Aviutl Tips, Tricks and Support thread

[poisondeathray]

Yes, I'm aware there are various workarounds... but it's more difficult to do, harder to control . Not to mention slower to render with more instances and layers. That's why I said it's too bad those other options cannot be controlled by midpoints, like the other parameters in the standard drawing menu

[MaverickTse]

yea, it is really difficult to use....
I'm nearly going to give up investigating it...OTZ

Perhaps there are easier script alternatives...

[poisondeathray]

It's still a pretty cool plugin/script . There are lots of creative ways to use this.

[MaverickTse]

Almost forgotten that I've included a "Calligraphy" script since EP v1.9 ...



(This GIF is created with the new GIF Export Plugin)

Better oriented for Asian characters, but you may find other use.

Attached: sample project that produce the above gif

[poisondeathray]

That's cool! What character is that? / What does it say ?

[MaverickTse]

That's cool! What character is that? / What does it say ?

It is a famous character in modern day Japanese (otaku) culture:
「萌え」(romaji IME: Moe)
(Note that actually I drawn it wrongly in the image.)

Translated to English means "being turned on by sexy/cute object or character"

[MaverickTse]

I decided to formally bump up the GIF plugin version to 1.0 after a few days of test on Win7 and Win8.1.
(Info Page: http://mavericktse.mooo.com/wordpress/archives/2028 )
The biggest improvements are the quality of Ordered-Dither, the Speed of optimization, and the package size.

The "DOWN" side is... Ordered Dither now produce much bigger file as it contains a lot more details,
and color accuracy suffer at low color count (Perhaps I need to tone down the threshold for scene-change detection)

Samples:

64 colors 109 frames 320x180 o8x8

OLD (299KB)<< >>NEW (691KB)


64 colors 109 frames 320x180 E-Dither
OLD-Stucki (1428KB)<< >> NEW (1124KB)


At really low color count (16 colors)

OLD-o8x8 (88KB) << >> NEW-o8x8 (180KB)


OLD-Stucki (308KB) << >> NEW (253KB)


Attachments: Original AVIs

[racer-x]

Here is a much better version of writing in sand using the Line (motion track) as a mask. Dang Paint.net is way too cool.....

[MaverickTse]

You guys seems really addicted in this kind of calligraphy presentation

[racer-x]

These examples are just animated mask-reviling Technics for before/after images. I think they're cool anyway........

[racer-x]

I learned how to make an Album in Paint.NET, so I decided to animate it on a Turn Table using Aviutl.

[racer-x]

Here's an interesting 3-D space scene I made today. All components were created with Paint.NET, then assembled and animated with Aviutl.

[MaverickTse]

Again optimizing on the inpainting stuff and produced this:
http://mavericktse.mooo.com/wordpress/archives/2053

IppRepair:: A successor to mtDeSub series

Dumped OpenCV completely and used Intel IPP instead. Also parallelized some code with Cilk Plus. The package is also built using Program-Guided-Optimization to further boost performance
This results in a significant performance boost and the plugin can now always work at native resolution (no pre-shrinking)

Changes from mtDeSub:
* More refined edge detection, especially for slender font
* Auto Canny threshold setting
* Selection area defined by two left-click on the main screen
(so that you can use it with Adv.Editing as it no longer requires dragging)

[racer-x]

I learned how to make bubbles in Pait.NET: http://forums.getpaint.net/index.php?/topic/30114-bubbles-making-lots-of-them/

They look good when used with the particle emitter in Aviutl.

[MaverickTse]

↑ more suitable on a "magical girl transformation scene", "bathing scene" or "love comedy" situation

[racer-x]

↑ more suitable on a "magical girl transformation scene", "bathing scene" or "love comedy" situation

You mean something like this?

[MaverickTse]

umm... adding some fog and fading out the fog towards the top will be even better

... Is it fine to skip the mosaic/blurring?

[racer-x]

Soccer Ball animation. All components were created in Paint.NET, then animated in AviUtl v1.10. I needed a reason to play with the new version.

[MaverickTse]

try to load some VFR video (e.g. WMV8), see if there is any sync problem.
this VFR support in L-Samsh Works should fix a lot of sync issues encountered in previous release

[budwzr]

↑ more suitable on a "magical girl transformation scene", "bathing scene" or "love comedy" situation

You mean something like this?

Ummm...I see some nipsters in there. Is that kosher? hahaha

[racer-x]

Marilyn Monroe is a classic, she looks good in anything. This is no different than the see through blouses the women wore back in the 70's. You remember those, don't you?

[budwzr]

Marilyn Monroe is a classic, she looks good in anything. This is no different than the see through blouses the women wore back in the 70's. You remember those, don't you?

