8a2117eb71b</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[nvidia]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Mon, 13 Sep 2021 15:21:19 GMT</pubDate>
            <atom:updated>2021-09-13T15:21:19.212Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tm94UUr8ocwMoEj3qkjj6Q.png" /></figure><p>As a creative technologist, I’m obsessed with making technology accessible. It doesn’t have to be a scary black hole filled with code and fancy terminology. Just like Bob Ross gave millions the confidence to grab a paintbrush, tools like Nvidia Canvas make it easy to create art with just a few clicks. So what better way to test out Canvas than by doing a Bob Ross tutorial?</p><h3>WTF is NVIDIA Canvas?</h3><p>Canvas is one of the latest pieces of software to come out of Nvidia. It uses AI to create landscapes with just a few clicks on your screen. It works by using what’s called a <a href="https://towardsdatascience.com/what-is-a-gan-d201752ec615">GAN</a>. This GAN was trained on a large collection of landscape photos that tell the computer how to interpret your brush strokes.</p><p>When you draw on the left panel what you’re actually doing is creating a segmentation map. That map is then used to influence where the specific features should live within the scene. Then the trained model is smart enough to take each element and blend them into a “photoreal” scene. Enough talking though! Let’s actually make something!</p><h3>Marking Up the Canvas</h3><p>Honestly, the most difficult part of this experiment was deciding which Bob Ross tutorial to follow. Ultimately I decided to follow the<a href="https://www.youtube.com/watch?v=gMEZp47VKC0"> Sunset Aglow Tutorial</a>. While watching, I noticed that Bob’s tutorial can be broken up into two core components: the sky and the landscape itself</p><h3>The Sunset Sky</h3><p>While Bob spent the first ten minutes applying a tasteful gradient across his canvas, I spent mine deciding on the style. Canvas offers nine styles to choose from. Each of the styles offers a different aesthetic to the final piece. After experimenting with each option, I found that style #6 was pretty spot on to what Bob had. Next up, each of the landscape features.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/335/0*UGPnMFCFAQBC1hX8" /></figure><h3>In-Scene Elements</h3><p>Just above the style selection pan, are different “brushes” to choose from. Each one is capable of creating a different element within your scene. To create a cloud, for example, just click, drag, and watch the magic happen.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*_kNy7F-LR1pCsEaK" /></figure><p>After creating the clouds I moved on to the trees, mountains, and river — click, drag, magic. Five minutes later, my masterpiece was complete.</p><h3>The Final Result</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*f9uJyx83E9OSRwt2" /><figcaption>Bob on the Left, and mine on the right</figcaption></figure><p>Canvas is no Bob Ross, but let’s be honest, neither am I. And it would have taken me a lot longer than five minutes to create a way less polished landscape.</p><p>Another cool thing about Canvas? All the content is procedural. So if I want to see what my mountains might look like at night, I just pick a new style and the GAN does all the work.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*QOtSEH7g4DduolRC" /><figcaption>Switching between a few of the available styles</figcaption></figure><p>If you think this is impressive, the future is going to blow your mind. This trend of using GANs is still in its infancy. The software is only going to get better and the tech more powerful. And it’s only a matter of time before industry-standard tools, like Photoshop, start to incorporate these features.</p><p>Final note, Nvidia Canvas does require an RTX series graphics card. Don’t have the hardware but still want to give Canvas a try? You’re in luck! Nvidia released a web-based tool that you can use right in your browser. You can check it out <a href="http://nvidia-research-mingyuliu.com/gaugan/">here</a>. Can’t wait to see the landscapes you create!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8a2117eb71b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/happy-little-ai-trees-bob-ross-meets-nvidia-canvas-d8a2117eb71b">Happy Little AI Trees: Bob Ross Meets Nvidia Canvas</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Epic Games is Building Something Special]]></title>
            <link>https://medium.com/madebymckinney/epic-games-is-building-something-special-bba386646bfd?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/bba386646bfd</guid>
            <category><![CDATA[epic-games]]></category>
            <category><![CDATA[unreal-engine]]></category>
