It is currently 25 Jun 2019, 14:29

All times are UTC [ DST ]


Search found 22 matches

Author Message

 Jump to forum   Jump to topic

Posted: 30 Apr 2013, 06:37 

(more edits coming: Go to page 10 for more current info and sample links, it's improved a lot since these early posts!)

--- original post

I've been working on decoding Laserdisc video directly from the RF signal, bypassing most of the player electronics and hopefully (eventually) getting cleaner pictures.

I've been using an LD-V8000 with the RF test point hooked up to a video capture card using cxadc for raw 8-bit 28mhz capture (which I updated for Linux 3.5) to get the RF signal, and wrote some very slow, dirty, code to decode the laserdisc image, and another program to do time base correction and NTSC color decoding (which amazingly hasn't actually been done by anyone yet, as far as I can tell... I was rather surprised by this)

I've posted the rough source to github: https://github.com/happycube/ld-decode - I did a little bit more work on this right now, before I head off for my real job ;)

There's a .jpg there with color bars with correct colors, albeit fuzzy.

The next steps are to clean up the software decoding (I'm thinking locking onto the analog audio signals, subtracting them from the RF, and then working out some sort of proper PLL for the video rf), combining the NTSC decoding phase using more information from the LD decoding, actually properly adjusting my players, and an AC3 mod but taken from the core RF signal.

(and/or purchasing an adjusted, modded player with a nice pickup... an LD-S2 might be out of my price range, but a CLD-1010 would be nice, and so would a gas tube model for Discovision disks...)

I'm surprised someone in Japan hasn't beat me to this, and if they have I'd definitely like to see it, it's almost certainly farther along than mine is!

(edited for a little more detail and clarity)

 Jump to forum   Jump to topic

Posted: 20 May 2016, 20:10 

Sounds like that's the video level - VR482 on the VDTB board adjusts that, and as long as that bit hasn't been bypassed that should fix it for you.

I suspect the original adjustment expected a voltage dip, and didn't expect such a high quality output circuit. ;)

If you can get a Pioneer test disk in Europe easily enough, that will help you get the levels correct since you can follow along with the adjustments in chapter 1 of the SM.

 Jump to forum   Jump to topic

Posted: 20 Sep 2017, 03:58 

Thanks for the schematic! Hopefully I'll have a chance to build it soon :)

Updated the simple test run - which is ntsc only. The mb and snw.raw are both too small now for the later tbc-ntsc.

updated: ./ld-decoder.py testpat.raw | ./tbc-ntsc - | ./comb-ntsc -O -d 2 - > testpat.rgb

And the PAL version of the simple test run:

./ld-decoder-pal.py -S testpal.raw | ./tbc-pal - | ./comb-pal -O > tmp-pal.rgb
display -size 974x576 -depth 16 -resize 1280x960\! rgb:tmp-pal.rgb

 Jump to forum   Jump to topic

Posted: 23 Sep 2017, 07:29 

This just went up yesterday and hit hackernews... figured some people here might like it :)

https://www.youtube.com/watch?v=0wxc3mKqKTk

(and yes he does have an LD player...)

 Jump to forum   Jump to topic

Posted: 14 Nov 2017, 05:34 

There are three very different 5xx mechs:

- 501/502 are stripped down 701/2's (IIRC at least the 501 had analog TBC)

- 503 is 703 without NR and a few other things... this is probably the one you're thinking about

- 504 is v1.0 of the late player series, very much like the 60x but with Karaoke instead of (basically useless) analog NR. 505/506 are v1.1 with some subtle mechanical parts changes... and AC3 output.

The 104 does have the advantage of being pretty easy to work on, since it's basically the same as the late dual-side players, but without the dual-side mech... https://www.youtube.com/watch?v=FIMBrbRPRro - and picturewise it might actually be quite good if you bypass the LPF after the time base corrector.

And I think the M series takes the cake for flimsyness ;)

 Jump to forum   Jump to topic

Posted: 24 Nov 2017, 05:38 

Close - CLV was the public name for both, but CLV continuously slows down the disk, while CAA uses discrete graduations to reduce crosstalk.

CAA disks can be spotted by the small number of CAV frames (and associated visible sync patterns) at the beginning of a side.

 Jump to forum   Jump to topic

Posted: 06 Jan 2018, 04:27 

If you haven't already recap the LV-20 board!

 Jump to forum   Jump to topic

Posted: 14 Jan 2018, 02:10 

That's a 1981 disk (basically Discovision In Name Only) made by Pioneer Japan - pre-1985 disks from that plant are very hit or miss. Some sides are very nice (I have a discovision dead side with ~37-38dB SNR) - others have much worse SNR and rot.

As for Discovision proper, the quality is hit or miss in different ways (the equivalent of level 1-2 rot should be expected - and on worse disks tracking issues), but the disks generally don't get worse over time.

I wouldn't import either type from the US unless the seller plays it to confirm it is playable with an acceptable-to-you level of rot first, or it's something you really want, quality be darned.

 Jump to forum   Jump to topic

Posted: 25 May 2018, 05:21 

You're right, the values are quite different with 50ohm termination on my 'scope:

- ~200mV ggv 115, ~312 ggv end
- ~408 AVS end

Looks like the attenuatior above a certain voltage lets too much of the audio signals (esp CD/EFM) get through, so I'll hopefully take it to the local Hackerspace tomorrow for a resistor swap.

Other random not-enough-coffee-and-slept-on-my-shoulder-wrong-again thoughts:

- @nemesis - looks like you're making good progress! Something that worked well for me was looking at a line and the two color bursts around it - my current TBC takes 2-3 passes.

- The DE10-Nano has an ARM SOC + FPGA, with HDMI output. With a lot of work it might be possible to do real time video decoding through that.

- I think a data stream format that reflects the EFM subchannels is going to be in order sooner or later. It would be good for [C/L]D-G. It doesn't have to be completely raw EFM but it could have a packet format that covers anything non-zero.

- CDV-R's existed in Pioneer's lab at least. They had special blanks... in theory someone could hack up a burner for it, but it would be a ton of work I doubt anyone will ever do. (but feel free to prove me wrong there!)

- If flash keeps getting smaller/cheaper we're not that many years off from affordable 256GB microSD cards that could store a disk's worth of 8-bit LD RF. Or four Blu-Ray rips. Or who knows how many hours of SD video with modern encoding ;)

 Jump to forum   Jump to topic

Posted: 07 Jun 2018, 17:17 

The people saying not to upscale don't have anything like a 2144. ;) If you're using an LCD or any other modern display you're scaling somewhere along the line. And you've got the best scaler, use it!

 Jump to forum   Jump to topic

Posted: 10 Jun 2018, 23:56 

I was able to reach in with the tray ejected once, but it's pretty easy to remove it, there are two tabs at the back of the tray and then it pops out. Just did that on a DVL-919 today, in fact. It's easier to put the tray back in when the player's unplugged ;)

(BTW The belt in the late CLD/DVL players also drives the tilt through CAM gear magic.)

 Jump to forum   Jump to topic

 Post subject: Re: CLD-R7G
Posted: 26 Jul 2018, 20:38 

210KVT is probably closest. However the late players don't have RF level adjustment. Thankfully donor pickups are pretty easy to find.

 Jump to forum   Jump to topic

Posted: 29 Jul 2018, 18:16 

Unfortunately there's just snippet view here, but I'm pretty sure this is a ref to the disk:

https://books.google.com/books?id=4nQfA ... AQ6AEILTAB

 Jump to forum   Jump to topic

Posted: 31 Jul 2018, 07:33 

I've got the disk! The signal quality is a bit off - there's a sort of internal staining effect that might be affecting SNR/signal quality quite a bit, but it's playable. (I saw a very similar effect last night on a Toshiba/EMI BGV demo disk which had a stained cover - I think it might be a water stain effect. water-rot?) I've got a rough DVD-R .zip I made using a Panasonic DVD recorder that's on it's way to Google Drive right now. It'll be a while until I have something better.

Link here (for a few days @ least): https://drive.google.com/open?id=1trZxSpJSEcRcfy4n-l2LGxMHvDKfI9xR

And yeah - I think Disclord would've loved this too.

Also, a couple of 1981 era MIT papers I found (not about this, but other stuff they were doing):

http://dspace.mit.edu/bitstream/handle/1721.1/15714/09134278-MIT.pdf?sequence=2
https://www.media.mit.edu/speech/papers/1981/bolt_SIGGRAPH81_gaze_orchestrated_dynamic_windows.pdf

 Jump to forum   Jump to topic

Posted: 19 Aug 2018, 15:59 

I think the 2009 batch of DVL's were the last ones Pioneer could economically build - there are several LD-only IC's which even if they still were available new wouldn't meet ROHS specs.

You might, just might be able to build 1000 for $1000 (edit: each!), but there probably aren't enough people willing to pay that much.

In a decade or two someone might bother to build a one-off machine to scan a Laserdisc and do the equivalent of a Kyroflux - assuming I'm still around I'd be up for helping with that. ;) (and if I'm not hopefully the code I've written will be useful regardless...)

 Jump to forum   Jump to topic

Posted: 05 Sep 2018, 06:53 

http://www.laserdiscarchive.co.uk/laserdisc_archive/philips/magnavox_vc-8010/magnavox_vc-8010.htm - I think that's the Pioneer-built unit you're thinking about. IIRC, Disclord and blam1 both liked it. Word is - and I forget the source - that Magnavox destroyed their own players after starting to sell these, to avoid warranty claims on them.

@rein-o - you're thinking of the 660 which is a dumb enough player to not look at the vertical blank data - and hence not worry about a whole group of Discovision mastering flaws ;)

On the other side of the pond, a slightly less crap version was later sold as the VLP 600 and 700. The Philips people could actually repair them, but they're still rather dodgy.

 Jump to forum   Jump to topic

Posted: 11 Oct 2018, 16:55 

Thanks :)

There's no magic in the CD RF, just metadata and error correction bits. With those you can make sure you have the exact bits that are intended to be there, which depending on one's setup can be difficult to do. With a RF sample you know when it skipped a sector and/or corrupted some data which a raw digital output stream won't say.

The variation in players is really how the digital data is processed - some cheap/early CD players do a bad job of making it analog, and a good CD player with oversampling produces a better waveform. There's more to it than that but eh, I'm not awake yet.

Also lossless compression (FLAC) is a thing and is even useful for Laserdisc RF signals - just not implemented by lddecode yet :)

 Jump to forum   Jump to topic

Posted: 14 Oct 2018, 23:07 

most LD's (that weren't pitched etc) are as good as the day they were made... but the Domesday disks were made at the PDO UK plant, so they're screwed in a decade or so as the edgerot creeps in.

 Jump to forum   Jump to topic

Posted: 15 Oct 2018, 00:06 

The problem is there isn't known to be one fully working one in existence. :( And even if there were shipping it to Japan might damage it.

Later production ones were less bad and I think I've seen one video of one *sort of* working but not properly.

 Jump to forum   Jump to topic

Posted: 26 Dec 2018, 04:02 

Most of the discussion is on the #domesday86 freenode channel now, along with the FB group, so while this thread's been silent there's still been plenty of good progress :)

New software release (rev3) today, which includes Simon's tools.

PAL quality is greatly improved: https://www.youtube.com/watch?v=qH0jnMLEJ64 is a 1-minute Domesday intro. Really pushing what the source material: a PDO UK disk from 1986... at the beginning of a CAV disk where higher frequencies are truncated. A MUSE PAL player would really help here, but they don't exist.

https://www.youtube.com/watch?v=qH0jnMLEJ64

 Jump to forum   Jump to topic

Posted: 23 Feb 2019, 20:01 

Phase errors are when the time base corrector is slightly off, causing the colors to wind up a bit off.

And that's an impressive capture chain, definitely worth doing the mod for! (As long as the phase offset is the same in the color burst, it'll correct it.)

 Jump to forum   Jump to topic

Posted: 19 Mar 2019, 05:31 

Did some digging just now - the second edition of the DVD Demystified disk has them - https://dvddemystified.com/files/DVDDemystifiedBarcodes.pdf

There's also a flag set to enable the compatible mode in the DVD-V player http://qa.pioneerelectronics.com/pio/pe/images/portal/cit_3424/28121644PIB152513_V72_VidTxtRet.pdf and http://qa.pioneerelectronics.com/pio/pe/images/portal/cit_3424/27945802LB2_vs_DVD.pdf shows the audio track mapping... I don't think there are more conversion specs available, and the information may have been lost with the closure of Pioneer's media division.

7100/7300 are PAL versions of the 7200/7400.
Page 1 of 1 [ Search found 22 matches ]


All times are UTC [ DST ]


Jump to: