VideoHelp Forum
+ Reply to Thread
Page 2 of 35
FirstFirst 1 2 3 4 12 ... LastLast
Results 31 to 60 of 1031
Thread
  1. Originally Posted by oln View Post
    Personally I bought it from one of the Domesday86 guys but I think he's out of boards now.

    There are some people working on making some batches for sale I think, I would suggest asking on the domesday86 project facebook page or in the #domesday86 IRC channel on freenode.
    What's the best time of day to reach them on the IRC channel to inquire about an order? I tried asking a few questions at different times but only one person responded. There are 26 people online. I know the UK is 5 hours ahead of me but I tried in the morning and afternoon as well. I don't use Facebook.
    Quote Quote  
  2. I decided to go with a CX23880 based capture card and cxadc instead. Can't beat 10 bucks! I've since seen some example decodes using this which looks as good as VHS is going to look. Heck, even LaserDisc captures look incredibly good at 28.6MSPS!

    If you want to capture RF from your VCR I would look for the service manual. I have a cheap Philips VCR and while I didn't find a service manual for the exact model, I did find one for the VR550 on manualslib.com. My machine, although it probably costed $80 new, tracks surprisingly well. In this case the test point is one of the many jumpers on the main board. I located the RF and ground test points after referencing the service manual's RF calibration section.

    Last edited by Titan_91; 3rd Nov 2019 at 19:08.
    Quote Quote  
  3. According to oyvindln nothing needs to be done with regards to signal level. It’s ideal for the Conexant chip input via the luma pin on the s-video port.

    https://github.com/happycube/ld-decode/issues/16
    Last edited by Titan_91; 7th Nov 2019 at 11:56.
    Quote Quote  
  4. Well you want to use it with a deck where you can tap the signal after the RF amplifier. Couldn't get strong enough signal on my JVC deck for the cx88 card (but the same worked with the DdD). My Sony deck gives a much stronger signal, so I presume the rf test point on that one may be after the rf amp, on the JVC I used (HR-J658) it's not. On decks up to like the mid 90s the rf amp was usually on it's own board, but on later decks the RF amp seems clumped together with other parts of the decoding chain on an IC so it may be harder to find a point to tap the signal after amplification.

    Also, you can use the composite inputs on the cx88 card, just have to specify the right input in the cxadc driver settings.
    Quote Quote  
  5. Thanks for the tip. I installed the card, built the driver from the GitHub repository, and used insmod to insert the module into the kernel. I rebooted and checked lsmod. Lsmod showed the module loaded before I rebooted but does not show it now. I guess I need to load it on boot and possibly blacklist the normal CX88 modules? I have little experience with using custom drivers on Linux, if anyone can help that would be great.
    Quote Quote  
  6. Formerly 'vaporeon800' Brad's Avatar
    Join Date
    Apr 2001
    Location
    Vancouver, Canada
    Search PM
    You do need to blacklist the normal drivers. At the time I used cxadc, I followed the instructions in the then-current ld-decode documentation, which was a Google Doc. Can't link it at the moment.
    Quote Quote  
  7. I forgot about the Google document. I remember a guide on how to use the driver and dd to capture samples but couldn’t find it. I’ll check and report back.
    Quote Quote  
  8. Awesome, thanks a lot.
    Quote Quote  
  9. I got the driver working. I ran the level adjustment tool using ./leveladj 31. I did a blank test recording connected to an unpowered s-video device. Besides the fact that my s-video cable sucks creating noise, the zero point is way off from DC center. Any suggestions?

    Quote Quote  
  10. I think I've put two and two together now. I hooked up a Roku box, ran the level adjustment tool with it turned on, chose level 19 with 0 low and high closest to 255, and captured a few fields of luma video. I opened this in Audacity and it looked good, but the lines were being clipped a bit at the top. At this point I realized "clipped" has to read 0. So I then chose level 6 which doesn't have as much dynamic range, but does not clip. Here are some Audacity screenshots, the first waveform being level 19 and second being level 6. My understanding of the goal is to get as much dynamic range as possible (low being 0 and high being closest to 255) without clipping. Do I have the gist of this?





    And here are both samples loaded respectively in the RawSource AviSynth filter running in Wine:




    With either case, the picture looks REALLY clean.
    Quote Quote  
  11. I hooked up a scope and capture card to the test point. I have to bottom out the volts per division to even see the RF envelope. I looked at the schematics for the main A/V block and the two amplifiers for the SP/EP heads are internal to the main IC. However, pins 78 and 79 break out to a capacitor that connect to what appears to be a switched amplified FM signal.

    https://imgur.com/a/E0392id
    Quote Quote  
  12. Yeah as noted you want to tap somewhere after the rf signal has been amplified.
    Quote Quote  
  13. Is there a current way to time base correct and decode a recorded composite signal from any video source using CXADC?

    Edit: Thanks oln for answering that on the IRC channel.
    Last edited by Titan_91; 22nd Nov 2019 at 11:00.
    Quote Quote  
  14. I have another RCA machine from 1994. I'm repairing the power supply and once I get it going I should be able to upload a usable NTSC sample. The motor controller and head amp have their own board. The FM circuit is also shielded which is nice. But it only has 2 heads. Will this affect the capture quality of LP/EP tapes?
    Quote Quote  
  15. May be slightly more noisy than an equivialent 4-head deck (provided it supports LP/EP), but how much I don't know, haven't done much testing with LP yet.
    Quote Quote  
  16. Amateur audiophile here - looking at Titan's posts I can't help but wonder if it might be possible to get the best of both worlds - since it's technically audio we're dealing with then there should be ways of recording/editing the audio so as to get higher dynamic range but without clipping. If I'm understanding this right, AVIsynth is interpreting the '19' with much better contrast than the '6', but with more glitches. Presuming the glitches are from the clipping, then the question is if there's some way of recording the signal at a much louder volume (per '19') but then normalising it so it does not peak on export (per '6'), thus getting a better dynamic range without the glitches... Or is AVISynth just reading better contrast because input is relatively louder, rather than being dynamically wider? Forgive my ignorance as footage and it's related programs are not my forte, but surely if it's the latter than it's just some thresholds that need tweaking in AVIsynth itself?

    If it's a case of simply fixing the clipping, it might be interesting to try fiddling with the 'audio' in a spectrogram editor such as iZotope. If you send over some of the test clips as above I'll work some magic on it and send them back, and we can see how that changes AVISynth's interpretations.

    Also has anyone done tests on getting the //actual// audio read off the tape directly in a similar fashion to the footage? Or is the usual aux-cord output 'close enough' to not warrant much effort trying to decode it directly?

    Oh, and just as a general note, I'm massively excited for this project and have been telling everyone I know about it - in all honesty this will absolutely revolutionise archiving as we know it, once it's perfected. Hopefully we can get more people working on it to collaborate further, and help bring the future of bringing back the past closer to the present, as it were...
    Quote Quote  
  17. Member Cornucopia's Avatar
    Join Date
    Oct 2001
    Location
    Deep in the Heart of Texas
    Search PM
    1. This is rf-modulated material. To get the best reproduction, you want to get the most modulation WITHOUT foldover. Foldover generates distortion. Even if it appears that you could further maximize dynamic range, it won't generate further meaningful data if distortion gets in the way.
    2. The audio that is part of the standard video rf signal is LOW quality. If you want much better quality, use the Hi-fi audio. Not sure if this setup is geared to getting that signal.

    Scott
    Quote Quote  
  18. Originally Posted by Cornucopia View Post
    1. This is rf-modulated material. To get the best reproduction, you want to get the most modulation WITHOUT foldover. Foldover generates distortion. Even if it appears that you could further maximize dynamic range, it won't generate further meaningful data if distortion gets in the way.
    2. The audio that is part of the standard video rf signal is LOW quality. If you want much better quality, use the Hi-fi audio. Not sure if this setup is geared to getting that signal.

    Scott
    When we talk about RF in this thread, we don't mean standard RF out on a VCR. We're talking about the RF tap point that's right next to the heads.
    Quote Quote  
  19. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    Obviously, That RF output is for TV only, it's low quality RF transmission of video and audio signal.
    Quote Quote  
  20. Member Cornucopia's Avatar
    Join Date
    Oct 2001
    Location
    Deep in the Heart of Texas
    Search PM
    "Next to the heads" - yes, I know this, but WHICH heads?...hifi audio follows a completely separate, different path from video. As does the low quality linear audio. The OT refers to (one of the) pairs of video heads, but that last question referred to getting audio. That was my response to him.

    Also, Regardless of whether you tap into it from right by the heads or after an amp - The only difference there is overall amplitude & SNR.

    Also, RF is RF, whether it is used as a carrier in broadcasting or a bias in recording. In both cases, they are simple, linear, 3-domain (amplitude, phase/frequency, time) waves, modulated by the intended "signal".

    Scott
    Quote Quote  
  21. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    I believe he was asking about the RF output on the back of the VCR that usually broadcast on CH 3-4.
    Quote Quote  
  22. Actually I was meaning more like.... has anyone tried testing the difference between the hifi audio from the usual speaker output on the player vs the hifi audio tapped off the tapehead-adjacent DDD, if it's been decoded? I'm assuming it would be slightly better but has anyone actually been able to check yet?
    Quote Quote  
  23. Originally Posted by MK Inst View Post
    Actually I was meaning more like.... has anyone tried testing the difference between the hifi audio from the usual speaker output on the player vs the hifi audio tapped off the tapehead-adjacent DDD, if it's been decoded? I'm assuming it would be slightly better but has anyone actually been able to check yet?
    I'm sure it would be better, but I don't think anyone has tested it. Presumably, all signals should be better if they're obtained directly from the source and without any processing done by the VCR itself. From what I've been told previously, it would require a few SDRs to decode linear audio separately from video (which would be a pain to sync back up to the video, but because it's the only option, that's fine), but perhaps someone would have to code something that's capable of decoding Hi-Fi with software. Just a theory so I might be wrong.
    Quote Quote  
  24. My plan is to record the video RF and the baseband audio output from the RCA jacks at the same time. That way if the tape drops out the audio dropout will be synced. Plus, who’s to say I would be able to play many of my 30 year old tapes back more than once before they break?
    Last edited by Titan_91; 30th Nov 2019 at 10:48.
    Quote Quote  
  25. Hello, I will leave some nice samples of VHS-Decode power: YT Playlist
    My recipe was to go CX2388X card (Avermedia 303).
    I ran everything on Linux Mint, installed the driver and software.
    I connected together the card through the composite and the output of head amplifier/switch IC - bypassing even one resistor.
    I chose the 10 times subcarrier capture (35,79MHz & 8-bit) previoisly modifying the python script - deemphasis filter values and the working frequency.
    My VCR is Panasonic FS-200. The SNR is about from 25dB for worst tapes to 40dB.
    Quote Quote  
  26. Nice videos, thanks for sharing. Looking at your other uploads I see you doing something with MPlayer. Were you using MPlayer to somehow decode and monitor the live CXADC feed?
    Quote Quote  
  27. Yes, I needed to preview the signal - I was wondering long time how the correct signal looks like. Such MPlayer display lets me to adjust the mechanism including the tracking, 2 guide rollers and a head swich moment (PG Shifter potentiometer). The stronger FM pattern the better signal I get.
    MPlayer itself does not include any means to decode VHS RF signal live. I was probing standard composite output with and without internal TBC plus the RF test points.
    For live preview I used following command:
    cat /dev/cxadc | mplayer -demuxer rawvideo -rawvideo fps=25:w=1833:h=625:y8 -aspect 16:9 -”
    Quote Quote  
  28. Which adjustments did you make to the deemphasis filter?
    Quote Quote  
  29. In vhs_formats.py:
    FParams_PAL_VHS['video_bpf_low'] = 4500000 #3350000 original
    RFParams_PAL_VHS['video_bpf_high'] = 7000000 #520000 original
    Besides the ability to live preview the results I had to wait for the frames to decode but I think it's good now.
    I looked at teletext patterns, I messed with the values until the dropouts were gone. These parameters vary depending on the condition of the tape.
    Other values I'll be looking at will be TBC sensitivity, video levels and the color filters.
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!