Introducing Celluloid: an AVISynth-based NLE

[RadicalEd]

One thing that's always bugged me is that AVISynth, for as powerful and flexible of an editing system as it is, has no graphical editing interface. AVISynth would make a competent editing system, with its vast colorspace support, massive array of plugins for every task, and strong editing features. Yet no one has ever been taken to task in compiling a true non-linear editor out of this fantastic project. Well, that's about to change. With the ongoing development into making AVISynth 3.0 cross-platform, and the inclusion of powerful new backends including FFMPEG and GStreamer, AVISynth will soon be fully ready for Mac OS X and Linux. Within the next few days, I'm going to begin development on an OS X-native, Cocoa-based NLE written on top of and using AVISynth as a backend. It'll be no momentous task, as all the hard work in getting an editing and rendering engine has already been done by the AVS team. The missing piece of the puzzle, then, is simply an interface, and an engine to translate interface interactions into script. Some features currently planned include: a built-in script editor with syntax highlighting and code-completion, full-fledged NLE mode or a simple, graphical AVISynth script generator, and automatic conversion of input formats to suit the project settings (a feature only implemented in Final Cut just recently). That's just a small fraction of much to come. And since the new AVISynth can utilize FFMPEG as a backend, support for encoding the output project to every format ever will be trivial to add. Get ready for the best editor ever, for free. I made a quick mock-up in Photoshop as a little glimpse into the future.

http://www.brutus-music.com/Mike/Images/celluloid.png

Suggestions, comments, and ideas are very welcome.

[DarkNite]

Amazing news! I wish you the best of luck, and a smooth development cycle.

[check]

Many have tried, all(?) have failed :) I wish you the best of luck, although *grrr* to the mac only part :p

[Soulhunter]

Just saw it on /g/ lol!

[RadicalEd]

Just saw it on /g/ lol!

Hahaha oh wow. I figured I'd get some more constructive responses here. /g/ is always hit or miss.

Many have tried, all(?) have failed

:( Any particular reason why? It doesn't seem that difficult...

[Richard Berg]

Avisynth is not optimized for interactivity. Every time you change something, you have to rebuild the whole filter graph.

Also, it's not easy to build a clip management system that's intelligent enough to avoid doing slow things. For example, if you want to apply a slow filter to just one scene, how would your program represent that to Avisynth? The easy way would make manipulating all scenes equally slow from then onward. You can workaround it but it requires more intelligence in your clip manager.

[Bidoche]

Richard is right, in order to get decent reactivity with non-trivial tasks, an incremental approach is essential.

ie you don't generate scripts but interface with avs at a lower level manipulating clips directly.

[RadicalEd]

ie you don't generate scripts but interface with avs at a lower level manipulating clips directly.
In other words, utilize the AVS libraries themselves, rather than building scripts. The problem with that would be that I was planning on having a script editor, and using .avs as the project format. But I suppose I could simply generate the script dynamically and still use the libs as the backend.

[Bidoche]

It all depends on what u want to do.

A script editor is good/useful too.

But if u want to be able to edit a filter properties in the rendered chain smoothlessly, it's better to skip the script part and talk to avs directly.

[RadicalEd]

Duly noted. I do plan to implement a script editor, but I can certainly do that while using the libraries as a basis for the real editing. I just hope this all isn't beyond my abilities :x Then again, that's why it's open source.
Speaking of which, I submitted an application to Sourceforge. Should get a response back in the next day or two. When I do, I'll post a link to the project page.

[tin3tin]

I can't help thinking, that you really are going to dig into this, then why choose a non cross platform language to code it in?

A cross platform gui with the strength of avisynth 3.0 could make it into the Blender of video-editng. :)

[RadicalEd]

I can't help thinking, that you really are going to dig into this, then why choose a non cross platform language to code it in?

A cross platform gui with the strength of avisynth 3.0 could make it into the Blender of video-editng. :)

A .NET port is possible. However, I do all of my development on a Mac, so it is far and away the most relevant platform for me, and could certainly benefit from such a project. Linux has Cinelerra and a few others, and Windows... well, Windows is lacking in this regard. However, friends have pointed out how difficult maintaining two separate codebases in two separate languages may be, which makes me hesitant to make any promises. It's a definite "we'll see" situation :/ Sorry I can't offer any more than that. On the other hand, it's open source, so anyone is free and in fact encouraged to start work on a port themselves.

[Zarxrax]

I would question the need for an advanced script editor to be part of the program. I know you've been out of the loop for a while, but there is an *incredible* script editor out there--AvsP (though it's not crossplatform). And I don't really see where most people using an an NLE would need to work with the scripts themselves at that level.

Best of luck with Celluloid, I think you may be greatly underestimating the difficulty of this though.

[Smetvid]

I have also been working on something like this but I am using a tool that is a little bit more cross platform. Well at least for PC and Mac it is.

I have been researching this project for a little over a year to see if it would work and I have done a bunch of small tests.

First of all I think AVISynth is better suited as a compositing application that has basic NLE functions built in. Due to speed restrictions a NLE based off of AVISynth would never cut it in the realworld. As a compositing program however the speed isn't all that bad. Most compositing programs work on a frame by frame basis so waiting to update the current frame isn't unheard of. I have worked with Flint, Combustion, After Effects and Shake and all of them require an update for each frame or everytime something changes. A AVISynth based compositing application could be made to work very much like any of the above programs minus some of the advanced keying plugins and rotoscoping tools.

For the program I have been working on there are 3 basic types of filters you can use.

1. A core filter
2. A plugin filter (use at your own risk)
3. And my own custom made filters that do not work with AVISynth itself such as advanced keying, rotoscoping, painting and motion tracking.

I have actually been working on 3 different programs that use AVISynth as a core engine that I use for my video production company.

1. Compositing
2. Conversion and enhancement tool
3. Rough edit storyboard editor.

Currently I am using the AVISynth 2.5 builds but I would love to be able to jump up to AVISynth 3.0 to be able to support higher then 8 bits per channel formats. Right now my tools are PC only since AVISynth 2.5 only works on PC. I could very easy with a click of a button create a mac version if AVISynth ever ported to mac.

I have also been thinking of testing out a different tool so I can create PC, mac and Linux version with no problem at all. I have not went this way yet due to only needing PC support right now but if that changes in the near future I will change as well.

I can tell you that this is a lot of hard work. I have been doing tests and research for over a year and I am only scratching the surface of porting every feature into any of my tools. I wish anybody else luck who takes on this project because it is hard work but at the same time it is a lot of fun.

[Blue_MiSfit]

Another vote of support here.

[RadicalEd]

Potential icon created:
http://www.brutus-music.com/Mike/Images/celluloidicon.png

[vinetu]

I see timeline here!!! Finaly!!! Amazing!!!

I think this is THE first NLE here and not "one off..."
Congratulations!

Now some constructive advises:
1. You can start the Celluloid as "Cut only" application - the AVsP can do the rest allready.
For example,there is nothing wrong if one use Premiere only for cut and move operations and then
continue with efects/transitions/keying in After Effects.
Later the project can become BIGGER and BIGGER in functionality and HDD space :)


2. About the interface - For me (when edit) It is very important to see as much as posible number of the _imported_ footage files/folders,
the time line and the "Output Video Window".Every other window or empty space is a wasted desktop space.
Think this way - "What is the most used/wached thing when you edit?" and then make those windows "On Top" in access priority.
Ask yourself The same question when planning "Hot keys" and the like. I remember the time, when the most important(for me) hotkeys in
Premiere 4.2 has been "hiden" in Premiere 5.0 - I was forced to use a 3 button combination hotkeys (was one in Prm 4.2),
plus the timeline zoom in/out butons(-/+) was removed from the NumPad forever...
User unfrendly interface has the ability to take your attention/energy on ONLY pushing the buttons/mouse instead of editing.

