DVD Pal to DVD NTSC

[FulciLives]

Well the process is complete... the video and audio sink and look great..
but every time i try to burn it using ulead movie factory. the video looks glitchy... shakes a little...

and ideas?


wesman

Don't use MF2

When I have AC-3 audio files I usually use SpruceUP since it is simple and I've been using it for a while now. But both the TMPG DVD Author and the DVD Lab programes (both popular now) should work and have free trail periods when you download them before you need to buy/register them.

- John "FulciLives" Coleman

P.S.
I have used MF2 here and there but it doesn't like ALL files created with TMPGEnc. It is a very tempermental authoring program.

[DJRumpy]

SpruceUp is a nice basic program. I use it myself. If your video is 16:9, you'll need to change the IFO to 16:9. SpruceUp only sets 4:3 on your IFO's, regardless of their true aspect ratio. It takes about 30 seconds using IFO edit. Just be aware that when you change the DAR, the display in IFO Edit will not update until you save each IFO. Just let it overwrite the originals. You can leave the MENU DAR the same, changing only the movie to 16:9.

[FulciLives]

SpruceUp is a nice basic program. I use it myself. If your video is 16:9, you'll need to change the IFO to 16:9. SpruceUp only sets 4:3 on your IFO's, regardless of their true aspect ratio. It takes about 30 seconds using IFO edit. Just be aware that when you change the DAR, the display in IFO Edit will not update until you save each IFO. Just let it overwrite the originals. You can leave the MENU DAR the same, changing only the movie to 16:9.

I respect you very much my Jedi friend but I remember saying this in aother thread not too long ago and you said that you had no problems using 16x9 source material with SpruceUP and that I MUST be doing something wrong to have to change the IFO file etc.

Still luv ya :P

- John "FulciLives" Coleman

[fabriceh]

This thread is kinda long and I did not read it all, so I may state something already covered before, but...

All I have to do is rip the PAL DVD using DVD Shrink and compress it to a smaller image (to fit on DVD5) and burn it on a DVD and I can then watch it on my NSTC equipment. I turn on the Region 1 bit in DVD Shrink as one of my DVD players seems to check for that flag and refuses to play the movie if the flag is not set.

Total conversion time - <1 hour (if you can burn at 2 or 4x).

Fabrice

[sendohtheman]

i have reencoded the video and it took 2 days to do it. now, i downloaded the ac3 delay corrector but i don't know what to put in the start delay and end delay? when i mux the encoded video and the audio demuxed by dvd2avi, it seems like the audio is like 5 seconds early and i think it gets worse as time goes on. the ac3 file itself says there is a -80 ms delay but i tried combinations of that in ac3 delay corrector but still no sync. if anyone can help, i really appreciate it. thanks again!!!!

sendohtheman

[mgh]

Have been doing NTSC VCD to PAL VCD using TMPGEnc and nothing more for ages. Got couple of NTSC DVDs with ac3 recently. Used DVD2AVI to load video in TMPGEnc-used DVD PAL template-muxed the m2v with ac3 demuxed from the NTSC. No problem.
I capture from TV and convert the video to DVD PAL format and store. Recently-have been demuxing audio-converting to ac3 (Besweet) and remuxing before authoring to mini DVD.
Tried to convert a 5 minute PAL mpeg2 video with ac3 into NTSC because of above discussion. Used DVD2AVI to load video in TMPGEnc-used DVD NTSC template-muxed the m2v with ac3 demuxed from the PAL. No problem.
I do not load long videos for conversion-from the NTSC DVDs-I stripped individual chapters with Smartripper and converted individually.
KISS (Keep it simple stupid) works for me!!

[FulciLives]

i have reencoded the video and it took 2 days to do it. now, i downloaded the ac3 delay corrector but i don't know what to put in the start delay and end delay? when i mux the encoded video and the audio demuxed by dvd2avi, it seems like the audio is like 5 seconds early and i think it gets worse as time goes on. the ac3 file itself says there is a -80 ms delay but i tried combinations of that in ac3 delay corrector but still no sync. if anyone can help, i really appreciate it. thanks again!!!!

sendohtheman

How did you convert/re-encode the video? Did you slow it down to 23.976fps or use the AVS script method to get a 29.97fps NTSC result?

If you went from 25fps to 23.976fps then you need to use BeSweet to slow down the audio.

This has been covered earlier in the thread.

You do not have to use AC3Delay because you will be converting the audio with BeSweet and there is a function in BeSweet to account for the delay which in this case is -80ms.

Please note when I say you need to convert with BeSweet that you need to use the framrate PRE-SET of PAL (25fps) to NTSC (23.976fps).

However, if you used the AVS script method then you should have a 29.97fps NTSC video that will MATCH the original AC-3 so in THAT case you would just run AC3Delay to change the delay value from -80ms to 0 (as in ZERO).

- John "FulciLives" Coleman

P.S.
When I say the AVS script method I am talking about the script that uses the SmartDeinterlacer stuff which we have talked about many times in this thread.

[DJRumpy]

Sorry for the delay in responding Fulci, I've been busy this week

Spruce has no problem importing 16:9 video, it just puts a 4:3 aspect ratio flag on any IFO's it creates. The dvd would still play without issue on a 4:3 television, it just has the wrong flag on the IFOs.

[FulciLives]

ANOTHER METHOD FOR INTERLACED 25fps PAL

I was thinking about this recently ...

What's wrong with deinterlacing the PAL source THEN converting it to 23.976fps PROGRESSIVE NTSC?

This seems that it might be better for those that have PROGRESSIVE TV displays (such as PLAZMA screens or HDTV sets) rather than the SmoothDeinterlacer method that Xesdeeni came up with.

Don't get me wrong Xesdeeni's method works and works well but you end up with a strange INTERLACED 29.97fps NTSC source. I say strange because you get a 3-3 pattern. 3 Progressive frames followed by 3 interlaced frames and on and on and on ...

It looks just FINE on a standard INTERLACED TV but I wonder how good it would look on a PROGRESSIVE display type TV display? When I play back such conversions on my computer (my computer monitor is the ONLY progressive display I have) I use WinDVD but the built-in on-the-fly deinterlacer still shows some interlacing artifacts.

That's what got me to thinking ... why not first deinterlace the PAL source at 25fps then convert it to NTSC by changing the frame-rate to 23.976fps and resizing to 720x480 from the original 720x576

So right now I'm trying the following AVS script on a short 25fps interlaced PAL source:

#==============================#
# -= Avisynth script by MovieStacker v1.1.1 =- #
#==============================#

LoadPlugin("C:\KVCD\MovieStacker\MovieStacker\Filters\MPEG2Dec.dll")
LoadPlugin("C:\KVCD\MovieStacker\MovieStacker\Filters\DecombLegacy.dll")
AviSource("D:\capvideo.avi")
FieldDeinterlace()
LanczosResize(720, 480, 0, 0, 720, 576)
AssumeFPS(23.976, true)

Now as you can see I used MovieStacker to help in the creation of this AVS file. I'm not good at the various methods of Deinterlacing so while I think this method in general will work and while I think this particular AVS script will work I'm not sure that I'm deinterlacing as good or as well as I can be.

Anyways I'm encoding the clip now. It's only about 11 1/2 minutes but I'm doing 2-pass VBR with bitrate settings of AVG 5700, MAX 8000 and MIN 2000 because the movie I want to encode is 90 minutes PAL (about 95 in NTSC) and this clip is from that movie so I wanted an idea of what the quality would look like with the final bitrate settings.

As my computer is SLOW as ass and I'm using TMPGEnc this short 11 1/2 minute clip will take something like 4 hours according to TMPGEnc although I notice it never takes as long as it first says ... at least on my system the estimated time decreases especially around the 50% mark of 2-pass VBR mode. It seems to run faster in the second 50% than the first 50%

So in other words I'll be back here in 4 hours or less to let you know how it turned out hehehe

In the meantime any suggestions on the best way to deinterlace the PAL before the conversion to NTSC would be most welcomed

- John "FulciLives" Coleman

[FulciLives]

Hello

Well I'm watching my clip now. I watched it on the computer and I'm now watching it on a TV and it looks really good to my eyes. I think this is the method I will be using from now on with INTERLACED PAL material.

I'd rather have a PROGRESSIVE NTSC as an end result even if it does mean converting the audio with BeSweet.

Again, although I think it looks good, I'd like to hear about other methods of deinterlacing if anyone (I'm thinking DJRumpy or Xesdeeni here) knows of a better method or methods to try.

Cheers

- John "FulciLives" Coleman

[DJRumpy]

Interlaced material looks fine on an HDTV. You could use either method, although you take a loss in quality by deinterlacing. It also doesn't make a lot of sense. Think about it. Your converting an interlaced video to progressive, and then performing telecine, which puts interlaced frames back in the video

[FulciLives]

Interlaced material looks fine on an HDTV. You could use either method, although you take a loss in quality by deinterlacing. It also doesn't make a lot of sense. Think about it. Your converting an interlaced video to progressive, and then performing telecine, which puts interlaced frames back in the video

Well I don't have an HDTV or PLAZMA etc. but I figure sooner or later I will so I'd like to encode NOW so it looks good LATER when I upgrade my TV display. I just thought that perhaps the odd 3-3 pattern of Xesdeeni's method might "fool" or "mess-up" the HDTV when playing it back. It is my understanding (perhaps wrong) that such a progressive display does some sort of deinterlacing upon playback so if that IS true then wouldn't it make sense to deinterlace the source on the computer first so it is PROGRESSIVE when encoding. In other words I would assume that a properly done slower-than-real-time computer deinterlacing method would work better than the TV which has to do it on-the-fly ... am I making any sense hehehe

As for converting to progressive and then performing telecine ... well now you got me confused ... hehehe

I deinterlaced the PAL source then changed the framerate to 23.976fps and encoded it that way with 3:2 pulldown. So if I'm playing it back on a progressive display with a progressive DVD player then there would be no telecine process right? That would only happen when watching it on a "normal" interlaced TV ... correct?

I have to admit the end result of my deinterlace PAL to progressive NTSC did look a tad soft (more noticeable on the computer monitor than on the interlaced TV I've watched the clip on) but then again I was using a PAL VHS capture as a source so I wouldn't expect it to look as sharp as a DVD source.

So maybe I'll try my "new" method with an interlaced PAL DVD source and see just how "soft" the deinterlacing makes it.

- John "FulciLives" Coleman

[DJRumpy]

I deinterlaced the PAL source then changed the framerate to 23.976fps and encoded it that way with 3:2 pulldown. So if I'm playing it back on a progressive display with a progressive DVD player then there would be no telecine process right? That would only happen when watching it on a "normal" interlaced TV ... correct?

You would think that's the way they do it. It turns out it's nothing like that. They do an odd 23.976 to 60 fps telecine-like conversion, displaying each frame multiple times. It seems like they make this stuff up just to complicate things. You'd think with the established framerates, they would simply make a television display cable of displaying the 23.976, or 25fps, or 29fps natively. Instead they come up with another method to take the progressive source, repeat the fields/frames a few times, to change the framerate to match the field display rate spec of a standard TV (60 fps for NTSC).

From what I understand (and I may have it all wrong..who knows? )It's somewhat of a new form of telecining, but not in the normal sense, since it's done at the frame level, rather than the field level. The hardware takes each top field and combines it with it's counterpart (a+b). It then repeats this a+b frame in a 3:2 pattern. So if we have a 23.976 fps progressive stream, and we look at 4 frames from this stream:

1 2 3 4

111 22 333 44

You end up with a 60fps progressive display. This is an over simplification, since you also have to deal with interlaced material at times, and the fact that each field has a different timestamp. The hardware handles those sources differently, doing a de-interlace for you. It can definately do a better job than a home-done deinterlace, assuming you have a good chipset.

In the above example, 1 2 3 4 are all made up of the upper and lower fields for each frame (i.e. the top and bottom fields for frame #1 are combined to form frame #1). This full frame is then repeated in the pattern indicated above.

Depending on how good the chipset is in an HDTV, it may attempt to perform IVTC on a telecined source, and then convert it to 60fps progressive, or for true interlace signals it could simple double each frame twice (converting from 30fps to 60fps).

Hopefully I haven't made this all as clear as mud.

[FulciLives]

DJRumpy:

I guess the question boils down to this then ...

What would Xesdeeni's method look like on an HDTV?

Will it look A-OK?

I'm worried that it might not since it has that odd 3-3 pattern (3 progressive then 3 interlaced repeated on and on) instead of the common 3-2 pattern that we normally have with NTSC material.

Unfortunatly I don't have an HDTV or PLAZMA or other progressive TV display to test what it would look like.

I do have a few PAL interlaced DVD discs that I've converted to NTSC using Xesdeeni's method (the one that produces the 3-3 pattern).

Maybe I will just have to take one of those discs to an electronics store and see if I can't play it back on a display model HDTV set but anymore all I have around me are mostly big box stores like BEST BUY who, unlike the days of old when you had more up-scale specialty shops ... don't like you mucking around very much with stuff. I was once told by BEST BUY that I couldn't hook up a display model DVD player to a display model TV to test it because they didn't have patch cords and it was too much trouble and if I didn't like the DVD player I had x number of days to return it anyways so "why the problem bud" ... hehehe

- John "FulciLives" Coleman

[DJRumpy]

I guess my answer here would have to be; it depends on the chipset. Since the signal is interlaced, and with an odd telecine pattern, then the chipset would have to work it's magic to deinterlace. Some do a great job. Some very noticeably bad. I suppose it would actually be a good test disc for a future tv

[koolio]

I'm trying to convert a PAL DVD to NTSC DVD format and am hoping someon on this thread would be able to help me out. The problem I'm having is that the .mpg file I created (after muxing the audio and video via TMPGEnc) is much larger than the movie on the original DVD (5.8GB vs. 4.6GB). Could someone explain to me why this happened? Also, what is the best way to compress the 5.8GB .mpg file I created so it will fit on one DVD-R?

Below is the process I underwent to create the NTSC .mpg file:
1. Used DVD Decrypter to extract the .VOB Files from the DVD (Ali G InDaHouse)
2. Used DVD2AVI to extract the audio and video files from the .VOB - resulting in a .D2V file and two .AC3 files.
3. Used BeSweet GUI to convert the .AC3 file to a NTSC .AC3 file.
4. Used BeSweet GUI to convert the newly created NTSC .AC3 file to a .MP2 file.
5. Used TMPGEnc to convert the .D2V file to a .M2V file with the following options selected:
- Audio: I selected Linear PCM (should I have chosen the MPEG-1 Layer II Audio (MP2) option?).
- NTSC DVD template.
- Video: I selected "Highest Quality (very slow)" setting.
- I selected the "Do No Frame Rate Conversion" option.
6. Used TMPGEnc to mux the .M2V file (created during step 5) with the .MP2 file (created in step 3) - resulting in the 5.8GB .MP2 file.

I've checked the .MP2 file and the audio and video seem to be in-synch. It's just that I can't fit the .MP2 file onto one DVD-R. Did I do something wrong during the conversion process that caused the .MP2 file to be larger than the original movie? All help is appreciated. THANKS!

[DJRumpy]

koolio, you should post this in your own thread rather than hijacking this thread.

To answer your question, yes, you should have used MP2, or AC3 for the audio, not linear PCM. The audio plus the AVG/CBR bitrate setting determines how large your output will be.

Post your question in it's own topic for better results

[koolio]

DJRumpy, thanks for answering my question and my apologies for "hijacking" this thread but I noticed that there were a lot of knowledgeable folks on this thread and thought my post would get the most exposure here.

Again, thanks for the answer - I'll try to re-encode with the mpeg2 setting.

[FulciLives]

DJRumpy, thanks for answering my question and my apologies for "hijacking" this thread but I noticed that there were a lot of knowledgeable folks on this thread and thought my post would get the most exposure here.

Again, thanks for the answer - I'll try to re-encode with the mpeg2 setting.

Not trying to be rude here but you keep confusing MPEG/MPEG2 with MP2. MPEG-2 is the video file whereas MP2 is an audio file. An MPEG file is a MPEG-1 or MPEG-2 video with AUDIO (it can be MP2 or AC-3 etc.)

- John "FulciLives" Coleman

P.S.
I would try using the WIZARD MODE in TMPGEnc as it has a built-in bitrate calculator so you don't make your final file(s) too large for a single DVD.

[DJRumpy]

Good catch FulciLives. I didn't even catch koolio's reference to MPEG-2

[Xesdeeni]

What's wrong with deinterlacing the PAL source THEN converting it to 23.976fps PROGRESSIVE NTSC?

Nothing, except you will lose temporal data. When you deinterlace to 25 fps, you get 25 "instants of time." When you play that on a TV that displays more than that number of frames per second (100Hz is common in Europe), you get 25 "instants of time" replicated (4x for the 100Hz display). Nothing is gained. But, if you leave the 50 interlaced fields in place, then play back through the progressive display, it will deinterlace to 50 progressive frames per second. You'll get 50 "instants of time," so the movement will be smoother.

The same holds true if you convert to 23.976 progressive fps.

What would Xesdeeni's method look like on an HDTV?

Admittedly I've only worked with a small sampling of HDTV material to date, but I noticed that 720p from film is 3:2 pulldown on the frame level, while 1080i is still a normal 3:2 pulldown on fields. But regardless of the original format (film or video), converting to DVD resolution doesn't even need to consider it. Conversion is a piece of cake. For the 720p, you just scale down to 720x480 (presumably anamorphic) and then take the even fields from one field and the odd fields from the next. For 1080i, you just separate the fields, scale down each field to 720x240, and then re-interlace.

I'll have more info in the coming months, as my HTPC with MyHD capture card is built and functional (although playback on my SDTV is having problems...see http://www.avsforum.com/avs-vb/showthread.php?s=189de124ca30b339ae3a33f179402178&threadid=287801). We're moving next week, so I won't be able to play with it much until after we're settled (and I've installed a decent antenna).

Xesdeeni

[FulciLives]

Xesneedi ...

I'm not sure if you follow my line of thought here or if it is me who cannot grasp what you are trying to say

What I am trying to say is this ... if you deinterlace the PAL at 25fps so you now have progressive 25fps PAL then convert that to progressive 23.976fps NTSC you will have a video that should look GREAT on a progressive display such as a PLAZMA or HDTV television. By GREAT I mean that you now have a progressive NTSC source tha does not need to be deinterlaced by the progressive display for proper playback.

I have limited understanding I admit of how a progressive type display handles interlaced material but I'm thinking that the process I just described would perhaps be better suited than your method (the AVS script with the smoothdeinterlacer change FPS thing) since that method creates an "odd" 3-3 pattern in the final NTSC video.

This 3-3 pattern is of no concern to me AT THE MOMENT because it looks absolutely fine on a standard interlaced NTSC TV which is all I have for now but I am worried (perhaps wrongly so?) that the 3-3 pattern might not look so good on a PROGRESSIVE display such as a PLAZMA or HDTV.

Please note that I in no way mean to discredit your method. I like it and have used it and it looks GREAT as I've said on an interlaced TV but like I said I'm afraid it might not look "right" on a PROGRESSIVE display.

BTW I have begun playing around with Donald Graft's SMART DEINTERLACER plug-in for VirtualDub (version 2.7 beta 2) and it comes with an HTML help file that talks about various method that can be used to restore INTERLACED PAL material to PROGRESSIVE if it was created from a FILM source. I must admit I don't much understand it. I recently got a TV tuner type capture card and captured a 5 minute clip of the Warner Bros UK VHS release of THE DEVILS (the only uncut widescreen release of the movie) and in VirtualDub it looks progressive even without loading the filter and yet this is from a VHS video? When I try to frameserve it into TMPGEnc (without applying any filters) it says that it is INTERLACED.

I don't understand hehehe

- John "FulciLives" Coleman

[DJRumpy]

It's probably telecined. How many frames did you examine? The only other thing I can think of is that you may have had the deinterlace filter loaded already. You'll get a live preview in virtualdub if that's the case. Be careful with NTSC vhs material. Most of it is telecined, not interlaced.

You should just be clear that when your output is intended for television, you should NEVER deinterlace. You only lose quality, as every television on the market is designed to display interlaced signals without issue. It shouldn't even be a question. Only deinterlace when you intened to watch the video on a display that cannot display interlaced signals as they were intented (like all new PC monitors).

[Xesdeeni]

What I am trying to say is this ... if you deinterlace the PAL at 25fps so you now have progressive 25fps PAL then convert that to progressive 23.976fps NTSC you will have a video that should look GREAT on a progressive display such as a PLAZMA or HDTV television. By GREAT I mean that you now have a progressive NTSC source tha does not need to be deinterlaced by the progressive display for proper playback.

I think we are mixing a bunch of things here and it's getting confused.

First, although my technique for standards conversion uses deinterlacing, the two actually have nothing to do with one another. I use the latter as a tool in performing the first.

Next, we need to reiterate the difference between deinterlacing and inverse telecine:

- Inverse telecine (IVTC): reversing the process of converting film to video. This involves reconstructing frames from the video. All of the information for every frame is present and distrubted across fields.

- Deinterlace: converting an interlaced video to a progressive video. All of the information for every frame IS NOT present and must be reconstructed.

When doing standards conversion, you need to convert both the frame dimensions and the frame rate (equivalent to the sampling rate in audio). I chose to deinterlace because converting interlaced video is unbelievably tricky (see http://www.snellwilcox.com/reference/pdfs/estandard.pdf). But as noted above, when the source is film, there is no need to deinterlace, since all the information is there. Instead, you should IVTC.

So we have two different discussions going on here. One regards deinterlacing and viewing the output on a progressive display. The other regards standards conversion.

With regards to deinterlacing, please understand that none of the PC deinterlacers are worth a crap compared to the deinterlacers in the TVs. These extremely sophisticated devices use motion estimation and a host of other techniques to reconstruct the missing information. If you have a choice between feeding 50 interlaced fps to a progressive TV and feeding 25 progressive Fps, always, always, always choose the former. If you have a choice of using a PC-based smart deinterlacer over the real-time deinterlacer of a PC playback software, choose the former.

As for standards conversion, that discussion is a bit more complicated. When you convert from video to video, my technique for conversion, using a deinterlacer, works extremely well. However, when the source is film-based, you have a number of choices:

1. Don't worry about it and just use the video to video method.

2. Use method 1, but IVTC instead of deinterlacing.

3. IVTC and then adjust the speed of the frames (25->23.976 or 23.976->25).

Method 1 will work pretty well, although it will not be as good as the other methods, and it will take longer due to the deinterlacing process.

Methods 1 and 2 will have artifacts (jumps or jutter, depending on whether you choose ConvertFPS() or ChangeFPS()) on smooth motion shots.

Methods 1 and 2 require no changes to the audio, while method 3 requires the audio to be sped up or slowed down proportionally to the frame rate.

This 3-3 pattern is of no concern to me AT THE MOMENT...

It's important to understand that when you pair the fields, you are NOT doing what the TV does. An interlaced TV NEVER displays two fields as a frame the way you are doing in VirtualDub. The concept of a frame is (almost) entirely arbitrary. The TV just displays a sequence of fields. The pairing you are doing is completely arbitrary. You could just as easily have paired one of the fields you are looking at with one from the next frame. You'll see something closer to what a TV does if you use the AVISynth command DoubleWeave(). Then you will see fields 0 and 1, followed by fields 1 and 2, and then 2 and 3, etc. This is closer to what the TV sees and what the deinterlacer will be dealing with. If you view this, you'll see that each and every frame shows up at least once as progressive.

So in answer to your concern, the pattern shouldn't confuse the deinterlacer. It will realize that the source was film and display accordingly.

Xesdeeni

[rabit]

I'm folling this guide :
http://www.geocities.com/xesdeeni2001/StandardsConversion/

to covert PAL (Video) DVD (25i Fps) -> NTSC DVD (29.97i Fps)

Tools I have been using:
AVISynth: tried both versions: 2.52, 2.08
MPEG2DEC.dll also tried MPEG2Dec3.ll
DVD2AVI: tried both versions: 1.76, 1.77
TMPGEnc 2.521
SmoothDeinterlacer.dll

No matter what version I have tried, error:
" SmoothDeinterlacer.dll is not a plugin of AVISynth2"
"Evaluate: Unrecognized exception! (E:\au3\object.AVS,line 3)

Or

"MPEG2DEC.dll is not a plugin of AVISynth2.5 "
"SmoothDeinterlacer.dll is not a plugin of AVISynth2.5)
"Evaluate: Unrecognized exception! (E:\au3\object.AVS,line 3)

Here is my AVS file:

LoadPlugin("MPEG2DEC.dll")
LoadPlugin("SmoothDeinterlacer.dll")
MPEG2Source("E:\au3\project.d2v")
SmoothDeinterlace(doublerate=true)
LanczosResize(720,480)
ChangeFPS(59.94) # or ConvertFPS(59.94)
SeparateFields()
SelectEvery(4,1,2)
Weave()
ConvertToRGB()

I have winxp

Please advise,
Many thanks

[DJRumpy]

You have either places the DLL's for an older version of AVISynth in your system path, or they are truley not for version 2.5.

Use the search function to search for those DLL names. You can also check your registry under HKEY\Local machine\Softare\AVISynth
Find the PluginDir key and see where it points. Ensure your new (avisynth 2.5) versions of these DLL's are placed in that directory (you can also change the directory by editing this key if you like). Delete all other instances of those DLL's.

The errors indicate that the DLL's being loaded are simply not compatible with the version of AVISynth that your using.

Once you place the correct DLL's in the PLUGIN directory, do not use the LoadPlugin command for those DLL's in your script.

[rabit]

Thank you for your reply, DJRumpy.
I followed your instruction above
-Delete other Avisynth paths, folders, just keep the 2.5 one
-Move those .DLL files from WINDOWS\SYSTEM32 to Avisynth2.5/plugins folder.
Then I run TMPGEnc, now error is:

1/Unable to load "SmoothDeinterlacer.Dll"
2/If I delete this line from the AVS file, error is:

"Unrecognized exception".E:\...line 3

I searched google and this forum for a SmoothDeinterlacer.Dll but all websites pointed to the same download place.

Do you have any idea?

Thanks

[Xesdeeni]

Grab the 2.5.x SmoothDeinterlacer from http://www.avisynth.org/users/warpenterprises/ ( http://www.avisynth.org/users/warpenterprises/files/smoothdeinterlacer_25_dll_20030218.zip ).

Xesdeeni

[rabit]

I just downloaded the new SmoothDeinterlacer.dll from the link you gave me above, but still get the same error. What's wrong actually?

How come Avisynth 2.5 then 2.08 does not work at all?
I have been searching, reading, following all the steps but nothing worked out.

Any solution, please.

I truly appreciate your help

[DJRumpy]

You should have uninstalled your old version, rather than deleting them. In either case, if the damage is done, make sure you re-install AVISynth 2.5 (or higher). Open your registry, to the key I indicated above, and ensure that your DLL files are placed in that directory
(usually C:\Program Files\AviSynth 2.5\plugins )

If possible, you should definately uninstall your old version. The new 2.5 version installs into it's own folder. Installing it again should move the pointers to the new install regardless.