Just a few notes. Nothing major

Hi, I’ve really been getting into shotcut. This forum is great. A special thanks to all involved :grinning:

I really like the way it’s efficient with resources. Is so easy to compile (git version or whatever) and doesn’t seem to use as many resources as other programs comparitively speaking. I never get any errors when compiling either.

I want to jot downs some notes because I’m a bit unclear on a few things.

  1. Firstly. How does shotcut manage memory? Just basically. Does it store all the files of a project in memory? My question really is, does shotcut use memory linearly as it uses media? Setting up a 60 min timeline vs a 3 min timeline? 40gb necessary vs 4gb? I’ve been able to use a 4gb computer with 4 threads for 1080p stuff (transcoded) with a few non-compositing filters with a bit of struggle but it works. This is necessarily getting upgraded now but curious to know how it actually works. Still, very efficient to what other programs seem to need.

  2. I’ve been watching Resolve videos on youtube and I can see on that program you can change a timeline into an asset much like what we have in the playlist. If 1. is true then would it be a big deal to make this a possibility? I’m only saying because I can see as projects become large then a project needs to be broken down into segments and easily moved around into sub-parts. I mean, I know it can be done on the command line, concat or joining files with ffmpeg or mlt… Anyway just thinking outloud here.

  3. I think it’s great that stability is the main driver for the program right now. I’m in it several hours at a time and respect that large amounts of data are flowing around and this is no easy task. I’m getting crashes mainly when I move clips around on the timeline, or between tracks. Some tracks don’t like having certain clips in them and they just refuse to work. I can work around those issues. And am really greatful for this software.

  4. Audio. Having this a big development push in this area is a nice to have and no big deal for me. On GNU/Linux, there is fantastic audio software and I know ladspa plugins are available but for me. A basic, mono or stereo mixing much like we have at the moment is well good enough. I know my needs are not the norm here (more Win users). Just sayin’ that’s all.

Thanks again, I really want to keep using this program for a long time. It’s been great, also I can see shotcut’s getting some real traction online. Bravo.

1 Like

Thank for you for the complements.

How does shotcut manage memory?

It is complicated, but basically through reference-counting and caching. Most people assume cache uses more memory. However, something not in a cache must be able to be restored. In our case, that is how recent things can be retained in memory for frame-stepping or iterating over tracks while other things can be removed from cache.

Does it store all the files of a project in memory?

Absolutely not! Imagine having long uncompressed or lossless compressed video files read entirely in memory. Basically, it is impossible. Digital photos, however, are held in memory but only as long as they are in the cache.

does shotcut use memory linearly as it uses media?

No. It can use up much memory in the first 10 or so pieces of media but then grows less than linear beyond that.

you can change a timeline into an asse

See File > Open MLT XML as Clip

Audio

You know about all of the audio filters, right? We are slowly adding more. Next version will add keyframes (aka automation) to the Pan and Balance.

1 Like

As someone who’s used free software for a long time, I’m always amazed that it exists. Big projects such as this are a feat of engineering and I am grateful for that.

This makes absolute sense. I bought 8gb of ram recently even though I know 16gb is recommended but had fairly good results even with 4gb. Was wondering, hence the question, if this was a silly idea because I plan to scale up my projects.

Did not know about this thanks.

I love LADSPA tools but as I understand it, shotcut can only export 2 channels. This means I can’t easily rough design the sound in a multichannel format to export to my DAW for further processing. I might have this wrong, it’s just what I read. However, I’m happy to do the basic audio track, just mono in this case, from on camera sources and then do the rest later.

Shotcut supports 5.1 since May, 2018; however, it is very limited on panning and channel mapping > 2 channels.

1 Like

With the resolution and fps, you need to make sure to set it the way you want it at the start of the project in Video Mode in the Settings menu. If it gets changed in the middle of the project or right before exporting it could negatively affect it. Is that the same case with sound meaning setting the sound (i.e. mono, stereo, 5.1) in the Settings menu as one thing at the start of a project then in the middle of the project or right before exporting changing the sound to something else?

1 Like

No

I prefer the most professional used VST2.0-Plug-Ins and lots are free.
Maybe in the future SC will have a REWIRE or ASIO-Interface. :smiley:

Have a look here about the VST-Topic: