It is currently 19 Mar 2024, 08:53




 Page 17 of 31 [ 606 posts ]  Go to page Previous  1 ... 14, 15, 16, 17, 18, 19, 20 ... 31  Next
Author Message
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 26 Dec 2017, 11:10 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
Short answer: It's digital, plug in the DA-1

Long answer:
The player band-passes the modulated RF signal to separate the digital audio part of the RF signal from the rest of the RF and then outputs it to the EFM port (the AUDB module in the player handles this - service manual page 25 of the PDF).

The AUDB module also decodes the digital signal and outputs the status of the L and R digital audio channels on the EFM connector too (pins 3 and 4).

So you have:

Pin 1: High if the disc is PAL and low if NTSC
Pin 2: GND
Pin 3: Audio L status (high if on and low if off)
Pin 4: EFM (this is the actual 'digital' signal)
Pin 5: Audio R status

Interestingly the EFM isn't really a digital signal; it's band-passed RF. This still requires some form of demodulation to get the data back followed by a DAC to get back the analogue sound; which is most probably what your DA-1 is for. In the player IC306 demodulates the RF (the "EFM decoder IC") and then IC308 acts as the DAC).
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 26 Dec 2017, 11:31 
Advanced fan
Advanced fan
User avatar

Joined: 08 Dec 2015, 01:43
Posts: 549
Location: Japan
Has thanked: 26 times
Been thanked: 56 times
thank you for the detail !!!!
That's very clear to me now !
signal passing through EFM is like a raw signal (RF ?) which is transformed to an other signal, some sort of digital (IC ?) by the player. DA-1 can regognise this signal and change it to Analog ?
_________________
LD-V4300D (PAL/NTSC)
CLD-959 (NTSC)
CXUHD (DVD/BD/UHD/SACD)
Lumagen 2144 (scaler)
Collection
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 26 Dec 2017, 11:38 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
Quote:
DA-1 can regognise this signal and change it to Analog ?


Yes, that's correct.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 26 Dec 2017, 12:06 
Advanced fan
Advanced fan
User avatar

Joined: 08 Dec 2015, 01:43
Posts: 549
Location: Japan
Has thanked: 26 times
Been thanked: 56 times
Thank you ! I will ask you many thinggs again when I will clean my player !
_________________
LD-V4300D (PAL/NTSC)
CLD-959 (NTSC)
CXUHD (DVD/BD/UHD/SACD)
Lumagen 2144 (scaler)
Collection
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 30 Dec 2017, 17:09 
Serious fan
Serious fan
User avatar

Joined: 26 May 2014, 19:25
Posts: 236
Location: United States
Has thanked: 65 times
Been thanked: 92 times
The work that has been done on this project is phenomenal. Sheer magic for most of us. The results look nothing short of incredible. Between the software development and the hardware development you two are really making the rest of us look dumb and lazy ;)

I, for one, have not yet tried this yet as there are too many steps involved and it is not yet quite perfected. Soon though, when the process becomes a bit more streamlined I will definitely look into this as a archival method.

But seriously, thank you for your hard work and time spent on this. Please keep going and keep sharing the results here with us.
_________________
Kevin
LD-S2|CLD-D704|CLD-D406|DVL-V888|LX-900U|Crystalio II|Yamaha APD-1|Sony XBR55X810C
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 30 Dec 2017, 17:20 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
Thanks for the kind words :) Happycube is working hard to get the ld-decode working with the new hardware (there are now two boards, one in Sweden with me and one in the USA with Happycube) - the software for the duplicator is gradually becoming stable and more user-friendly too.

