Current status of ld-decode / vhs-decode (true "backup" of RF signals)

[Titan_91]

I never touch my tapes either. I thought the article only mentioned rotating the reels but maybe I read it wrong.

One of our 30 year old home videos was jammed in a machine about two decades ago. If I remember correctly we spooled out some tape and let it kind of relax. Then after some time we wound it back up and put it on a shelf where it sat for 20 years since. It plays but it the sound of the wrinkled section moving past the head drum makes you cringe.

I'm glad I captured the RF as the tape really isn't holding up well at this point. I also don't want to screw up the heads in my modded VCR.

[oln]

Latest git now decodes NTSC. Didn't finish up color for it yet.

Screen from vaporeon800's spiderman sample.
The odd vsync area is due to head switch being set very low.

[Attachment 51709 - Click to enlarge]

The zig-zag patterning is chroma interfering with the luma signal I think. It was quite prominent here, maybe due to the frequency spectrum of theluma signal being 400 KHz lower down on NTSC (525-line strictly speaking) than PAL. Will need to work on the filters to reduce it, and ill have to see if it is as prominent on other recordings or if some of it is due to bad Y/C separation during the tape recording process.

[lordsmurf]

Was wondering if someone could get one of these devices and reverse engineer them for VHS-decode, Nothing better than re-inventing the wheel:
https://forum.videohelp.com/attachment.php?attachmentid=47487&d=1544778291

The DPS dropout compensation doesn't work well at all.

[Titan_91]

Latest git now decodes NTSC. Didn't finish up color for it yet.

Screen from vaporeon800's spiderman sample.
The odd vsync area is due to head switch being set very low.

[Attachment 51709 - Click to enlarge]

The zig-zag patterning is chroma interfering with the luma signal I think. It was quite prominent here, maybe due to the frequency spectrum of theluma signal being 400 KHz lower down on NTSC (525-line strictly speaking) than PAL. Will need to work on the filters to reduce it, and ill have to see if it is as prominent on other recordings or if some of it is due to bad Y/C separation during the tape recording process.

Awesome!! I cloned the latest release and tried it on one of my 35.8 MSPS samples. It's not finding any sync pulses for me. It also can't find any in the Spider-Man sample. What am I missing? Do I need to build something first like with ld-decode before I run the script?

python3 vhs-decode -n --cxadc3 NTSC_test_pattern.raw NTSC_test_pattern.tbc

[dellsam34]

I'm glad I captured the RF as the tape really isn't holding up well at this point. I also don't want to screw up the heads in my modded VCR.

The quality of picture of the VHS decode is not up to the standard yet so you may want to hold to that tape for now.

[oln]

What am I missing?

Try renaming the file extension to r8, otherwise it's interpreted as 10-bit. Need to add an option to set bit depth/type, but haven't done that yet.

[Titan_91]

The quality of picture of the VHS decode is not up to the standard yet so you may want to hold to that tape for now.

I still have it just in case, but the important thing is we have RF captures now regardless of the state of vhs-decode.

Try renaming the file extension to r8, otherwise it's interpreted as 10-bit. Need to add an option to set bit depth/type, but haven't done that yet.

That's probably it. Will verify.

[Titan_91]

Renaming the files was it. Here's what I have from ld-analyse so far:





I tried to load in a short clip of Arthur to compare 28.6 and 35.8, but a lot of frames were dropped and ld-analyze shows a wrong output. The others I have decoded fine for how the software is right now. Oln do you want the Arthur samples to help troubleshoot?

[MrKenSan3]

Hello folks,

I'm new here and just catching up on what you all are doing. Looks pretty amazingly cool!

A friend has been working on digitizing some old VHS tapes, and together we've been reading through threads here and on http://www.digitalfaq.com/ to learn best practices and understand more about the hazards involved. When we learned of the Domesday Duplicator and the branch dedicated to decoding VHS RF signals, we thought this might be a great way to do something even spiffier than spend gobs of cash on gear to do the few tapes and and I have. It's understood that this is at an early stage now, but it seems like once it matures, there is a lot that could be done to create quite high fidelity results since we have all the data and can look forward and backward in time. The potential for very fine quality like in the main Domesday project is promising.

The hardware for the Domesday Duplicator looks like a very nice implementation to allow for the collection of the LD RF signal. That it's being repurposed for VCR recordings is really cool.

My friend pointed out some speed bumps for people who want to try this out for the RF from the VCR head...
1. Audio capture sync with the video. Since the video RF is captured independently,
the audio track has to be manually synch'd later.
2. Many of the people who would like to use the decoder by ole via the Domesday hardware
don't always have the means to easily assemble the HW on their own.

My background is electronics, and I have a bit of spare time at the moment. If there is sufficient interest, perhaps we could enhance the Domesday hardware for VHS capture...

* By adding a capability to capture audio in parallel to RF/Video info.
This would require:
- Some FPGA and Driver work for the capture
- Audio capture HW Interface (a little electronics design)
- Adaptation of the decoding SW to parse a new data stream format

* With sufficient interest, potentially we could assemble some number of these boards professionally.
There are a number of local assembly vendors near my location, so we could have boards
(which might also include an optional audio interface) fabricated and populated to make
this technology more accessible to a wider audience.

So, wanted to pass the idea by the folks here to see if there might be some interest?


Thanks for the great and very interesting work!

[oln]

We've talked a bit on discord/IRC about an updated Domesday duplicator with more channels for hifi, head switch signal, other audio etc. I think simon has even designed an updated one. Feel free to hop on discord to talk about a designs though, could use someone with electronics knowledge. Manufacture would also be nice, not everyone has the means and equipment to populate a board themselves.

Oln do you want the Arthur samples to help troubleshoot?

That would be great yeah.

[Titan_91]

An updated more accessible Domesday Duplicator would be great. For me the appeal would be multiple channel input for HiFi and ability to use USB3 as opposed to an old PC for PCI. In addition if PC-less storage to a USB3 hard drive could be done that would be awesome. That and maybe additional RF amplification and impedence matching for low level VHS RF signals.

[dellsam34]

The potential is there but the only way to find out for sure is through a crown funding account, Maybe one of the mods here or at digitalfaq can setup that page so we can go there and contribute, I will be the first to chip in.

[Titan_91]

Arthur samples. For me these turned out skewed with what looks like the wrong fields being combined. Also the dropped frames were my fault, I was using a blank section of the captures before with no picture information.

https://www.mediafire.com/file/4dfn28hcstnqks4/Arthur.7z/file

[Gibson's Squares]

This project is so exciting. I only just learnt about LD-Decode today when I watched a great hour-long interview with two of the guys working on the Domesday86 project. (https://www.youtube.com/watch?v=klK4UZ5nlqs)

Regarding the comments earlier about the results being comparable with (or currently worse than) traditional digitizing methods, I think the bigger picture needs to be looked at. Ie if/when this VHS-decode system meets current digitization standards it will be a much much cheaper solution for many folks than buying/maintaining hi-end VCRs, TBCs, capture cards etc. One of the Domesday86 project objectives (that this whole RF recovery thing sprang from) is for the gear and software to be very cheap and open source.

Also, thinking even more widely, once acceptable quality is achieved for VHS recovery it can also be applied to other formats for which there is no (or even scarcer than VHS) high-end gear, eg Betamax, V2000, N1700, etc...

I wish I could fast forward to the end of this story. So excited to see how it turns out.

[Titan_91]

Coming back to the audio discussion, I have one tape with warbling Hi-Fi audio. The volume frequently goes in and out. I think I'll try that with Zcooger's GNU Radio flow chart to see if anything can be done in software.

[Cornucopia]

not sure what you mean by warbling in ref to hifi audio, but most issues w hifi audio are due to mis-tracking, which should be fixed before capture. that kind of thing is almost always not fixable in post.

Scott

[fsquared]

So here's what I want to do:

- Get multimeter

- Solder some RG58 to the pin

- Put the RG58 through the S-Video luma pin

Is this how I should go about doing it?

[Titan_91]

You will need a ground connection. Connect the s-video port ground to the VCR's isolated ground.

[oln]

I tried to load in a short clip of Arthur to compare 28.6 and 35.8, but a lot of frames were dropped and ld-analyze shows a wrong output. The others I have decoded fine for how the software is right now. Oln do you want the Arthur samples to help troubleshoot?

Latest git decodes them fine now, needed to increase the maximum offset the code that looks for the hsync line starts allowed (there won't be a lot of deviation on that on laserdiscs obviously). The head switch being in the vsync area made the difference between one hsync position and the next compared to the normal line length very large and messed up he rest of the field.

[fsquared]

You will need a ground connection. Connect the s-video port ground to the VCR's isolated ground.

Oh, right. How could I forget about the ground? [emoji2357] Should I also use RG58 for that?

[Titan_91]

Latest git decodes them fine now, needed to increase the maximum offset the code that looks for the hsync line starts allowed (there won't be a lot of deviation on that on laserdiscs obviously). The head switch being in the vsync area made the difference between one hsync position and the next compared to the normal line length very large and messed up he rest of the field.

Yep works great for the A Bug's Life sample too. Thanks for fixing that so quickly.

Oh, right. How could I forget about the ground? [emoji2357] Should I also use RG58 for that?

Since you already have it, might as well to best shield the ground as well.

[dellsam34]

Latest git now decodes NTSC. Didn't finish up color for it yet.
The zig-zag patterning is chroma interfering with the luma signal I think. It was quite prominent here, maybe due to the frequency spectrum of theluma signal being 400 KHz lower down on NTSC (525-line strictly speaking) than PAL. Will need to work on the filters to reduce it, and ill have to see if it is as prominent on other recordings or if some of it is due to bad Y/C separation during the tape recording process.

Any progress on eliminating that zig zag pattern? By the way this thread should be a sticky, can one of the Mods look into it?

[oln]

Haven't looked into it yet, been trying to get chroma to work first.

[Titan_91]

Haven't looked into it yet, been trying to get chroma to work first.

Is there anything else I can provide or test to help? Thinking about hitting up Goodwill again for any unusual tapes.

[oln]

Maybe it was posted already, but do you or anyone else have captures of NTSC plain colors or colorbars? Trying to get NTSC chroma to decode correctly.

[Zcooger]

NTSC samples were put under this link: https://www.mediafire.com/file/p8gzklumiv3d8z1/CXADC_Zenith_XBV342_35.8MSPS_NTSC_test_...ptures.7z/file

[dellsam34]

Is there a webpage showing a step by step hardware and software setup specifically for VHS? Can any of the members make a thread for this purpose?

[Titan_91]

Maybe it was posted already, but do you or anyone else have captures of NTSC plain colors or colorbars? Trying to get NTSC chroma to decode correctly.

Hopefully the samples I posted will be useful.

[oln]

Yup, made a little progress on NTSC color today. I spent way to long not realizing that the color under frequencies of PAL and NTSC are not the same, but now that I accounted for that the phase of the color signal was as expected. Still needs way more work to look good though.

[dellsam34]

If there is a time where this project could have a big boost to materialize it is now where everyone is indoor due this pandemic.