Image issues on poorly stored tape, can it be fixed?

[zinkatres]

I recently retrieved a box of old family movies from my parents and unfortunately many of them are in poor shape.

After a good amount of cleaning, I captured this tape, only to notice it has quite a lot of issues, most notably the rolling horizontal lines. From what I watched, sound did not seem to be affected, and I didn't notice any dropped frames or anything getting out of sync

I have attached a small clip of what I am seeing. The actual issues varies over the length of the tape and is not present on the entire tape.
Physically the tape doesn't look to be damaged as far as edge curling or anything like that.

I did try running this tape through both my JVC S7600 and AG1980, and the results were slightly different. The attached clip is from the JVC. The AG1980 image was much worse, almost looking like snow over most of the frame

Just wanting to get some opinions on if this is something that is able to be corrected fully or in part with an avisynth script, or if it's a lost cause and I should just enjoy what I was able to get out if it.

I appreciate any info you all can provide. Thanks!

[lordsmurf]

Quote:

Originally Posted by zinkatres

After a good amount of cleaning,

This concerns me. What, exactly, precisely, does that mean, did that entail?

Quote:

I have attached a small clip of what I am seeing.

Since you get same results on more than one VCR, it's physical tape damage. Hard to fix in hardware. At best, Avisynth to restore.

[zinkatres]

Quote:

Originally Posted by lordsmurf

This concerns me. What, exactly, precisely, does that mean, did that entail?

Mostly cleaning of the outside casing as they were pretty filthy. To deal with the mold, I did also run the tape through a cleaning pass as outlined in this video / and webpage

https://www.youtube.com/watch?v=uVq0...l=DustinKramer

https://lunchmeatvhs.com/blogs/blog/...ad-chip-parton

I have used this method before on other tapes with good amount of success. Also to note, these same issues were present on the tape when I ran it through my test vcr, before any cleaning took place, so I don't believe it's anything I caused.

If there is a better method of removing mold buildup, I am welcome to the idea

Quote:

Originally Posted by lordsmurf

Since you get same results on more than one VCR, it's physical tape damage. Hard to fix in hardware. At best, Avisynth to restore

At this point I would agree a software solution is necessary, but I am so far not well versed in what avisynth is capable of or what other methods might be viable

[lollo2]

Lots of shifted fields and many lines errors. Difficult to fix. Something to try just as experiment:

- from the hardware point of view: disable lineTBC in JVC 7600 and add a Panasonic ES10/ES15; replace Blackmagic capture card with something else

- from the software point of view: just a quick attempt to realign the shifted fields and reduce horizontal stripes (destructive operation)

video_dir=".\"
video_file="tape-errors.avi"

video_org=AviSource(video_dir+video_file).trim(0,34).convertToYV16()

# plugins directory
plugins_dir="C:\Users\giuse\Documents\VideoSoft\MPEG\AviSynth\extFilters\"

	# shift fields GMa
Import(plugins_dir + "shift_fields_GMa.avsi")

	# FixRipsp2
Import(plugins_dir + "FixRipsp2.avs")
	# RgTools
loadPlugin(plugins_dir + "RgTools-v1.0\x86\RgTools.dll")
	# DePanEstimate
loadPlugin(plugins_dir + "depanestimate110\DePanEstimate.dll")
	# FFTW
loadPlugin(plugins_dir + "LoadDll\LoadDll.dll")
loadDll(plugins_dir + "fftw3_20040130\fftw3win32mingw\fftw3.dll")
	# DePan
loadPlugin(plugins_dir + "depan1131\DePan.dll")
	# MaskTools2
loadPlugin(plugins_dir + "masktools2-v2.2.23\x86\masktools2.dll")
	# MVTools
loadPlugin(plugins_dir + "mvtools-2.7.41-with-depans20200430\x86\mvtools2.dll")

# shift fields
video_org_shift=video_org\
.shift_fields_GMa(8,0,2)\
.shift_fields_GMa(9,0,2)\
.shift_fields_GMa(12,0,1)\
.shift_fields_GMa(13,0,1)\
.shift_fields_GMa(15,-1,0)\
.shift_fields_GMa(17,0,1)\
.shift_fields_GMa(27,0,1)\
.shift_fields_GMa(33,0,3)

# repair fields

### separate fields
video_org_shift_sep=video_org_shift.AssumeTFF().separateFields()

### select even fields
video_org_shift_sep_even=SelectEven(video_org_shift_sep)

### select odd fields
video_org_shift_sep_odd=SelectOdd(video_org_shift_sep)

### repair
video_org_shift_sep_even_rep=video_org_shift_sep_even.FixRipsp2().FixRipsp2()
video_org_shift_sep_odd_rep=video_org_shift_sep_odd.FixRipsp2().FixRipsp2()

### interleave
video_interleaved=interleave(video_org_shift_sep_even_rep,video_org_shift_sep_odd_rep)

### weave
video_restored=video_interleaved.Weave()

stackhorizontal(\
subtitle(video_org,"video_org",size=20,align=2),\
subtitle(video_restored,"video_restored",size=20,align=2)\
)

Short video of the results: y.avi

y.jpg

[hodgey]

The lines look like dropouts in the signal on the tape that the VCR is compensating for yeah. The JVC vcrs tend to mask dropouts for many more lines than panasonic decks so maybe that's why it looks a bit better in the JVC, though as noted a Panasonic ES10/ES15 might be helpful here. Those are generally able to line things like these long dropouts up a bit better horizontally than the internal TBC in the JVCs.

[latreche34]

This type of severe tape drop outs can be fixed by modern DOC that use adjacent frames to get the clean lines out of them, A script should do something similar as shown above, but nothing magical here, Tape degradation cannot be fully reversed you have to keep that in mind.
Once you do the first pass capture, pull the magnetic tape out of the cassette shell to expose the affected area gently clean it with warm water and a safe detergent solution, Do not use alcohol or any other chemicals, rinse with distilled water preferably and let dry and try a recapture. Do not attempt this if unsure how to do so, I have to add this warning because some people will always find a way to screw things up.

[zinkatres]

Quote:

Originally Posted by lollo2

Lots of shifted fields and many lines errors. Difficult to fix. Something to try just as experiment:

- from the hardware point of view: disable lineTBC in JVC 7600 and add a Panasonic ES10/ES15; replace Blackmagic capture card with something else

Not a bad idea. I'll see what I can track down as far as the ES10 goes and try a few different combinations, hopefully I can get a cleaner capture that would require less post work.

I did already ditch the Blackmagic card and pulled out my old Pinnacle 710usb on a dedicated windows XP machine. That has been running great so far on other tapes.

Your sample restored clip looked pretty good, Is that a ready to go script or will I need to adjust it over the course of the tape? I wish I understood the avisynth scripting better.

Quote:

Originally Posted by latreche34

This type of severe tape drop outs can be fixed by modern DOC that use adjacent frames to get the clean lines out of them, A script should do something similar as shown above, but nothing magical here, Tape degradation cannot be fully reversed you have to keep that in mind.
Once you do the first pass capture, pull the magnetic tape out of the cassette shell to expose the affected area gently clean it with warm water and a safe detergent solution, Do not use alcohol or any other chemicals, rinse with distilled water preferably and let dry and try a recapture. Do not attempt this if unsure how to do so, I have to add this warning because some people will always find a way to screw things up.

I have no issues disassembling and reassembling the tapes, that's the easy part . Do you have any suggestions on how to set up for the cleaning if it's the entire tape that needs it? I thought about building some kind of 3d printed manual winding mount to wind and clean tape by hand, but don't really want to re-invent the wheel if someone already has a good method outside of using a VCR as I have before.


Again I appreciate everyone's input here

[latreche34]

There are designs and homebrew solutions but VHS tapes don't get that dirty unless the storage conditions are that bad, Fast forward cleaning is almost useless because it will never get the surface cleaned as desired.

If the tape is dirty in certain spots it is much better to address them manually by pulling the tape and cleaning those areas, If the entire tape is affected there is really no perfect way of achieving it without dismantling the cassette and do a slow cleaning and that can take longer especially if there is more than one tape involved.

I'm always tempted to design a system for any type of cassette using automated slow speed cleaning without removing the tape from the cassette. A cleaning solution bath is used, followed by a soft roller sponge pass, rinse bath and finally vacuum dry on both sides of the tape, The system stands vertical with the cassette on top and the tape opening facing down and a tape path contains all the items described above. But I have yet to see the need for such system.

[timtape]

Successfully cleaning videotapes, especially wet cleaning, rinsing and drying is not simple. If it was many more would be doing it and sharing the results.

[zinkatres]

Quote:

Originally Posted by latreche34

There are designs and homebrew solutions but VHS tapes don't get that dirty unless the storage conditions are that bad, Fast forward cleaning is almost useless because it will never get the surface cleaned as desired.

That makes sense. Unfortunately, these tapes were stored in an unfinished cellar/basement for the last 20 years and the rubbermaid container they were in got busted a some point along the way so they were not even sealed up in that regard. My parents are also pretty heavy smokers so there's a lot of build up from that to deal with as well.

So far most of the dirt and grime seems to be on the outside cases. Most of them do show mold on the spools that you can see through the plastic window, but overall the tape surface seems to mostly be in good shape.

If I get ahold of an ES10/15, do you think I would be better off sacrificing a mid range VCR and do an initial capture as is without any cleaning and see where the tape stands? Would this be a better route than starting with an initial fast forward cleaning?

[lollo2]

Quote:

Is that a ready to go script or will I need to adjust it over the course of the tape?

It is just a quick experiment for the first 34 frames.

For the "shifted fields" part you have to find youself all of them, long and boring work. There are specific scripts to detect them, but they do not find all of them, and create many false positive. Let me know if you need one of them.

The "repair" part can be applied as it is to the whole video, although is better to select the range where you want to correct the error.