VideoHelp Forum
+ Reply to Thread
Page 3 of 4
FirstFirst 1 2 3 4 LastLast
Results 61 to 90 of 116
Thread
  1. Im following this thread because im coming from other thread seeking VHS and Hi8 capture card, someone here said Magewell is best capture for VHS. I have VHS (not S) tapes and JVC player, I wonder if I use Magewell will I loss picture quality and which model is recommended.
    And thank you for answering.
    Quote Quote  
  2. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    It seems I have been the first to post information related to this magewell card . I have been capturing VHS, video8 and MiniDV for about an year as I have a large collection and in this period I tried or bought a significant quantity of players and capture cards and also tried several workflows variants to get he best possible quality.

    Just before I bought this magewell card I got the best results capturing analog tapes with the workflow developed along the initial part of this thread .

    VHS captures

    SVHS JVC Deck (1)->svideo->DVD Recorder Panasonic DMR-EH65->HDMI-> HDMI capture card ( startech usb3hdcap)->Virtualdub-> captured file (usually using a lossless codec ( Huffyuv, lagarith or UT video) in an AVI container.

    (1) I have 2 SVHS JVCīs decks an HR-DVS3 ( combo mini DV and SVHS ) and a JVC HR-9500. The SVHS players even for VHS tapes provide better results and they come with line tbc ( 4MB buffer ) and DNR. From these two the 9500 provide better results , the 9600 is marginally better you can see a list of most suitable players at
    http://www.digitalfaq.com/forum/video-restore/1567-vcr-buying-guide.html

    In this workflow the Panasonics act as an ADC and TBC ( line TBC and as a limited frame synchronizer as it do buffers some frames) .

    When capturing analog tapes you are going to confront yourself with lost or inserted frames and itīs almost mandatory to have an external full frame TBC (frame synchronizer). However the ones that are usually recommended the AVT TOOLBOX AVT-8710 and the DataVideo TBC-1000 are both already difficult to find , expensive (between 200-500 USD ) and most of the times have unpredictable results and they may significantly worsen yours captures ( I had terrible results with an AVT-8710 that costed me new 320€). The Panasonic replaces the need for the TBC and avoids one extra D/A cycle ( on the external TBC ) if capturing in hdmi .

    As you can see from the data and samples provided the magewell performs very well and it is the better capture card I have found so far for analog tapes VHS, video8 , HI8, Pal or NSTC ( I tested more extensively in Pal but I have made some test with ntsc sources and it performed equally well ). It includes a line TBC that works well and mainly a frame synchronizer with 256 MB on board memory which is extremely important for analog tapes capture). Even if the price is high you donīt have to buy a separate external TBC and you donīt risk degrading yours captures quality and your workflow will be much simpler and easier and with less cables and fuss .

    SVHD deck-> svideo-> Magewell -> VirtualDub-> Huufyuv

    I got the best results with the player setting indicated in the posts ( basically TBC/DNR ON , Edit ON, BEST on, all the others OFF)

    The card has extensive presets capture formats and supports a large number of color spaces and can delivery 8 bits or 10 bits. Also you are not limited to windows XP or windows 7 it will work in any windows version also linux and MAC. Being a pcie 2.0 card you will have also a better bandwidth. Also you can upscale or downscale directly through the card hardware without limitations.I would say the card as a lot of features you can see the data sheet from the card itself and the one from the ADV7842 chip.

    I think a card model similar to the one I bought may be appropriate. It does 2K ( you have also 4k versions) it does have direct hdmi capture and composite, svideo and component capture through a breakout cable.

    With the exception of the very expensive cards like Kona I have tested most of the currently used or proposed cards with exception of the ATI AIW AGP cards but I tested for instance the ati usb 600 also much recommended and the results are much worse then with the Panasonic or the Magewell and you do still need an external tbc.


    So far after experimenting a lot of cards this one is the probably the best I have tested and having an internal frame synchronizer makes a big difference. You are not going to find easily another one with this features for this price.
    Last edited by FLP437; 19th Oct 2016 at 23:36.
    Quote Quote  
  3. Formerly 'vaporeon800' Brad's Avatar
    Join Date
    Apr 2001
    Location
    Vancouver, Canada
    Search PM
    Forgot to respond to this post:

    Originally Posted by FLP437 View Post
    In the workflow using the Panasonic the hdmi stream seems to be YcbCr 4.4.4 would I have any interest to capture in 4.4.4 instead of 4.2.2, I could do it with the magewell if it would add any value
    Absolutely not. VHS chroma resolution is well below 4:1:1, even.