            <category><![CDATA[photogrammetry]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Wed, 07 Jul 2021 16:17:41 GMT</pubDate>
            <atom:updated>2021-08-05T21:08:00.912Z</atom:updated>
            <content:encoded><![CDATA[<p>And I’m not talking about Fortnite.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KvpEu_xU1PPNqLoD.jpg" /></figure><p>Epic Games has been working on an empire! From groundbreaking updates to their game engine to several strategic acquisitions, they’ve built a software suite that can change the way we work.</p><p>To demonstrate what I’m talking about, I’ve gone ahead and put together a quick example. Take the inflatable dinosaur below. What would it take to completely digitize this little buddy and drop it into a virtual world? It’s surprisingly simple to do, thanks to our good friend Epic Games.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6BPsVMvDF_hoRC1whHS7Qw.jpeg" /></figure><h3>Photogrammetry 101</h3><p>Photogrammetry is simply the act of taking photos of an object from all different angles. Then those photos are stitched together to create a 3d model. Sounds complicated, but new tools have made it incredibly easy. And to answer your next question: yes, Epic Games already has this tool in their arsenal. In fact, just this year they acquired Reality Capture, one of the best photogrammetry solutions on the market.</p><p>How does Epic’s photogrammetry tool work? Below I present the result! In all honesty, my expectations were low but it only took a few clicks and the quality of the final product blew me away.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*VqlkQE8o5gd4aV8X2kJ_0g.gif" /></figure><p>There are a couple of inherent limitations to photogrammetry, the primary one being that the model mesh is extremely dense. Aka, it’s really high quality, which equals a large file size that most computers struggle to manage. Traditionally the next step would be to bring this into another piece of 3d software and clean it up, but for the sake of our demo, I’m going to skip this step. Instead, we’re going to use it straight away in Unreal Engine, Epic Game’s game engine.</p><h3>Nanite is King</h3><p>Yes, Unreal Engine is called a game engine, but that’s a vast understatement. It’s become a real-time visualization tool that can build entire worlds.</p><p>Unreal Engine 5 early access was released about a month ago and it came with a couple of groundbreaking new features. The one we’ll use with our little dinosaur friend is called Nanite.</p><p>Nanite is Unreal’s new geometry system that renders high poly meshes within a scene. With the proper computer, it basically removes any of the limitations you’d traditionally have with high-density meshes. I’m not kidding when I say that this is some of the most groundbreaking tech I’ve ever seen.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JxySJXOSKmQqwSaXKq4Trw.png" /></figure><p>In roughly ten minutes I turned this inflatable dinosaur toy into a 3D model and used Unreal Engine to place it in a photo-real scene. In my opinion, the result is pretty incredible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*74mqq2tVYUwOnhIyOYSE0g.gif" /></figure><p>The technology Epic Games offers has the potential to change the entire field of 3D virtualization. If you have the means (and the GPU), I highly recommend giving this workflow a try. I’m a firm believer in the metaverse promised by Epic Games and am convinced Unreal Engine 5 is a crucial step in that direction.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bba386646bfd" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/epic-games-is-building-something-special-bba386646bfd">Epic Games is Building Something Special</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Metahuman Creator —  First Impressions]]></title>
            <link>https://medium.com/madebymckinney/metahuman-creator-first-impressions-142243d44177?source=rss-29d4aee17786------2</link>
            <guid isPermaLink="false">https://medium.com/p/142243d44177</guid>
            <category><![CDATA[unreal-engine]]></category>
            <category><![CDATA[meta-human]]></category>
            <category><![CDATA[virtual-production]]></category>
            <dc:creator><![CDATA[Alec Kunkel]]></dc:creator>
            <pubDate>Wed, 21 Apr 2021 17:18:22 GMT</pubDate>
            <atom:updated>2021-08-05T21:08:16.880Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Metahuman Creator — First Impressions</strong></h3><p>For those of you who haven’t heard, Metahumans is the latest and greatest innovation from Epic Games. The promise: Create real time photorealistic avatars in a matter of minutes. And, like most, when I saw the <a href="https://www.youtube.com/watch?v=6mAF5dWZXcI">first demo footage</a>, I was in awe.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aJyIcWLSVNUoJ-UhkZke_w.jpeg" /></figure><p>I’ve played around with other character creation tools in the past — <a href="https://www.adobe.com/wam/fuse.html">Adobe Fuse</a>, <a href="https://www.reallusion.com/character-creator/">Character Creator</a>, and <a href="https://blendermarket.com/products/humgen3d">Human Generator</a>. These tools are great, but the Metahuman demo made them look like child’s play.</p><p>Being able to have a high-def 3D character completely rigged in minutes would be a game changer! So when the first official beta dropped for Metahumans, I had to apply.</p><p>Does Metahumans live up to the hype? Is Metahumans all bloat or is this really a step closer to the metaverse? Here’s what I’ve learned after playing around with it for a few days.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*F7Xz0eRf3GzBpQMdR32Uag.gif" /></figure><h3><strong>Simple to Pick Up</strong></h3><p>Yup, the Metahuman character creator is surprisingly easy and intuitive. First off, it’s all cloud-based. Meaning, you can make these characters right in your browser. No high-end computer necessary. You can catch me outside on my college Chromebook ushering in the next evolution of virtual humans.</p><p>When you open up the site, it only takes a few minutes to get your bearings. Metahumans lets you blend, move, and sculpt characters with predefined parameters. Each tool is easy to understand and actually use. Unfortunately the simplicity that makes it easy to dive in becomes a frustrating constraint on truly unique creations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qad8ea2qh4fLcvGVw2bQRg.png" /></figure><p>The options are designed to work well in a tightly controlled sandbox where all the face structures are guaranteed to look good. This means the silhouette of the face will always live in a pretty consistent box. This restriction becomes most apparent when trying to recreate recognizable people with distinct face shapes. So don’t expect to make a Brad Pitt look-alike.</p><p>Here’s the thing though: It’s in beta, and I get that! I’m not going to hold this against them. As the tool evolves, I’m sure we’ll get more options.</p><p>Now, if you do want to actually use these characters in your work, know that you’re limited to use within Unreal Engine and Maya only.</p><p>Honestly, this isn’t necessarily a bad thing. It really feels Epic has fine-tuned the workflow. Your model from Metahumans is instantly available in Bridge and can easily be exported to UE4.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LqRBRlP4DaxYUjjgHzLB1g.png" /></figure><h3><strong>Breathing Life into Your Metahuman</strong></h3><p>The other exciting feature of Metahumans is that the characters come fully rigged. Not only does the face come pre-rigged and ready for animation, but the body maps neatly onto the standard Unreal skeleton and is ready to pair with existing animations or fresh mo-cap data.</p><p>It only takes three clicks to get your first face mocap up and running. It really is something else to see your creation come to life so quickly. I couldn’t help but smile watching my Metahuman mirror my movements.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*M_ZRm9TCx5WZM30KL5fDSQ.gif" /></figure><p>Basic head tilt, eye movement, and mouth shapes seem to map easily. However, nuances of speech quickly show the need for more advanced fine-tuning. Specifically, small lip and palate movements can be lost or mis-positioned. For example…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VYujPULnEC18v8Y5vxAwRQ.png" /></figure><h3><strong>What’s the Verdict?</strong></h3><p>Is Metahumans going to replace human actors? Not yet.</p><p>There is a distinct and consistent “feel” to your Metahumans. Between the challenges of performance capture and the “sameness” that comes from the beta limitations, it’s not about to replace traditional actors.</p><p>However, looking at this as only a human-replacer would be missing its potential as a creative tool. There are a million use cases for Metahumans. From pre-vis to game dev there are plenty of reasons to get excited about Metahumans and try it out.</p><p><a href="https://metahuman.unrealengine.com/">https://metahuman.unrealengine.com/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=142243d44177" width="1" height="1" alt=""><hr><p><a href="https://medium.com/madebymckinney/metahuman-creator-first-impressions-142243d44177">Metahuman Creator —  First Impressions</a> was originally published in <a href="https://medium.com/madebymckinney">Made by McKinney</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>