Help with filtering pictures problems from VCR capture

[gil900]

Hello,
I have a VCR tapes and DVD recorder.
I decided I wanted to convert the tapes into digital format.
So I connected the VCR to DVD And started to record the tapes in the DVD.

Then I put the disc that the DVD did in my laptop And took the files from the disk.

Files were:
VTS_01_1.VOB
VTS_01_2.VOB
And so on ..

Then I downloaded the plugin MPEG2.vdplugin for the Software VirtualDub so I can load the files in the software.

So far so good.

I used some filters to video but I'm looking for a filter to remove frames with this disorder:
http://ufu.co.il/files/9uzzwb4aajla451mnq1e.jpg

These problems in the picture Appear in split seconds, so I want to remove or fix These split seconds frames With a filter or something else..

How to do it?

Another type of problem is this:
http://ufu.co.il/files/ohnpojf729y1y8tor2hd.png

That happens for a split second. But these split seconds I do not want to delete.
I want someway to fix this kind of problem.

Thank for the helpers!

[sanlyn]

Can't help much with still photos. You should upload a few seconds of the part of the video that has these problems. About 12 MB of mpeg could be up-loaded, oughtta be enough.

[gil900]

Can't help much with still photos. You should upload a few seconds of the part of the video that has these problems. About 12 MB of mpeg could be up-loaded, oughtta be enough.

The following video:


Have this problem:
http://ufu.co.il/files/ohnpojf729y1y8tor2hd.png

and this problem
http://ufu.co.il/files/9uzzwb4aajla451mnq1e.jpg
In less serious as in the picture ..


EDIT:
NEWvideo about the second picture::

[sanlyn]

In order to correct your mpeg videos (I assume you recorded to standard DVD, which is MPEG2 format), the mpeg should be converted to lossless compressed AVI (using huffyuv or Lagarith lossless compressors). Because the video has already been subjected to lossy compression (MPEG) there would be considerable data loss, noise, and compression artifacts, besides the problem of damaged frames. Further conversion with MPG -> Flash video incurred additional loss. I don't think this problem can be fixed in MPEG. You would have to work in AVI with Avisynth.

[gil900]

In order to correct your mpeg videos (I assume you recorded to standard DVD, which is MPEG2 format), the mpeg should be converted to lossless compressed AVI (using huffyuv or Lagarith lossless compressors). Because the video has already been subjected to lossy compression (MPEG) there would be considerable data loss, noise, and compression artifacts, besides the problem of damaged frames. Further conversion with MPG -> Flash video incurred additional loss. I don't think this problem can be fixed in MPEG. You would have to work in AVI with Avisynth.

So you're telling me I need to convert it to AVI format?
That's no problem .. I have converted AVI format without compression (60GB Size )

So what to do?
Use this conversion with Avisynth?

If so, what next?
How do I fix it?

EDIT:

EDL (oops, sorry. Interrupted a small emergency here at home). So, much depends on how much you're willing to learn. Members here can help with many ideas, and a few chuncks of edited MPEG would be helpful. If you need help or tutorials in editing MPEG, decompressing to AVI, etc., feel free to ask. You did mention that you used "some filters" with your videos, but we need more detail about that.

Yes, I prefer to edit the MPEG with VirtualDub.
I loaded the VOB file (MPEG) in VirtualDub with plugin And started working on it directly without conversion to AVI.

Already used in several filters.
this is my filters settings:
http://postimage.org/image/wqacd4zep/

EDIT:

The following video example:
https://www.dropbox.com/s/9oicxontsq8gctq/test%204.avi
I've done to define filters in VirtualDub and it is almost ready.
I could not fix is this:
http://ufu.co.il/files/9uzzwb4aajla451mnq1e.jpg

This phenomenon occurs many times in this video example.

[poisondeathray]

If they are only a few bad frames you can probably replace the bad frames or fields

Either with duplicates, blends or motion interpolated frames/fields using mvtools2 and avisynth

[gil900]

If they are only a few bad frames you can probably replace the bad frames or fields

Either with duplicates, blends or motion interpolated frames/fields using mvtools2 and avisynth

This automated process? Or I have to choose any frame which is not good either manually?

I know only use VirtualDub

I can do it in VirtualDub?