    Vaporeon800 the way you present the data on yours screenshots is useful can you post the avysinth script .thanks
    It's attached to this post (beware the large MBs of auto-loading pictures in that thread).

    Originally Posted by FLP437 View Post
    In this workflow the Panasonics act as an ADC and TBC ( line TBC and as a limited frame synchronizer as it do buffers some frames) .
    I'd like to see evidence from anyone who claims that the Panasonic frame sync is "limited" in any way. This seems to be a meme spawned somewhere on the forums. The DMR-ES15 completely prevents downstream capture devices from dropping frames, according to me and Sanlyn. PAL model(s) do the same, according to Skiller.

    Lordsmurf refuses to call anything that doesn't remove Macrovision from the output a "full frame TBC", which is IMO nonsensical. I posted images proving that my Panasonic & Toshiba DVD recorders regenerate and replace the vertical interval. They deliberately add a new Macrovision signal back into the output, since they are consumer electronics that play ball with Hollywood. But that doesn't mean they don't do "full" frame sync.
    Last edited by Brad; 17th Oct 2016 at 17:35.
    Quote Quote  
  4. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    I just posted some samples with my new Sony HI8 XR CCD-TRV87E Pal from 1999 but I do think the results from the digital 8 are still the best
    https://forum.videohelp.com/threads/378172-D8-Camcorder-Vs-High-End-Hi8-Deck-video8-capture post 12
    Last edited by FLP437; 23rd Oct 2016 at 11:32.
    Quote Quote  
  5. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    "The DMR-ES15 completely prevents downstream capture devices from dropping frames, according to me and Sanlyn. PAL model(s) do the same, according to Skiller."

    I do agree with vaporeon800 my Panasonic DMR-EH65 also completely prevents downstream capture devices from dropping frames and this as been a fact in all my captures
    Last edited by FLP437; 19th Oct 2016 at 23:28.
    Quote Quote  
  6. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Component video for analog capture

    Reading some texts in video capture and preservation sites I remarked references about analog video capture advantages using component as an input signal

    “ Superiority of analog video capture using component output (separate RGB) and SVHS players with 4 heads and TBC”

    That gave me an idea, I donīt have SVHS players as for instance the JVC BR-S622DXV that have TBC, DNR and component outputs or similar players, however I remembered that I had a Samsung DVD/ VCR Combo that was a SVHS 6 Heads HIFI player that was able to output RGB to the Scart plug (not only the DVD but also the VCR) and that used to provide very good image on modern television sets.

    The VCR apparently doesnīt have TBC or DNR ( or at least this is not referred in the manual ) however this should not be really a problem as I intended to capture using the Panasonic DVD recorder DMR-EH65 using the Scart RGB input and capturing with the Magewell over HDMI from the dvd recorder deck and both have excellent TBCīs and also frame synchronizerīs so in theory everything should run smoothly.

    So my test workflow has been the following

    Samsung DVD/VCR SV-DVD1E scart (RGB)-> scart(RGB) Panasonics DMR-EH65->HDMI-> Magewell Pro HDMI->VirtualDub->HuufYuV

    I captured YUV2 full gamut (1-254) over HDMI, after I will use in restoration PC601 matrix to convert to RGB in avisynth for filtering purposes, avoiding any black or white clipping.

    Most of the DVD/VCR combos do a sort of technological bridge between old and new technology and most of them have component or RGB output over Scart including from the svhs player. Most of them also provide up conversion and progressive scan technology and some even HDMI output beyond component and Scart RGB.

    My Samsung player is far from a top of line in the DVD/VCR series and I wanted to see if even in this case I could see any benefits. The results have been quite interesting and I think is easier to detect a reasonable benefit even with a machine that as I have said is not a top of the line.

    I think this approach for VHS capture can provide a significant benefit namely if the player is a top one

    TEST9_HR-S9500_tbc_on - previous best capture JVC SVHS HRS9500 (TBC/DNR on) ->svideo-> magewell pro HDMI-Virtualdub->HuufYuv

    Test_Sansung_RGB - new capture with the workflow previous indicated
    Image Attached Files
    Last edited by FLP437; 29th Oct 2016 at 18:28.
    Quote Quote  
  7. Member vhelp's Avatar
    Join Date
    Mar 2001
    Location
    New York
    Search Comp PM
    How is the Magewell Pro HDMI with commercial movies ?

    In terms of picture quality, I prefer the Hauppauge HVR-950Q and the ATI-600. The 950Q ignores macrovision (MV) but the 600, after about 3 seconds, mv kicks in and distorts the image. I use a Grex (mv stripper) in-between to remove the mv.

    https://forum.videohelp.com/attachment.php?attachmentid=39157&stc=1&d=1477776598
    Image Attached Files
    Quote Quote  
  8. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    In what concerns the Hauppauge HVR-950Q I donīt know as I donīt have it ,but in what concerns the ATI 600 I got much worse results they have nothing to do with the quality provided by the two last workflow’s as you can see by some samples in the beginning of this thread, also itīs absolute mandatory to have a full tbc for using it and the experience ( and cost ) I have had with the toolbox AVT-8710 for instance was quite bad.

    Both the Panasonic and the magewell have frame synchronizers so this problem doesnīt arise. I have not used this or the previous workflow with commercial titles , the Panasonics is extremely sensible to macrovision and HDCP , however the hdmi splitter strips the HDCP without problems I will try when I have time a VHS commercial title and see what happens in respect to macrovision .
    Quote Quote  
  9. The brights are very blown out in the component capture. Can you adjust the video proc amp while capturing to prevent that? The component capture suffers less from oversharpening halos and rainbows. It's colors are more saturated but you should be able to crank up the saturation in the composite capture with the video proc amp.

    Adjust to similar levels and saturation (simple bob()):

    Image
    [Attachment 39160 - Click to enlarge]
    Quote Quote  
  10. Member vhelp's Avatar
    Join Date
    Mar 2001
    Location
    New York
    Search Comp PM
    The component capture suffers less from oversharpening halos and rainbows.
    I agree. That's why I asked how it is with commercial movies also because I saw another demo image of elsewhere (I think it was the same card) and it did not have those artifacts but it was cartoon and not a very good representative for touting the card. I still would like to see a demo of a commercial tape.
    Quote Quote  
  11. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Yes I can adjust the video processor amplifier during capture. This test has been made with default values.
    With the magewell I can also capture with limited quantization and I will get exactly between 16-235, full quantization and saturation and I get values between 0-255 and full quantization and full gamut saturation and I get 1-254.
    Quote Quote  
  12. Oversharpening halos with VHS are mostly caused by the VHS player. Since the tape has very low bandwidth they boost frequencies near the upper bandwidth limit.
    Quote Quote  
  13. Originally Posted by FLP437 View Post
    Yes I can adjust the video processor amplifier during capture.
    The blown out brights didn't effect this particular shot much -- just a few things like the white t-shirts and the tan building in the background. But it might be catastrophic for some shots. The top 25 percent of the luma range is crushed.
    Quote Quote  
  14. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    So I must adjust contrast during capture or could it be related with my option to capture with full quantization and extended saturation gamut (1-254)
    Quote Quote  
  15. I'm not sure what the extended gamut setting is going to do. Your cap already has lots of brights above Y=235:

    Image
    [Attachment 39162 - Click to enlarge]


    After adjusting to match the slightly adjusted composite capture you can see the brights are capped around Y=190. None of the missing bright details can be restored.

    Image
    [Attachment 39163 - Click to enlarge]
    Quote Quote  
  16. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Ok. thanks I going to see if this as to do with the several quantization options offered for capture output or if it does need video processor amplifier adjusted
    Quote Quote  
  17. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    I am trying to see what is going wrong with this component capture experience. My previous goal was to capture in YUV but with full luma spectrum ( 0-255). After I will be doing in Avisynth RGB conversion using pc601 matrix ( as almost all filters in Avisynth and VirtualDub are in RGB ) so no clipping in theory should occur.
    Before I was only using histogram levels mode so I was happy as depending of my options the limits were respected (0-255), (1-254) or (16-235) but I was not aware of the clipping.

    The problem is that seeing the classical histogram and comparing with the previous Y/C capture it seems itīs occurring a kind of translation from the blacks to the whites with the latter surpassing 255 and being clipped. In this case obviously I will lose information. However when the frames are rich in mid tones with few whites the results from the component capture seem clearly better.

    I was trying to identify if this is a problem posed by the DVD/VCR player when converting to component or if it is some wrong doing by myself.

    In order to control settings on the fly I made in graphstudio a "workflow" to be able to adjust contrast and brightness in real time however I must be doing something wrong because if I change the video proc amp settings within the module signaled as 1 ( the one doing the capture) in the graphstudio model I see the changes immediately on the rendered image however if I change the settings in capture module signaled as 2 (for VirtualDub ) nothing happens, I should also be seeing the results immediately. I must be doing anything wrong, there are for instance within utvideo VCM and DMO codecīs versions, may it have something to do with this , probably not?

    I am somewhat convinced that an analog capture from component could provide better results so Iīm trying to see if I can find a workflow that could works ( without having to buy a SVHS VCR already with component output ), I can also try a direct component capture by the Magewell converting the Scart RGB signal to component ( lindy active scart RGB /component adapter) I will try it later and see what happens.

