Tracking lines in video; how to fix these VHS tapes?

[deter]

Many times when dealing with VHS or Betatapes you get tracking lines in the video. Most of the time it is very bad at the start of a tape.

Was wondering what methods you use to fix this.

I think it is possible.

One of the ideas is tape baking for really bad tapes. Just wondering if you have tried this and does it help.

The other idea is to tighten the tracking or tension in the VCR itself.

Can you adjust the tape, like a tape transplant to get better tracking in a video?

These lines are annoying and hinder the enjoyment of watching an old video. On some tapes it happens so much that it almost wrecks the video.

It seems, the tapes recorded in SLP mode, have the most issues.

Adjusting the tracking can make a tape play different. Still you may get different errors in that place were the tape is damaged. Different VCR's may play the tape different.

[NJRoadfan]

If the tape isn't damaged, fast forward and rewind the tape completely to re-tension the spools. The beginning of the tape will usually have problems as thats the exposed end (assuming one properly stores their tapes completely rewound). It also gets beat up from the machine threading. The pros avoid this by recording a leader with color bars at the beginning of the tape.

Tracking problems will be more of an issue with slower tape speed. Moving the tape slowly and consistently is a precision operation, not much room for mechnical "slop", like SP mode. Having multiple VCRs on hand helps, as some track EP/Beta-III tapes better then others.

[admin]

The tape-start issue is as NJR states -- exposure to elements is often heightened here, assuming the tape was fully rewound. Technically, that's bad storage technique. The first several minutes of any tape should be used for some kind of non-sense at home, or a color bar test pattern by pros (with accompanying shrill pitch tone for broadcast purposes).

It's generally not a tracking issue, but one dealing with cleanliness and integrity of the tape itself. In some cases, it's how well the tape interacts with the heads of a VCR, where the VCR is known to be good. In others, it's a bad or dirty VCR head.

Baking is only something that you should do for moisture or mold. With mold, first you clean, then you bake. Mold generally means moisture has permeated some part of the tape. Baking can only be done in a chemist oven, not a kitchen oven. It's very easy to destroy/melt a VHS tape this way, if done wrong. A blow-dryer is a horrible idea, as it will unevenly warp the tape -- something that's not necessarily visible to laymen.

Most of the time, it is a simply issue of using another VCR, or manually twiddling the tracking dials. Sometimes it helps to have an ancient VCR that does not do digital tracking, but rather mechanical tracking.

Beyond that, you have to rely on software editing, restoration and forensic recovery techniques -- some of which have their own weakness and trade-offs.

[deter]

For this purpose it is not worn heads...

For the most part, the tape(tapes) will react the same in just about any VCR. With different VCR's the end result may be a little bit different. If u know what part of the tape has errors and u know what u are looking for.

With worn heads u may tend to get more of a drop out effect.

Now if you have really tight tracking or tension it may iron out the tracking streak line in the video. It is like the tape runs over it self for just a few frames causing a flicker or line thru a small part of the video. An idea was that maybe tape baking could iron out these small crackles. The picture is normally not 100% damaged it just has a few lines of errors which only last a few frames. If it is really bad you may get lots of lines running thru the video.

Pretty much most VHS tapes u will tend to get some kind of tracking error in the video. If u really look u can find them.

The JVC decks have heavy NR so sometimes these errors get blurred out of the picture and they are harder to see.

Bad storage or whatever the result of the fact that certain tapes have more scan line or tracking line errors in the video than others, could be a reason.

The fact is you have this tape and you want to get it to digital format. It is rare, you can't go out and buy the DVD. So u want to get it fixed.

It just seems SLP tapes have a heck of a lot more of these errors than other tapes.....Some of the store purchased tapes also get these types of errors.

You just have to deal with it or just dump the project.

Still think it may be possible to fix these crackles or skips in the tape. Again this is all theory....

It is like a shirt that needs to be ironed......

The tape rewinding does help...

[admin]

Baking, no.

Ironing? Maybe -- only if it was physically crinkled. I've considered this myself many times, but not (yet) run into the need to try a method. You'll want to use an iron and secondary heat transfer -- VHS tape melts easily, so no direct contact.

Quote:

It just seems SLP tapes have a heck of a lot more of these errors than other tapes

It's helical scan compression -- not much room for error, unlike the uncompressed/spaced SP writes.

Quote:

It is like a shirt that needs to be ironed......

This is physical tape damage. The tape was "eaten" at some point in the past.

[lordsmurf]

I've been following the other thread at Videohelp, and while some interesting ideas have been posted there, most of them are ultimately limited in their usefulness. (That's over at http://forum.videohelp.com/threads/3...nes-from-Video)

There are two kinds of tracking errors: constant and variable.

Constant tracking errors are relatively easy to fix -- simply adjust the tracking on the VCR. In the worst cases, the tape was recorded on a deck so out of alignment that you have to modify the VCR. Your best bet is to use the same VCR that made the tape. When that's not an option, as is often the case, you have to essentially "break" the VCR temporarily (move it out of proper alignment) to get closer to where the tape can track.

Constant tracking errors generally don't move around much. The partial-swap Avisynth replacments suggested on VH will only work here. That's where you capture the tape twice, with the constant bad-tracking noise in different places, and then use Avisynth to remerge.

More on that at http://forum.videohelp.com/threads/3...ues-and-Damage

However, constant tracking errors are not very common.

Variable tracking errors are the most common kinds of tracking noise caused by misaligned VCRs, cheap tapes, and other aged-related problems. The tracking bars move all over the place, often making HiFi audio flutter and buzz. It's generally impossible to capture the tape twice, with the tracking lines simply moving location. You tend to get a slight unstable signal or a really unstable signal, and that's it.

In the best cases, variable tracking errors come and go only at the top or bottom (or both top and bottom) of the image, and you're able to capture/record the mono linear audio track without noise. The solution to "fixing" this video is to simply mask the bad top/bottom of the video with black. In some cases, you may have to crop so much video that you're able to create a 16:9 widescreen version.

Also...

There were some discussions of tracking-like errors, which were actually cases of magnetic dropout. Dropouts are problematic, too, and require frame-averaging (or median) methods to remove or reduce. Some of the most advanced professional hardware TBCs have drop-out compensation, but those are still very limited in how well they work. Our tests with a DO TBC failed, as it still missed most issues.

There are some limited software options available for correcting this error, however it requires progressive video -- meaning you'll have to deinterlace the video first. However, VirtualDub's median filter works much worse on deinterlaced video compared to the interlace video. But you can't do it, because it butchers the interlacing after median processing, and doesn't look right even if you deinterlace afterwards.

There really isn't a useful Avisynth median filter, however, and filters like Despot and DeVCR are supposed to address exactly these errors. However, none of these really work all that well. In the case of DeVCR, it's pretty much undocumented on how it should be used, so we're left to guess at values. So far, even in the hands of very skilled Avisynth script'ers, none of this produces consistently excellent results without causing major side effects.

More on that at http://forum.videohelp.com/threads/3...Avisynth/page2

Some many years ago, Video Finesse had filters in Adobe Premiere 6.0 and 6.5, but those have long ago been discontinued. Their usefulness was also minimal, as everything else appears to be.

This is one of my "cuss projects" right now, because at the end of every test I tend to cuss at the screen. Someday I hope to have a stable and consistent method for removing dropouts.

The best dropout compensation would theoretically also reduce tracking error noise, possibly remove it entirely.

In the meantime, the only guaranteed method to repair videos like this is to hand-by-hand repaint each frame -- many thousands of them -- in an editor. VirtualDub can export a video as a series of still images. You can also edit with overlays, inside Adobe Premiere CS4-CS5. The actual painting is still done in Photoshop, with the single extracted still from the Premiere timelines.

This isn't the answer you were hoping for, but it's the only one available right now.

[deter]

If one can stop the error before it happens, aka hardware that is normally the best fix.

Not the real world, cause u are going to have these problems.

Tracking lines have become a major hurdle over the last few months.

Tape Baking is still something that needs to be tried.....

Have created different methods to solve tracking lines in the video. However it takes a long time. It is a hell of a lot of work. Two or three hour video can take hundreds of hours.

Yea Frame Painting is an option, you also have to be a little bit skilled in graphic design. Not sure if the avg. person can Frame Paint, or actually restore old photographs.

Granted people want things now, today, and they don't ever want to pay any money for anything. My problem is on personal projects, however if someone required my service for fixing a video, pulling out the odd tracking line and drop outs, the fee would be high.....

Here is the method...

1) You have to find them (Not Easy)
2) You have to deal with them (Have about 10 methods for fixing tracking line problems)
3) You have to test the video

(if the method didn't work, you try another one, you keep doing this until the problem is gone)

5) Than you have to re-scan the entire video. Guess what? you find more bad frames

6) Than you repeat the steps above....

Once you get to the final copy...You have to watch it again and make sure the entire film/video works....

Have done this more than a few times and it works like a charm. The end product is very good.

However is it really worth the time....

Kind of feel like a low budget George Lucas when I am doing this.......

[lordsmurf]

Quote:

Originally Posted by deter

5) Than you have to re-scan the entire video. Guess what? you find more bad frames
6) Than you repeat the steps above....

Sometimes you can fix a video in tiny pieces, using all sorts of varied methods, and then reassemble it in an editor like Premiere. It sounds like you may be going that route. I don't envy that project -- BEST OF LUCK TO YOU!

I've had to do this in the past, but it's always been on videos of less than 5 minutes. Anything longer would pretty much require a time investment comparable to building the pyramids. (Okay, maybe not quite that long, but you surely understand the hyperbole!)

Quote:

people want things now, today, and they don't ever want to pay any money for anything

Or worse, they think they can do it themselves, with decisions based on marketing horsepuckey. "I can buy this for $29.95, and click the 'Clean It Now' button and it will be as perfect as a CSI episode." I get rather disgusted with the number of people who buy $25 capture cards and $75 film scanners and think they're equipped and qualified to offer services to others. Ugh.

But I rant.

[jmac698]

Hello,
I'm one of those technical people from Avisynth I can probably help. First of all, all these noises you talk about are in fact, extremely easy to fix. This might seem a revelation to you, but I have a tape now that won't play in 5 VCR's (only shows a blue screen), but one I did get to play, and was able to remove all white lines 100% and reduce noise dramatically.
Please capture your tape 3 times, with the settings below, then you can use a script to automatically remove the noise (kind of like a majority vote across captures).
Yes, capturiing reliably is an issue, until you know the secret settings.
VirtualDub
capture timing options
Only the following should be checked:
correct timing
auto disable sync
I have a diagnostic video and analysis system to verify that you can capture with no lost frames.
See http://forums.virtualdub.org/index.p...T&f=6&t=19460&

Note that the sample is a TRUE restoration, with no enhancement of any kind, I simply replaced the noise added by the tape. I didn't need any fancy equipment to do this, the software is all free, there are no adjustments to be made, and yes, you click the fix it button (more or less...)

[deter]

jmac698,

I would be interested in knowing what scripts you are using in Avisynth. (Haven't been able to even get the program to work) But I think I need it.

Video drop outs from oxide tapes is another issue that one has to deal with. Having them removed helps the video, had a friend use Remove Dirt on a video and it worked really well pulling out some of the drop outs.

Next a tape that plays like that picture in your post. From my XP, that can mess up the heads on a VCR. It is a royal pain, trying to re-clean the heads. On something like that tape, I would say have it baked or cleaned before putting it in to a VCR.

From using this site, don't capture anything via the PC, everything is done with an older JVC recorder, recorded in FR80 mode on DVD-RW disks, never have drop frames. It is than ripped to the computer via VOB2MPG v3.

(Don't really issue stuff back to DVD) kind of a pain, rather just use a hard drive that can be used with the computer and played on the TV.

the VHS / Betamax picture is improved a bit by the chip set in the recorder. (use a Panasonic Digital Tuner as a pass thru to improve the picture....(Plus a few other things) Only use JVC & Panasonic SVHS VCR's....in addition have a prog amp & external TBC.

Created this picture to kind of show what kind of tracking errors I am speaking of. Basically, the picture rolls over itself, creating a scan line or error in the picture. Not talking about really crapy tapes that track horrible. This problem normally lasts about 4 frames to longer. If it is only 2 frames, that can be fixed pretty easy. Anything over 4 frames, is pretty hard to fix. Since the scan lines roll thru the frames, normally working top to bottom not sure if Avisynth would have a script to work with this.


Untitled66665.jpg

[jmac698]

Yep, that is probably quite easy to fix, however can you record a small section 3 times (I know it's a bit tedious) and upload it, I can make the script around that and tell you how to install and run it. It's not hard to use when made.
As for vob2mpg, fine if you already bought it, but the free version is extremely slow. There is a free program which does the same thing with all the features. It's called pgcdemux.
Select VTS_01_0.IFO for input,
by PGC
in dropdown, select long PGC
check only Create PGC VOB, then One File per VID
Process
The next step uses dgdecode.. but I'll explain that with the avisynth stuff.

Now that I can capture reliably on PC, I would prefer that option as I can record in uncompressed quality, and also adjust the recording levels to capture the full range of signal. It also saves a step, obviously. I can use passthru to make use of some of the Pany hardware. Whatever majic the chip chip does can be done in software, except the TBC part.

I just wanted to add that removedirt, descratch, devcr, depulse, depulsec, and despot are the filters I've tried, and in my sample depulse worked best. Removedirt can be used in addition, but that's a pretty heavy filter and to remove a lot of noise you have to be left with some blockiness.

[lordsmurf]

Quote:

Originally Posted by jmac698

removedirt, descratch, devcr, depulse, depulsec, and despot are the filters I've tried, and in my sample depulse worked best. Removedirt can be used in addition, but that's a pretty heavy filter and to remove a lot of noise you have to be left with some blockiness.

I'd like to see some specific script examples of what you've done.

[jmac698]

L.S.
That's too easy

loadplugin("DePulse.dll")
loadplugin("descratch.dll")
loadplugin("despot.dll")
import("deVCR.avs")
avisource("video.avi")
#Test various destreak algorithms
#best is 0, 3 has ghosting issues
destreak(0,16,35,false)

function destreak(clip v, int mode, int lumathresh, int chromathresh, bool show) {
  return select(mode, \
  \
  v.converttoyuy2.depulse([<"h", 80, 235, 80>],[<"l", 16, 79, 79>],0,show).converttoyv12, \
  deVCR(v, lumathresh), \
  v.Turnright(). \
  DeScratch(12, 10,5,1, 5, 41, 0,1, \
    keep=0, border=1,modeY=3,mark=show). \
  Turnleft(), \
  v.DeSpot([<"p1", 1, 20, 12>], [<"p2", 1, 20, 1>], pwidth=41, pheight=1, 25, interlaced=false, \
    ranked=true, p1percent=0, dilate=16, fitluma=true, blur=1, motpn=false, seg=0, sign=0, tsmooth=0))
}
#for descratch:
#minlen 0-5 are no different and best, minlen=6+ increased streaks
#maxlen 41+ is best
#blurlen=1 was best, increases otherwise
#maxgap=5 is best, increases otherwise
#mindif=12 is good
#border=1 is best
#asym=10 is best

That script just wraps the various plugins into one function and supplies some defaults. The funky numbers in [] are for AvsP, to translate just keep the last number only, e.g. depulse(80,79,0). And you should probably set interlaced=true where it appears above. As for RemoveDirt, remember it needs the old 0.9 version of removegrain (or possibly the separated removegrainT?)

Also, for very strong denoisers, take your pick, all these work about the same but I haven't tested which is faster:

tnlmeans,frfun7(3),vaguedenoiser(threshold=7,chromaT=0,Wiener=true), DeGrainMedian(limitY=8,limitUV=3,mode=1),TBiLateral

[admin]

Don't forget that you can use the "code" bbcode.

I've fixed your post.
Otherwise the forum can mess up various Avisynth scripts, DOS commands, HTML code, etc.

[admin]

Quote:

Originally Posted by deter

Created this picture to kind of show what kind of tracking errors I am speaking of. Basically, the picture rolls over itself, creating a scan line or error in the picture.

This Spiderman created image shows a timebase error -- horizontal jitter -- and not a tracking problem. It's not really the same sort of error going on here. Timebase correction, or even basic frame synchronization, would more likely fix this -- assuming this isn't an embedded error.

Knowing your source, it may be an uncorrectable embedded issue. By "embedded", I mean that you have a copy of a tape that had a video error, and that error was recorded into your secondary copy as part of the new image, and is no longer a signal error. Thus it becomes apparent why "copies of copies" made without any timebase correction is a bad idea.

This is where "software TBC" (using image-based algorithms, not signal-based ones) would be useful, however such a beast does not really exist. What is currently available adds just as many errors as it fixes, as it's still just theoretical that I've seen to date. Similar forensic methods are always done on a per-frame basis, to the exact error. Repaint would be less work, with simple geometric shifting of the scanlines. (Some of the forensic methods are simply automation of the repaint, algorithms made for the exact video -- not something that can be used on any tape.)

Still probably not the answer you were hoping for.

[jmac698]

Deter,
That Spiderman jitter problem can be fixed. People are thinking that software TBC is the only approach, but there's other ways as well.
I need to see a video clip to be able to help further. One idea would be to look for huge jumps in motion for a small set of contiguous lines, and replace them with interpolated picture. This isn't nearly as hard as TBC, because we are using the surrounding stable picture as a reference, and the error is only in one spot that's easier to detect.
A full software TBC is hard because common attempts don't have any reference as to how the lines should be corrected (there is an answer, something like a 2nd order derivative).

[admin]

Quote:

A full software TBC is hard because common attempts don't have any reference

In my mind, software TBC would operate in much the same manner (in terms of complexity and what's essentially a base form of AI) as a high-end DSLR's internal database, which is used for any number of things -- focusing, composition, auto exposure, white balance and maybe even HDR. A basic definition can be found on the Wikipedia Canon DiGiC III page. Nikon cameras have the same thing (EXPEED, I believe), as do other brands.

The term "software TBC" is mostly a derivative term and explanation from hardware, but anything that achieves those goals would be what we're shooting for. It would have to be an intelligent system that can sense video has been skewed within the image, and adjust it accordingly. It doesn't necessarily need a reference beyond what it sees in front of it.

Quote:

One idea would be to look for huge jumps in motion for a small set of contiguous lines, and replace them with interpolated picture.

Since unstable timing is what likely caused this, and such an error tends to be constant across the picture, I don't know that such a reference would work well. It's not really a "jump" or even a change in the video -- not usually. There can be some motion and movement in the skew, but all positions are incorrect in relation to the rest of the image.

Have you ever read this book?
VCR Troubleshooting & Repair, Third Edition by Robert Brenner and Gregory Capelo.

There are several copies available quite cheaply on Amazon right now:
http://www.amazon.com/gp/redirect.ht...reative=390957

That may give you some in-depth knowledge on how analog signals work, especially from VCRs, VHS tapes and similar formats. You seem to have a decent working knowledge of video so far, but since it was stated to be a hobby, I'm not sure where your knowledge falls off. Hence the suggestion to push through this excellent tome. If you already have it, and have read it, then excellent!

I'm by no means an advanced expert in this area either --- well, to some I am, but I know where my weaknesses are, and my definition of "expert" differs from that of most laymen -- but I've found this book to be one of the most helpful things I've ever read.

With your heavy background into engineering and/or programming, and a deep understanding of video, I'm betting you could create some excellent breakthroughs. Or at least, that's what us non-programmer non-engineers always hope for! We want to use the gear, not necessarily create it. That's your job to make the tools!

[deter]

KP,
You are the darth lord and I am the apprentice don't you remember.
I purchased every devise you ever said to buy.

The tape I sent you many moons ago, the one that would not play on any of your VCR's
This is still the strangest thing. On one of my AG1980's it plays perfect, so perfect that the tracking is pulled out of the video.
Some of the capacitors were shot on this machine so I took it to be repaired. (wouldn't get full color, it was dull and a tad bit off.)
I told them to not touch anything in the VCR, just replace the caps. They didn't listen, they fixed a pinch roller charged me $125 and I flipped. I tested this tape again. It played even better. For some reason, the video now has drop outs in picture.

I have taken in 5 AG1980's in to the shop. On 4 of them after replacing the caps, it would have video drop outs in the picture. These drop outs were never coded in to the video. On other machines the tape plays with zero drop outs.....So the error was in something they did in the VCR....

I would take the VCR's back, look you messed up the machine. I would fight with the guy, he would give me some BS, than 2 months later, they fix it.
They will not tell me why this happens or what caused it.
He would just replace more caps and make up some new problem and charge me more money....

- I haven't tested any of these tapes that I speak of above with the super tracking AG1980 cause it has been in the shop for the last 3 months.

I found on the Super Tracker AG1980, if you rewind the tape than hit tracking ASAP, it pulls everything out of the picture...Now with the bloody pitch roller fixed it just tracks everything really good....
Sadly now I only get it see her like 1 time per month and than tell the repair guy, nope, you messed up again...
- I have all the hardware you speak of above

May even have more VCR's than you....LOL
(More than likely not, but I have a lot)

None of these tapes are 2nd hand copies....
These are embedded errors. They are not time based errors.
(However sometimes playing with tracking can fix it)

Even on the Super Tracker AG1980, I have looked at some of the old recordings and if these errors are in the video they will still show up.....

Any error or streak that lasts over 4 frames will normally always happen.
Depending on the VCR, the error in the tape may play back and show up in a different form.
Trust me I can find these things on just about any recorded VHS tape.

Again, still think tape-baking could be the answer but I have never done it....
http://www.tangible-technology.com/tape/baking1.html

This statement is pretty much spot on...
"There can be some motion and movement in the skew, but all positions are incorrect in relation to the rest of the image."

Again the avg person watching these videos are not going to pick up on them. With a JVC deck with NR, it blurs them out, they are still in the picture.

You watch them on a Non-HD TV more than likely since the picture is not as clear or sharp, you may not see many of these errors.

-- merged --

Right now, I have 5 AG1980's that are perfect. Spent a lot of time and money on this.

2 are still in the shop...I have another AG1980 with worn heads....I have a lot of these machines

One has herringboning noise, which is more than likely something in the power supply, it didn't have this before.
When I had the caps replaced the 1st time, the VCR had drop outs, after the drop outs were fixed now it has herringboning noise.

Super Tracker now has drop outs...
(one of the reason why this had to get repairs was cause of un-natural color shifts in the playback) Just bad color tint I could fix this with a prog amp....

The other 5 machines are MINT !!!!! (I may start a repair business for AG1980's)
You buy these on EBAY you are going to have issues so PM my account.....

However, each machine tracks tapes a little bit different.
On one of the machines LP recorded tapes track with minor noise lines to the right, if u track it, they go to the left. If you play with it long enough it goes off the screen.
The tracking on another machine on SLP tapes the color at the bottom of the screen is lighter. It depends on the tape you just have to hit the tracking button...

Again for me it is all about the detail, trying to achive the best possible results. Things have been good. If you look at the end results of a video vs someone recording VHS to DVD, it is not even close.

I think KP once said to me, it would be like DVD2, just lower than a DVD, pretty much can get most of the recordings to this level.

Some of the betamax stuff, with full frame cleaning looks better than regular TV.
As stated before, I use HDMI outputs to the TV, not S-Video......

The FR80 on the MacroBlocking has also been really good. But FR180 still not sold on and don't really use it....

This is like my 2nd full year or season doing VHS work, the stuff is outdated, and wish I could get the stock recordings, but you can't get them. So pretty happy with the results....Spent a lot of time with this, I would say compaired to others, pretty close to being at higer than PRO level, but not at the JEDI MASTER level yet....

With soundtracks and my soundboard mixer, I can really do a lot of soundtrack repairs...

Invented a method to repair FLV files using hardware, this is not something I want to do, or fix but sometimes u find a rare recordings online. I was told u can't fix heavy pixelation in a video. 1) you can fix the soundtrack on an FLV file 2) you can fix some really bad pixelation....If you don't believe me...Just record an FLV file with pixelation to a VHS tape....This is not my method but it is the concept...JVC SVHS PLAYERS HAVE NR......Passthrough filters are the key to FLV

Would like to thank KP for all the help over the years, it has really helped.....

[kpmedia]

Quote:

For some reason, the video now has drop outs in picture.

If the tape is developing this over time, especially as it's played more, then it points to oxide shedding or other wear damage on a very small scale. Although head condition can affect playback quality, it tends to be far more erratic, and almost impossible to reproduce the same error in the same place on the same tape. That points to a tape issue. Why some VCRs (same model, different unit) may gloss over the error is the point of curiosity, and something I don't presently have an answer for. There's no difference in processing by the VCR, if the models are the same, and all components are functioning.

Quote:

These drop outs were never coded in to the video. On other machines the tape plays with zero drop outs.....So the error was in something they did in the VCR....

I've seen similar odd errors on a Samsung SV5000W unit, which is a re-branded Panasonic. The fix there has always been to jam a "cleaner tape" into the deck, which isn't good for it long-term but seems to work temporarily.

Quote:

Invented a method to repair FLV files using hardware

I think I invented that method about a decade ago (more, actually -- well into the mid/late 1990s) when I recorded rough-quality MPEG-1 MediaBase streams from an online college class to VHS. The VHS version looked better. I used this trick again in the RealMedia and VCD eras.

I'm actually in the middle of restoring an FLV to more-watchable quality. It's the only available copy of a pilot from a cancelled series that popped up on Google Video several years ago. With Stream Transport, it was easy to download. The "hard part" has been with its demux, re-encode to uncompressed format, and slow filtering. It's not super-duper great, but it's much better, and can at very least soon be burned to a DVD and watched on a large HDTV with some degree of enjoyment.

Quote:

Again for me it is all about the detail, trying to achive the best possible results.

As long as you remember that restoring is about "making it better" and not "making it perfect" then you'll remain sane. Trying to achieve the magic of the CSI TV shows is about as realistic as creating a Flux capacitor, popping it into a DeLorean (which are still sold new, FYI) and going back in time to re-record it with a DVD recorder.

Quote:

Would like to thank KP for all the help over the years, it has really helped.....

Awesome.

[deter]

After doing a ton of research on this topic...

This is what I found.....

People really don't care about problems with VHS tapes....

Script editing more than likely will not work...
(Could be wrong, if it was written correctly)

We kind of want to clean up this problem on playback also...If we can.....

Tape Cleaning maybe...

Tape Baking? Could help.....

Time Base Corrector? No.....It doesn't help

I believe the data or recorded video is still on the tape, however do to many different reasons the pulse can no longer be read correct thru the VCR. It can be how it was recorded, it can be mother time, it can be the VCR, or a lot of other factors. aka bad blank tapes, poor storage....

The end results are the same, damaged video.....

Tried out a few weird things, like playing the tape backwards, over a video tear, guess what sometimes it removes it. Played the entire tape backwards. Has made some of these errors go away.

Paused the video over the tracking issues. Ran the tape a lot of times over that area in slow mo. What did it do? Burned a hole in the video creating a massive drop out....

It is understood also that some parts of the tape could be damaged beyond repair....

What we need is a Pulse Signal Corrector, to play back the recording with out too many errors.....

The signal is not being read correct thru the VCR...

Read in another post how the HM-DH40000U was able to help with video tearing...Now idea how much these things cost...Or if it would work....

Also read on this site about the ES10, not really sure if that would fix these problems.

Than looked at this mac daddy of a VCR

http://pro.jvc.com/prof/attributes/f...l_id=MDL101540

That is a hell of a lot of money to test out a theory....

Than thought about renting this machine for a week...

http://www.datadev.com/computer-tape...pe-rental.html

End Results are:

We need to get the tape back to the best shape we can before we play it back. Cleaning and baking could help but on the hardware end of things we need a VCR to be able to read the pulse correct....or better yet, get the pulse correct....