WISH YOU LUCK RadicalEd!!! ...and a windows version too ;) if possible...

[RadicalEd]

Made some pretty significant updates to the interface mock-up (the screenshot in this thread is updated). There's now a filter browser with search, a filter settings pane, and live previews of assets, among other little touch-ups. Development will probably start tomorrow. Still waiting for the SourceForge page to get greenlit.

[Richard Berg]

In other words, utilize the AVS libraries themselves, rather than building scripts. The problem with that would be that I was planning on having a script editor, and using .avs as the project format. But I suppose I could simply generate the script dynamically and still use the libs as the backend.
Using .avs as the project format can work. Especially if you don't mind rebuilding the filter graph for parameter updates, giving performance similar to compositing apps.

What you need to avoid is the 2nd problem I cited: letting a slow filter slow down the entire project. That means you have to be careful about combining clips. For example, it's tempting to have the timeline be represented by something like "clip1_filtered + clip2_filtered + ... + clipN_filtered". It's a completely general solution in just 1 line! However, it will have perf problems, because the "+" operator is just shorthand for the Splice filter, thus you're putting all filters (slow & fast) into the same graph. What it boils down to is that your script-generating backend needs to have some optimization/intelligence built into it.

[RadicalEd]

What you need to avoid is the 2nd problem I cited: letting a slow filter slow down the entire project. That means you have to be careful about combining clips. For example, it's tempting to have the timeline be represented by something like "clip1_filtered + clip2_filtered + ... + clipN_filtered". It's a completely general solution in just 1 line! However, it will have perf problems, because the "+" operator is just shorthand for the Splice filter, thus you're putting all filters (slow & fast) into the same graph. What it boils down to is that your script-generating backend needs to have some optimization/intelligence built into it.Gotcha. I'll have to work something out then. Speaking of which, I'll have XCode 3.0 up and running in about an hour, which means I can finally get started. On that note, does anyone have any precompiled binaries of AVS 3 for Mac? :| I'm having some problems compiling.

[Bidoche]

However, it will have perf problems, because the "+" operator is just shorthand for the Splice filter, thus you're putting all filters (slow & fast) into the same graph.I fail to see why it will cause a perf issue.
Even if slow and fast filters are spliced together, you are still rendering one frame at a time.
If she ends up belonging to a fast filter, it will be fast, no ?

[Richard Berg]

You're right, in my example the slow filter shouldn't be in the callstack for a fast frame.

I know I've run into this perf issue before while running slow filters on sections of a video. Maybe the scenario isn't as common / hard to avoid as I thought.

[RadicalEd]

Sourceforge page just went live at http://sourceforge.net/projects/celluloid

We need all the help we can get, so if you code, have a Mac, and know Avisynth, hop on board!

[RadicalEd]

Check out the new project website!
http://celluloid.sourceforge.net

[RadicalEd]

We have a new IRC channel on FreeNode!
irc://irc.freenode.net/#celluloid

[RadicalEd]

I should have asked beforehand, but may I have permission to display the Avisynth logo on Celluloid's project page? If not, I can whip up a custom copy pretty fast. Thanks.

[Guest]

As we say in the swimming world, "all splash, no dash". Have you written any code yet?

[RadicalEd]

As we say in the swimming world, "all splash, no dash". Have you written any code yet?I'll admit: it's vaporware for the time being. We're recruiting developers and ramping up to begin development. Since I've been advised not to try compiling Avisynth 3 for a while, we're going to be focusing on developing the interface, and the script translation engine for the time being.

[RadicalEd]

The site (http://celluloid.sourceforge.net)'s undergone a big update including a MASSIVE update to the mock-up (http://celluloid.sourceforge.net/Screenshots.html). You'll find a list of changes here (http://celluloid.sourceforge.net/News/EE0AA2D1-C311-403C-B0E5-CC4B2D94FB00.html), but I want to highlight one particularly mind-blowing feature never seen before in a video editor that I think you'll like a lot:
Celluloid is bringing the ubiquitous “Magic Wand” tool from the image editing world over to the world of video editing. Want to apply a Tweak filter to your source? Oh, but wait, darn, you’ve already sliced and diced it into dozens of clips, and command-selecting them all is going to be a real pain. That’s where the Magic Wand tool comes in. Magic Wand automatically selects all clips from the same source. So if you’ve got moving or filtering to do on a specific set of clips, worry not. Magic Wand is here to help. This is a breakthrough feature unique to Celluloid, not featured in any of the pro editing suites available.

Enjoy :)

[Yobbo]

Wow! Never seen software pimped so enthusiastically without even existing! :p

[check]

I think before you add any more magical unicorn features that have never been seen in the wild before you should at least find one developer who can begin work on it. There are plenty of entrepreneurs who have planned an avisynth NLE, and none have succeeded because they leapt before they looked.

[Yobbo]

There are already very good & very functional Avisynth-based video editors out there - eg avsFilmCutter, etc.

What is missing in them, I think, is the ability to mix in new audio tracks, whether it be background music, narrations, sound FX etc. Obviously this would require a reliable preview/visual cue to keep the sync tight.

The video side of things is already down, I feel. I wish an audio multi-tracking feature could be added to one of the existing tools. Can, for example, Audacity code be integrated in?

Anyway, good luck with Celluloid, but like they say, don't count your eggs before etc...

(BTW, limiting your software to Mac potentially alienates most of us anyway, I'm sure)

[smok3]

so it will run on iphone as well?

[honai]

If anyone took a look at http://celluloid.sourceforge.net/Screenshots.html he'd get the impression that those are, well, actual screenshots since the term "screenshot" is defined as, well, a shot of the actual screen - and not the mock-up that it is, even though the caption reads "mock-up", thus directly contradicting the headline. This ambiguity is misleading at least. Other terms come to mind.

The page http://celluloid.sourceforge.net/About.html further supports this suspicion, giving the impression that there were indeed a functional application available ("The World’s First Avisynth-based NLE" and other buzz), whereas in actual fact no single line of code exists yet.

Again one has to wonder what the actual motivation of the maintainer of the project is, and why he chose such misleading language and buzz to promote a virtually non-existent application ...

[CruNcher]

Hehe yeah indeed it sounds like some Viral Marketing for a Product that's not yet exisitng, but i doub't that RadicalEd just wan't to see the reaction of how many people wan't it to market it later commercialy, im sure that's not RadicalEds intention, maybe to get reaction from the Pro guys (Adobe,Newtek,AVID,Sony), but since some of theses have very high requirements like 10 bit support i doub't they would be very shocked about this news at least not for the Pro Consumer Divisions but on the other side their and other Average joye Consumer Divisions (Pinnacle,Ulead,Intervideo,Cyberlink,Nero and others) should be worried a Free Crosplatform NLE with Avisynth behind it's curtain working jesus that's Great (tough to call it celluloid is a little bit off i think) :D

[TheFluff]

Using Avisynth 3? Nice shops of iMovie as "screenshots"? "Development starting soon"? Hugely impressive plans but no code?

I guess this means we'll see you with a potentially usable alpha version in 2017 then. Maybe.

(Hint: don't start megalomanical projects like this. Start small and work your way up, and preferably base your new program on existing, working tools, not with stuff that doesn't even work yet and has had exactly two posts in its "progress" thread in the last nine (9) months, and approximately zero significant code changes on its SVN in the last three months (Avisynth 3, I'm looking at you))

[honai]

I see that in the meantime the "author" has removed the term "mock-up" from the "screenshots" page and thus is now implying that there were an actual application and actual code.

I think that finally crosses the line from "honest but futile attempt" to "misleading" since the author already acknowledged in this thread that no single line of code exists yet.

[Yobbo]

Is it finished yet?

[RadicalEd]

Man, lots of negative replies here for no reason. For your information, the only photoshop in the interface screenshots currently is the iMovie video controls and timeline, and that can be easily fixed. Everything else is real, live Interface Builder, including the entire single-script mode.

As for developers, sure, we need them, but if you've bothered to take a look at the team page (http://celluloid.sourceforge.net/Team.html), you'll notice that we have four developers currently.

As for the release, no, it won't take 'till 2017. My guess is that the single-script mode will be functional before the fully fledged NLE mode is, due to its simpler nature. If I could give an estimated time for the first alpha, I gladly would, but I really can't determine that at the time. The project only just started about two weeks ago. We're also limited by the progress of Avisynth 3.0's development, and currently I've been asked to not even try compiling it on Mac yet.

Speaking of Mac, you might notice that I've added mention of an EFL and .NET port of the interface. This should put to rest all the cries for a Windows version. It is my belief that the structure of Celluloid should make it relatively easy to port (this is explained in detail on the about page), mainly because 2/3 of the program is already platform-independent. The only thing that needs porting is the interface, and this shouldn't prove too difficult a task.

As I said, yes, a cynical critic might currently refer to the project with the word "vaporware," and they'd be right in a sense. However, I'd like to ask you all to have some patience and some faith, and appreciate the project's goal rather than ragging on its hypotheticality.

There are already very good & very functional Avisynth-based video editors out there - eg avsFilmCutter, etc.Sorry, but this (http://www.videohelp.com/toolsimages/avsfilmcutter_824.jpg) isn't even close to what we're attempting with Celluloid. That looks like a glorified video player (no offense to the creator).

i doub't that RadicalEd just wan't to see the reaction of how many people wan't it to market it later commercialyYou'd be right about that. Plenty of people have mentioned the marketability of such a project commercially. However, it is not my intent to repackage open-source software as proprietary, and also, it is my belief that we're more likely to attract a decently-sized development team as an open-source project on Sourceforge.

Those unexpected rebuttals out of the way, look for a big update to the site very soon.

[Leak]

As I said, yes, a cynical critic might currently refer to the project with the word "vaporware," and they'd be right in a sense. However, I'd like to ask you all to have some patience and some faith, and appreciate the project's goal rather than ragging on its hypotheticality.
Sorry, but the impression you made on me with your very first post in this thread (and first impressions are what's important) was "100% vaporware". That's not something you should do, ever.

Show some interesting working code (and no, the interface is not interesting if the video part is not there yet at all), get cookie.

Show mock-ups and marketing blather reeking of snake oil, get sneered at.

It's that simple.

If you had just said you were looking for developers for a project idea you have, I'd have had no problem. But starting a thread with a sensationalist title of "Introducing <foo>: the world's first <bar>" without anything but a mocked up screenshot to show for it is just about the poorest form possible.

(Oh, and it doesn't really help that all the navigational links on your "team" page aren't clickable at all in Firefox. You might want to not put images above the links by abusing z-order and absolutely positioned divs... :rolleyes:)

np: The Orb - Ultraworld Probe Eight - Outlands (Adventures Beyond The Ultraworld (Disc 2))

[RadicalEd]

If you had just said you were looking for developers for a project idea you have, I'd have had no problem. But starting a thread with a sensationalist title of "Introducing <foo>: the world's first <bar>" without anything but a mocked up screenshot to show for it is just about the poorest form possible.

(Oh, and it doesn't really help that all the navigational links on your "team" page aren't clickable at all in Firefox. You might want to not put images above the links by abusing z-order and absolutely positioned divs... :rolleyes:)What can I say? The Apple bug got me.

I'm aware of the website bug, but I haven't had a chance to fix it yet. Probably today.

[DSP8000]

Oh yeah, and you could just use pure CSS for the rollovers and java for the reflection images. All you do is just use normal images and control the reflection with java. Leak is right, z-index is broken.

