Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
FrenzyTheKillbot
Jan 31, 2008

Good Hustle
Ok, I just built a new computer and got MeGUI running again. I set it up like the guide said, but the first video I ran through it was messed up. The video itself runs slower than it should, maybe half speed, but the audio is fine. Could that have something to do with MeGUI bugging me to add ".ConvertToYV12()" to the end of my script? And me going along with it like a sucker?

Also, I tried to get FFMS working instead of using DirectShowSource, but I flat out couldn't figure it out. If it really is better and someone wants to let me know how to do it, that would be great.

Adbot
ADBOT LOVES YOU

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

MEAT! posted:

The convert to yv12 is fine. All video you see on the internet, DVD, and television are in YV12 colorspace. The desync might have to do with a change in framerate. The best thing to do is to post your script.

For FFMS just extract the ffms2.dll into your avisynth plugin folder. It should autoload just fine as long as you aren't using 100+ plugins.

code:
#Call all source videos
source1=(DirectShowSource("E:\BD\BD-01-1.mp4")).SelectEven().ChangeFPS("ntsc_video")
source2=(DirectShowSource("E:\BD\BD-01-2.mp4")).SelectEven().ChangeFPS("ntsc_video")
audio=WAVsource("E:\BD\Audio\BD-01-mixed.wav")
#audio=WAVsource("E:\BD\Audio\BD-01-game.wav")
#

#Trim into separate clips 
clip1=Trim(source1, 180, 2674)
clip2=Trim(source1, 3216, 12630)
clip3=Trim(source1, 12713, 33586)
clip4=Trim(source2, 1887, 10205)


#Put together using the correct transitions

video0=clip1++BlankClip(clip1, 45)++clip2++BlankClip(clip1, 45)++clip3++clip4++BlankClip(clip1, 75)

video0=video0.Crop(2, 10, -0, -8).Lanczos4Resize(1280, 720).Tweak(cont=1.1).AudioDub(audio).ConvertToYV12()

return video0
It's worth noting that this is the exact same script template that used to work before getting my new computer. Didn't change anything except adding the YV12 thing.

I also must have clicked the wrong link for the FFMS thing, cause I got a whole folder full of files I didn't recognize.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

MEAT! posted:

Don't use Selecteven orchangefps ntsc video.

Use this link for ffms2

https://github.com/FFMS/ffms2/releases

My capture device records at 60fps (can't change it) and basically just gets duplicate frames. Is there a better way to do it?

I got the FFMS working. I'll switch to it if the concensus is that it's better than DirectShowSource.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

MEAT! posted:

Basically what I told Niggurath. If you want 30 fps, just put fpsnum=30 in your ffvideosource call. If you want NTSC, fpsnum=30000, fpsden=1001.

Alright, I'll give it a try. Thanks for the help.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
Ok I've got a bit of a question about capture devices.

Right now I'm using the AverMedia Game Capture HD which works great for me. It's got a little remote and it records video files straight to an external hard drive. However, it doesn't accept an HDMI input and I'm looking to start using that. All the ones recommended in the OP that accept HDMI seem to require a program running on a PC in order to record. This is a problem for me since my PC and game consoles are not close together. Does anyone know of a device that can record HDMI and doesn't require the PC nearby to work?

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

Xenoveritas posted:

Maybe? The Hauppauge HD PVR Rocket may do that. I don't own one and have never tried it, so I have no idea if it actually does what you're asking, but it's worth looking at. At the very least you could be our guinea pig. :v:

Hm, it certainly seems like it should be able to. In fact the only real limitation it seems to have is that it can't record 60fps. I'll have to look a little deeper into it. Thanks for the recommendation.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
Hey everyone, I am having an audio desync problem that is driving me a little crazy. I have videos from two separate recording sessions, but both recorded with the same utilities under the same circumstances. However, while videos from the first session are fine, the videos from the second session all exhibit gradual audio desync. When I open the videos themselves in VLC, they play fine and with no desync. When I import them into an Avisynth script and open it in VirtualDub, the desync appears. The video length appears the same as the source video file, and the audio gets "longer". Again, this is only happening to one set of videos. Anything I do to compare videos from the different sessions seems to show their attributes are the same, so I have no idea why this is happening.

The code I'm using to import the files is:
code:
LoadPlugin("C:\Program Files (x86)\AviSynth\plugins\ffms2.dll")
Import("C:\Program Files (x86)\AviSynth\plugins\FFMS2.avsi")

source1=FFMpegSource2("E:\MEC\Videos\MEC-02-1.mp4", vtrack=-1, atrack=-1, fpsnum=60)
Any ideas?

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

Xenoveritas posted:

Are you recording with ShadowPlay? Even if you aren't, because it works in VLC, this solution may help. Basically there's a known issue with some FFMS builds that cause it to desync audio from the video with certain MP4s - such as the ones ShadowPlay generates.

I am recording with ShadowPlay and that was probably the issue since your suggestion seems to have fixed the problem (once I figured out how to use that drat thing). It's weird that one set of videos had the problem and the other didn't. Maybe there was a driver update I forgot about. Thanks so much for the quick reply.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
So, I've been making a few gifs using GifCam but I have the feeling they're too big. As an example, one gif is 960x540 and 7 seconds long and it comes out to 34 MB. Is that alright, or are there any suggestings for optimizing them better? I'm basically sticking with the default settings at the moment.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

vilkacis posted:

The only sensible advice I can offer is "don't make seven second 960x540 gifs".

Video that poo poo.

I'm doing it specifically instead of video clips for the One Shot SSLP Challenge.

Which might make the advice :justpost: since it's a gimmick challenge anyways.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
So here's a question. I've been working on a new LP, recorded at 1080p and the plan was to output at 1080p. However, I've decided to try outputting the final product at 3840x2160 to reduce how much YouTube screws with it. I figured this would be as easy as throwing a resize at it in Avisynth and encoding as normal, except every time I try I get a whole ton of artifacts. It doesn't seem to matter which resize function I use so I figure it must be the encoding, except all I have is MeGUI set to the slower preset and constant quality with a cf of 18.5. Can anyone tell me what I can do to fix this, or what the best method of upsizing a final video for upload would be?

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

Touchfuzzy posted:

Just from first glance, I'd imagine the resize is just quadruple-magnifying the artifacts that were already there in your video. Do these artifacts show up in the preview of the video after the resize but before you encode? Or is it only after you encode? Do you have screenshots of the same frame from both videos to look at?

nielsm posted:

Blocking? Blur? Wrong colors? What kind of artifacts?
If you view the output of your AVS script before encoding it, does it look right?

Alright well I'm a dope. In trying to prove that the artifacts only showed up after encoding I managed to prove that it was an obvious issue I was missing. They were caused by VLC player which I was using to view the final video. If I open it through an AVS script, no artifacts show up. Thanks for the help. I'm dumb.

Maple Leaf posted:

code:
loadplugin("K:\Program Files\AviSynth 2.5\plugins\ffms2.dll")
a=ffaudiosource("1.mp4")
v=ffvideosource("1.mp4",fpsnum=60000,fpsden=1000)
audiodub(v,a)
dissolve(trim(18495,26280), trim(10093,17374), trim(273,9625), 60)
Any ideas what could be causing this?

So even though I've just shown my incompetence, I'm curious. Is there a reason you're importing the audio and video separately and then jamming them together again? Just as an example, in my scripts I pull them together as such:

code:
LoadPlugin("C:\Program Files (x86)\AviSynth\plugins\ffms2.dll")
source1=FFMpegSource2("E:\video1.mp4", vtrack=-1, atrack=-1, fpsnum=30)

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

Maple Leaf posted:

When I try to run that script on my video, avisynth tells me that FFMpegSource2 (or FFMpegSource, for that matter) doesn't exist as a command, so. I guess I might be using an older version of FFMS2?

To more directly answer your question, that's just how I've always done it and how I was taught to do it. No real special reason, and if your template worked for me, I'd switch to it immediately.

Yeah that's fair. Most of the things I do are just because that's how it was done in an example I found online. But that's super weird that your FFMS2 doesn't work the same as mine.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
Regarding desync from using Shadowplay, you might want to check out this solution that I was given when I was having the same issue with my Mirror's Edge Catalyst LP. It's kind of finnicky and annoying but it worked.

Xenoveritas posted:

Are you recording with ShadowPlay? Even if you aren't, because it works in VLC, this solution may help. Basically there's a known issue with some FFMS builds that cause it to desync audio from the video with certain MP4s - such as the ones ShadowPlay generates.

edit: Actually, Maple Leaf, you could always give that a try to deal with your audio not working issue too. Unless you figured that out already.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
Anybody have a good microphone/audio capturing guide? I've been doing the same recording and post-processing in Audacity for like a decade, except for tweaking filters and upgrading microphones periodically to the Blue Yeti I have now. I spend so much time scrubbing my audio for mouth noises and stuff, and then I'll see a random clip of some teenage streamer who just has their mic in front of them and the audio sounds great without any of that work. Makes me feel like there was some big leap forward in audio capture that I just missed the memo on.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle
It's mostly things like lip smacks when people go to talk, the occasional tongue click, and exhales after people finish speaking. I can remove them but I'm just curious what the trick is for people who seem to be able to just stick a mic right in front of their face and have it sound good.

For reference, here's the steps I take in post with my audio tracks:

-Noise Reduction
-Noise Gate
-Click Removal
-Compressor
-Normalize
-manual scrub for noises/coughs/etc.

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

IGgy IGsen posted:

In goon LP there's a lot of strange and outdated practices still making the rounds as best practices. And hey, if it works for you, then do whatever. But in this case, you can simply skip an unnecessary step. No real reason to do it

Hello. I still do all my editing using VirtualDub and Avisynth through Notepad++, and then encode using MeGUI.

If someone ever wanted to do an effort-post on more current best practices, you'd have an interested reader here.

DMorbid posted:

The best thing to ever happen to video editing, if you ask me, is hardware encoding. Being able to render a 30-minute video in less than ten minutes is *so* nice.

I was under the impression that hardware encoding was still trading off quality/filesize for speed. I software encode all my stuff and it takes a lot longer than I'd like to admit.

Adbot
ADBOT LOVES YOU

FrenzyTheKillbot
Jan 31, 2008

Good Hustle

DMorbid posted:

Maybe, but in practice any difference is minimal especially once YouTube re-encodes the whole thing. I'm sure there are ways to really optimize the quality and filesize, but I'll take the convenience of GPU rendering anytime.

Meh, long encoding times are a one-time cost and storage is cheap so I'm not overly worried about filesize. But I think I might be fixated on the idea that my final product needs to be at the best possible quality before letting YouTube have their way with it. The "compromised" quality is probably completely negligible, but when I look back at videos I made 10 years ago I really wish they were better quality and there's no way to fix them now.

I should maybe set up a hardware encoding process for review copies though. Every time I make a mistake that requires a re-encode it fuckin hurts.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply