|
Mush Man posted:While YouTube regularly messes with 480p, it doesn't always. Recommending a target of 720p+ isn't always necessary. I used to have two presets in MeGUI - one for SD video and one for HD video. When I encoded 480p video with the SD settings, it would look horrible on Youtube, so I went with the prevailing advice at the time and double-sized to 1280x960 using the HD settings. One time, I forgot to resize (or rather, didn't bother because I didn't care about the quality of the video I was uploading) and left the settings at HD quality because it had been so long since I'd used SD video, and the 480p version on Youtube looked fine. Maybe it's because of changes they made to their encoder since the last time I tried it, but perhaps just having higher quality in the source video made the difference. I should look at writing a new guide to Avisynth, particularly going over the structure of my scripts and how I use functions for repeated operations, and some of the more common mistakes I've seen people make. I should also finally hand over some of the helpful functions I've written - is there a good place on the Wiki for those yet?
|
#
¿
Dec 2, 2013 14:11
|
|
|
|
| # ¿ May 2, 2024 02:35 |
|
|
StandardVC10 posted:Is anyone familiar with an error where AVISynth will insist upon the absence of a decompressor for HFYU? I just installed the relevant codec but it doesn't seem to matter, I get the error in both VirtualDubMod and MeGUI. Where did you get the video from in the first place? Can you play it directly in any video player? Does it load in VirtualDub/VirtualDubMod if you're not using Avisynth? If so, you may try transcoding to something lossless like Lagarith and working from there. You should be able to delete the original version once you have a version you can work with. One thing that may be worth trying is using a different Source filter to ingest the video - if AviSource can't find it, perhaps another filter will.
|
|
#
¿
Dec 3, 2013 19:18
|
|
|
ADTerminal posted:I'm in this exact same situation, and me and a friend do use Skype screen sharing to stream games to eachother. Yes... think of a better VAC setup that doesn't do that. I use Line 1 for game audio, Line 2 as my Skype input device, and Line 3 as my Skype output. Line 1->Line 2: Guest can hear game audio Line 1->Headset: I can hear game audio Microphone->Line 2: Guest can hear me Line 3->Headset: I can hear guest I think that should do the job.
|
|
#
¿
Dec 4, 2013 19:02
|
|
|
ADTerminal posted:
Also add your Skype audio device settings and your Windows playback device settings - this image isn't giving the full story.
|
|
#
¿
Dec 4, 2013 21:56
|
|
|
ADTerminal posted:
Huh. What are your "listen to" settings for Line 1 and Line 3? Also, the stupid question I should have asked first... are you using headphones or just playing the sound out your speaker and into your microphone?
|
|
#
¿
Dec 4, 2013 22:50
|
|
|
Well, then, I'm stumped. I use VACs for everything and not the Listen To, and everything works the way I tell it to. I also don't stream on Skype - the webcam could be grabbing sound from the wrong place somehow. I'll leave it to someone who uses Windows 7 to explain what could be causing your problem. Of course, it could also be on your friend's end.
|
|
#
¿
Dec 4, 2013 23:57
|
|
|
JerikTelorian posted:So, you may have actually answered my question more than I thought, but I don't actually understand what "mux" is -- is it multiplexing, or combining the audio into one stream? "Mux" is an abbreviation of "multiplex", which is any operation that combines multiple inputs into a single output. It looks like what you want to do is mux two audio sources together while still retaining the ability to listen to them separately on audio devices. VAC can do that for you. I don't know about the other program, but it looks like that one depends on the Windows 7 audio forwarding, which is a bit less robust than VAC.
|
|
#
¿
Dec 5, 2013 18:43
|
|
|
JerikTelorian posted:That is exactly what I want to do. Is VAC the only thing that can manage this, or are you familiar with anything else? VAC is what I use, and I know it works for that purpose. That's all I can say. I've found it to be a pretty great multi-purpose tool, particularly since I use Vista on my LP computer for better backward compatibility with the games I want to play and don't have the OS-level audio routing tools that 7 provides. Considering that most of the setups I've read about for streaming still require at least one software repeater to get all the sounds where they need to go, I don't think Win7 alone will do the job anyway. I don't know anything about any other software methods because I have one that works for me.
|
|
#
¿
Dec 5, 2013 20:07
|
|
|
The last time I tried to install an nVidia update, it failed and left my graphics properties in an unusable state. I think even System Restore failed and I had to uninstall my graphics drivers and reinstall from an old version. I don't think they actually support my card anymore and are just too lazy to remove it from the list of supported cards.
|
|
#
¿
Dec 5, 2013 21:20
|
|
|
Why not just transcode the raw videos to a smaller lossless format and trash the huge ones, then do your edits from that? Keeping the lossless (unedited) AVIs and junking the overly huge raws should save you plenty of space. Another thing I do when an LP runs long and I'm worried about space is remove the intermediate files - I usually have a couple of AVIs and WAVs sitting around as well as the source videos, scripts, and Audacity projects that spawned them, so I can safely delete them once I don't think I'll need them anymore, and in an emergency, I can recreate the final product using the same method I used in the first place. Personally, I hate junking any original source, even if I only think I'll ever need the edits, but that's just me, and I've got a 2TB backup drive that I'm actively using and another 2TB drive still in the box. Considering that I'm just getting around to throwing out boxes from electronics that I've had for years (as well as the boxes from the ones I've already gotten rid of), I may not be the best person to take advice from regarding keeping old things you don't plan to need again.
|
|
#
¿
Dec 18, 2013 14:41
|
|
|
Suspicious Dish posted:A while back in old of the old threads I swore I posted a Star Wipe thing for AviSynth, but I can't find it anymore. Can anybody help me search through the old threads for it? I'm lazy and don't want to rebuild it. Well, I found you saying that you'd made it in the old TSF, but I can't find anywhere that you actually posted a link to it.
|
|
#
¿
Jan 18, 2014 14:48
|
|
|
Since we're talking transitions, here's one I wrote up ages ago and just don't recall ever having occasion to post. Remember those sitcoms where they used to do flashbacks by zooming in until the screen was a pixelated mess, then zooming back out into the new scene? This is sort of like that, but without the zooming - you could combine it with a zoom easily enough, I suppose, but I don't know how that would look. This is a simple function that just resizes the clip to make it look pixelated, then resizes back to the original size with the giant pixels. The optional parameters are there to adjust the effect to work better with your specific clip. "left" and "right" are obviously the before and after clips, "numframes" is how many frames of each clip to use in the transition, "minratio" optionally sets the minimum resize (aim for as close to 4 pixels in each dimension as you can get for the best effect - too big and it won't get very pixely, too small and it'll sit at 4x4 for a noticeable time), and "rate" is an exponential factor to control how quickly the effect happens - higher numbers decrease the amount of time the image just looks blurry at the start instead of properly pixelated. It works best for less than a few seconds' worth of numframes. Let me know if there are any problems with it, and I'll try to improve it.code:
|
|
#
¿
Jan 18, 2014 18:18
|
|
|
It's possible to have Avisynth make the mask for you automatically if you have a clip of the character model you want as well as a clip of the same background without the character in it. If that's the case, I can provide a walkthrough for how to extract a clip of the character alone, and it should be pretty easy to grayscale and overlay it from there. It's rather a special case, though, and probably won't apply to most games.
|
|
#
¿
Jan 19, 2014 23:54
|
|
|
Psion posted:You know, that didn't occur to me. I now feel dumb. I'll give that a shot, thanks. I usually run all of my videos through a ChangeFPS, and any color conversion or resize functions, right at the start just to make sure there are no mismatches along the way. If you write a function at the top of your script and call it on every input, you guarantee consistency.
|
|
#
¿
Jan 20, 2014 05:32
|
|
|
I'd recommend doing a System Restore first, if you have that option - go back to a configuration that you know worked and start from there. When it comes to diagnosing capture device problems, I tend to start with VirtualDub's capture program. That will let you play with the device's source, compression and encoding options, and lots of other stuff that might help you figure out what's going on.
|
|
#
¿
Jan 20, 2014 22:41
|
|
|
If there are syncing issues and you have the game audio captured with the video, you can still fix it later. I have no idea why you would think that you couldn't. It's possible to export the audio from the file, do whatever you want to it, and then reimport. You can also fix certain kinds of desync during the editing of the video. The only thing that tends to interfere with resyncing (in cases where it would otherwise be possible to do so) is if you record live commentary along with the game audio and don't separate them into different files. I think it's actually MORE likely that the tracks will desync if you use separate programs to capture them, because each program would be unaware of the other program's unique hiccups or delays. If you do capture game audio in VDub, I'd check out those Resync settings - they seem to be there for the specific purpose of minimizing desync. Do several test recordings with each setting to see which one seems most reliable.
|
|
#
¿
Jan 25, 2014 14:15
|
|
|
Your Dead Gay Son posted:Edit 2: Okay, another dumb thing I can't figure out... If you do multiple edits, the effects of each edit apply to the input to the next edit. So if you did something like: trim (100, 0) trim (500, 14395) then by the time it does the second trim, it's already chopped 100 frames off the front. If you're doing this, then you'll need to recalculate the correct frames by exporting after each step. This is why I never do an edit from an edited video and never do an edit without assigning it to a variable, unless I've specifically exported my video to recalculate frame numbers. And even then, it limits what I can do afterward because any change I make to an upstream video will affect the later videos in the sequence. Just don't do this. Ever.
|
|
#
¿
Jan 27, 2014 03:00
|
|
|
Geemer posted:Wouldn't that limit him to only cutting on keyframes? It's been a while since I've used pure VirtualDub, but I seem to remember something along those lines. Yep, keyframes only. Depending on how frequent the keyframes are, that may or may not work for the video. It sounds to me like someone hasn't actually investigated the quality options for their encoding program or has picked a program that doesn't provide enough options - I've never had a problem with MeGUI and 264 at an appropriate quality setting, especially for DOSBox recordings.
|
|
#
¿
Mar 1, 2014 01:55
|
|
|
Perfect quality and small file size pretty much automatically means the built-in ZMBV codec. I didn't think anyone recorded from DOSBox using external tools.
|
|
#
¿
Mar 1, 2014 03:06
|
|
|
dis astranagant posted:Also repeating a previous edit since no one probably saw it: Looks like my source video is missing the last 2 seconds of audio and that's pissing off the encoder. What can I do to fix that? Use vdubmod to cut that part out? In Avisynth, you can either pad the audio track with silence (Blankclip) to the right length and AudioDub that onto the video from the original clip, or cut the silent frames from the video and AudioDub the original audio onto that - it's probably best to AudioDub either way to avoid some weird case where cutting the silent video also somehow cuts into the audio from prior to that. You could also just AlignedSplice a one-frame Blankclip onto the end to force it to pad with silence.
|
|
#
¿
Mar 8, 2014 13:52
|
|
|
dis astranagant posted:Trimmed the silent frames out (and a few extra to be safe) and dubbed the audio on and the latest attempt will sometimes go out of sync by a second or so then resynch somehow I'm going to suspect that the editing you've done has gotten out of order. Could you post the script you're using, just for reference? Also, are you sure that didn't happen before making this change?
|
|
#
¿
Mar 8, 2014 22:56
|
|
|
The Unholy Ghost posted:Here's where the problem comes in. I'm trying to connect everything together, but the USB Live 2 Device only has female AV Cord Inputs. The 3-Way Splitter also has female AV cord inputs, so there's no way to connect the thing. It did come with a dual male AV Cord, but I have to use that to connect the TV. Is there something else I need to buy now, or is there some complicated wiring technique I need to learn? Gender changers are pretty standard A/V equipment, and I always have a few spares lying around, particularly when I'm using splitters. Just run out to your local Radio Shack or equivalent and pick up the appropriate connectors. They're cheap.
|
|
#
¿
Mar 9, 2014 01:29
|
|
|
Artix posted:So I have a 40 minute video that I decided to break up into two parts. I set up all my edits in AviSynth, tell it to only render the first few parts of it, feed it into MeGUI, and I get a nice, crisp 20 minute video. Great, exactly what I wanted. Now I go I to render the other half, so I comment out the line that had the stuff for the part 1 video and add in a line with the stuff for part 2. I have not changed anything else about the script aside from adding this line. Now, VirtualDub crashes every time I try to load the script. If I go back and try to load the entire video from the beginning it's perfectly fine, including the part of the video that I want in the second half. It just absolutely refuses to load anything that doesn't start from the beginning. The script in question: You left out the part of the script that defined vid, but I can't imagine that matters. If you can create a video that contains everything you want for your second part, why can't you just do that, then Trim and Fade that?
|
|
#
¿
Mar 10, 2014 23:52
|
|
|
Kaubocks posted:...however when I try to open my code, I get the error "Layer: image formats don't match". I'm not sure what's causing it because I've never done something like this before. For reference, here is my code: That means that the image formats to the clips you're trying to layer don't match. Intuitive, I know. To fix it, you'll need to figure out what video format your "video" clip is and convert your "annoying" clip to that format. You can either "return video.Info" and look at the colorspace line, or "return video", open it in VDub, and use the File Information menu option to see it. Of course, you can also just comment out the line that has the layer command, since you never actually put that into the returned clip... unless "image" is supposed to be the same clip and you forgot to change it?
|
|
#
¿
Mar 16, 2014 01:15
|
|
|
slowbeef posted:Projected filesize 666 GB? Yikes. Uh, I guess I have to look into this frameserver business! Did you choose a lossless codec or just leave the video uncompressed? I've made that mistake a few times.
|
|
#
¿
Mar 19, 2014 22:34
|
|
|
dis astranagant posted:What's the best way to sync things up for post-commentary? Maybe I'm overthinking things but hitting record in Audacity then switching to MPC and hitting play sounds impossible to sync up and would make some editing headaches later. I put my microphone near the mouse when I click the play button, so the click comes through nice and loud in the audio. If you can't match the click sound to the start of the video, there's something wrong with one of the programs involved, or your sense of timing. There is a slight additional delay, but that's going to be true of any method you use, and if it's noticeable in the final video, you can just slide the audio to the left a bit to compensate.
|
|
#
¿
Mar 25, 2014 11:01
|
|
|
Silver Falcon posted:Huh. Well it exported just fine as a Flac. Thanks! Now, can I WavSource a Flac in AviSynth, or do I just have to shove that Flac right into MeGUI with my script and use it as the audio file? You can only WavSource a WAV (or possibly an AVI with a WAV-compatible audio track), but BassAudio should have a source filter for flac. BassAudio handles just about everything. (I grabbed that link from the Avisynth wiki - the wiki itself provides few details for most external filters, so it's usually easier to go straight to the source.)
|
|
#
¿
Mar 28, 2014 14:01
|
|
|
skoolmunkee posted:What I did was mute the game audio and export the commentary tracks as an mp3, then re-import that and autoduck against it. Then you can export the final audio track with the individual commentaries and the autoducked game. It's probably easier just to merge the tracks in Audacity, provided that the relative volumes have all been adjusted properly. That collects them into a single track that you can autoduck against without compressing them further.
|
|
#
¿
Mar 30, 2014 03:06
|
|
|
Tin Tim posted:Hey, folks. I have a little problem with a game that I'm currently trying to record. So far, I recorded all my stuff with FRAPS at 30fps, and never had trouble. But this game slows down hard when I use 30Fps, and the control input lags a lot. When I just look at the Fps counter without recording, the game ranges from 90 to 100something Fps on average. So I tried recording with 60Fps for an experiment, and while the game ran smooth and without issues, the raw footage was rather choppy when I played it back with the Vlc player. Have you tried transcoding the 60ish FPS video to a flat 30 FPS and running that through VLC, or encoding a preview video with the same framerate but a lossy codec? 60 FPS with lossless compression is pretty chunky, and not many programs will handle it smoothly. When I work with 720p video in Avisynth, I always have to export in a lossy codec so I can preview it at normal speed. All the edits are the same, frame for frame, so I can throw away the preview version when I export my final video.
|
|
#
¿
Apr 12, 2014 20:42
|
|
|
TheMostFrench posted:The converted AVI plays fine with audio in stuff like VLC media player, but when I try to edit it with an AVS script (I've tried VirtualDub and AvspMod) I get an error 'Error initialising audio stream output size'. KillAudio couldn't possibly work, because you have to load the audio before you can kill it. If it can't load... Also, general Avisynth note: You're loading the same video twice in that script. Put it into a variable the first time and use the variable if you want the raw clip in later lines. 1) Can you open it in VirtualDub without a script? If so, try re-encoding the audio and Direct Stream Copying the video and see if that works. Not ideal, since that's a lot of re-encoding and MP3 is lossy, but I think the problem is that Avisynth doesn't natively work with MP3. 2) You might want to jump back a step in your process and skip ffmpeg, use MeGUI's File Indexer to work with the original MOD or MPG file, and see how that works. There should be a tutorial in the OP for the HD-PVR that includes the necessary steps. 3) Another option if VDub works is to save a Direct Stream Copy of the audio as an MP3, then download the BassAudio extension and use that to import the MP3. Save a separate Direct Stream Copy of the video with No Audio that you can AviSource, and AudioDub them together in the script. 4) As an alternative to any of the above, figure out how to use ffmpeg to output different formats. This part, I don't know myself, but you don't want MP3.
|
|
#
¿
Apr 13, 2014 13:47
|
|
|
Tin Tim posted:So here's my follow up on that. When transcoding the 60fps recording to a 30Fps video, I end up with a video that looks like it has constant max payne bullet time, and an audio track that's shorter than the video You did something wrong, then. It's supposed to take the 60 fps video and drop about half of the frames to make it a 30 fps video running at the same speed, not double the duration of each frame.
|
|
#
¿
Apr 13, 2014 21:25
|
|
|
I found a Bob deinterlace to be adequately effective at removing the scanlines when I did the Toonstruck LP. If you check the opening video, you can see how the result looked. I'd have to check whether I used any further editing to brighten it afterward, though.
|
|
#
¿
Apr 15, 2014 03:42
|
|
|
StrangeAeon posted:It may have been mentioned somewhere already, but has anyone noticed a bit of a purge of LPs going on at the Archive? Blip and Viddler videos seem to be missing in numerous archived LPs. You mean the one that's been going on for the past year or so? I think everyone's had time to notice by now. Click the big red banner at the top of the page with the missing videos and you can download them. I don't know how everyone manages to miss the giant red banner, but they do. We need a big red banner at the top of each page that says "HEY, LOOK AT THE BIG RED BANNER AT THE TOP OF EACH PAGE" or something, but that's not Tech Support Fort material. Try asking in the LP Archive thread.
|
|
#
¿
Apr 21, 2014 04:02
|
|
|
The T posted:Is there a way to figure out what the stretch ratio is, though? I'm bad at math. With the commentary track exported at exactly the length it should be, pull the video up in VirtualDub and from the Audio menu, pick Audio From Other File and open the commentary audio. Now, in the Video menu, pick Frame Rate and next to "Change so video and audio durations match", it will tell you what AssumeFPS value to use if you're changing the video speed. If you want to change the audio speed in Audacity instead, the ratio will be 1 divided by that number. I do have to ask the stupid question, though - are you commenting over a 30-ish FPS exported version of your video, or just playing the Avisynth script and recording over that?
|
|
#
¿
Apr 28, 2014 23:02
|
|
|
There was an Audacity Crash Recovery utility that I believe automatically did all the manual work of combining the audio clips into a single track - it's gone out of fashion with Audacity 2.0's recovery feature, but you might find it helpful.
|
|
#
¿
May 4, 2014 17:03
|
|
|
Here, again, is my go-to format for all of my video editing. Do it in this style, and you should rarely, if ever, get confused.code:
|
|
#
¿
May 16, 2014 23:35
|
|
|
If you've got constant desync throughout the video, can't you just shift the audio and call it good?
|
|
#
¿
May 20, 2014 01:36
|
|
|
Try encoding a short clip (do a Trim(0, 50) or so on your final video) and see whether the problem is still inherent in the final output.
|
|
#
¿
May 21, 2014 23:18
|
|
|
I've also had success with fading text using Animate to transition from a color with 255 alpha to the same color with 0 alpha, or vice versa. It's a bit more work, but I think it would render more quickly, if that matters - dissolves are pretty slow.
|
|
#
¿
May 26, 2014 22:11
|
|
|
|
| # ¿ May 2, 2024 02:35 |
|
|
turtlecrunch posted:1) Can I just copy and paste this over and over again in the same .avs file while changing the number of the avi (so 2.avi, etc) then import the script to MeGUI and it will batch encode all my clips? Or do I need a separate script for each one? If so, is there a way to make MeGUI process multiple scripts in a sitting? I'm surprised nobody answered this, so... yes, you can do batch work with MeGUI, and in fact that's how I do most of my processing anymore. You load each script and hit Queue under each of the video and audio sections to queue all of the encodes, then go to the Queue tab and hit Start to make things happen. You can even tell it to shut down the computer when it's done if you want it to run overnight. (And unlike VirtualDub, it resets that to "Do Nothing" each time you start the program.) Once the encodes are all done, you'll need to use the Mux tool to finish the job, but you can queue all of the mux jobs as well, and they don't take very long. I recommend using this method in general because I usually end up finding something in the audio track that I need to fix post-encode, so I like being able to set the audio and mux jobs to Waiting and hit Start to quickly re-encode without having to reload the script and re-queue jobs. You can empty the queue once you know you like what you have.
|
|
#
¿
May 29, 2014 23:39
|
|