Just for fun I ran some frames from the PAL Jason and the Argonauts disc through the process end-to-end (Pioneer LD-V4300D RF -> Duplicator -> ld-decode -> photoshop cropping and filtering -> Premier Pro video editing and encoding) and popped the results on YouTube as an unlisted video). Bearing in mind that things are far from perfect in the decoding (at the moment), the results are still extremely impressive...

https://www.youtube.com/watch?v=o1HeOj0pJ_s

Once we are a little closer to the calibration required for the BBC Domesday Project discs (they are a lot older and degraded than a modern disc like Jason) then I will start to run more test captures.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 31 Dec 2017, 11:16 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
I just came here to see if by any chance there was a minor update to this project, and I can see there's been a massive one! As I mentioned a few pages back, I'm working on getting full rips of games released for the Pioneer LaserActive. I'm very interested in this new FPGA-based data capture solution, and I want to join in with testing and to help develop this and move it forward. I'm a software engineer by profession, and I can actively help to improve the decoding software. I ran into a wall with the hardware capture process though, not being able to get the original capture solution proposed here up and running, and then running out of time to do more work on it due to life getting in the way. I'm back up though, and looking to finally finish this project.

What I want to know is, can I get my hands on one of these prototype hardware boards? I know it's open source hardware, but since it's very much in the prototype stage I'd feel more confident having a known working one supplied rather than "going it alone" as it were, especially as I don't have another working solution on hand to troubleshoot against. All my problems so far have been in simply generating that basic analog RF waveform, which has meant I haven't been able to do anything on the software side. Do you have any spare boards simoni? If I can get my hands on one, I'll be able to help get the software side where it needs to be. I've got a strong interest in the EFM decoding of the digital data too. I've already got a capture solution for that in place that allows me to get 2448 data/subcode data per sector in uncorrected form, for the full digital disk area including the lead-in with the TOC data. I can use that proven ripping process to build an EFM decoding process to extract the raw frame data for the digital data, and compare with my existing results. I'd also work on porting/cross compiling the toolchain to Windows, to make it more accessible. Let me know if I can hop in on this, I'm very keen to move it forward!
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 31 Dec 2017, 14:20 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
Thanks for the offer of help nemesis, it's highly appreciated.

Right now I don't have any spare boards (and it's quite expensive in time and money to make them and ship them around - given that it's a prototype which isn't really proven yet (although the current results look very promising)).

I'd rather be a little more sure of the hardware before making more so, as a compromise, if you let me know what type of discs are most interesting/useful for you to work on the decoding - I can prepare some RF samples for you and send you a link to download them from my owncloud server (as the samples are usually fairly large). I have a selection of PAL and NTSC laserdiscs with digital audio which should be suitable for use in testing EFM decoding (and soon Happycube will be able to make samples too). I realize this isn't as much 'fun' as a real duplicator board but, if we make some progress in the decoding, then by that point I'll be in a better position to supply a board without the current risk of wasting both your and my time/funds :)

Does that sound like a workable starting point? If so, let me know what you need and I'll make up some samples.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 31 Dec 2017, 18:59 
Absolute fan
Absolute fan
User avatar

Joined: 18 Apr 2012, 18:02
Posts: 1614
Location: United States
Has thanked: 71 times
Been thanked: 88 times
What issues were you having with the original capture board setup? I can probably help with that.

If you can work on the EFM side that'd be awesome - I made a little progress years ago but never got the RF filtering quite right. There's plenty of docs on the CD decoding - to plug a scan I made, the CLD-900 technical reference on LDDB's manual site is a good place to start. It would probably be best to start with a CD audio sample, and I can email you an 8MB one to get started. (there are 75 frames per second, so it only takes a MB or two of data)
_________________
Happycube Labs: Where the past is being re-made, today. [meep!]
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 01 Jan 2018, 11:57 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
simoni wrote:
Thanks for the offer of help nemesis, it's highly appreciated.

Right now I don't have any spare boards (and it's quite expensive in time and money to make them and ship them around - given that it's a prototype which isn't really proven yet (although the current results look very promising)).

No worries. If you're stamping out another revision though which is getting closer to where you want, and you'd like another pair of eyes/hands testing it, let me know.

Quote:
I'd rather be a little more sure of the hardware before making more so, as a compromise, if you let me know what type of discs are most interesting/useful for you to work on the decoding - I can prepare some RF samples for you and send you a link to download them from my owncloud server (as the samples are usually fairly large). I have a selection of PAL and NTSC laserdiscs with digital audio which should be suitable for use in testing EFM decoding (and soon Happycube will be able to make samples too). I realize this isn't as much 'fun' as a real duplicator board but, if we make some progress in the decoding, then by that point I'll be in a better position to supply a board without the current risk of wasting both your and my time/funds :)

I'm primarily interested in NTSC from a CAV disk (irrelevant when ripping?), as that's the format for all(?) of the LaserActive disks produced. I'd very much appreciate some known good samples of this, but from one with a digital audio track if possible (just straight stereo digital, no DTS). I'd also like one sample of a PAL rip with digital audio too. I don't have a PAL player or any PAL disks, so I'll need a supplied rip to look at that at all.

Thanks for your offer to provide samples, it's much appreciated.

happycube wrote:
What issues were you having with the original capture board setup? I can probably help with that.

The main ones I ran into were just the basic dumb issues around setting up a machine for this purpose, and getting the Linux environment configured. In the end it was mostly the fact my circumstances changed and I lost all time to work on it, so I wasn't able to work through things until I got it working. I also had some issues with the card I'd sourced, but I have two or three more now from when I last tried, so I'll have more test cases there. I've got an RF point tapped on my player, I just really need a configured linux software environment ready to drive the capture hardware. I'll dig out a drive I can wipe and start again with a fresh Ubuntu install on my test machine, and see how I go. If I get stuck I'll post here, with your help I'm sure I'll get it off the ground.


One quirk I mentioned briefly last time is that I'll be wanting to decode the video (in my case at least) to a raw image which shows all 525 lines of the NTSC frame, something like this (but for NTSC):
Image
This will allow me to capture all the data in the vertical blanking region, which is important for emulation. The disks for the LaserActive also often stored completely different video streams in each field, as the system had a digital buffer and the software was able to select even/odd field only and have it line doubled for output purposes, so this encoding isn't only more "raw", it's more useful for my purposes both for archiving and for use later. I assume it'll be easy enough to modify ld-decode to support this output if it doesn't currently, as this format is literally just a direct representation of the underlying composite signal.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 01 Jan 2018, 12:44 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
I've PM'd you a link to a capture file containing frames 10315-10800 from Disney's Fantasia disc 2 which is CAV, NTSC with digital audio.

It's over a gigabyte of capture, so it should provide you with plenty of frames to play with :)
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 24 Jan 2018, 00:34 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
Hey guys, here's where I'm up to. I've got a capture rig configured, and I was able to capture some nice composite video signals using cxadc. Kudos for that alone, I've been wanting a way to simply continuously capture composite video at a high enough data rate, and this is the first time I've been able to do that without issue. I was able to capture the RF signal from a Laserdisc, but I definitely need some amplification as the signal amplitude is way too low, and the decoder can't recognize it. I'll be getting parts in the next day or two to build the signal amplifier you provided specs for simoni (cheers for that!)

In terms of compiling ld-decode, happycube, I need your help. I've got ld-decode pulled, but I can't build it properly, at least with your latest code. First up, it looks like you're missing a makefile under "app\tbc", or at least the root makefile tries to call make in there, and there's no makefile to use in that directory. Could you take a look at that? I also had trouble in the root makefile, clang complained about the "-lfann" argument and wouldn't compile where it was used. I couldn't find that documented as a valid argument, seems like it's a GCC flag. Could you doublecheck and advise?

On another angle, there's the code itself. I'm really finding it quite difficult to orient myself with it right now. I know this has just evolved as a rolling prototype, but right now I think it'll be hard for anyone except you to work on it happycube, and if you stop work on it, I'm not confident anyone else will be able to pick it up in the future. I'd like to get involved, but a massive lack of code comments, and especially lots of "magic numbers" in various places make it hard for the uninitiated to follow. There's also plenty of duplicated code between files with slight variation, and I'm not sure what's current or obsolete at this point. Is there any chance you could do a "reboot" of this from scratch in a clean repo, taking just the code you intend to carry forward, with some comments added for the main functions, and in particular for each magic number explaining what it's for and why that number was chosen? If you can have a crack at that when you get the chance, I'll be able to help contribute. This code doesn't need to end up "production quality", but if you get hit by a bus tomorrow, the code really needs to be understandable in and of itself so that someone could use it as a reference to re-create the entire process with no additional information. You've achieved something really incredible here, but unless the code ends up in a state where it's largely self-documenting, people might end up having to re-invent the wheel here in the future if it can't be understood.

