I am using a Roxio VHS to DVD card, but I am using OBS instead of ROXIO.
My plan was to capture and edit in MP4 x264 and then encode to MPEG2 for the DVDs
Should I :
1. Capture in MPEG to to make DVDs, then convert to MP4?
2. Capture to MP4 to make video for modern TVs?
3 Capture in both MPEG2 and MP4 and edit twice?
Which makes more sense to you?
+ Reply to Thread
Results 1 to 30 of 34
OBS is a screen capture it doesn't manage well digital signal coming from a capture card, and most importantly it is designed for HD and 4K streams not SD.
As mentioned already, lossless all the way until the final formats.
One has suggested I use virtualdub
One thing is for sure. I don't want to Use Roxio. I'm frustrated and confused.
The virtualdub suggestion concerns me because it produces AVI files, which aren't around anymore.
I want to produce video that a default windows installation would play without having the people who view my video forced to install HuffnPuff.
My goal is to use standard formats readable on every computer and TV for years to come. I don't want to use ancient file formats.
I see people on youtube using OBS and it looked easy. The video looked at least as good as the one roxio captures.
My motto is, if a file cannot be played on the most popular devices by default, then it is lossy because only some video guru with a specific configuration can play it.
So I want something that can be played on every TV. Played on every computer. Played on every DVD Player. EDITED ON ANY PROGRAM.
MP4 seemed like the standard that does all this in the modern age. Mp4 was from God. HuffnPuff sounds like the Devil.
So what software editor reads HuffnPuff and logorithm? What do I use to convert it into MP4 and MPEG2?
So what software editor reads HuffnPuff and logorithm?
I don't want to use ancient file formats.
My motto is, if a file cannot be played on the most popular devices by default...
So I want something that can be played on every TV. Played on every computer. Played on every DVD Player.
I get it.
I'm not funny.
Abysmal Ignorance. This is true. I have not done this before.
I have to hand it to you. You've handed my my rear properly.
So I think I have figured out how to get virtualdub communicating with the capture card.
I installed lagarith and it shows up as an option in video.
What codec do I use for lossless audio?
Don't compress the audio while capturing. Compress it later to AC3 (for DVD) and AAC (for MP4).
For the audio setting during capture, use 48000-16bit. Both video and audio will be captured into the same file. Any video editor can read AVI---though you may need to get the Lagarith plugin, if it is not already installed. On "dead" video formats, you may be thinking of FLV, which was the old Adobe Flash format. Always edit and save in a lossless format. Once your video is ready for presentation, make your compressed copy. You will need to go with MPEG-2 for a DVD, and MP4, if you wish to upload to Youtube.
For compression, I have heard that "HandBrake" is pretty good---and it's free. I have yet to try it, so some others here might have better suggestions. For playing your uncompressed video files, you might want to look into "VLC Media Player". It's basic, and doesn't take-up a lot of memory or processor resources. It's also free. You will recognize it by the traffic-cone icon.
One other note: You will find that in saving video files, you must always save as a new file. Video editors will not let you save to the same filename, as the one currently open. Video files are a "streaming" format---so, they cannot save over themselves.
Since this post, I have tried capturing with virtualdub and both the huffyuv and lagarith codecs, The results are worse than with OBS capturing to MP4. The video and audio freeze intermittently. I have a clean installation of windows 10 and virtualdub 2. This is why it seems easier to use OBS. I think lossless is not an option for me. The tapes are damaged also I think, but the audio in particular is much better with OBS. I can't have audible dropouts when there are none on the tapes. The software is what is causing this. Lossless should not be worse than x264. I have already had my lecture on the evils of not using lossless.
VirtualDub is hard to get set up for video capture. Be sure not to play the audio while capturing, Audio -> Enable Audio Playback (unticked). That is one of the big issues with VirtualDub. Also, don't compress the audio while capturing, just save the audio as uncompressed PCM. If you still have issues go to Capture -> Timing... and try the different sync options, especially Do Not Resync. Beyond that read this: https://forum.videohelp.com/threads/104098-Why-does-your-system-drop-frames
Last edited by jagabo; 12th Apr 2021 at 13:49.
According to virtualdub, it isn't dropping any frames. But I know the tape isn't causing it because when I capture with other software and codecs, the audio is fine. I would think that adding the overhead of compression would be worse than with lossless.
I am thinking maybe I should capture once for mpeg2 and once for mp4.
The settings I was following from Tim Ford's article on capturing VHS with OBS direct you to set the canvas to 720x480 and the output to 720x540.
I was using Kdenlive to transcode the MP4 to MPEG2 VOB files, but he also said use 720x540 on the output for DVD.
I am leery about using a weird aspect ratio, But he suggests this to overcome the fact that modern devices have rectangular pixels.
For the “Output (Scaled) Resolution” setting, type in 720×540. The reason for this is that your old-school VHS tapes use a resolution that will not look correct when played back on a typical computer screen (it will look a bit stretched). By changing the output resolution to 720×540, you’ll be using a 4:3 ratio which will look correct on more modern devices. If your intent is to burn the files to a DVD for use in a DVD player, you can use 720×480 as the output.
Whoever that Tim Ford is I can guarantee you he knows nothing about VCR other than insert a tape and press play. There is more to analog video than just a video signal, It has horizontal timing signal, vertical timing signal, chroma signal, luma signal and a specific capture resolution for each color standard. If any of the mentioned signals is weak or missing you get problems.
No Monamo, I didn't make it up, at least not intentionally. It's a typo. I should have typed 720x480 as he clearly stated in the article.
The part you are referring to is what is puzzling me.
Here's the source of my confusion.
He recommends capturing and outputting in 720x540 for computers and other devices but 720x480 if you are creating DVDs.
My intention is to do both. So should I follow the 720x540 advice and then encode to MPEG2 720x480 or follow the 720x480 advice, but create another MP4 from it that is 720x540?
Or should I just capture twice for each separate purpose?
Always capture NTSC video at 720x480, YUY2 of other YUV 4:2:2 chroma subsampling, with a lossless codec. Filter, resize, and encode later as suits your purpose(s).
So my most recent experiment is to use Virtualdub, Capturing at 720x480 using h264 lossless for video and PCM for audio. Still getting intermittent blue screens due to the tape. But they occur at different points that were successfully captured before. I will see whether they are freezing and having audio glitches, which is why I was so stubborn about OBS, which wasn't perfect either.
I do have a couple more question and I don't know if they belong in another thread.
The VCR I am using has a progressive scan switch set to on. Mediainfo reports progressive scan. Would that mean that the video I am capturing was never interlaced? Do some camcorders record progressive or are they all interlaced?
Also I am wondering if capturing directly from the original camcorder might help some of the problems from the tape. What are your thoughts?
FFmpeg supposedly has a H264 "lossless" mode but it has been proven to be not so lossless in reality - certainly less artifacting than a great many other codecs but not completely "lossless".
Also please keep in mind that when recording audio with exactly the same specifications as the ORIGINAL AUDIO bit for bit as PCM your result is a lossless copy of the original. Record that same audio using a LESSER recording specification (i.e. less sample rate and bit depth) and you now have a "lossy" copy of the original regardless of the fact that it's recorded in a PCM format or not.
Your editor is deinterlacing the h.264 video, but not the YUY2 video. And you've resized them, further messing them up. Use AviSynth to open them and you'll see they're identical:
v1 = LWLibavVideoSource("yuy2.avi", cache=false, prefer_hw=2) v2 = LWLibavVideoSource("x264lossless.avi", cache=false, prefer_hw=2, format="YUY2") Subtract(v1,v2) Levels(127, 1.0, 129, 0, 255)
[Attachment 58465 - Click to enlarge]
[Attachment 58466 - Click to enlarge]
Last edited by jagabo; 17th Apr 2021 at 22:52.
That sample is not the perfect example to show compression artifacts, If you look at the bottom text you can see them clearly.
Those are deinterlacing artifacts, not compression artifacts. Once again, the software you are using is deinterlacing the h.264 encoded video, but not the uncompressed video. There are no compression artifacts. The x264lossless video is truly lossless -- unless you don't know how what you're doing. Which is probably a good reason for jpvoodoo to avoid it and stick with Lagarith or UT Video.
Last edited by jagabo; 18th Apr 2021 at 07:35.
Well ok.. Once de-'lacing was turned off on the one they were more in line with each other, sure..
Let me ask you this. Say you have a transport stream container (.TS file) of 1 audio and 1 video stream -
the audio stream == AC3 5.1 surround
the video stream == MPEG4 AVC H.264 9Mbps 720P
File size == 2.8GB
Can this be put to a LOSSLESS (i.e. no loss in quality from the original) H.264 matroska file of lesser file size?
If not, there is no point to this particular conversation...
If yes, how? That doesn't take 1-2 hours of encode time to complete.
Last edited by thghgv; 18th Apr 2021 at 13:10.