DVD Recorder and interlacing

[Grumpycat]

Hi,

I am getting a new Sky+ box and want to archive my existing recordings before swapping boxes. I have a Toshiba HDD & DVD Video Recorder (RD-XS25) which I plan on moving the recordings to. I then plan to create DVD's which I will then rip using handbrake to create MP4s.

Now I have already done a few transfers but I am puzzled why some of the DVD recordings show strong interlacing artifacts when played back on a PC but others don't. Even episodes of the same series differ when no settings are being changed.

I'm pretty green when it comes to this sort of thing, so is there something basic I am doing wrong?

Thanks

[usually_quiet]

Hi,

I am getting a new Sky+ box and want to archive my existing recordings before swapping boxes. I have a Toshiba HDD & DVD Video Recorder (RD-XS25) which I plan on moving the recordings to. I then plan to create DVD's which I will then rip using handbrake to create MP4s.

Now I have already done a few transfers but I am puzzled why some of the DVD recordings show strong interlacing artifacts when played back on a PC but others don't. Even episodes of the same series differ when no settings are being changed.

I'm pretty green when it comes to this sort of thing, so is there something basic I am doing wrong?

Thanks

Interlacing is more obvious when there is movement.

When playing interlaced video, use a software player that de-interlaces automatically, like MPC-HC, or MPC-BE, or which can be configured to do it, like VLC (automatic deinterlacing has to be turned on in "Preferences" to be permanent). If you are using Windows Media Player, it doesn't seem to deinterlace uniformly.

[hello_hello]

I'm guessing but maybe sometimes the interlaced flags are missing so the player doesn't know de-interlacing is required. Depending on the player you're using you may be able to force de-interlacing and if the problem goes away that's likely to be the cause. You should be able to force de-interlacing when re-encoding.

I don't know anything about Sky or how you transfer the recordings but are they all standard definition? If some are HD being resized down to DVD resolution that might cause it if they're interlaced as you can't just resize interlaced video. That's one of the primary reasons for de-interlacing when re-encoding.

Maybe attach a small sample to a post. You can open a DVD with DVDShrink, use it's re-author function to drag a title from the right pane to the left, then the edit function to edit the title down to a small section. When you use the backup function, the edited title should be a single vob file. You won't need much. Just a minute or so where there's obvious artefacts. DVDShrink has a preview to make it easy to work out what you're editing.

[Grumpycat]

I'm guessing but maybe sometimes the interlaced flags are missing so the player doesn't know de-interlacing is required. Depending on the player you're using you may be able to force de-interlacing and if the problem goes away that's likely to be the cause. You should be able to force de-interlacing when re-encoding.

I don't know anything about Sky or how you transfer the recordings but are they all standard definition? If some are HD being resized down to DVD resolution that might cause it if they're interlaced as you can't just resize interlaced video. That's one of the primary reasons for de-interlacing when re-encoding.

Maybe attach a small sample to a post. You can open a DVD with DVDShrink, use it's re-author function to drag a title from the right pane to the left, then the edit function to edit the title down to a small section. When you use the backup function, the edited title should be a single vob file. You won't need much. Just a minute or so where there's obvious artefacts. DVDShrink has a preview to make it easy to work out what you're editing.

Thanks, here's some examples. (I hope they are suitable).

They are the taken from the intro of Episode 1 and 2 of a programme, both of which were burnt to the same DVD.

Episode 1 is fine, episode 2 shows interlacing artifacts. I can't remeber if they were in HD or not, however other HD recordings don't exhibit the same issues.

I also plan on changing the widescreen flag on the files, as I have an issue there too.

[orsetto]

This "interlacing" artifact seems common with the DVD recorder workflow. Most of these units munge their recordings in some subtle way that makes ripping/re-encoding a bit of a problem. The DVDs they record from a decoder box or VHS play well enough, but once ripped to MPG files and converted by utilities like Handbrake to another format, the converted videos may display added combing or interlacing issues that are not visible on the original DVD. As usually_quiet mentioned, this is especially noticeable in scenes with movement. It doesn't happen with every recorder-based DVD conversion, and I'm sure there are tricks for avoiding it when it does, but I haven't yet found a definitive solution (other than using CloneDVDMobile to make m4v files directly from these DVDs, but CloneDVDMobile has other issues).

[usually_quiet]

Thanks, here's some examples. (I hope they are suitable).

They are the taken from the intro of Episode 1 and 2 of a programme, both of which were burnt to the same DVD.

Episode 1 is fine, episode 2 shows interlacing artifacts. I can't remeber if they were in HD or not, however other HD recordings don't exhibit the same issues.

I also plan on changing the widescreen flag on the files, as I have an issue there too.

I downloaded the samples. Both are flagged interlaced, according to MediaInfo. If I play the files in VLC with interlacing turned off, both samples do show interlacing artifacts, but they are much more pronounced in the 12.94 MB sample. The interlacing artifacts disappear in both samples if I turn on deinterlacing.

As far as the widescreen flag, often times widescreen flags aren't set by a DVD recorder, especially when using an analog input for recordings.

[Grumpycat]

Thanks, here's some examples. (I hope they are suitable).

They are the taken from the intro of Episode 1 and 2 of a programme, both of which were burnt to the same DVD.

Episode 1 is fine, episode 2 shows interlacing artifacts. I can't remeber if they were in HD or not, however other HD recordings don't exhibit the same issues.

I also plan on changing the widescreen flag on the files, as I have an issue there too.

I downloaded the samples. Both are flagged interlaced, according to MediaInfo. If I play the files in VLC with interlacing turned off, both samples do show interlacing artifacts, but they are much more pronounced in the 12.94 MB sample. The interlacing artifacts disappear in both samples if I turn on deinterlacing.

As far as the widescreen flag, often times widescreen flags aren't set by a DVD recorder, especially when using an analog input for recordings.

Thanks for taking a look

What I don't understand is that they have both been produced in the same manner. That is the playback from the satellite set top box hard drive is recorded to the DVD recorder's hard drive. Then they are dubbed to a DVD by the recorder.

Obviously when encoding to .mp4 in Handbrake, I can add a decomb filter, but I'd prefer if all my recordings were like the first example which looks acceptable to me.

[LMotlow]

The first VOB is progressive video encoded as interlaced. The interlaced fields are duplicates.
The second VOB is progressive video encoded as interlaced, but the duplicate "interlaced" fields are out of sync.

[Grumpycat]

The first VOB is progressive video encoded as interlaced. The interlaced fields are duplicates.
The second VOB is progressive video encoded as interlaced, but the duplicate "interlaced" fields are out of sync.

Okay, thanks.

Is it possible to fix the issue?

[hello_hello]

Simple field matching seems to do the trick, then you'd encode it as progressive. I don't think you can do it with Handbrake, so you'll probably need to use an Avisynth based GUI for re-encoding. If you've never used Avisynth this probably won't mean much to you, but after adding field matching to the script

TFM()

your second example goes from this:



To this:


You shouldn't need to de-interlace. Just match the fields up.
That's the sample displaying in MeGUI's preview. I didn't worry about aspect ratio issues so it's not resized correctly.

Although if I did resize I'd probably crop a few extra pixels from the sides and resize to exactly 16:9 dimensions for encoding. In this case 14 pixels left and 14 right and 4 pixels top and bottom to remove the black, then resized to 960x540 (or any 16:9 dimensions).



When it comes to re-encoding time try MeGUI. Use the file/open menu to open the first vob file in the set and follow the prompts. MeGUI has a bit of a learning curve but if you need help creating a script, adding field matching to it and then encoding, I think you'll find it here easily enough.
You might also need to use DVDShrink to prepare the DVD for encoding (the video needs to be contained within a single set of vob files) but that's not hard.

[jagabo]

Composite or s-video signals are always interlaced in the sense that fields are transmitted one at a time (never entire frames). Both videos are interlaced in the sense that the MPEG encoder treated them as interlaced video and encoded them top field first. And the actual source is progressive (probably film) for both. In the first video pairs of fields from the same film frame are paired together so the frames look progressive. In the second video pairs of fields from different frames were paired together so the frames look interlaced whenever there is motion. Ie, the first video was broadcast top field first and capture top field first. The second video was broadcast bottom field first (for some unknown reason) and captured top field first. Both will play properly when using a player that deinterlaces for display. The second video can be "fixed" by re-pairing the fields properly. As hello_hello indicated, you can use TFM() in AvisSynth. Or, if you know the entire video is this way you can use SeparateFields().Trim(1,0),Weave().

[Grumpycat]

Thanks guys.

The example hello_hello has posted looks very promising. I've never used Avisynth before, so it looks like I have got some learning to do!

Would the telecine filter in Handbrake achieve the same thing?

[LMotlow]

Thanx to jagabo for another simple but brilliant explanation.

I did this 3 different ways. But TFM() was simplest. First I made a .d2v project file with DGIndex, then used this script (adjust the path statement for the video's location):

MPEG2Source("drive:\path\to\video\VTS_01_1.d2v") 
TFM()

The re-encoded result is attached. I used a slightly higher bitrate to re-encode and remixed the audio later. Sorry, my h264 encoders aren't on the PC I'm using at the moment. Encoded as progressive. Only question now is, what the does the rest of the broadcast look like?

Would the telecine filter in Handbrake achieve the same thing?

No way. This video isn't telecined.

[Grumpycat]

Composite or s-video signals are always interlaced in the sense that fields are transmitted one at a time (never entire frames). Both videos are interlaced in the sense that the MPEG encoder treated them as interlaced video and encoded them top field first. And the actual source is progressive (probably film) for both. In the first video pairs of fields from the same film frame are paired together so the frames look progressive. In the second video pairs of fields from different frames were paired together so the frames look interlaced whenever there is motion. Ie, the first video was broadcast top field first and capture top field first. The second video was broadcast bottom field first (for some unknown reason) and captured top field first. Both will play properly when using a player that deinterlaces for display. The second video can be "fixed" by re-pairing the fields properly. As hello_hello indicated, you can use TFM() in AvisSynth. Or, if you know the entire video is this way you can use SeparateFields().Trim(1,0),Weave().

Okay, so the fact that some programmes are okay, but others not due to the actual broadcast, something which I have no control over?

[smrpix]

Okay, so the fact that some programmes are okay, but others not due to the actual broadcast, something which I have no control over?

Frequently, yes.

[hello_hello]

HandBrake's decomb filter does IVTC (I'm fairly sure), and then it's supposed to de-interlace any residual combing, so logically it's got to do field matching. You could try enabling the decomb filter while telling Handbrake to output a constant frame rate of 25fps. I don't know whether that'll get Handbrake to do field matching or if it'll just try to de-interlace. It's log file might be informative enough to tell you. If it's the latter the de-interlacing will probably cause blurring so that's not ideal. You can only try though. I don't use Handbrake myself so I'm not sure how it's filters work exactly. If not, Avisynth is the better method anyway as it gives you more control, once you've learned a bit about using it.

[Grumpycat]

HandBrake's decomb filter does IVTC (I'm fairly sure), and then it's supposed to de-interlace any residual combing, so logically it's got to do field matching. You could try enabling the decomb filter while telling Handbrake to output a constant frame rate of 25fps. I don't know whether that'll get Handbrake to do field matching or if it'll just try to de-interlace. If it's the latter the de-interlacing will probably cause blurring so that's not ideal. You can only try though. I don't use Handbrake myself so I'm not sure how it's filters work exactly. If not, Avisynth is the better method anyway as it gives you more control, once you've learned a bit about using it.

Thanks, I'll give it a try. I have also downloaded Avisynth, but it looks a bit daunting! I'll have to do some reading up before using it.

Thanks again for your help.

[manono]

HandBrake's decomb filter does IVTC (I'm fairly sure)....

Based on their usage handbook, it's a conditional deinterlacer (deinterlaces when it spots interlacing). It doesn't field-match (or IVTC) and uses either Yadif or a blur (as near as I can tell) when there's interlacing. You don't want that.

https://trac.handbrake.fr/wiki/Decomb

The Detelecine filter may or may not work. I don't know (and don't care). Although it doesn't specifically mention using it for PAL sources, what they say might mean it can be made to just field-match. AviSynth is the way to go here (and everywhere, as far as I'm concerned), and Handbrake doesn't use AviSynth.

https://trac.handbrake.fr/wiki/Telecine

There are encoding programs that use AviSynth that don't require the user to know much, if any, AviSynth. At most you might have to edit a script to add in 'TFM()'.

[LMotlow]

I have also downloaded Avisynth, but it looks a bit daunting!

Yep, "daunting" is a pretty good word for it, first time around. When you run the installer it makes a program group in your program listings named "Avisynth 2.5" (that's the name, even if you install 2.6. Don't worry about it). Open that program group and you'll see a subgroup with "Avisynth Documentation" in the list. Click that, and you'll see the online help that begins with "Getting Started". Has some quickie sample scripts to get you used to things.

First thing the Getting Started page tells you is to create a test script. Most people use Notepad to type the script. Save that in Notepad by clicking "File..." then "Save". Then at the bottom of the Notepad window give it a name (Test.avs) and click "Save as type..."and choose "All types". If you don't specify "All types", Notepad saves it with a .txt file extension, which won't work.

Then they tell you to open that file in Windows Media Player. That's the first thing that really threw me. Most versions of WMP won't do a thing with it. It's enough to turn you off for good. Folks the world over open that file with VirtualDub, which is your friend when it come to Avisynth. Open it in VirtualDub like any other media file.

Avisynth Help has a table of contents in the left-hand margin of the Help page.

The script I used that had MPEG2Source and TFM() in it uses two external utilities. Those are not Avisynth commands or functions. MPEG2Source is a function in an Avisynth plugin named DGIndex.dll. You get that .dll by downloading the free DGMPGDec utility. Download that to a new folder (call it DGIndex to make it easy. Keep track of it. You'll use DGindex a lot). Unzip the download and you'll find a bunch of files. One of them is DGIndex.dll. Copy DGIndex.dll to the Avisynth plugins folder, which is located in the Avisynth 2.5 program folder.

TFM is another external function in a plugin named TIVTC.dll. You get that plugin by downloading TIVTCv105.zip here: http://bengal.missouri.edu/%7Ekes25c/TIVTCv105.zip. Create a folder named TIVTC and download the .zip to that folder. When you unzip it you'll see a bunch of files, one of which is named tivtc.dll. Copy that tivtc.dll plugin to the Avisynth plugins folder. TIVTC doesn't just remove plain telecine, it also works on other stuff.

You'll be using DGindex and TITVC until the cows come home. Very useful stuff.

To run TFM in a script you need to create a d2v index file, which is an index that lets Avisynth decode VOB, MPEG, and m2v files. Look inside that DGindex folder you created and you'll see a stand-alone program named DGIndex.exe. Double-click to run DGIndex.exe. You get a small window on your desktop. Click "File..." in the top menu bar and you'll get a blank window with the title "File list" at the top. Not to worry. On the right hand side of that window click "Add", then navigate to the location of your VOB file. Click on that VOB to select it, then click "Open". You'll be back at the File List window again and that VOB will be listed as an included file. Down at the right bottom of that window, Click "OK". You'll see that movie file open in the DGIndex main window. You're almost in business.

In the DGIndex window, click "File...", then click "Save project file and demux video". For a big VOB it'll take a few minutes to build that index. When DGIndex is finished, go to that VOB's original folder and you'll see some new files. One of those files will be an audio file with a long name ending in "AC3" or whatever audio format was in the movie. The other file is a ".d2v" project file. That's the decoding index for your VOB. Keep the VOB, the audio file, and the .d2v together. You'll use the audio file later.

You can see in the script I posted that Avisynth doesn't open the VOB itself. Instead, MPEG2Source opens the .d2v file. When you run that in VirtualDub, you'll see the movie in VirtualDub, frame by frame, fixed up and reorganized by TFM.

You can save that file in VirtualDub as a decoded YV12 AVI. I used Lagarith lossless compression to save it and feed it to an encoder. There are other lossless codecs. It's been so long since I used Handbrake, I don't even remember what I did with it. It's also possible to run that script in other encoders like MEGui, which can read Avisynth input and can remux the audio and use HCenc to encode to MPEG. You don't mention how you're encoding, so we'll have to work from there.

Best of luck.

[hello_hello]

HandBrake's decomb filter does IVTC (I'm fairly sure), and then it's supposed to de-interlace any residual combing, so logically it's got to do field matching. You could try enabling the decomb filter while telling Handbrake to output a constant frame rate of 25fps. I don't know whether that'll get Handbrake to do field matching or if it'll just try to de-interlace. If it's the latter the de-interlacing will probably cause blurring so that's not ideal. You can only try though. I don't use Handbrake myself so I'm not sure how it's filters work exactly. If not, Avisynth is the better method anyway as it gives you more control, once you've learned a bit about using it.

Thanks, I'll give it a try. I have also downloaded Avisynth, but it looks a bit daunting! I'll have to do some reading up before using it.

Thanks again for your help.

Edit: I see LMotlow has posted alternative instructions for doing it all the same way, only differently (no GUI vs GUI). Either way.......
The only thing I'd add is he mentions Avisynth installs in a folder called Avisynth 2.5 even if you install version 2.6. That's no longer true. The very latest 2.6 release candidate (RC3, I think) now installs in a folder simply called Avisynth. It's only been out a few weeks. If you've installed an earlier version, uninstall it before installing the latest 2.6, let it delete the registry entries/plugin folder location and delete the Avisynth 2.5 folder yourself when it's finished if you need to before installing the new version.

According to manono, Handbrake isn't the way to go, so try MeGUI. There's quite a few Avisynth based encoder GUIs. The object of the exercise is generally for the GUI to let you crop and resize and/or apply de-interlacing etc in the usual manner, and the GUI creates the script required for encoding. Aside from MeGUI (which is the one I use), StaxRip, RipBot & Hybrid are other Avisynth based GUI's that come to mind. If you want to modify the scripts manually, you want a GUI that makes it easy for you. I think all the GUI's I mentioned will let you do that, I just prefer MeGUI.

Open the first vob file in the set using the File/Open menu and MeGUI should offer to index it and extract the audio. Add the indexing job to the queue and run it. When it's done the script creator will open with a preview for you to crop, resize and apply filtering. To apply TFM() you could enable de-interlacing, select "film" as the source type and "IVTC" as the de-interlacing method. When you switch to the Script tab you'll see a couple of lines like this:

LoadPlugin("C:\Program Files\MeGUI\tools\avisynth_plugin\TIVTC.dll")
tfm(order=-1).tdecimate()

That's MeGUI loading the required plugin and applying IVTC. The tfm() part does the field matching and the tdecimate part would then remove the duplicate frame (for IVTCing NTSC to 23.976fps). In your case, all you want is the field matching part, so you'd delete .tdecimate()

LoadPlugin("C:\Program Files\MeGUI\tools\avisynth_plugin\TIVTC.dll")
tfm(order=-1)

Make it the last thing you do before saving the script because MeGUI re-writes the script any time you change something in the script creator GUI so manual changes will be lost. You can use the preview button before saving though to see the result in the video preview.

Oh, and the video is being seen as 4:3. I assume it's 16:9 so when the script creator opens the first thing you should do is change the Input DAR to "ITU 16:9 PAL" so then when you apply cropping and resizing MeGUI's aspect ratio error calculations will be correct.
Also, and I wish it was the default because I'm so sick of typing instructions, if you want to resize the video "up" (ie 1024x576 or 960x540 etc) instead of resizing "down" (720x404 or 640x360 etc) you need to enable "upsizing" in the script creator's Avisynth profile configuration. It's monumentally retarded that resizing a PAL 16:9 720x576 DVD to 1024x576 square pixels for the correct aspect ratio is considered upscaling, and even more retarded that the option is disabled by default and hidden away in the script creator's profile configuration, but the MeGUI developer thinks it's a good idea for some reason. Of course if you use anamorphic encoding instead of resizing to square pixels none of that probably matters.

That should get you started. You don't need Avisynth installed because MeGUI uses it's own portable version but it's still a good idea as then you can open scripts and encode/preview them using other programs. I use MPC-HC for previewing scripts running on my TV before I encode them. If you've installed Avisynth, hopefully you've installed the latest version 2.6 release candidate. Version 2.5.8 is the last official stable version but it's outdated.

[newpball]

It's monumentally retarded that resizing a PAL 16:9 720x576 DVD to 1024x576 square pixels for the correct aspect ratio is considered upscaling,...

It really is not.

In simple terms, PAL has not enough bandwidth to render 1024 discrete values for each discrete line.
When you go from 720 (or 704) to 1024 you really do upscale!

By the way a horizontal resolution of 720 lines (or 704 which is more common) may be kind of stretching it already. But "stretching" things is common in the industry, makes it look better.

[LMotlow]

I'm not one for upscaling SD video. Always looks kinda goofy, if not soft. Anyway, I boo-boo'd and posted 4:3 earlier. You can't go BluRay with 720x576 progressive, but mkv can take advantage of high BluRay-style h264 bitrates for progressive SD 720x576. So, attached is a rework at 16:9 DAR.

[hello_hello]

It's monumentally retarded that resizing a PAL 16:9 720x576 DVD to 1024x576 square pixels for the correct aspect ratio is considered upscaling,...

In simple terms, PAL has not enough bandwidth to render 1024 discrete values for each discrete line.
When you go from 720 (or 704) to 1024 you really do upscale!

"It's not true, it's not true, newpball is wrong, he is wrong" (repeat 1000 times).

It's nothing to do with bandwidth. Open a PAL 16:9 DVD with pretty much any software player and count the number of pixels on your PC monitor that are being used to display it at the correct aspect ratio. If it's something like 1024x576, then how can encoding it at 1024x576 be upscaling? You could argue the storage aspect ratio is being upscaled, but not the display aspect ratio. You'd be the first to interrupt a thread and fill it with chicken little predictions if someone suggested encoding it while reducing the height because that would be downscaling. Or are you going to argue that as long as the width exceeds 720 it's not?

"It's not true, it's not true, newpball is wrong, he is wrong" (repeat 1000 times).

There's an obvious line between what's considered upscaling and what isn't when it comes to sources with square pixels, but when the source is anamorphic the position of that line is less clear and in my opinion it should be where the resolution of the height starts to increase for anamorphic 16:9.

"It's not true, it's not true, newpball is wrong, he is wrong" (repeat 1000 times).

You tell people to resize to 1024x576 because it retains the most detail. Sometimes I think you just like to contradict to cause arguments. Resizing that way should be considered resizing, not upscaling, and MeGUI should be able to distinguish between the two, in my opinion.

By the way a horizontal resolution of 720 lines (or 704 which is more common) may be kind of stretching it already. But "stretching" things is common in the industry, makes it look better.

That doesn't even make sense. Are you trying to say industry practice is to stretch a video beyond it's correct aspect ratio, or was that just generally meaningless?

I'm not one for upscaling SD video. Always looks kinda goofy, if not soft.

I don't understand the "soft" remark. I find, at least when using a sharp resizer such as spline36, a PAL DVD resized to 1024x576 and re-encoded that way tends to look a tad sharper than a 720x576 anamorphic encode. In fact most of the time, even resizing to 960x540 and running that fullscreen will tend to look a bit sharper than the original video.

Here's some comparison screenshots from some comparison encodes I made when first deciding whether to switch to square pixel encoding. Both are running in MPC-HC fullscreen at 1080p so they're both being upscaled the same way. The difference is the first is being upscaled from 720x576 while the second was resized to 1024x576 first.









The attached sample (a re-encode of Grumpycat's second sample) was resized to 1048x576 for easy comparison. Normally I'd crop and resize to exactly 16:9. I can't really pick the difference between it and the anamorphic encode when they're both running fullscreen.

"It's not true, it's not true, newpball is wrong, he is wrong" (repeat 1000 times).

[hello_hello]

Here's the samples from which the previous screenshots were taken. No other filtering was applied as I recall. Unless I applied the same filtering to both (ie same noise removal) but I doubt it as that wasn't the purpose of the excercise. I encoded these a fair while ago though. January 2014 according to the file dates.

[newpball]

I really don't want to get involved in this discussion so I give my two cents, take it for what it's worth:

- Upscaling SD makes a picture soft but at least the damage done is minimum.
- Sharpening an upscaled image ruins detail and gives halos.

Only sharpen a video if:

a) you downscale it.
b) the display/distance ratio is relatively small for the given resolution.

[hello_hello]

I really don't want to get involved in this discussion so I give my two cents, take it for what it's worth:

Giving your two cents worth is getting involved, although I'd tend not to argue with the value you're putting on it.

- Upscaling SD makes a picture soft but at least the damage done is minimum.

Therefore by your own definition resizing to 1024x576 can't be upscaled because the resizing didn't make it softer.

- Sharpening an upscaled image ruins detail and gives halos.

Only sharpen a video if:

a) you downscale it.
b) the display/distance ratio is relatively small for the given resolution.

Generalisation which don't always apply. Much of it depends on the quality of the sharpening. I can show you examples of sharpened video that wasn't resized and just looks a little sharper. But that'd be a waste of time because you'll once again go bum up and pretend they're not there if they don't suit your argument.

"It's not true, it's not true, newpball is wrong, he is wrong" (repeat 1000 times).

[Grumpycat]

HandBrake's decomb filter does IVTC (I'm fairly sure), and then it's supposed to de-interlace any residual combing, so logically it's got to do field matching. You could try enabling the decomb filter while telling Handbrake to output a constant frame rate of 25fps. I don't know whether that'll get Handbrake to do field matching or if it'll just try to de-interlace. If it's the latter the de-interlacing will probably cause blurring so that's not ideal. You can only try though. I don't use Handbrake myself so I'm not sure how it's filters work exactly. If not, Avisynth is the better method anyway as it gives you more control, once you've learned a bit about using it.

Thanks, I'll give it a try. I have also downloaded Avisynth, but it looks a bit daunting! I'll have to do some reading up before using it.

Thanks again for your help.

Edit: I see LMotlow has posted alternative instructions for doing it all the same way, only differently (no GUI vs GUI). Either way.......
The only thing I'd add is he mentions Avisynth installs in a folder called Avisynth 2.5 even if you install version 2.6. That's no longer true. The very latest 2.6 release candidate (RC3, I think) now installs in a folder simply called Avisynth. It's only been out a few weeks. If you've installed an earlier version, uninstall it before installing the latest 2.6, let it delete the registry entries/plugin folder location and delete the Avisynth 2.5 folder yourself when it's finished if you need to before installing the new version.

According to manono, Handbrake isn't the way to go, so try MeGUI. There's quite a few Avisynth based encoder GUIs. The object of the exercise is generally for the GUI to let you crop and resize and/or apply de-interlacing etc in the usual manner, and the GUI creates the script required for encoding. Aside from MeGUI (which is the one I use), StaxRip, RipBot & Hybrid are other Avisynth based GUI's that come to mind. If you want to modify the scripts manually, you want a GUI that makes it easy for you. I think all the GUI's I mentioned will let you do that, I just prefer MeGUI.

Open the first vob file in the set using the File/Open menu and MeGUI should offer to index it and extract the audio. Add the indexing job to the queue and run it. When it's done the script creator will open with a preview for you to crop, resize and apply filtering. To apply TFM() you could enable de-interlacing, select "film" as the source type and "IVTC" as the de-interlacing method. When you switch to the Script tab you'll see a couple of lines like this:

LoadPlugin("C:\Program Files\MeGUI\tools\avisynth_plugin\TIVTC.dll")
tfm(order=-1).tdecimate()

That's MeGUI loading the required plugin and applying IVTC. The tfm() part does the field matching and the tdecimate part would then remove the duplicate frame (for IVTCing NTSC to 23.976fps). In your case, all you want is the field matching part, so you'd delete .tdecimate()

LoadPlugin("C:\Program Files\MeGUI\tools\avisynth_plugin\TIVTC.dll")
tfm(order=-1)

Make it the last thing you do before saving the script because MeGUI re-writes the script any time you change something in the script creator GUI so manual changes will be lost. You can use the preview button before saving though to see the result in the video preview.

Oh, and the video is being seen as 4:3. I assume it's 16:9 so when the script creator opens the first thing you should do is change the Input DAR to "ITU 16:9 PAL" so then when you apply cropping and resizing MeGUI's aspect ratio error calculations will be correct.
Also, and I wish it was the default because I'm so sick of typing instructions, if you want to resize the video "up" (ie 1024x576 or 960x540 etc) instead of resizing "down" (720x404 or 640x360 etc) you need to enable "upsizing" in the script creator's Avisynth profile configuration. It's monumentally retarded that resizing a PAL 16:9 720x576 DVD to 1024x576 square pixels for the correct aspect ratio is considered upscaling, and even more retarded that the option is disabled by default and hidden away in the script creator's profile configuration, but the MeGUI developer thinks it's a good idea for some reason. Of course if you use anamorphic encoding instead of resizing to square pixels none of that probably matters.

That should get you started. You don't need Avisynth installed because MeGUI uses it's own portable version but it's still a good idea as then you can open scripts and encode/preview them using other programs. I use MPC-HC for previewing scripts running on my TV before I encode them. If you've installed Avisynth, hopefully you've installed the latest version 2.6 release candidate. Version 2.5.8 is the last official stable version but it's outdated.

As a newbie to these things, when I first looked at your post, I thought that there's no way I can do that. But carefully following your detailed instructions and with a bit of trial and error I have managed to replicate your earlier examples and 'fix' my files.

So a big thanks for all the advice on this thread.

[hello_hello]

You're welcome.