[poisondeathray]

You have to manually specify the affected frames or fields with numbers. There isn't any algorithm that can distinguish effectively between what is a "good" frame or what is a "bad" frame.

You can't do it in vdub alone

There are some examples on how to do this in this forum if you use search

[gil900]

You have to manually specify the affected frames or fields with numbers. There isn't any algorithm that can distinguish effectively between what is a "good" frame or what is a "bad" frame.

You can't do it in vdub alone

There are some examples on how to do this in this forum if you use search

Thanks but I do not know how to search with my bad English.
Can you give me a link please?

[poisondeathray]

I'm not sure if it will be helpful; You're probably not going to understand it unless you put in a few good hours learning avisynth basics first . Even if English was your first language it's difficult to learn avisynth at first

https://forum.videohelp.com/threads/343296-Transfer-of-real-bad-tape-to-pc-and-restore

In this example, there was a bad middle section that was interpolated (replaced frames from generated new frames) . (there was more done to it , like dropout removal, but just to illustrate frame replacement)
https://forum.videohelp.com/threads/343296-Transfer-of-real-bad-tape-to-pc-and-restore?...=1#post2139402

There are other ways using duplicates eg. badframes.dll . Avisynth is free but you can even do this sort of thing in NLE's , but motion interploated frames usually require FX type software like after effects and various plugins

[jagabo]

It's possible to replace individual frames (or groups of frames) in VirtualDub using the Cut/Copy/Paste functions. But working with interlaced video makes it a little bit more difficult.

[sanlyn]

Sorry, I got called away. gil900, repairing bad frames will likely require that you learn to use Avisynth and work in different colorspaces other than RGB. You wouldn't have to be an Avisynth "engineer" (gosh, wish I was one!), but if you could upload a suitable video sample so others could take a look, they could see what would be involved. If you've used VirtualDub you apparently know what a "video" looks like, but caution: members here can help with specific tools, but we'd need a short sample of your unprocessed MPEG. Getting that far isn't difficult, if you'd like a quickie list of steps to take.

As it stands, you did OK loading your video into VirtualDub. But it has been converted to a colorspace that can't be used with most Avisynth tools. It has also been filtered with an autogain/autowhite plugin that would change luma and chroma values from frame to frame, along with other filtered changes that would make repair more difficult. It's deinterlaced rather than its original interlaced form, which means the original frames (including broken ones) have been altered and resized. For frame repair, most Avisynth techs would use the original format and field placements. Working with the unprocessed original in its original colorspace would help ensure accuracy and predictable results.

You can use something like free DGIndex to extract about 10 seconds of your original video without altering it. At this point, you don't even have to have the audio track.

[gil900]

I have the original files that the DVD recording.
But I do not think those original files are not compressed.

I think that the DVD also did compression ..
But that's all I have.

If it's okay then I'll upload sample video tomorrow (by time in my area)
Now I must sleep ..
So I will send a sample later.

[sanlyn]

Getting a little late here, also. I'll try later to find some simple instructions for DGIndex. There are other ways to do this, anyone can make suggestions.

Yes, your DVD video is compressed MPEG. That's not the best way to transfer tape to DVD, but it's workable. From this point, though, the idea is to avoid as much recompression and overprocessing as possible. Getting MPEG into AVI isn't difficult, using the right tools.

[gil900]

Getting a little late here, also. I'll try later to find some simple instructions for DGIndex. There are other ways to do this, anyone can make suggestions.

Yes, your DVD video is compressed MPEG. That's not the best way to transfer tape to DVD, but it's workable. From this point, though, the idea is to avoid as much recompression and overprocessing as possible. Getting MPEG into AVI isn't difficult, using the right tools.

OK. hare is a small part of the original file:
https://www.dropbox.com/s/k0dfe4ivsy7ns4k/Example1.m2v


The original file was *. VOB but somehow it has to *. m2v
But I do not think that the program touching video encoding or something ..
i removed the audio.
I hope it's okay.

Video has the two phenomena described in this topic.

[sanlyn]

Thanks for uploading. I don't know what program was used to create the .m2v (they are video-only files), but it's only 24kb in size. That's not enough data for even a single frame of MPEG. I'll invite you to try using the very popular DIGIndex, which you'll find being used everywhere in this forum.

To cut a chunk of MPEG from your video, you need DGIndex.exe and DGDecode.dll. They are part of the free DGMPGDEC package. It is a small stand-alone application. It does not require an installer.

1. Click this link to download DGMPGDEC: http://hank315.nl/files/DGdecode/dgmpgdec158.zip and place it in a folder on your hard drive. You can name the folder whatever you want, but "DGINDEX" is good.

2. Inside the new folder, unzip the downloaded .zip file. You will see several files (including two html user manuals). Two of the files are DGIndex.exe and DGDecode.dll . These two files must stay together in that folder.

3. You can make a desktop shortcut to DGIndex.exe by right-clicking on the file name, then select "Create Shortcut" in the popup menu. When the shortcut is created, copy and paste it to your desktop. (You don't need a shortcut to DGDecode.dll. When DGIndex runs, it will find the .dll in this folder).

4. Double-click your DGIndex shortcut, or double-click on the DGIndex.exe file. Either method will start the DGIndex program.

5. (below) The DGIndex program window. This image is from an XP computer. To open a video file, click on "File" in the top menu, then click "Open . . ."

[Attachment 13287 - Click to enlarge]

6. When you click "Open", you'll see the 'Open" dialog. It looks like any other Window dialog that is used to locate and open a file in Windows. Locate your MPG or VOB video file, and click on it to select it, then click "Open" or "OK", whichever applies.

7. When you open a video file in DGIndex, you will see the "File List" window (below).

[Attachment 13288 - Click to enlarge]

8. You will see your MPG or VOB listed. At this time DO NOT click the "ADD" button. This would simply add more files to the list. In this case, we want only one video file. So click the "OK" button to view the video.

9. Preview and Edit window (below).

[Attachment 13289 - Click to enlarge]

There are several controls on the window, at the bottom. Along the bottom you'll see the sliding Scroll control. Use this to move quickly back and forth in the video.

The "<" button moves the video backward. The ">" button moves it forward. These two keys move the video in groups-of-pictures (GOP). A group might contain 9 to 15 frames or so, depending on how your video is encoded. When you make cuts in MPEG or VOB, you want to cut the at the start of frame Groups. The > and < buttons always stop at frame groups.

The "[" key is used to mark the START position of your cut. The "]" key marks the END position of your desired cut. In the video sample shown here, look at the area marked in BLUE along the bottom of the video. This indicates the range of frames that has been selected using "[" and "]". The range you select should include some of the damaged frames of your sample.

10. Scroll to a part of the video that you want to edit and submit. Include a few frames groups of video that are before and after the frames in question. About 15 to 20 frame groups should be enough (this would be just several seconds of video). The edited cut would be 5 to 12 MB in size, which is acceptable.

11. When your selection is made, click on "File" in the top menu, then click "Save project and Demux video".

12. You will see the standard Windows "Save" dialog box (below), a standard Windows dialog that looks similar to the standard "Open" window.

[Attachment 13290 - Click to enlarge]

The save location will default to the folder where your original video is located. You can change the location if you want. The saved video will have a default name that is the same as your original video. You can change the name, too. When you have set the location and file name, click "Save".

13. When DGIndex is finished, you will see the Information panel (below). Because you are saving only a few seconds of video, this happens very quickly. Near the bottom of the info panel you'll see "FINISHED".

[Attachment 13291 - Click to enlarge]

Look into the folder where you saved your video edit. You will see two new files. One will be a .d2v file (basically an "index" to your video that other programs can use). The other file will be the .m2v -- the video-only portion of your sample file -- actually, it's an MPG in disguise.

The .m2v is the video file you want to upload. Don't upload .d2v's; they are small project files with information about the video that is specific to your computer, and can't be used elsewhere.

To upload a file to this forum, do this:

1. Go the the "Quick Reply" window at the bottom of the thread. In the Reply window you can type your message or other text.

2. Just below the Reply window you'll see a large button labeled "Upload files/Manage attachments". Click on that button. (DO NOT CLICK "My files")

The upload window will usually appear to the left of the forum window. Near the top are input panels to locate and find the file you want to upload. When you find and enter that file's location, click the "Upload" button. DON'T CLOSE this window until the upload completes. You'll see a message that the upload has finished. Close that window, then go back to the Reply window and click Post Quick Reply or Submit Reply, or whatever button you see that tells you to post your reply.

The video will appear as an attachment in your post.

[gil900]

Thanks for uploading. I don't know what program was used to create the .m2v (they are video-only files), but it's only 24kb in size. That's not enough data for even a single frame of MPEG. I'll invite you to try using the very popular DIGIndex, which you'll find being used everywhere in this forum.

...................................

no..
Guide you wrote very invested and I am very grateful for the help.
But you didn't need to write all this.
I know how to use the computer and how to make shortcuts.

And that's because I do use the software DIGIndex.
And this file does not 24KB size. The file is 94MB.

you didn't downloaded the file right.
Try to download this link:
https://dl.dropbox.com/s/k0dfe4ivsy7ns4k/Example1.m2v?dl=1
(This is a direct link and not a link to the download page)
If it didn't downloaded good then try downloading the link with a download manager.

Or try the link I created with Web proxy:
https://pierce-traffic.cyberghostvpn.com/go/browse.php?u=https%3A%2F%2Fdl.dropbox.com%...&b=7&f=norefer

[sanlyn]

Wow! Well, my error. I don't know what I did the first time, but don't ever say that people here don't make mistakes . I hate Internet Explorer anyway, I'm going back to Firefox on my regular PC.. I'll leave the DGIndex post up there, others might find it useful.

Second try worked, and thanks for the nice video. Yes, some familiar playback problems indeed. The interlaced video is in YV12, average bitrate in the mid 4000's -- a bit low for full-frame MPEG, but the soft image might not need higher bitrate. More on that later. There's some jitter from unsteady tape movement. Some of that can be fixed in Avisynth. Some people try to correct camera movement as well; I wouldn't worry about it, as trying to fix it often gives strange results.

Just a quick note for now, I note in an earlier post you used Border Control to fix the border noise. There are other ways to do it, and to center the image by working only with the borders without masking some of the "real" video at the top. In Avisynth I usually do something like this:

vidpath="E:\forum\whatfilters\"

MPEG2Source(vidpath+"Example1.d2v")
Crop(8,0,-16,-16).AddBorders(12,8,12,8)
ConvertToRGB32(matrix="Rec601",interlaced=true)

The term "vidpath" is a variable that I invented to contain the character string for the drive and folder path to the video on my PC. You could do the same thing, but change the characters in "vidpath" to match your own system. I do this because my working videos are on external hard drives, kept apart from the operating system. When I change drives, all I have to modify are the drive letters.

Or you can write the path in the usual way:

MPEG2Source("E:\forum\whatfilters\Example1.d2v")
Crop(8,0,-16,-16).AddBorders(12,8,12,8)
ConvertToRGB32(matrix="Rec601",interlaced=true)

Because it's likely that you'll have to do some color or other work in VirtualDub, the last line in the code shows the preferred way of converting YV12 to RGB. More on that later. I'd say that some frame repair and other work has to be done in YV12 first, before moving to RGB. Going back and forth between color spaces is not a good idea. Do what you must in the original YV12, move to RGB later.


[Attachment 13295 - Click to enlarge]

Will return in a while. I'm being interrupted again at home.

Again, thanks for the upload.

[gil900]

Ok thanks for the help.
I'm waiting for the solution of these phenomena ..

And I do not know how to use AviSynth 2.5 software.
for That I i need explain how to use like as you did with DGIndex.

I do not know how to use the code you gave me ..
Is there a way to do the same thing (fix the border noise) in the same way with VirtualDub?
I prefer to do most things you would recommend in a single step And not convert to a new format (with filters) and then work on what was coded before.

I saw a way to load scripts in VirtualDub ..
I thought this way I am going to slove the problem.

EDIT v1.2:

In any case if it is impossible do it in VirtualDub, or recommended to do it with other software then I would love to learn to use another software (AviSynth) and fix with your way if it is the Right thing to do.

In this case feel free to fix those problems in your way and tell me how to do it step by step.

And the other problems I will fix in VirtualDub.

[sanlyn]

Hm. These are notes I used earlier for other users, they might be helpful here:

And I do not know how to use AviSynth 2.5 software.
for That I i need explain how to use like as you did with DGIndex.

Installing Avisynth is no big hassle. It's a small utility that doesn't clutter your PC with toys. When it's installed it creates a few internal links so that Windows and programs like VirtualDub will know where Avisynth is located. It also has an easy uninstaller that removes those links.

I found a website about installing Avisynth, with pictures. This site is out of date a little, so ignore their download links. For now, Avisynth 2.5.8 is all you'd need. Download Avisynth_258.exe from here: http://sourceforge.net/projects/avisynth2/files/AviSynth%202.5/AviSynth%202.5.8/. This is the official 32-bit release, which works in 64-bit Windows. There are 64-bit versions coming, but not ready yet. Most plugins you would' use are 32-bit anyway. The installer instructions are at: http://www.m2ts.org/util/avisynth .

I do not know how to use the code you gave me ..
Is there a way to do the same thing (fix the border noise) in the same way with VirtualDub?

jagabo and poisondeathray mentioned this, but in VirtualDub it's more difficult and limited. The script was a simple example. Fixing those borders would come later. The AVS script fixed borders and centered the image with no resizing, which should be avoided with VHS.

I prefer to do most things you would recommend in a single step And not convert to a new format (with filters) and then work on what was coded before.

Actually, when you load your VOB into VirtualDub you did convert -- VirtualDub converted the colorspace from YV12 to RGB to run its filters and it expanded the video's image organization. It removed key frame markers (which you don't need in AVI) and expanded compressed groups-of-frames into full-scale images. This subject is complicated, but not every frame in an MPEG will contain a complete image. It does look that way when you view the video, but only because your playback software expands those partial images. This is why, if you decompress a 4-GB MPEG, you end up with a 25-GB AVI. Changes you make in your video have to be re-encoded to MPEG. That, after all, is what most videos "look like" before they're converted to MPEG, BluRay, DivX, or whatever.

Repairing damaged video usually takes multiple steps, if only because breaking up complex processing into smaller steps is less of a drag on your system. If you want to change one of the intermediate plugins in your one-step job, you'd have to run the whole thing again.

In any case I would love to learn to use another software if it is recommended (AviSynth)

So feel free to fix it your way and tell me how to do it step by step.

Well, it's good to see you understand why you'd need something like Avisynth for repairing video. But don't get the idea this can be done in a weekend. Members here can help a lot, and there are so many examples in the forum it would take years to see them all. Avisynth is not that difficult if you know how to use a PC, and knowing something about VirtualDub does help. But you'll need patience. Don't get into a passionate rush: some of this stuff is ridiculously easy, but some of it can tax your nerves. You do have something of a problem video here. But that's the only way to learn about restoration. You won't learn much from working with "perfect" videos.

I took a quick lunch break just now, but I'll keep an eye on this thread. I've learned a few things about Avisynth and VDub, but other members here are more expert than I when it comes to some aspects of repairing bad frames. I've run into some bad ones myself. If I could learn to fix them, almost anyone can. MvTools is likely to be the major plugin. I'm still learning that plugin, and I'm just as anxious to see how other members handle some of the aspects I haven't learned yet.

When you get Avisynth installed, remember that it's not a utility you run directly. It's run with scripts composed in Notepad, and usually loaded and viewed in VirtualDub. There are other viewers, but I find many of them rather buggy and difficult to learn.

To run the sample script I posted: copy the text of the script into a text file in Notepad. Don't forget to change the path to the .d2v file you created for Example 1. Note that the script opens the DGIndex .d2v file, not the .m2v. When you save the script, look down at the bottom of the "Save" window. You'll see a small window for "Save as type...", which asks for the type of file you want to save. The default is "Text documents" (with a ".txt" file ending). Don't save it as .txt, even if it technically is just plain text. Choose "All files" instead. Choose a location, then give the file a name that ends in ".avs". Then save.

Don't save an AVS scrip-t as ".txt":


[Attachment 13301 - Click to enlarge]

Save it as ".avs":


[Attachment 13302 - Click to enlarge]

Next, start VirtualDub. Click on "File", then click "Open video file...". In the dialog window, navigate to the the location of your .avs script. When you get into that folder, .avs file will show up because VirtualDub has learned to recognize ".avs" as a video file (if you have Avisynth installed as shown earlier). Select your avs script and let VirtualDub open it. It may take a few seconds to load. At that point your video will be in an RGB colorspace. If you save it as-is, it will be saved as uncompressed AVI in the wrong colorspace. I'll show you how to save it without making those changes.

Most people keep saved AVI work in losslessly compressed files. Of several lossless compressors available, I'd pick Lagarith. "Lossless" means just that: unlike Divx, MPEG, JPG, etc., the compression is not that tight because no data is discarded or downgraded to get a smaller file size. Lagarith is popular and fast. It doesn't have multi-version problems like huffyuv. You can get Lagarith here: http://lags.leetcode.net/codec.html. It's free. Download the "Lagarith Installer (v.1.3.27)". Save it to your hard drive. Go the to the save location and double-click the installer. Very small installer, very fast.

After Lagarith is installed, it's available for use in VirtualDub after VirtualDub is restarted. If you want to save an AVI version of your sample1.d2v as original YV12 video with no colorspace changes, do this:
1. In VirtualDub's top menu, click "Video".
2. In the expanded menu click "Color Depth..."
3. In the list of colorspace choices, on the right, click the box for "4:2:0 planar YCbCr (YV12)".
4. Click "OK", then click "Video" again.
5. In the expanded menu, click "Compression..."
6. In the list of compressors, click on "Lagarith lossless codec"
7. After you click on Lagarith you'll see a small window to the right, and a "Configure" button.
8. Click the "Configure" button.
9. In the configuration window, you'll see a few small check boxes on the left. Leave them unchecked.
10. In the "mode" window on the right, choose "YV12". That is your video's original colorspace.


[Attachment 13303 - Click to enlarge]

11. Click "OK", then click "OK" again to close that window.
12. Again, in VirtualDub click "Video", then click "fast recompress".
13. Finally, click "File", then "Save As AVI...". Give it a location and then a name with ".avi" on the end. This will save your video as AVI, unmodified from its original colorspace except for expansion into fully complete images for processing.

After you save this way a few dozen times, it goes faster.

It will take about 45 seconds to save your original video. It began as example1.m2v at 90.3 MB. It will be saved in Lagarith lossless compression as a new AVI at about 649-MB. That size difference gives you some idea of how much an MPEG is actually compressed. That's why MPEG is called "lossy".

[sanlyn]

gil900, if you install Avisynth and want its documentation (and have about a month to read all of it!), click Start, go to your All Programs list, click on the "Avisynth 2.5" program group, and click "Avisynth documentation". It starts at a basic level, then gets complicated. But interesting. . . .for short periods at a time.

[gil900]

Ok I will read all the post #20 soon ..
Meanwhile, a preliminary reading of the post, I realize you're not familiar enough to filter these phenomena.

But your explanation very much contributes to me.

I deleted the download to Example1.m2v because I thought everything was under control and you know how to fix it ..

So I will bring the download to Example1.m2v back.

Thanks for your help

[sanlyn]

With your VOB still on your hard drive, you can always make more samples for yourself with DGIndex and just save the .d2v file instead of the entire edited clip. Then all you need with Avisynth is to have the original VOB and any .d2v files in the same folder. The .d2v is used to read the original MPG or VOB. Move either of them away, and the MPEG2Source command in the scripts won't work any more.

I'll be able to help with some of the issues in your sample video. I just don't know all of it. I don't know anyone who does, I guess, and some members are more expert in some areas.

By the way, if you start reading Avisynth documentation, they give you a quick sample scripot called "test.avs" and they say it plays in Windows Media Player. It doesn't. Just open it with VirtualDub.

Back later. Have to repair a computer.

[jagabo]

The .d2v is used to read the original MPG or VOB. Move either of them away, and the MPEG2Source command in the scripts won't work any more.

Turn off Options -> Use Full Paths in DgIndex (before making your D2V files). Then you just have to keep the source and D2V file in the same folder.

Also, D2V files are plain text. You can edit them with NotePad. So even with full paths you can edit the D2V file to reflect the new path if you move the files.

[2Bdecided]

jagabo's post is my "you learn something every day" moment for me today!

[gil900]

[QUOTE=sanlyn;2177267]Hm. These are notes I used earlier for other users, they might be helpful here:

............
Next, start VirtualDub. Click on "File", then click "Open video file...". In the dialog window, navigate to the location of your .avs script. When you get into that folder, .avs file will show up because VirtualDub has learned to recognize ".avs" as a video file (if you have Avisynth installed as shown earlier). Select your avs script and let VirtualDub open it. It may take a few seconds to load. At that point your video will be in an RGB colorspace. If you save it as-is, it will be saved as uncompressed AVI in the wrong colorspace. I'll show you how to save it without making those changes.
................

I did not have the file d2v for Example1.m2v.
Then I understood myself that it is a project file Of DGIndex ..
So I opened the Example1.m2v in DGIndex and then I clicked on:
file -> save project -> save

Then I got the file Example1.d2v.

I hope that's okay.

I installed already the AviSynth 2.5 and I noticed i need the plugin MvTools.
"MvTools is likely to be the major plugin."

So I downloaded the plugin is here:
http://avisynth.org.ru/mvtools/mvtools.html # download

Then I put the files in the archive on
\AviSynth 2.5\plugins

I hope that's okay.

I put the files;
Example1.d2v
Example1.m2v

On Location:
D:\


Then I created a avs file As you've explained.
Avs file content is:

MPEG2Source("D:\Example1.d2v")
Crop(8,0,-16,-16).AddBorders(12,8,12,8)
ConvertToRGB32(matrix="Rec601",interlaced=true)

But when I load the avs file in VirtualDub, I get the following message:

http://ufu.co.il/files/9s3hskc20ne68jqrjeu8.png

What did I do wrong?

[sanlyn]

The function MPEG2Source is inside DGDecode.dll. It depends on where you placed the DGIndex download and where you're running it from. That could be highly variable, so do this:

Find where your main Avisynth 2.5 program folder is located (the folder where you installed Avisynth 2.5). Its "plugins" subfolder will be there. If you installed it in your main Windows partition (the C drive), it will be in one of the main Program folders.

Next, go to the folder where you downloaded and unzipped the DGIndex files. Make a copy of DGdecode.dll from that folder and place the copy in "Avisynth 2.5/plugins".

[gil900]

The function MPEG2Source is inside DGDecode.dll. It depends on where you placed the DGIndex download and where you're running it from. That could be highly variable, so do this:

Find where your main Avisynth 2.5 program folder is located (the folder where you installed Avisynth 2.5). Its "plugins" subfolder will be there. If you installed it in your main Windows partition (the C drive), it will be in one of the main Program folders.

Next, go to the folder where you downloaded and unzipped the DGIndex files. Make a copy of DGdecode.dll from that folder and place the copy in "Avisynth 2.5/plugins".

Ok.
Now it works

And I really see your fix about the border noise!

I see you wrote me that a guide is included with AviSynth 2.5.

I messed around abut programming (I know a same language to create scripts)
So I've done a little programming.. And this is about the same thing. (I guess)

Before this post, I never thought that there is a "video programming".

This thing is new to me.

Ok so I'm waiting for a script that will solve these problems (I said about them the beginning of post)
And of course, another script that will solve all other problems will be welcome.

Thank you

[jagabo]

You can replace individual frames or fields (or short segments) using motion compensated techniques. Your video is interlaced making it a little more difficult:

function ReplaceFramesMC(clip Source, int N, int X)
{
    # N is number of the 1st frame in Source that needs replacing. 
    # X is total number of frames to replace
    #e.g. RX(101, 5) would replace 101,102,103,104,105 , by using 100 and 106 as reference points for mflowfps interpolation
 
    start=Source.trim(N-1,-1) #one good frame before, used for interpolation reference point
    end=Source.trim(N+X,-1) #one good frame after, used for interpolation reference point
 
    start+end
    AssumeFPS(1) #temporarily FPS=1 to use mflowfps
  
    super = MSuper()
    backward_vec = MAnalyse(super, isb = true)
    forward_vec = MAnalyse(super, isb = false)
    MFlowFps(super, backward_vec, forward_vec, blend=false, num=X+1, den=1) #num=X+1
    AssumeFPS(FrameRate(Source)) #return back to normal source framerate for joining
    Trim(1, framecount-1) #trim ends, leaving replacement frames
  
    Source.trim(0,-N) ++ last ++ Source.trim(N+X+1,0)
}

# open the video
Mpeg2Source("Example1.d2v")

# deinterlace it
Bob()

# replace frame 154 (the top field of frame 77)
ReplaceFramesMC(154, 1)

# reinterlace the video
AssumeTFF()
SeparateFields()
SelectEvery(4,0,3)
Weave()

You need to download and install mvtools for this to work.

You can greatly reduce the chroma noise with:

 MergeChroma(last, McTemporalDenoise(settings="very high"))

You'll need to download and install McTemporalDenoise.AVSI (along with the aforementioned mvtools).

[gil900]

You can replace individual frames or fields (or short segments) using motion compensated techniques. Your video is interlaced making it a little more difficult:

function ReplaceFramesMC(clip Source, int N, int X)
{
    # N is number of the 1st frame in Source that needs replacing. 
    # X is total number of frames to replace
    #e.g. RX(101, 5) would replace 101,102,103,104,105 , by using 100 and 106 as reference points for mflowfps interpolation
 
    start=Source.trim(N-1,-1) #one good frame before, used for interpolation reference point
    end=Source.trim(N+X,-1) #one good frame after, used for interpolation reference point
 
    start+end
    AssumeFPS(1) #temporarily FPS=1 to use mflowfps
  
    super = MSuper()
    backward_vec = MAnalyse(super, isb = true)
    forward_vec = MAnalyse(super, isb = false)
    MFlowFps(super, backward_vec, forward_vec, blend=false, num=X+1, den=1) #num=X+1
    AssumeFPS(FrameRate(Source)) #return back to normal source framerate for joining
    Trim(1, framecount-1) #trim ends, leaving replacement frames
  
    Source.trim(0,-N) ++ last ++ Source.trim(N+X+1,0)
}

# open the video
Mpeg2Source("Example1.d2v")

# deinterlace it
Bob()

# replace frame 154 (the top field of frame 77)
ReplaceFramesMC(154, 1)

# reinterlace the video
AssumeTFF()
SeparateFields()
SelectEvery(4,0,3)
Weave()

You need to download and install mvtools for this to work.

You can greatly reduce the chroma noise with:

 MergeChroma(last, McTemporalDenoise(settings="very high"))

You'll need to download and install McTemporalDenoise.AVSI (along with the aforementioned mvtools).

Thank you.
But something did not go well.

I added your function to the avs file and put the file - mvtools.dll to the folder:
AviSynth 2.5\plugins

But when I load the script in VirtualDub, I get the following message:
http://ufu.co.il/files/k5xo9h9p3efeq8r1oxiu.png

This is the code of my avs file:

MPEG2Source("D:\Example1.d2v")
Crop(8,0,-16,-16).AddBorders(12,8,12,8)
ConvertToRGB32(matrix="Rec601",interlaced=true)

function ReplaceFramesMC(clip Source, int N, int X)
{
    # N is number of the 1st frame in Source that needs replacing. 
    # X is total number of frames to replace
    #e.g. RX(101, 5) would replace 101,102,103,104,105 , by using 100 and 106 as reference points for mflowfps interpolation
 
    start=Source.trim(N-1,-1) #one good frame before, used for interpolation reference point
    end=Source.trim(N+X,-1) #one good frame after, used for interpolation reference point
 
    start+end
    AssumeFPS(1) #temporarily FPS=1 to use mflowfps
  
    super = MSuper()
    backward_vec = MAnalyse(super, isb = true)
    forward_vec = MAnalyse(super, isb = false)
    MFlowFps(super, backward_vec, forward_vec, blend=false, num=X+1, den=1) #num=X+1
    AssumeFPS(FrameRate(Source)) #return back to normal source framerate for joining
    Trim(1, framecount-1) #trim ends, leaving replacement frames
  
    Source.trim(0,-N) ++ last ++ Source.trim(N+X+1,0)
}

# open the video
Mpeg2Source("Example1.d2v")

# deinterlace it
Bob()

# replace frame 154 (the top field of frame 77)
ReplaceFramesMC(154, 1)

# reinterlace the video
AssumeTFF()
SeparateFields()
SelectEvery(4,0,3)
Weave()