Feature Requests

Hello guys. I've been testing out the newest version that still under test. I laid out several MP4 files (11 in total) on my timeline specifically in the Ibrahim Traore video and was having trouble making the green screen understand my instructions. I asked it to apply the effect to all the videos on the timeline, but it kept on forgetting to do the effect on some. I have a suggestion if you guys can make underlord understand an instruction based on a section that one highlights on the text, or lets say between 2 book marks so as to make it understand the instructions based on a specific section. Using a highlighted text section command would be awesome. Another option is the time on the timeline (From this time to that time). I think that way we would edit a long video in small section commands to make the Ai understand what we are specifically trying to do better. Also the apply a B -roll to a background of a main video character that has had a green screen effect applied, command (Needs to be refined). Add options like whole background, Top right, top left for example. So i think a small time based, or highlighted text based commands would be awesome.

It will be great if we can choose our own threshold at which the code will decide certain parts of the recording is silence or not. For example, a slider to choose anything below -40db to be silence, and anything above to be voice. As of right now, very very often, I'll have one of two problems: 1- Shorten the gaps using the settings I prefer but then end up with most of the words at the end of my sentences cut off, and I have to manually properly edit that. 2- Increase the gaps between each sentence, which doesn't create as many problems with the words being cut, but it makes the pace much slower than I want, and the gap between each sentence much longer than I want it to be. If there is another fix to this problem that I'm not aware of, please feel free to let me know.

Pls allow us to fine tune edits. like Extend by frames. I'm also experiencing a lag when i try and edit quickly. I am really not able to--it's a little clunky.

Where is the timeline, how do I get it working?

I forgot to click record button and already missed it 2 times. If there are 2+ people on the call, please remind to click on record button after 3 mins or when a new person joins, so that I will not forgot to click the record button Descript Room

Every time the software is updated is reverts to factory preferences. For example, I might not be aware that there is an update and when I upload a new audio file, it automatically levels the audio and adds "room tone." I typically have these settings turned off because I'd rather do that audio work myself as these settings in Descript are destructive and I cannot make changes after the audio is uploaded. Please do not do this as it wastes my available minutes if I were to re-upload the audio after changing the settings back to my own preferences. Please do not revert preferences back to factory settings during each new update. Thank you.

Please add support for mouse side buttons and global media keys, or allow full customization of keyboard and mouse shortcuts for playback control. This is a standard feature in many professional tools and would dramatically improve productivity and user experience in Descript. Currently, Descript does not recognize side buttons on high-end mice or dedicated media keys for basic playback functions like Play/Pause — even though these inputs work flawlessly in other professional software such as Premiere Pro, VLC, and even browsers like Chrome. This limits editing speed and accessibility for users who rely on programmable mice or hardware media controls as part of their daily workflow. As of now, Descript does not allow mapping of mouse or media keys to functions, and no workaround is available.

I cannot correct text in desript anymore. it presents an error

Since you are moving towards more AI help in our workflow, in the future, if AI could suggest from all the available libraries the right B-Roll for the scenes we are active in, that would be amazing.