I guess you're right.

[MaverickTse]

The first Chinese book on AviUtl, "視頻免費編輯指南2014", is now available in Hong Kong from 大眾書局(PopularHK)

[PhotoCat]

Congrats Maverick! Can you give us the link of your marvelous book in the book store?

[MaverickTse]

Congrats Maverick! Can you give us the link of your marvelous book in the book store?

See my blog post here:
http://mavericktse.mooo.com/wordpress/archives/2046

The first link in the post points to "Commercial Press" page, while the second link points to "SuperBook City" where you may purchase online.
I don't know if they ships overseas however.

Bookstores here will not be operating normally until 23rd Feb 2015, due to Luna New Year holiday.

[eddman]

Hello,

Can anyone tell me how can I merge/combine two audio layers?

The first audio layer will not be edited, but the second one will be cut in certain parts. Then I want to combine the two and have the second audio overwrite the first one in the existing parts.

Hope this image can show it better:

Also, is it possible to export the audio in other formats besides WAV, or without re-encoding altogether?

[MaverickTse]

@eddman

If you want the highest quality of audio that you even want to skip re-encoding, you'd better handle the audio in specialized audio-editing software.
To achieve the effect in the graph, you need to SPLIT the first audio object and MUTE(either deleting the section or turn down the volume) the overlapping parts.

AviUtl do not have a standalone audio export function except WAV. The encoder GUIs such as x264/x265/ffmpegout can encode the audio stream into a variety of audio formats however, if encoding to MKV container. These includes AAC(using qaac or NeroAAC), Opus, Ogg, mp3, mp4alac

Again, if you do not need to sync closely with the video stream, do that in audio editor. After all, AviUtl is not designed for audio editing.

[eddman]

@eddman

If you want the highest quality of audio that you even want to skip re-encoding, you'd better handle the audio in specialized audio-editing software.
To achieve the effect in the graph, you need to SPLIT the first audio object and MUTE(either deleting the section or turn down the volume) the overlapping parts.

Is there a way to not split the first audio and simply have the overlapping parts from the second audio to mask the first one? Is there an audio editor that can do that? I'm not sure if I have enough time to split and cut the entire first audio.

AviUtl do not have a standalone audio export function except WAV. The encoder GUIs such as x264/x265/ffmpegout can encode the audio stream into a variety of audio formats however, if encoding to MKV container. These includes AAC(using qaac or NeroAAC), Opus, Ogg, mp3, mp4alac

That's too bad. The second audio is a rather low quality MP3 and I was hoping to avoid re-encoding it.

Again, if you do not need to sync closely with the video stream, do that in audio editor. After all, AviUtl is not designed for audio editing.

Actually, I do. I didn't mention it because I thought it didn't matter in regard to audio layer merging.

This is the thread I made before posting here: https://forum.videohelp.com/threads/370321-Looking-for-a-solution-to-create-a-video-w-o-re-encoding

I tried asking here to get replies from Aviutl pros.

[MaverickTse]

I don't get why people starting some where from last year starts keep asking for "editing without re-encoding" thing.
You guys have to aware that, most media format are not designed with lossless editing in mind. The only easy way to get things going without re-encode is using AVI video and WAV audio. If your source is not in a lossless format, forget about the "no-reEnocde" thing.

Since one of your audio data is MP3, it is even more not worth the effort to hack for lossless edit. Just convert it to WAV for further editing. The quality loss should be negligible.

I can't think of any way that does not involve manually muting Track 1, even in audio editor like Audacity.
http://manual.audacityteam.org/o/man/mixing.html

So... either take your patient to mute by parts or ask this question on a audio forum/board

[eddman]

I don't get why people starting some where from last year starts keep asking for "editing without re-encoding" thing.
You guys have to aware that, most media format are not designed with lossless editing in mind. The only easy way to get things going without re-encode is using AVI video and WAV audio. If your source is not in a lossless format, forget about the "no-reEnocde" thing.

Since one of your audio data is MP3, it is even more not worth the effort to hack for lossless edit. Just convert it to WAV for further editing. The quality loss should be negligible.

I can't think of any way that does not involve manually muting Track 1, even in audio editor like Audacity.
http://manual.audacityteam.org/o/man/mixing.html

So... either take your patient to mute by parts or ask this question on a audio forum/board

Thanks for the replies. I wasn't going to edit the video at all; just wanted to use it as reference during audio editing. I do know that most video codecs are not meant to be edited, but I do edit them sometimes based on I frames without encoding.

Anyway, I edited the first audio in Aviutl, since that was the only option, and then exported both streams as a combined WAV file, so no loss in quality. Later, I'll convert it to flac.

Thanks again.