On a different track, I'm really keen to experiment with using this code to process straight composite video signals. I think this has a wide range of applications, and frankly would make this software with a decent capture device the perfect analog video capture card. The current tool chain here seems quite LD-specific though. Can you see any potential problems with trying to use the TBC and comb filtering stages you've designed on an arbitrary composite video signal? I'd like the RF demodulation from a LD to be an optional stage you can bypass, where you can just feed in any composite video signal instead. Do you think this is achievable, and what would need to be done to make it happen? This is what I'm most interested in experimenting with first on this codebase, and I think it'd be interesting to compare the quality of a software demodulated signal from RF with a straight composite video capture.

Back on laserdisks, I do have some concerns about capturing CAV disks, which all of the disks I need to capture are. I note that you mention the signal strength increases as the track advances, and we're supposed to sample the signal level near the end to figure out a suitable "upper" limit for the capture. Isn't that going to result in poor resolution near the start of the track though? A higher resolution capture like the 10-bit one your device does simoni would mitigate this somewhat I'm sure, but in terms of getting the best possible capture, is more work required here? How large is the difference in strength from the start of the disk to the end?

On another track, have you thought about the idea of taking multiple analog captures (IE, three or four), and combining them to filter out noise and potentially even increase effective resolution? It strikes me that this should be possible to achieve, and could be helpful on rare disks, especially to deal with defects. If you have more than one copy of a disk and all of them have rot for example, you could do a virtually perfect error correction as a pre-processing stage by combining multiple independent samples. I do this when capturing subcode data for the digital data I rip from the LaserActive games for example, and it's proven highly effective in eliminating errors. I can also confirm bit errors in the master from this, as if the same error is present in the same place from three separate sources, you can be sure it was there on the master. This is a bit more complex on an analog signal, but I think it should be possible. Just some more food for thought.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 24 Jan 2018, 02:37 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
Hey simoni, I've got a few questions about your Domesday Duplicator. I'm considering stamping out a board and assembling it myself. Do you feel the hardware is far enough along for that to be worthwhile? Software revisions totally aside, I'm just interested in the hardware itself at this point and whether you anticipate any major revisions in the near future. Also, even if that's the case, are you expecting the USB and FPGA boards you've chosen to stay the same into the future? If you are, I'll at least probably order them in, as it costs around $180AUD for me to source them and will probably take a few weeks to arrive. Once I have my hands on one, I can help on the software side for that too. Windows-based software with a general purpose capture focus comes to mind, as I'll have many other applications for a device like this.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 24 Jan 2018, 05:57 
Absolute fan
Absolute fan
User avatar

Joined: 18 Apr 2012, 18:02
Posts: 1614
Location: United States
Has thanked: 71 times
Been thanked: 88 times
I'm (slowly) working on a cleaner RF decoding program, and simoni is redoing the TBC. So there should be better code soon :)
_________________
Happycube Labs: Where the past is being re-made, today. [meep!]
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 24 Jan 2018, 06:28 
Honest fan
Honest fan
User avatar

Joined: 11 Sep 2017, 14:24
Posts: 88
Location: Sweden
Has thanked: 2 times
Been thanked: 36 times
The makefile is currently missing a 'qmake' command on the app/tbc - it's written in Qt, but I missed the qmake (which generates the Makefile called by the root Makefile). As happycube said, this is all WIP at the moment. The new tbc app was experimental, but I hit the same problems as you state nemesis; the original code lacks comments and is full of magic numbers - so I was unable to simply port the existing TBC to Qt is a useful manner.

So; I stopped working on app/tbc... what I then did was generate a new application and ground-up rewrite the TBC - I've got quite far with this now and I hope to have something working in the next few weeks (at which point I will merge the new code with happycube's repo). Until then, you will need to use the original tbc-ntsc and tbc-pal code (as the new version is not yet fully functioning). The new TBC isn't based on the old one... I reinvented the code as I needed it to be object-orientated so that all the individual functions of the TBC were cleanly separated. This also makes it possible to multithread (as I can make the code non-scalar by pipelining the data through threads).

