Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Capturing and Editing Video > Avisynth Usage

Reply
 
Thread Tools Search this Thread Display Modes
Old 24th September 2015, 04:34   #1  |  Link
rs008f
Registered User
 
Join Date: Aug 2003
Posts: 129
VHS restoration.

http://www.mediafire.com/watch/v316e...whitevideo.avi
Black and white video of just text. A lot of static/video noise. Background audio noise. Garbage at bottom
How can i restore this video?
rs008f is offline   Reply With Quote
Old 24th September 2015, 04:40   #2  |  Link
feisty2
I'm Siri
 
feisty2's Avatar
 
Join Date: Oct 2012
Location: void
Posts: 2,633
Tell Santa u wanna magic video restoration machine this xmas
feisty2 is offline   Reply With Quote
Old 24th September 2015, 04:52   #3  |  Link
jmac698
Registered User
 
Join Date: Jan 2006
Posts: 1,867
How long does it go on? It might be better to just recreate it. For audio, you can use the denoising available in Audacity. For the noise at the bottom, you can just mask it out. For changes in brightness, I could write a script just to look for the fade-ins and outs and adjust the text in between to have even brightness. For the wiggling, I could match the position line-by-line and stabilize it.
jmac698 is offline   Reply With Quote
Old 24th September 2015, 06:55   #4  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,685
Wow, I think some people just enjoy making unhelpful comments. I really don't see the point in doing it.

Do you still have the videotape, and can you re-capture it? A lot of the video problems in the text might be corrected by using a time base corrector (TBC) in the capture chain. The noise at the bottom can be masked, or cropped (I favor masking, in order to not have to re-size the video). The horizontal noise dropouts can be eliminated using any one of several plugins. I'd need to see more in order to know which one might work best.

The linear audio track is very low fidelity, and you can't create the lost high frequencies, so you are out of luck there. However, there are many plugins that can do a great job on the hiss, without requiring a noise gate, something which I definitely dislike because it makes the audio seem to pop in and out. Unfortunately, the tools that do a good job on the audio are not cheap. I use iZotope RX3 Advanced, which is a $1,200 tool. I do not know of any shareware or freeware tool that can even begin to touch what you can do with this. It is as close to magic as any software I have ever used. But, just to be clear, it will not be able to add "brightness" or fidelity to your audio; it will only reduce the hiss.

[edit]I just saw the post about Audacity, and while I haven't used it, I have heard it is a really good tool. I just scanned the features, and while I didn't see noise reduction mentioned, it sounds like there are some denoising tools available. You should try it. Also, if you want to post an MP3 of the audio track, I'd be happy to take a shot at denoising it with iZotope RX3.

Last edited by johnmeyer; 24th September 2015 at 06:58. Reason: Added line about Audacity
johnmeyer is offline   Reply With Quote
Old 24th September 2015, 07:01   #5  |  Link
jmac698
Registered User
 
Join Date: Jan 2006
Posts: 1,867
There is a way to add higher frequencies, it's called spectral replication, it basically follows the pattern of the lower frequencies but in higher frequencies. This has been around for a long time, for example the Aphex Aural effects boxes. I used one in an effect pedal and it made a sample sound very nice.
jmac698 is offline   Reply With Quote
Old 24th September 2015, 07:18   #6  |  Link
rs008f
Registered User
 
Join Date: Aug 2003
Posts: 129
OK Thanks for help guys. I used a Canopus ADVC110 with WinDV software and a SONY SLV-N88 VCR to capture. Looks like there's nothing to be done, garbage in, garbage out, I guess.

Last edited by rs008f; 24th September 2015 at 07:31.
rs008f is offline   Reply With Quote
Old 24th September 2015, 07:47   #7  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,685
Quote:
Originally Posted by jmac698 View Post
There is a way to add higher frequencies, it's called spectral replication, it basically follows the pattern of the lower frequencies but in higher frequencies. This has been around for a long time, for example the Aphex Aural effects boxes. I used one in an effect pedal and it made a sample sound very nice.
Is that like the "audio extrapolation" in the old Nero Toolkit? I used that a few times on 6-hour linear VHS audio, and it did add a little "brightness" to the audio that was a little better than just "turning up the treble" and then de-noising. However, it was not quite in the same "FM" (f- magic) as some of the hum reduction, spectral repair, de-noising, de-reverb, and pop & click reduction that I do every day with RX3.
johnmeyer is offline   Reply With Quote
Old 24th September 2015, 07:56   #8  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,685
Quote:
Originally Posted by rs008f View Post
OK Thanks for help guys. I used a Canopus ADVC110 with WinDV software and a SONY SLV-N88 VCR to capture. Looks like there's nothing to be done, garbage in, garbage out, I guess.
I haven't used the ADVC110, but I just skimmed through the manual, and the only thing I see that you should check is DIP switch 5. It changes the setup level, and if I am reading the manual correctly, it should be set to ON.
johnmeyer is offline   Reply With Quote
Old 24th September 2015, 11:34   #9  |  Link
jmac698
Registered User
 
Join Date: Jan 2006
Posts: 1,867
I used the wrong term, it's called exciter
https://en.wikipedia.org/wiki/Exciter_%28effect%29

And yes, that looks like a great program, however things like de-reverb can be done for free, for example it's a demo script in the Matlab DSP toolbox, and the technology has been used in conference phones for a long time, then cell phones. And with stereo, you can separate a voice from background quite easily, basically, what's loudest in both channels? Keep that and remove the rest.

Here's a list of plugins
http://www.musicradar.com/tuition/te...lug-ins-538311

Last edited by jmac698; 24th September 2015 at 11:40.
jmac698 is offline   Reply With Quote
Old 24th September 2015, 17:03   #10  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,685
Quote:
Originally Posted by jmac698 View Post
I used the wrong term, it's called exciter
https://en.wikipedia.org/wiki/Exciter_%28effect%29

And yes, that looks like a great program, however things like de-reverb can be done for free, for example it's a demo script in the Matlab DSP toolbox, and the technology has been used in conference phones for a long time, then cell phones. And with stereo, you can separate a voice from background quite easily, basically, what's loudest in both channels? Keep that and remove the rest.

Here's a list of plugins
http://www.musicradar.com/tuition/te...lug-ins-538311
I'll have to check out all of those. Thanks for the links!

De-reverb is really tough, and is one of the weakest portions of RX3 (where it was first introduced) and, I assume, the just-released RX4. However, there is a reason it doesn't do too well. The problem with "dereverb" is that you only have two microphones in many camcorders (unless you are dealing with 5.1), spaced at an unknown distance apart, often moving around the room as the camera moves. One reason the Polycom conference phone works so well (they pioneered the echo cancellation) is that they have many microphones, pointed in all directions, in a configuration and distance that is precisely known, and the recording instrument (i.e., the phone) is not moving. This gives the engineers a lot more "stuff" to work with when they design their cancellation algorithms.
johnmeyer is offline   Reply With Quote
Old 24th September 2015, 19:44   #11  |  Link
jmac698
Registered User
 
Join Date: Jan 2006
Posts: 1,867
I think the movement is the worst part, if everything is stationary is only takes a few seconds to learn (ie train the filter coefficients through feedback), at least from the demo I remember playing with. When things move, the sound is changing too fast to "lock on" to

btw, I don't see a feature for track alignment. This is where you use another device to record audio then automatically line it up to the camcorder audio, so you can use it for a replacement.

Last edited by jmac698; 24th September 2015 at 19:52.
jmac698 is offline   Reply With Quote
Old 24th September 2015, 20:57   #12  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,685
Quote:
Originally Posted by jmac698 View Post
btw, I don't see a feature for track alignment. This is where you use another device to record audio then automatically line it up to the camcorder audio, so you can use it for a replacement.
I'm not sure whether you are talking about not seeing the track alignment feature in RX3/RX4, or in some other program or plugin. The iZotope products definitely do NOT have this feature.

If I understand what you are asking for, I use PluralEyes. It is fantastic.

I just shot an outdoor wedding two weeks ago with four cameras, plus a Zoom recorder connected to the P.A.'s soundboard. I put everything on five timelines in my NLE, pushed the button in PluralEyes, and came back ten minutes later. All the timelines were synced, with the stuff that didn't have any match pushed off to one side where I could deal with it manually. Most of these un-matched segments were from the reception where I sometimes only used one camera.

I've used PluralEyes before, but this was particularly impressive because the cameras that were a long ways from the speakers had really weak audio, mixed with lots of ambient outdoor noise, including wind. Despite that, PluralEyes still got everything right. Pretty impressive product, and not much money.

BTW, if you ever decide to use it, get the older version 2.x. The 3.x version broke a lot of things.

Last edited by johnmeyer; 24th September 2015 at 21:02. Reason: Added the PluralEyes version 2 suggestion to the end of the post
johnmeyer is offline   Reply With Quote
Old 25th September 2015, 07:17   #13  |  Link
sven_x
Registered User
 
Join Date: Oct 2011
Location: Germany
Posts: 39
I gave the audio part a short try, using Wavelab 6 and some plugins. (Sorry, attaching mp3 files is not possible in this forum).

- Oxford denoiser with default settings, + increased sensitivity in the high range (it is helpful to use the "difference" function to only listen to what would be removed from the audio track and than adjust the various parameters)
- hyperprism dx exciter, "air" setting
(Only use it with care. It may bring back the noise which was removed in the first part)
- also added a "quasi stereo" effect, which is not necessary

blackwhiteaudio-pseudo stereo.mp3 (html)
blackwhiteaudio.mp3 (html)
sven_x is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 05:25.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.