[trythil]

To RadicalEd:

I'd like to also ask that the marketing be put on hold until some more workable project artifacts are produced. It's fun, yes, but it does nothing good from a software engineering perspective.

By "workable project artifacts" I mean materials like use cases and prototypes: materials to more clearly define what the project is, the problems it is meant to address, and what it needs to do. (Is that even well-defined? All I've learned from reading the existing literature is "it's an AVISynth-based editor", and that's all. I've yet to see a real concrete mention of initial feature set, target audience, or anything of the sort, and I'm technically on the development team!) Mock-ups and honesty are steps in this direction. Marketing isn't.

Finally, on a personal note: All this spin makes me and the other recruited developers look bad by association, and I don't want that. (In open source and free software, reputation is an enormous force for one's benefit or detriment -- as I'm sure you're noticed, or at least are discovering.) I'd like to help out with this project because I think it's an interesting idea, but if I'm going to get tainted by all the hype and resulting (justified) ill will I don't really want to be a part of it.

Thanks.

[honai]

For your information, the only photoshop in the interface screenshots currently is the iMovie video controls and timeline, and that can be easily fixed. Everything else is real, live Interface Builder, including the entire single-script mode.

Amazing - so where's the code that did all of that, including the Interface Builder? Since you're doing an open-source project it'd only be natural if you released the existing source-code right now. Also, by doing that - which would take probably 5 minutes of your time - you'd be refuting the vaporware impression quite easily, don't you think.

[Ebobtron]

Sorry, but this (http://www.videohelp.com/toolsimages/avsfilmcutter_824.jpg) isn't even close to what we're attempting with Celluloid. That looks like a glorified video player (no offense to the creator).
:D "glorified" :D Thanks, better than I thought.
There are already very good & very functional Avisynth-based video editors out there - eg avsFilmCutter, etc.

What is missing in them, I think, is the ability to mix in new audio tracks, whether it be background music, narrations, sound FX etc. Obviously this would require a reliable preview/visual cue to keep the sync tight.

The video side of things is already down, I feel. I wish an audio multi-tracking feature could be added to one of the existing tools. Can, for example, Audacity code be integrated in?

Anyway, good luck with Celluloid, but like they say, don't count your eggs before etc...

(BTW, limiting your software to Mac potentially alienates most of us anyway, I'm sure)
What is missing is a convenient way to edit the audio in avisynth. What we need is an audiotrim() or some kind of audioplay(vframe start, vframe end) so we can start an audio track at frame number x.

Then we could build audio tracks with something like

MixAudio(audioclip).audioplay(x, x+xx)

something easier than

MixAudio(audio clip).DelayAudio(frames *(1/framerate))

maybe that is good enough if I had a way of triming the sound without dumby frames.

Until it is part of AviSynth adding Audacity capabilities to FilmCutter are just going to give you a two or three stage process that we already have. So just use Audacity, hand your script a wave file and your done.

FilmCutter is a GUI and an NLE (based on a narrow definition) but the goal was to reduce the number of tools that needed to be opened individually to get a task done. Utility has been the goal not wizard style script creation. FilmCutter has always been able to apply a filter or process to all clips at once, just put the filter in the source clip.

Creating an AviSynth based GUI that mimics Premier or Vegas to me seems repetitive and confining. Anything I automate takes the magic away from AviSynth, that magic being the endless combinations and configurations.

Scripting brings together the creative magic of Fizick's plug-ins and Ian's clever programming skill, DG's mpeg2 decoder. I can go on and on and still leave someone important out.

Skill and inspiration don't always track so if someone is still willing to take a crack at it, great.:thanks:
Sorry, they changed my medication again.

[wonkey_monkey]

and java for the reflection images.

Youch. Why rely on the user having Java (and having the patience to wait for it to load) when you can just add 5-10% to the height of your image?

David

[Ebobtron]

MixAudio(audioclip).audioplay(x, x+xx)

something easier than

MixAudio(audioclip).DelayAudio(frames *(1/framerate))The examples above are bogus.

The first should look like:audioclip.audioplay(x, x+xx)
MixAudio(audioclip)


The second like:audioclip.DelayAudio(frames *(1/framerate))
MixAudio(audioclip)

Not to worry there is no "audioplay" function. But there is good news, the DelayAudio(frames *(1/framerate) method of starting a sound track will work out and beyond 200,000 frames. So all we need is an audio chopper.

:)

[TheFluff]

Everything else is real, live Interface Builder, including the entire single-script mode.
Excellent, a script mode for a scripting language that doesn't even have a functioning parser yet (IIRC the Avisynth3 script parser was "almost done" 3 months ago, but still isn't fully functional). How incredibly useful, a pretty GUI with buttons that do absolutely nothing!

As for developers, sure, we need them, but if you've bothered to take a look at the team page (http://celluloid.sourceforge.net/Team.html), you'll notice that we have four developers currently.
I hope all of them are working on getting Avisynth3 working, because without that your pretty program is going to absolutely nothing.

We're also limited by the progress of Avisynth 3.0's development, and currently I've been asked to not even try compiling it on Mac yet.
... seriously, why did you even start this project? How about you start the "finally get AVS3 working after these 5+ years of development" project instead, so this project can actually become possible to implement? Because right now, no matter how many shiny interface buttons you have, none of them are going to do anything because the underlying video architecture simply does not exist in any working form whatsoever!

[RadicalEd]

Because right now, no matter how many shiny interface buttons you have, none of them are going to do anything because the underlying video architecture simply does not exist in any working form whatsoever!This is simply untrue. Avisynth 3 currently has implemented a cross-platform, GStreamer backend that works quite nicely. Just because I've been asked to hold on compiling for a little while doesn't mean boo regarding Avisynth 3's ability to function.
Amazing - so where's the code that did all of that, including the Interface Builder? Since you're doing an open-source project it'd only be natural if you released the existing source-code right now. Also, by doing that - which would take probably 5 minutes of your time - you'd be refuting the vaporware impression quite easily, don't you think.Fine by me, I'll commit the single script interface to cvs first thing in the morning (need to crash, I got five hours of sleep yesterday). I wouldn't have really thought of doing so until the interface actually functioned, but if you want a ghost interface as proof that .nib files exist, that's fine by me.

I'd like to also ask that the marketing be put on hold As always, I appreciate your input, and I greatly appreciate your contribution to the project, but I've been working on the latest website and interface updates too damn hard to just ditch it for fear of marketing strikeback. The good news for you is that after this big, mouseoverable update, I don't have any more html tricks up my sleeve and we can really get down and dirty and give these critics some functioning code. I dislike the objections to the "marketing" because every project needs to bring attention to itself. I'm just trying to get some buzz that will hopefully attract a few more developers. I don't understand the negative reactions at all. I especially can't imagine it making any of the developers "look bad by association."
I've yet to see a real concrete mention of initial feature setHave you read the about page? The features are bulleted right on the project's main page.

and no, the interface is not interesting if the video part is not there yet at allYou've lost me here. The video part is Avisynth 3 and it is there. How can the interface be uninteresting if it's essentially the core component of the whole project? Please explain.

[Gerard V]

Whereas the sales pitch may have been premature, I for one can only encourage the original poster to have a go. The negative tone of many of the above posts seems, to me at least, to be inappropriare for a community with substantial open-source development credentials. Surely we don't require every potential developer to run this gauntlet of criticism. Regardless of how people begin, is it not in our best interests as a community to encourage each and every development? And to support it with ideas, advice and encouragement, rather than a list of reasons why it won't work? I am sure many initiatives will fail, but every now and again someone produces a useful tool, or even a real gem. So even if the pitch was premature, that is the only feedback the poster needs on that issue. After that - we should be discussing ideas and how-to's.