On the duplicator board; the current version is the version I intend to use and I don't foresee any major modifications to it. There is a small issue in the RF front-end (which is documented on the project page), but the net-effect is that the ADC has a 1.8v peak-to-peak sensitivity (instead of 2v). This could be fixed, but it really doesn't make any practical difference. The USB3 board and FPGA board are fixed now - I have no intention of switching to another set of boards.

The duplicator software is written in Qt C++, so it's quite portable. There are some differences in the way that USB is handled between Windows and Linux though, so that part would need to be modified for Windows support. Having said that, the project is documented and the code is (imho) structured and well commented; so modifying it should be a fairly straight-forward task.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 25 Jan 2018, 03:32 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
I'm ordering a run of Domesday Duplicator boards. I'll have spares, so if anyone's interested drop me a PM.

I've had a chance to take a closer look at the design, and I'm really interested in what you've come up with here simoni, and especially on all the other applications this could have. You've done a great job of keeping the interface board cheap and simple, and I think there's a lot of potential to make a set of general purpose analog/digital capture boards. I think it'd be interesting to allow for up to a 32-bit input (hell, 64-bit if possible) in a general form in the firmware, and design some general purpose analog/digital input boards. In terms of cost, the current PCB design just nudges over 100mm width. If it's possible to limit it to a 100mm width, it's possible to have the boards fabricated at a rate of $5 for 10 at pcbway. At that price point, it's practical to design a whole set of specialised and interchangeable input boards. It should be easy enough to design a single digital input board, but you could have a range of analog boards built around different ADC chips at different pricepoints, which vary from relatively low bit 1 channel boards, up to 14 or 16 bit single channel or even multi-channel. There's actually a huge market for this kind of device. Even though the FGPA and USB evaluation boards cost a decent amount, if there was enough interest it'd be easy enough to manufacture specific versions of those with just the required hardware for a much cheaper price.

Being open source, I hope this project evolves in that direction. I've been looking for a device like this for years now. I even thought seriously about having a crack at designing one myself, and I looked at that same USB evaluation board, but my lack of experience in FPGA design and analog electronics was too high a barrier.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 25 Jan 2018, 06:13 
Absolute fan
Absolute fan
User avatar

Joined: 18 Apr 2012, 18:02
Posts: 1614
Location: United States
Has thanked: 71 times
Been thanked: 88 times
One possible cost reduction would be to use an FPGA+ARM+GigE board like the DE0-Nano-SoC. In theory that should handle the full capture bandwidth while keeping the flexibility (and buffering) of the FPGA. If you can put everything on one header that should get the board down to size.

When dealing with very low volumes, the real expense is design time, so using whatever you're comfortable with to get a quick effective solution is best and PCB size/cost isn't a big deal really. Designing an original FPGA and/or USB3 board isn't easy at all, and the sandwich approach is quite clever indeed.
_________________
Happycube Labs: Where the past is being re-made, today. [meep!]
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 29 Jan 2018, 04:37 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
With quite a bit of pain over the weekend, I managed to get ld-decode working with a resampled version of 32MSPS capture data from Fantasia, and with a repository rolled back to b6d97a6 from January last year (after various make/script fixes). Since I don't use Linux on a daily basis, I also did a quick and dirty port of what I needed to Windows, so I can now decode video from RF on my primary machine. I must say, the quality of the output is very impressive! I'm really keen to see how this looks at a higher sample rate with 10-bit resolution! I haven't seen too many samples put up for this project, so here's one for anyone who might be interested (click image for video):
Image

In the capture space, I didn't have any luck capturing raw RF. The amplitude of the RF signal seems too low, and I didn't have a VGA amplifier. I built the amplifier circuit you provided on the weekend simoni, but it didn't seem to function properly. I seem to get a distorted and non-amplified output. I checked everything over about 50 times and I believe I followed the schematic exactly, but with little experience in analog electronics I don't claim to understand how the circuit is supposed to work or how to diagnose what's going wrong. Do you have any suggestions on what I could do? Is it possible there's an error in the schematic? I'm moving forward building a Domesday Duplicator device anyway, but it'd be nice to get RF captures working in the interim.

Not being able to get an RF capture right now, I started experimenting with processing direct composite video captures using ld-decode. That's something I'm very interested in getting working, and I'm able to get good captures of the composite video without amplification, so it was easy to play around with. The TBC code (from January last year) doesn't like a straight composite video signal, and fails to detect the sync properly. I don't actually need the signal to have TBC applied in this case as it's already prepared for direct display on a monitor, but I do need the sync detection and framing to occur. Due to the failures at this stage, I currently get a vertical and horizontal rolling output like this:
Image

Since this code is on the way out, I'm going to try and build the new TBC code you've been working on and play around with that. Maybe it already works, I'll find out when I've got it built. I'll also be working on adding support for a full frame output mode through ld-decode, to produce an image like I showed above. If you guys wanted to cross-test, here's a short sample of composite video test data I'm currently working with, which I used to generate the image above:
https://mega.nz/#!MZVziIaZ!1Hx8RoXaGD0oLa5WkOPAhRBDqez84S4a5_gq46FsGE8

This data is a simple largely static image which should be easy to decode. The data has been converted into a 16-bit unsigned format consistent with what's output by lddecode.py, at the same sample rate (from a 28.8MSPS capture). This data was originally taken from the following 8-bit unsigned capture:
https://mega.nz/#!UEcCGYBa!EssOiS7dLIaTZ0GFFY61apc3IqcEl9bIw0XVbHF5kLc

This is a much larger sample (700MB compressed, around 60 seconds IIRC), and it contains a variety of large transitions, and some interlacing for part of it too, so it's a pretty good test case for taking straight composite and processing it. In the long term, I'd like to see some tool in ld-decode eventually be able to take in any composite signal that a multi-sync monitor could handle, and correctly process it too. Laserdisc-specific modes and tools certainly have a place, but from my perspective, after demodulation has been done and a composite video signal extracted, the following steps in the pipeline should be able to work with any composite signal from any source, and I'll be trying to push in that direction. This will make these tools applicable to a wide variety of tasks.


Last edited by nemesis on 29 Jan 2018, 06:17, edited 1 time in total.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 29 Jan 2018, 05:24 
Shows curiousity
Shows curiousity
User avatar

Joined: 18 Jul 2016, 05:18
Posts: 28
Location: Australia
Has thanked: 0 time
Been thanked: 7 times
To help out anyone else who might be interested in experimenting with the software who isn't able to generate their own captures right now, I've uploaded the raw RF capture I used to generate the above Fantasia clip at the following location:
https://mega.nz/#!EQtwVJxC!Mlx5rzTZ-aGg_-kYmWXZSEqq3XdYMhtUQDsEkIJQMgM
This is a 10-bit 32MSPS capture of an NTSC CAV disk.
Offline
 Profile  
 
 Post subject: Re: (WIP) Laserdisc software image decoder from raw signal
PostPosted: 29 Jan 2018, 09:13 
Absolute fan
Absolute fan
User avatar

Joined: 18 Apr 2012, 18:02
Posts: 1614
Location: United States
Has thanked: 71 times
Been thanked: 88 times
Nice sample! I haven't gotten around to tweaking (my old) ntsc tbc to deal with the 32mhz/10bit samples properly quite yet, so good job on getting it to work. It's great that you guys are looking at this with much fresher eyes than mine - and taking it to the next level :)

And here are some sample videos I did, but nothing with the new tech (yet?): https://www.youtube.com/user/happycube/videos
_________________
Happycube Labs: Where the past is being re-made, today. [meep!]
Offline
 Profile  
 
Display posts from previous:  Sort by  
 Page 17 of 31 [ 606 posts ]  Go to page Previous  1 ... 14, 15, 16, 17, 18, 19, 20 ... 31  Next


Who is online

Users browsing this forum: No registered users and 6 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to: