2000 old VHS tapes to convert?

[shanedownfall]

This plea for help feels fitting for DFAQ. Anyone based in the UK that can help out?

https://takes.jamesomalley.co.uk/p/my-vhs-nightmare

[ChrisW6ATV]

That "recover and archive Teletext" project sounds very cool.

When I visited the UK and Ireland in 2000, I did not watch much if any TV but I loved Teletext and looked at it constantly. As we stayed in different hotels with different TV models, I was fascinated by the range in quality/features of the Teletext functions on each set. Some were very convenient, others basic and harder to use efficiently.

I hope those tapes go to good use in the Teletext project.

[latreche34]

It has to be a very different way to process 2000 tapes, These processes of teletext and VHSdecode are for very specific tapes with historic value, not for an average tape especially when the lot number is that huge. An arduino module that can connect to the RF points of the VCR, processes and decodes the data from those specific scan lines and spits out BMP images into an SD card, The process is real time so each tape takes two hours for SP or more for low speed recordings. The process can be multiplied by having more VCRs and modules.

If video is also to be captured, The best way with an acceptable quality is to get a legacy PCIe capture card with MPEG-2 chip on board, So the video is saved directly as 576i MPEG-2 4:2:0, Some USB devices also were made with a MPEG-2 chip such as Pinnacle and I believe Roxio and Plextor made some models, DVD recording is an option but you don't want to waste discs plus the process is twice as long as it requires ripping the discs.

[lordsmurf]

Quote:

Originally Posted by latreche34

process 2000 tapes

People never understand how long that process is. The actual capture time may be 1:1 (ie, 2 hours of programming = 2 hours of conversion time), but you have file movement time, organization time, error time, etc. It quickly doubles and triples. If we assume at least 90 minute of SP per tape (extremely likely), that puts us at 3000 hours. But at least double that, so 6000 hours. Those are extremely conservative numbers.

6000 hours = 750 8-hour days
5-day work week = 260 days
You'd literally have to work full-time, for 3 years, to convert it all.

Adding a VCR doesn't halve the time, because there's still per-human hour limits. But let's assume so, just for fun. So if you spin up 3 units working in tandem, it'd "only" take 1 year of full-time 8-hour 5-day weeks.

Even if you added 5 volunteers, you'd still have 8 hours per week of converting tapes, for a year.

Nobody has time for this. (Well, archivists do, but they get paid. That's the actual 9-5 job.)

Those 2000 tapes were a mistake, because nothing was ever labeled. I have at least 2000 tapes in my collection, but I have logs of what's on each tape. I never have, and probably never will, convert all of it, because it's not all convert-worthy.

This is hoarder mentality. It's not history, or archiving, or anything else. I've dealt with this in the hobby community for 20 years now. Those people burn out, never finish what they start. What they do do is rushed sloppy quality, and it's just a waste of time, too crappy to enjoyably watch.

If enough people have enough time to waste, then only then is it feasible. Of course, what are the odds that all of the volunteers have quality standards? I've always found that too few do. Some just "want to be a part" without having any idea what they're doing.

Yes, I just pissed in Cheerios.

Neat idea, but sometimes ideas need a cold dose of reality.

[latreche34]

Teletext can be fun but like I said, if an automated system is made where human intervention is just to remove and insert another tape, Then the teletext pages are all stored in a file by page number and an app can be made to browse those pages like you would normally do on the TV with the remote control, This is a European thing by the way, the US did not have teletext, just caption.

Video on the other hand is not worth much especially if the tapes are recorded in 6 hours per tape mode, the quality is just not worth it, most movies are scanned @ 4k and up already, why capture it on VHS in SLP quality, Commercials maybe, but searching 6 hours files hunting for commercials is time consuming.

Having said all that, Even if this is achievable, who is going to watch those captures, I see websites hosting some of those so called VHS rips and I can't stand 2 seconds watching them due to the very bad quality on both the playback and capture levels.

[Blazil]

Hello all,

To start off, I would like to say that I am new to this forum and don't want to overstep any bounds. The reason I made an account is because I found this thread by chance as I was searching the internet for something else, namely fellow teletext archivists.

I could not help but notice that the idea of archiving teletext is quite simply put squashed in this thread. And I don't understand why.

Firstly, I would say people should be free to decide what they spend their time on as a hobby, even if it is useless. Also, I know a lot more useless hobbies than archiving teletext.

Secondly, most teletext broadcasts were never saved by the studios who made them. That means that any random VHS tape out there could hold some teletext information that doesn't exist in any other copy anymore. Also, VHS wasn't supposed to have a high enough bandwidth to be able to store teletext. So any information that is recovered in this way is basically "brought back from extinction."

Quote:

Originally Posted by lordsmurf

This is hoarder mentality. It's not history, or archiving, or anything else.

I disagree, obviously.

Quote:

Originally Posted by lordsmurf

What they do do is rushed sloppy quality, and it's just a waste of time, too crappy to enjoyably watch.

The teletext archive (https://teletextarchive.com/) does contain examples of some questionable quality of pages recovered, but the process can't always fully compensate for bad tape quality. Bringing back stuff that is not supposed to exist anymore comes with a price.

I would say this of the process of recovering teletext: you can learn a great deal of skills, ranging from repairing VCRs to programming python to learning how to set up a proper archive.

Having an interest in older technology, I always felt there was nothing I could do with my old VCRs and tapes, because I didn't really care about some old TV program. But somehow when I learned of the teletext recovery process, everything just fell in place. I could satisfy my interest in history and in the old technology, have a reason to have my VCRs running, find a purpose for an otherwise useless 10 year old pc, and actually have something be produced as a result.

You can argue that the average teletext broadcast was nothing special, and maybe this is true. The actual information in there isn't always super interesting. News of those eras is likely archived in different manners. I would say a teletext archive is an archive of culture, and specifically of a culture that could in some ways be considered a predecessor to the internet.

And you know what the fun part is? (Yes, surprise, there could be fun in this). It is always fun to me not knowing what I am going to find: a broadcast from an unusual station, or a broadcast more than 40 years old, or maybe some minutes at the end of a tape of a half-taped over broadcast, still allowing its teletext to be recovered.

I would say that going through 2000 tapes will take a long time, but I will also tell you this: having recovered the teletext of 200 tapes myself from my own collection within a year, I actually felt I had to slow down at a certain point lest I be running out of tapes. And I do this only as a hobby outside of a vastly different day job.

[dpalomaki]

Another factor to keep in mind is the time between needed major servicing of the VCR(s). things such as belts, friction material, lubrication, cleaning, alignment, etc

For example, the Panasonic Ag-1980 and AG-1970 recommend upper cylinder (heads) replacement every 1000 hours. Perhaps over kill but something to keep in mind given that most machines available today are aged and parts scarce.

That is not to say doing 2000 tapes is a dumb idea (it is after all a personal choice) but recommend that the project prioritize to do the most important stuff first.

[lordsmurf]

Quote:

Originally Posted by dpalomaki

That is not to say doing 2000 tapes is a dumb idea (it is after all a personal choice) but recommend that the project prioritize to do the most important stuff first.

That's the problem here, with this exact "call for help". There's no logs. Literally no idea what on the tapes. It'll be "quicker" to just convert it all, and look later. But again, 2000 tapes will take literal years.

Uncatalogued VHS is rarely a "treasure hunt", but more like wading through a dumpster. I've seen these sorts of situations before, for decades, and the tapes generally end up trashed/binned. In years past, my pre-MS days, before health issues, I'd take on "mystery tape" projects, generally just a sampling of 10 tapes max. It was 99%+ crap on the tapes. Sometimes you'd find neat interstitials, sometimes neat commercials. But mostly just worthless footage, most of it now available commercially. I always hoped for 80s toons WOC, but it rarely happened.

Anybody that wants to enter that sort of "unknown tape" hobby just needs to know reality, know what to expect. Literally have zero hopes.

[latreche34]

It's hard to sort through teletext contents, it takes more time per tape to go through hundreds of pages of teletext than to hit capture and search later, It's like microfilms, take a snapshot and worry about the contents later, because it takes time to go through the contents of a microfilm to decide whether to press the button or not. Besides Who decides what's important in teletext? jokes, sayings, news, trivia games, sports, graphics, it changes from one person to another.

[Blazil]

Quote:

Originally Posted by latreche34

It's hard to sort through teletext contents, it takes more time per tape to go through hundreds of pages of teletext than to hit capture and search later, It's like microfilms, take a snapshot and worry about the contents later, because it takes time to go through the contents of a microfilm to decide whether to press the button or not. Besides Who decides what's important in teletext? jokes, sayings, news, trivia games, sports, graphics, it changes from one person to another.

Which is exactly why the form of teletext capture used by the people with the 2000 tapes and by myself involves recovering the entire teletext broadcast. The only thing needed to speed up the process is knowing where one recording begins and the other ends.

Sometimes a tape contains 2 TV recorded movies, which you can easily distinguish after digitizing the analog video first. A teletext capture of a single broadcast takes around 15-30 minutes and requires some processing afterwards, but that is just putting in commands in a terminal. You could even use a script for that.

Sometimes its a tape with 40 brief fragments of soccer match summaries in which case the entire tape length better be captured to make it more efficient. The teletext data can always be sorted out later, and deleted if you really must save those few MBs of space.

[latreche34]

So it is not possible to just capture the teletext on the fly while the tape is playing without having to save the video to the HDD?

[Blazil]

Quote:

Originally Posted by latreche34

So it is not possible to just capture the teletext on the fly while the tape is playing without having to save the video to the HDD?

It is possible to capture the teletext on the fly without first digitizing the analog video. This is that step I meant that takes 15-30 minutes for one broadcast. Basically 15-30 minutes is taken real time, so that the data can later be used to interpolate complete pages from the many partial copies in that 15-30 minute stream.

But in order to mechanically spare the VCRs from continuous winding/rewinding/playing/stopping as one searches for where various recordings start and end, I prefer to first digitize the analog content so that I can skip through it on a PC.

I use a broken DVD recorder for this, because it still functions as a rudimentary TBC and readily converts composite video into component video. Then I use one of those gaming HD MP4 recorders to record the component video.

[lordsmurf]

Quote:

Originally Posted by Blazil

I prefer to first digitize the analog content so that I can skip through it on a PC.
I use a broken DVD recorder for this, because it still functions as a rudimentary TBC and readily converts composite video into component video. Then I use one of those gaming HD MP4 recorders to record the component video.

That method will create a lousy quality video conversions (high compression, messes up the interlace), but I'm guessing you only do it for browsing purposes, correct? In order to find the teletext topics you want?

[Blazil]

Quote:

Originally Posted by lordsmurf

That method will create a lousy quality video conversions (high compression, messes up the interlace), but I'm guessing you only do it for browsing purposes, correct? In order to find the teletext topics you want?

What if I don't use it just for browsing purposes?

I don't think that recording in 3 Mbit H.264 results in lousy quality at all. For one, the result has a very good contrast and detail level, even getting all the analog noise. Also it never results in audio de-synchronization or frame drops. I suppose if you take individual frames under a magnifying glass, you might be able to spot artifacts from de-interlacing, but I don't think they are that noticeable.

It is a stable and reliable method. I am not claiming it is the best method. But its quality certainly beats recording directly onto PC via TV capture card.

I am curious now what you think would be a better method.

[dpalomaki]

Quote:

I don't think that recording in 3 Mbit H.264 results in lousy quality at all.

Bottom line boils down to who is the client (you? or someone else?) and what is their expectation. That is the opinion that matters most.

The information provided on this site will help you achieve the best practicable end result. Only you can decide whether or not it is overkill for your purposes.

[latreche34]

Quote:

Originally Posted by Blazil

It is a stable and reliable method. I am not claiming it is the best method. But its quality certainly beats recording directly onto PC via TV capture card.

No it doesn't, not even close, Try capturing lossless AVI using a capture card. It is even very useful for teletext, caption and PCM audio encoded video tapes, because one, it captures interlaced (the fields are kept separate for further analysis), second the luma is at its highest quality which is where the data is stored, and last, the chroma is kept pristine if the video is to be saved.

[lordsmurf]

Quote:

Originally Posted by Blazil

What if I don't use it just for browsing purposes?

Then it's non-archival butchered (non)quality video.

Quote:

I suppose if you take individual frames under a magnifying glass, you might be able to spot artifacts from de-interlacing, but I don't think they are that noticeable.

No. Half-blind grandmas see deinterlacing noise. Literally.

Quote:

I am not claiming it is the best method. But its quality certainly beats recording directly onto PC via TV capture card.

No.

Quote:

I am curious now what you think would be a better method.

Lossless captures -- later compressed to 4:2:2 high-CRF H.264, if space is really an issue (and in an era of 22TB drives costing $300 USD, it's really not).

At worst, MPEG-2 captures, with 15mbit minimum. 4:2:0 is lossy, but not terrible.

- Always leave interlaced.
- Never H.264 capturing, it creates artifacts and mush, at any bitrate.

[aramkolt]

If it was me, I'd probably have to make some compromises over capturing one tape at a time and using capture PCs. I could see a setup where 10 VCRs (with line TBCs) are going at once each to their own SDI or Component outputting frame TBC each going into an AJA Ki Pro capturing in anything from ProRes Proxy to 422HQ.

You'd also have to come up with some way to automate the actual stopping of each capture on each AJA device when the content is done too, which generally would be fast forwarding to the end of the tape (usually the counter stops if its unrecorded after the end of the content if it's unrecorded after that) reset the tape counter, then do a full rewind to show how many minutes are on any individual tape. You could mark the tapes in advance with how long they are.

From there, you'd need a digital timer that you could start for each capture box that would trigger the AJA to stop recording when the timer runs out. They do have RS422 communication protocol and also support LANC protocol. Guessing there has to be a sort of LANC controller that you program in a countdown time and then it emulates pressing the "stop" button once it ends? If you didn't want to mess with all that, you could probably 3D print some sort of servo controlled actuator that physically pushes the stop button on the AJA after a kitchen timer runs out.

If you did all that, you could probably capture at least 20-30 tapes a day by yourself, with probably 1 hour of actual time required by you. Just start one batch of 10 tapes when you leave for work, one batch of 10 when you get home, and one batch when you go to bed.

It'd be pretty expensive to buy all that stuff, but you'd be done in something like 3 months with the capture part anyway.

Poor man's way of doing this without TBCs with a bit less quality would be to use consumer VCRs + DV capture. DV isn't ideal, but should blow the pants off of 3Mbit H264 capture and it will preserve the interlacing. Most DV conversion chips have some degree of line TBC effect and are relatively tolerant to consumer video sources. On a budget, the ADVC-110 (or 100, or 55) plus older Windows laptops (you'll want them to have SATA drives for ease of 500GB or more SSD installs) with firewire ports would be the way to go. Dell D630's or around there should be well suited, run XP, and then capture with WinDV. MPEG2 could also be done with a similar laptop or a device that supports MPEG2 capture. Those Dells also can support a second hard drive via a CD drive caddy that installs in place of that drive which may be handy as well in such a setup. I believe most MPEG2 hardware conversion chips also do a decent job with line timing errors as well, but I haven't really tested them myself yet.

[lordsmurf]

Quote:

Originally Posted by aramkolt

If you did all that, you could probably capture at least 20-30 tapes a day by yourself, with probably 1 hour of actual time required by you.

Nope, not a chance. You're not accounting for everything in that simple calculation.

[latreche34]

DV is a good candidate for this type of data extraction, It fully preserves the luma, which is where the data is stored. But the automation part is not that easy, at least for a regular person. I can see potential of automation using the sync pulses, with an arduino module connected to the sync pulse head of the VCR and into USB and stops the capture when no pulses are detected, the VCR will keep playing the tape until the end and rewinds it automatically.