Hi There, Apologies for the long delay here. We are on the cusp of releasing an update to Templater and needed some heads down time with the code :) If you have pre rendered elements already setup, that’s a good first step! The less calculation that AE has to perform the faster it will speed through the frames. However, please note that the pre-rendered files you import to AE as layers are best saved as uncompressed files. Why? Because for every frame that AE reads in a compressed video file, it has to decode the compressed image before it can composite it—this slows down AE considerably. So, it saved a good deal of time to have AE skip any decoding. You might try to export as a PNG sequence or an uncompressed Quicktime. This poses another problem—that of space on your drive. Since you have a 1:20 video, any pre-rendered element is likely to be HUGE if you store it as uncompressed video. So, just be sure you have enough space on your drive to handle a huge file like that. The other factor that impacts speed is output resolution. In your output module, you can set it to render at half resolution so that AE doesn’t process an inordinate amount of pixels for every frame. Of course, you need to consider your final projection platform to really take this into consideration. If you are making mobile only, then rendering at half resolution can save you a considerable amount of time. Another technique involves post-processing the output with a command line program like ffmpeg. In this case, you can render out segments of your master composition that have absolutley NO dynamic frames. That is, render out the frames that do not change and place them on your drive. Then, when running through the 500 rows, you can render out the segments that change. Finally, you can use a post-job event script to “stitch” or concatenate the pre-rendered video files with the dynamic output that Templater creates. This is an advanced technique, but you can see how to use ffmpeg with a NodeJS wrapper on our github repository. The catch here is that all the video files you stitch together must be created with the same output module out of AE. Otherwise, you will see visual artifacts that just aren’t desirable. In addition, if you want to have audio on the video, you’ll need to mix it in that ffmpeg process. FFmpeg is a very complex program and does require some level of expertise to use effectively. Another way to handle speed is to have multiple machines with aerender installed on them. Then have Templater replicate the master comp with the “Send Replicates to Adobe Media Encoder” preference checked. This will effectively create an project file with a replica comp of the master comp. Use the post-job script to copy the project file to a machine that has aerender on it. Use aerender’s command line options to have it render out the replicated composition. The catch on this one is to make sure the machine running aerender has all drive and folder mappings exactly the same as on the machine that you created the template. I know this is alot of information. I definitely want to make these techniques into a video, and that will likely come towards the end of the year. Thanks for your patience with us as we grow Dataclay! Looking forward to hearing about your outcomes!

@wst-shorty You only need to start AE once for each video, not for each render. For a single video, have more than one object / row that each specifies which compositions you want AE to render. That is, each object / row will have a different target value. Then, specify those rows / objects in row_start and row_end values in the templater-options.json file. You should also register a script for the post-batch event that handles the “stitching” together of the outputs. Go through the README file on our github repo for event-scripts. Take note of the NodeJS concatenate sample script that does this. You’re using PHP so it will be different, but if you have node installed on the server, you can use it straight away. Please let us know if this helps. Thanks!

@BusterCherry Apologies for the late reply on this. We don’t have this feature currently implemented, but we do intend implement a similar design that you’ve proposed. The Templater Settings effect should have an additional parameter that is perhaps labeled something such as “Fallback Target Layer”. This param would essentially allow a user to specify different ways to “search” for a layer target if the one selected is not enabled / visible. We have this in our queue for feature improvements and your post validates our need to have this in there. In the meantime, you might be able to use an “invisible” video of varying lengths that can act as a proxy to what you want (i.e. a layer shifting in the timeline a certain amount of seconds / frames depending on your re-versioning data). Please let us know if this helps or if we can provide more clarification. Many thanks again for your patience! Best, Arie

Hi there! Yes we have confirmed this as a bug manifested in the 2.7.0 release. It comes up when you invoke the “Preview” task, but have no specified target composition. We have implemented a fix, but for now, the workaround is to be sure you’ve specified a target composition in any of the three methods. Please read this knowledge base article about specifying a target composition with Templater. Thanks, Arie

Sorry you’re having that issue. We’re working hard to get the docs and GitHub repo updated to reflect all of the new features of 2.7. We’re hoping to have those up in the next few days. In the meantime, we added a new setting to the Templater Preferences. Can I have you try adding this line into the prefs object of your templater-options.json: "supress_ae_dialogs" : true We’re going to be issuing an update soon, so for 2.7.1 that option will no longer be required. Also, if you are launching the Bot with the CLI, make sure that you use a -ui argument in your shell script.

@gjkoble Yes this is absolutely possible by enabling Bot. The trial actually gives you all features from all editions of Templater. So, technically, you could not acheive what you are asking with Templater Rig, you would need Templater Bot. However, since you are still within your trial period, you could still do this. You can point to any URL that responds with JSON object arrays. Each object in the array would effectively be a new video. It’s up to you how you want to setup the server that Templater requests data from at regular intervals. You can see Bot in action by watching the following video: https://youtu.be/_BJEu7coCsI However, that video shows a Google Sheet as a Templater data source. You can do the same thing except instead of using a Google Sheet, you can use a URL feed. Learn more about using URL feeds with Templater on our support site. Note the following from that article, however: When using a URL feed, The Bot cannot update the value of the “render-status” key because the data source is specific to the user’s system. Therefore, users should register custom scripts to update “render-status” values at appropriate events to avoid re-rendering “ready” data. Hopefully this helps, and please do not hesitate to ask any more questions! Thank you for giving Templater a spin!

@rens Unfortunately AE’s API access to the character palette is extremely limited, so your best bet would be to create a pre-job script that copies your master AEP and opens the project in AE, to run Templater off of the copy instead of the master. This will leave your master always untouched. In the instance that a missing glyph breaks one project, it won’t break all of the subsequent projects. You could also try running an ExtendScript that updates the font on that layer on the post-update event every time, but I think that could be more problematic. Bot events.

@kdav82 Are you using a local JSON file, or a JSON feed/URL? If you’re using a JSON feed, have you registered a script with the post-output or post-job event in Templater’s Preferences, and if you have, what is the script/incantation that you’re using? When you are using the CLI, are you launching the Bot, or are you launching a batch render job? If you’re using a render job, it will ignore the render-status. The downside of using the CLI to run the Bot is that it won’t ever spin down-- the process needs to be ended manually, where a batch render job can close on completion.

@rivalpro I wrote a forum post a while back for this type of “slideshow,” with a variable number of slides. You can find it here. There is one minor tweak that you can do with this method. We’ve added comp switching via a flag in your data (wrapping the comp name in double curly braces) to the latest version of Templater, so instead of swapping with a video asset, you could swap with an empty 2-frame comp that is double the framerate of the containing comp.