Vertical Wiggle

[Okiba]

Hello everyone!

I'm working on a project converting around 50 VHS-C tapes with family footage taken from around 1991 to 2003. The source device is an old Sony Video Camera my father had. I noticed that all the footage from 2001 to 2003 has vertical lines that distort the image. Earlier footage from the same source, does not suffer from this issue. So this let me believe that something happened to the Video Camera around 2001.

Here's a sample.

Note both the vertical shutter, and how the edges wiggle and wrap. The guys at DigitalFAQ were kind enough to help me and let me know that a VCR with TBC/DNR *might* solve the issue. Professional VHS with TBC/DNR are quite expensive those days and hard to find on my country (so it also includes expensive shipping cost). I wouldn't want to spend big buck on a new setup, only to figure a better VCR will not help with the above issue. So:

1. Is there any way to know how much better the video will become a better setup? or perhaps because we assume the original video camera was to blame nothing can really solve that issue?

2. Is there anything I can do post capture to improve it? I didn't do any post-capturing modification to the videos. I do have them all on Loseless format to do it later when I'll have more time (HuFyuv).

My setup is VirtualDub for capture (1.9 on Windows 7), Diamond VC500 capture dongle with an LG Combo VCR/DVD device (the manual for the device can be found here with technical spec).

Thank you!

[lordsmurf]

No, nothing can be done post-capture.
This must be resolved before capture, using VCRs with TBCs.

[jagabo]

Some old DVD recorders can clean up the horizontal time base in pass-through mode. Like a Panasonic ES10 or ES15. Those are more easily and cheaply available.

[johnjonh]

what happens if you play those tapes with the original camera?

[Okiba]

Thank you for the answers everyone (Hello again lordsmurf ).

Does every DVD record can work as pass-through? or it's a rare feature found specifically in Panasonic ES100/ES15? I have another DVD Record here at home. Maybe it's worth the try?

Sadly, the original Video Camera no longer working. The original idea was perhaps use it as a source instead of VHS. But I guess not using it for 20 years had it toll.
I'm still curios though - why time-base issues started to show only later on the Camera life? I was sure time-base issues were SETUP releated, and not a source related? Out of 52 VHS-C tapes I capture, only 4 suffered from this problem, and only those taken post 2001. My setup is not completely bad I assume, because 46 other tapes from the previous 9 years looks decent. I know it's not 4 problematic tapes, because a single tape had couple of events in it. 3 events from 2000 looked perfectly fine, and the final event on the tape from 2001 had time-base issues.

You can find a short demo of lossy footage (ff-mpeg with constant bit-rate of 3500kbps) taken before 2001 here - which to me looks fine. But I guess it due to lack of training - I would actually love if you can point me to the problems a better setup will solve for educational purposes (I know there's some chroma issues I think they are being called on the top and right section?)

Thanks!

[oln]

I have another DVD Record here at home. Maybe it's worth the try?

What model is it?

Pretty much every dvd recorder will digitize and re-create the signal when passed through, but only some have filters to correct bad wiggling. The Panasonic ES10/15 are the most commonly used ones and have especially strong correcting capabilities though most other panasonics can do great job as well (at least for PAL). Newer Sony and Pioneer models based around NEC chips also work ok for most tapes, as well as a few toshiba models based around the same chipsets. I think some funai-made ones can work as well (they use chips from panasonic), but I believe they suffer from the automatic gain control causing a lot of flickering.

In those LG DVD-recorder/VHS combos the video goes through the internal digitizer, which can reduce wiggling a little bit (otherwise it would look much worse on all tapes with your capture card.), but the A/D chips in those don't have the same horizontal stabilizing functionality (aka line-tbc/velocity correction and similar names) as the other mentioned models. If you are to get one of those, you would also need a different VCR as the one you got already digitizes and re-creates the video signal on playback as mentioned, which prevent any further stabilization. I would guess the reason it's more prevalent on the later tapes is the electronics and mechanics of the camera getting more old and worn which could result in it not recording quite as smoothly as when it was new.

[lordsmurf]

(Hello again lordsmurf ).

Hello.

Does every DVD record can work as pass-through?

No.

or it's a rare feature found specifically in Panasonic ES100/ES15?

Yes.

I have another DVD Record here at home. Maybe it's worth the try?

Doubtful.

Sadly, the original Video Camera no longer working.

Doesn't matter, cameras are almost never good for playback anyway.

I'm still curios though - why time-base issues started to show only later on the Camera life?
I was sure time-base issues were SETUP releated, and not a source related?

Errors were present the moment the tape was made. You're misremembering.

My setup is not completely bad I assume, because 46 other tapes from the previous 9 years looks decent.

Decent isn't necessarily good. Samples needed to confirm.

Out of 52 VHS-C tapes I capture, only 4 suffered from this problem, and only those taken post 2001.
I know it's not 4 problematic tapes, because a single tape had couple of events in it. 3 events from 2000 looked perfectly fine, and the final event on the tape from 2001 had time-base issues.

Not uncommon for home-shot footage to look as time progresses, as the gear is aging (and often not properly cleaned/maintained).

taken before 2001 here - which to me looks fine. But I guess it due to lack of training - I would actually love if you can point me to the problems a better setup will solve for educational purposes (I know there's some chroma issues I think they are being called on the top and right section?)

The timing wiggles are obvious in the date. It should be solid, not jello. It even hard jerks left from time to time, very unstable. Those same issues will exist all through the picture, not just the date. The chroma offset is ugly, shifted down at least 4 pixels, maybe 1 px to the left. Otherwise all I see is overscan. And are you using composite?

[Okiba]

Not uncommon for home-shot footage to look as time progresses, as the gear is aging (and often not properly cleaned/maintained).

I would guess the reason it's more prevalent on the later tapes is the electronics and mechanics of the camera getting more old and worn which could result in it not recording quite as smoothly as when it was new.

Ha! That's what I was missing. I didn't know time base issues is also a factor of aging gear. Thanks for clearing it up.

What DVD Recorder model is it?

I need to check. It at my Father's home. Will get back to you, but I think It's Pionner (PAL).

If you are to get one of those, you would also need a different VCR as the one you got already digitizes and re-creates the video signal on playback as mentioned, which prevent any further stabilization.

Ohhh. That's a very important feedback. Thank you. I'm not sure I can find a different VCR. I will concentrate on that before I'm grabbing a DVD Recorder. So DVD Recorder MUST be used with a normal VHS and not DVD Recorder.

The timing wiggles are obvious in the date. It should be solid, not jello. It even hard jerks left from time to time, very unstable.

Now when you pointed it out, I do see it. But very mildly and I have to focus, its hard to catch it with the woobly camera on a boat. I assume it just takes more training to notice it. I remember I used to not care about 30 FPS vs 60 FPS in Video games when I was younger. Now everything below 60FPS seems jerky to me.

The chroma offset is ugly, shifted down at least 4 pixels, maybe 1 px to the left.

I'm sure I understand entirely what you mean by "Chroma Offset"? I couldn't find anything about it on Google either? can you perhaps explain it or share information about it online I can research?

Otherwise all I see is overscan

This is a raw file encoded directly from the Loseless Huffyuv file. Normally I apply overscan corrections (I normally end up triming around 12 pixel from the bottom and 16-20 from each side left/right - while never going beyond 40 on both sides toghther).

And are you using composite?

I'm using Svideo. I found a recommendation in DigitalFAQs for Acoustic Research PR-121 Pro Series II S-Video Cable - And I found this one by luck on a local retail.

Otherwise all I see is overscan

So it's TBC issues and Chroma Offset (which I'm not yet sure what is and if it can be fixed with a better setup or requires post-processing).

By the way, I didn't played with the "Timing" option in VirtualDub. Correct me if I'm wrong, but those effect Audio/Video sync, not Time based wiggles like the VCR suffers from? Also - is the TBC issues is a side effect of the capture? or It's the VCR and I will experience the same connected directly to CRT?

Thank you both!

[lordsmurf]

chroma offset = "color bleeding" (sort of, not exactly, but related)

Timing settings in VirtualDub are mostly for a/v sync settings, dropped/inserted frames setup.

TBC issues?

[restore-chappy]

I tried experimenting rotating the video 90° then running QTGMC in progressive mode and then adding deflicker in Virtualdub
and utilizing "Spot removers" pixel locking noise reduction to stablise the image.

Not really a cure as such but slightly easier on the eye, Here's the side -by-side example.

[Okiba]

chroma offset = "color bleeding" (sort of, not exactly, but related)

Oh, I see. Yes, I can see color bleeding. I think last time I was doing some research about Avisynth and post-processing, there was a way to shift pixel like you said. So I assume that's something you can fix during post-processing. Is chrome offset is a VHS problem? does it mean that once I figure the correct Shifting values for a single video, rest of the videos will have the exact values (given I'm
using the same source/setup ofcourse)? or it's based on the type of content being filmed?

TBC issues?

Ops, I'm sorry. I used the wrong term. I mean Time base issues.

Not really a cure as such but slightly easier on the eye, Here's the side -by-side example.

I agree. Much more easier to the eye. And I suspect that if I would have showing the modified video to the untrained eye (like my father) - I'm not sure he will notice a problem (where the original file is clearly damaged).

I tried experimenting rotating the video 90° then running QTGMC in progressive mode and then adding deflicker in Virtualdub
and utilizing "Spot removers" pixel locking noise reduction to stablise the image.

Up until now, I was busy with capturing all the tapes I own for archiving - as I don't have much time to sink into Post-Processing right now (new baby) hich seems like a complex and big world. So while I heard about QTGMC - I'm pretty far off from how to do what you mentioned. I'm guessing I need to do some reading. Just to make sure I'm on the right track here:

- Rotating the Video 90" should be simple
- QTGMC is a AVisynth plug-in. I need to figure how to run it, and make sure I run it in progressive mode (even though the original content is Interlaced?)
- I should Rotate the video back, and use a deflicker filter on it in VirtualDub. Everything above should be done on the Loseless HuffYuv file and it should stay Loseless across all the process.

Is that a right direction to follow? and if you aware of already existing guide - it would be amazing if you can share a link (or maybe share the AviSynth script you used/VirtualDub Deflicker Filter settings).

Thanks again!

[jagabo]

I can see color bleeding.

Here's a colorful ball on a grey background:

[Attachment 54946 - Click to enlarge]

Here's the same ball with the chroma shifted to the left and down by 100 pixels:

[Attachment 54947 - Click to enlarge]
The colors are still sharp but they are in the wrong location. Obviously, that's a much bigger shift than in your cap.

VHS has another problem: chroma (colors) are saved at a much lower resolution horizontally, than the luma (grey scale signal):

[Attachment 54948 - Click to enlarge]
That causes the colors to bleed horizontally and small colored objects become desaturated.

Here's the chroma both blurred and shifted so you can more easily see how blurry the chroma is:

[Attachment 54952 - Click to enlarge]

I was doing some research about Avisynth and post-processing, there was a way to shift pixel

ChromaShift(), ChromaShiftSP().

[restore-chappy]

QTGMC is a AVisynth plug-in. I need to figure how to run it, and make sure I run it in progressive mode (even though the original content is Interlaced?)

The sample you uploaded is progressive so i presume you deinterlaced the orignal file in Handbrake?

Yes you should do any process lossless all the way until final encoding/delivery.

I'd be happy to share my experiemental process if you think it would be worthwhile for your videos?

[Okiba]

The colors are still sharp but they are in the wrong location.

Thank you Jagabo for the detailed explanation of what Chroma Offset is! Is Chroma offset happening because the setup I'm using? or because the source? (the Video Camera?). Also, will I see the identical offset in all the footage I have been converting? (taking in mind it's all coming from the same Video Camera and being captured with the same setup)? If so, I wonder if I can take all the videos I have captured so far and correct them all at once.

What about "Bleeding"? Is this again happening because of the Setup/Source I'm using? Can you fix that also like you can fix an Offset?

The sample you uploaded is progressive so i presume you deinterlaced the orignal file in Handbrake?

Oh, I'm sorry. That's a mistake on my side. Right now, I stored all the captures as HufYuv Loseless (which I assume is interlaced). I only used Handbrake so I can upload a smaller file for you to inspect. I didn't know it change it from Interlaced to Progressive. I plan the final Encoding to happen on avidemux and keep it interlaced.

I'd be happy to share my experiemental process if you think it would be worthwhile for your videos?

That would be very worthwhile. As for now, I'm not sure I'm going to be able to put my hands on a better setup any time soon. It might happen, it might not. Therefore - as for now - I think it's a good idea to apply your method to the 4 problematic tapes I own. It's better compared to the current situation. Thanks!

[restore-chappy]

stored all the captures as HufYuv Loseless (which I assume is interlaced)

I don't use Handbrake so I don't know if it deinterlaces by default- but if it doesn't then it sounds like your Lossless Huffyuv's
might be progressive ...and thats not the best scenario you want.!

Could you please cut out a short section in Virtualdub making sure you select "direct stream copy" under the video settings etc
The we can check if its been captured interlaced and advise further.

If you haven't used QTGMC and scripts before then you can read up on some of the excellent guides here as it can be quite tough getting it all running,

What are the durations of the tapes?

[jagabo]

Handbrake deinterlaces by default. It doesn't even support interlaced encoding.

[lordsmurf]

Handbrake is newbieware garbage that makes the best random decisions for you. It's often wrong.
If quality matters, never use it.

For x264, use Hybrid.

[restore-chappy]

Handbrake deinterlaces by default. It doesn't even support interlaced encoding.

Thats good news (in one sense) that the posters Huffyuv files are indeed interlaced,

I dont understand why Handbrake can't support interlaced video, Other like Xmedia Recode can etc.

[Okiba]

Handbrake deinterlaces by default. It doesn't even support interlaced encoding.

Whattt. Good to know. I only used because I wanted a smaller print file. But now I see the forum supports to up 500MB files. So I upload 10 seconds loseless versions of clips I previously upload - EpcotRaw.avi with major time-based issues and SwissRaw.aviwhere time-based issues are less visible (at least to my untrained eye).

For x264, use Hybrid.

Oh, I never heard about Hybrid. I was planning to follow the conversion Avidemux guide. (Mpeg2 FF, Interlaced, TTF with constant Bitrate and AC3 Afen). But I wasn't sure what was the correct Bitrate to use, and it seems like the article was speaking mainly about DVDs.

If you haven't used QTGMC and scripts before then you can read up on some of the excellent guides here as it can be quite tough getting it all running,

I'm a software engineer on my daily life, so I believe getting all the setup part will be OK. The harder part for me will be figuring out how to use it all properly, as I'm a newbie in everything Video/Audio (and everyday I'm being amazed from scratch how much there's to learn). I will be doing a quick search around the forum see if I can find any QTGMC guides. But correct me if I'm wrong, isn't QTGMC is a de-interlacer? I want to keep the video interlaced because I don't mind the TV doing the heavy lifting for me. Or were you only using it because the original file I uploaded was in progressive mode?

What are the durations of the tapes?

It depends on the tape, but each small VHS-C tape is around 45 minutes.

Thanks!

[restore-chappy]

But correct me if I'm wrong, isn't QTGMC is a de-interlacer? I want to keep the video interlaced because I don't mind the TV doing the heavy lifting for me. Or were you only using it because the original file I uploaded was in progressive mode?

Although QTGMC is a deinterlacer it has a progressive mode that can cleanup badly deinterlaced video and fix shimmering.

My test repair was only a experiement so if you plan to keep the files interlaced then any repair method will have to take that into account
and a re-think on the processing stages and filters.

Some processes and filters do require you deinterlace first though so sometimes its unavoidable.

Thanks for the uploaded test files they are indeed interlaced and they look to be captured correctly.


Are your final encodes for DVD video spec? You can go higher on the bitrate in Mpeg2 video if you dont plan on encoding to DVD video specs etc.

[Okiba]

My test repair was only a experiement so if you plan to keep the files interlaced then any repair method will have to take that into account
and a re-think on the processing stages and filters.

This whole capturing/restoring thing is a bit overwhelming. There's a lot to learn (and I'm in a sleeping hours battle against a new toodler ). That being said, I have been delaying capturing our family history for far too long. So I settled in doing research just about the first part - Capture. The goal is to capture all the footage, store it in a Loseless format (HufYuv) - and later - when I have more time, sink time into post-processing and restoring. So the only reason I keep it interlaced, is because that's "easier" (as It saves me another step along the road), and I'm planing to watch this content on the TV anyhow - so the interlacing is taken care by it. But I'm not against progressive mode. And if that's what needed to clean couple of problematic videos - so be it. Correct me if I'm wrong - but there's no difference between interlanced and progessive mode quality wise? a interlanced video will look identical to progressive mode video on a TV that supports both?

So I probably going to invest some unplanned time to read about restoration just so I can fix those 4 problematic videos. But nothing beyond that for now.
I still need to understand if the problem with the Chroma offset identified by Lordsmurf happening because the video camera/setup. Because if it does, I can probably apply the same correction values to ALL the videos I captured. But if it's per footage, I will postpone that too for now.

Are your final encodes for DVD video spec? You can go higher on the bitrate in Mpeg2 video if you dont plan on encoding to DVD video specs etc.

No, I do not plan on using DVD at all. Once the restoration process is done, I'm assuming I will get rid of the loseless form and encode it to a smaller file. Following one of the guides - the original plan was to use Avidemux and set video output to be Mpeg-2 with constant bitrate (I don't know what the recommended bitrate if I'm not specifically going for DVD spec. I know it's 1500kbps for DVD). Interlaced TFF. The Audio to be AC3 with bit rate of 384 and the output format to be MPEG-PS. But this guide was specifically for DVD. I might need to check Hybrid as mentioned above, or tweak the Avidemux settings. While I do want to encode it to a smaller file - I don't want to hurt the quality.

By the way. The original question for this post was answered. It sound to me like we talking about restoration now? perhaps I should open a dedicated topic there?

Thank you restore-chappy!

[jagabo]

DVD supports up to 9800 kbps MPEG 2 for video. But why use MPEG 2? h.264 or h.265 will give better quality per bitrate. And for video that's going to be watched I'd deinterlace with QTGMC -- that gives better quality than any DVD player or TV delivers with interlaced MPEG 2. Archive the interlaced huffyuv source.

[restore-chappy]

DVD supports up to 9800 kbps MPEG 2 for video. But why use MPEG 2? h.264 or h.264 will give better quality per bitrate. And for video that's going to be watched I'd deinterlace with QTGMC -- that gives better quality than any DVD player or TV delivers with interlaced MPEG 2. Archive the interlaced huffyuv source.

This ^

Some folk say you throw away half the quality by deinterlacing- but i dont think its anywhere near half when done by QTGMC,(help me out here LS)
If there's a lot of motion then perhaps deinterlacing with QTGMC to 50fps will help.

Here is a short test between Yadif and QTGMC, which is which?

Notice the improvement to the vertical wiggles on the title captions.

[lordsmurf]

Some folk say you throw away half the quality by deinterlacing- but i dont think its anywhere near half when done by QTGMC,(help me out here LS)
If there's a lot of motion then perhaps deinterlacing with QTGMC to 50fps will help.

Deinterlacing interlaced source can be half.

However, the "half" argument I make is 59.94 > 29.97. Don't get those confused.

The deinterlace itself is always destructive, but how destructive depends on the source and the method. Part of that method (not just picking QTGMC) is choosing 59.94 or 29.97 (50/25 PAL). Some people insist that 59.94 is fine, while 29.97 is bad (and they consider that to be "half").

But 59.94>29.97 is not half. The simpleton view of math, 59.94 / 2 = 29.97 = half, is well... simpleton.

You have 720x240+720x240, as two fields per second. But to create 720x480 you have to extrapolate/interpolate to 59.94. Much of that data is redundant, cloned from neighbors. It's fake data to pad out the palette. While you lose some motion data going from 59.94>29.97, it's just not half. Some frames will definitely show halving, while equally as many will show redundancy.

Beyond mere math, there are problems with creating 59.94 from interlaced video. For example, I had a recent Avisynth issue, trying to solve vibration in 59.94: https://forum.videohelp.com/threads/398713-QTGMC-image-vibration-motion. That was a consequence of interpolation. The fix was either a long process-intensive script ... or 29.97fps. I did both. Sometimes the motion deserved max frame rate, sometimes not.

I have to do 59.94 and 29.97 on a case by case basis. 50/25 for PAL.

Deinterlace is far too misunderstood.

[jagabo]

Don't forget: everything you see on a modern TV screen is progressive. So your interlaced video is going to be deinterlaced one way or another when you watch it. The question is whether QTGMC is superior to what your player/TV can do. In most cases it is. Most TV's I've seen do something like Yadif.

QTGMC isn't perfect -- no deinterlacer ever will be. But maybe someday TVs will get better than QTGMC (probably not soon, and probably not significantly) -- so save your interlaced source.

[Okiba]

A lot of good information here. Had to go seek more information before replying.

So first of all, I knew H265/254 exists, and that's it's more efficient. I just didn't realize the reason people pick MPEG2 was because DVD limitation. Good to know. What would be the preferred bit-rate and is there a big different between 264/5?

Here is a short test between Yadif and QTGMC, which is which?

As you mentioned, it clearly visible the right video is way better. Is that QTGMC?

Don't forget: everything you see on a modern TV screen is progressive.

Actually, I wasn't aware of that. I was sure modern TVs CAN display Interlace. So they do the conversion to progression on real-time?

so save your interlaced source.

Yep. All the files are captured, masked (to remove overscan) and saved.

I have to do 59.94 and 29.97 on a case by case basis. 50/25 for PAL.

This requires me to experiment. As I mentioned, I remember from the times I played video games, Higher FPS was always "better" because everything was smoother. I probably need to take a video I captured, and try to de-interlace it once on 25FPS and then on 50FPS - To see if perhaps I can figure the "Rule of a thumb" for when to use 25 and when 50 (I'm using PAL).

So it seems like the next goals are pretty clear to me:

- Figuring out how to identify Chroma Offset better - and figure out if all the captured videos suffer from the same Offset (and if so - apply the correction to all).
- Taking a step forward and getting better with AviSynth and QTGMC, as de-interlacing is the first step to try to get those faulty videos "better" (following restore-chappy restoration process).

For future readers - this is a good blog post about how to use QTGMC (including installation of AviSynth+ and all the QTGMC dependencies).

[restore-chappy]

Definately a lot of good info from LS and Jagabo.

As you mentioned, it clearly visible the right video is way better. Is that QTGMC?

Yes the QTGMC 25fps render is on the right.

QTGMC isn't perfect as mentioned and not the delfacto deinterlacer for every interlaced video,
Sometimes it can look a little posterised or plastic so try Yadif and others as well, Sometimes I deinterlace with Yadif
then i might run QTGMC in Progressive mode- 1/2 as it sort of cleans up the artifacts and noise caused by Yadif -but doesn't seem
to make the image as plastic- although it can soften.

Each video can need its own treament and method and order of filters so I usually cut a small 30sec section out of the video
and try out different methods.

good luck and hope you can find the time inbetween the new toddler

[Okiba]

Ohh, by the way. Can you please explain the thinking process so I will be trying it out later on when I got QTGMC sorted?

I tried experimenting rotating the video 90° then running QTGMC in progressive mode and then adding deflicker in Virtualdub
and utilizing "Spot removers" pixel locking noise reduction to stablise the image.

- How rotating it effects QTGMC?
- I'm assuming I will be running QTGMC on the interlaced loseless and not progressive (as the progressive file was an error).
- Deflicker is a VirtualDub plugin. Will read about it.
- What do you mean by "and utilizing "Spot removers" pixel locking noise reduction to stablise the image". How did you "utilized" it? is spot remover is a VirtualDub plugin?

Thanks!

[jagabo]

- How rotating it effects QTGMC?

TurnLeft().QTGMC().TurnRight()

Normally QTGMC() cleans up artifacts created on near-horizontal lines/edges. Sometimes you have vertical lines/edges that need cleaning. Hence the 90 degree rotation.

- I'm assuming I will be running QTGMC on the interlaced loseless and not progressive (as the progressive file was an error).

He was talking about using QTGMC as a post processor for another deinterlacer, QTGMC(InputType=1/2/3). In those modes it just cleans artifacts created by the other deinterlacer. The frame rate remains the same.

- What do you mean by "and utilizing "Spot removers" pixel locking noise reduction to stablise the image". How did you "utilized" it? is spot remover is a VirtualDub plugin?

He's talking about various temporal noise reducers like RemoveDirt(), TTempSmooth(), RemoveSpotsMC(), etc.

http://avisynth.nl/index.php/External_filters#Denoisers

[restore-chappy]

- How rotating it effects QTGMC?

TurnLeft().QTGMC().TurnRight()

Normally QTGMC() cleans up artifacts created on near-horizontal lines/edges. Sometimes you have vertical lines/edges that need cleaning. Hence the 90 degree rotation.

- I'm assuming I will be running QTGMC on the interlaced loseless and not progressive (as the progressive file was an error).

He was talking about using QTGMC as a post processor for another deinterlacer, QTGMC(InputType=1/2/3). In those modes it just cleans artifacts created by the other deinterlacer. The frame rate remains the same.

- What do you mean by "and utilizing "Spot removers" pixel locking noise reduction to stablise the image". How did you "utilized" it? is spot remover is a VirtualDub plugin?

He's talking about various temporal noise reducers like RemoveDirt(), TTempSmooth(), RemoveSpotsMC(), etc.

http://avisynth.nl/index.php/External_filters#Denoisers

Thank you jagabo for helping out explaining the processes, Its appreciated.
The "spotremover" i mentioned was this Vdub plugin

http://spotremoverfilter.com/index.html

Whilst i dont think its spot removing abilities are as good as AVIsynth removeDirt() and such like, its noise reduction can be useful
for locking pixels- which gives the impression of stablising the image.
The author mentions it was based on the idea of Vdubs Temporal Nose reduction and there are various controls for locking the colour separately etc.

I think "Smart Smoother" (Donald Graft) and "VHS Filter" (Flaxen) are alternatives that have some degree of pixel locking, So these can be tried.

The process i did on Okiba's file was to deinterlace first to 25ps with QTGMC then rotate 90° then QTGMC progressive mode, then rotate -90° ,then noise reduction
(i used spotremover) Deflicker (Donald Graft) and i was going to add Camcorder colour denoise 1.6MT but forgot.

I left QTGMC sharpness control on default to slightly sharpen the image as its default setting is a little too strong normally and i find you have to turn it down.

I would think all these processes could be easily wrapped up neatly in a AVisynth script but i don't possess the skills....NudgeNudge Jagabo !

I work mainly in video/film work with products like Pfclean and only have just skimmed the Avisynth surface myself.