    However having tried with limited and full Magewell outputs quantization settings and with codecs that allow full 0-255 ( as UtVideo RGB- ULRG ) and standard 16-235 (which can however be exceeded by Magewell output settings) Iīm beginning to think that it could possibly be related with the DVD/VCR player expanding luma and doing it somewhat asymmetrical when converting to RGB or some sort of interaction between the DVD/VCR player and the DVD recorder with similar final results , however I'd like to confirm if possible if itīs some wrong doing or a real workflow problem.
    Image Attached Thumbnails Click image for larger version

Name:	image1.jpg
Views:	478
Size:	2.43 MB
ID:	39177  

    Click image for larger version

Name:	image2.jpg
Views:	246
Size:	215.0 KB
ID:	39178  

    Click image for larger version

Name:	image3.jpg
Views:	264
Size:	872.2 KB
ID:	39179  

    Quote Quote  
  18. Originally Posted by FLP437 View Post
    The problem is that seeing the classical histogram and comparing with the previous Y/C capture it seems itīs occurring a kind of translation from the blacks to the whites with the latter surpassing 255 and being clipped.
    Yes. Each 8 bit sample can only hold values from 0 to 255. Any sample that would have been outside that range is clamped to be within that range.

    Originally Posted by FLP437 View Post
    However when the frames are rich in mid tones with few whites the results from the component capture seem clearly better.
    Unless you plan to capture in segments with different proc amp adjusments for each segment, or adjust the proc amp continuously while capturing, you're better off capturing with one compromise setting and making adjustments in software later.

    I'm not exactly sure what you're doing with GraphStudio. In the past I've just captured (previewed) normally with VirtualDub and opened GraphStudio (or the older GraphEdit, or GraphEdt), added the capture module (or whichever module had the proc amp) and adjusted the proc amp settings there. The changes appeared live in VirtualDub's preview/cap. I don't know if that will work with the Magewell device.

    It looks to me like you simply ran Histogram(mode="levels") (true histogram) after TurnRight().Histogram().TurnLeft() (a waveform monitor). If so, the histogram contains all the blacks and dots in the waveform monitor. If you want to see both properly you need to create each from the original video:
    Code:
    function WaveAndHist(clip vid)
    {
      waveform = vid.TurnRight().Histogram().TurnLeft().Crop(0,0,-0,256) # just the waveform monitor
      hist = vid.Histogram(mode="levels").Crop(vid.width,0,-0,256) # just the histogram
      both = StackHorizontal(scope, hist) #stack waveform and histogram horizontally
      vid = vid.AddBorders(0,0,both.width-vid.width,0) # add a black border to side of the video to match the width of "both"
      StackVertical(both,vid) # stack the analytics over the video
    }
    Then just call that from the body of your script:

    Code:
    import("C:\Program Files (x86)\AviSynth\plugins\WaveAndHist.avs") 
    WhateverSource("filename.ext")
    WaveAndHist()
    If you change the extension from .AVS to .AVSI it will import automatically into every script.

    Do you understand what a waveform graph is? If not, this may help
    https://forum.videohelp.com/threads/340804-colorspace-conversation-elaboration?p=212156...=1#post2121568
    Quote Quote  
  19. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Hi jagabo,

    I just tried the code you provided and Iīm getting some errors, Iīm doing for sure something incorrectly.

    I put the function code as provided without any changes ( are there any adjustments to make?)
    in a file with an avs ( or avsi) extension and put it on the avisynth plug-in directory. The second code piece I adapted to my particular case I made a second avs file

    Code:
    import("C:\Program Files (x86)\AviSynth\plugins\WaveAndHist.avs") 
    AviSource("E:\New captures\test_12.avi")
    WaveAndHist()
    However it gave me an avisynth error " histogram: levels mode only available in PLANAR line 4 ,so I included in the last script that calls the function an additional line
    ConvertToYV12(matrix="Rec601",interlaced=true), this error disappeared but a new one just popped out " I donīt know what "scope" means line 5

    What am I doing wrong ?

    Related to the component capture the idea was interesting but Itīs proving almost impossible to get captures without no clipping at all, doesnīt matter if Iīm capturing between 16-235 or 0-255. Using the video proc amp and the magewell quantization limits I can always adjust the luma limits to whatever I need, but some sort of clipping seems to always occur as it seems the luma boundaries have expanded and the capture card doesn’t know how to adjust so even if the image seems cleaner than through svideo directly or indirectly (using the Panasonics dvd recorder )apparently itīs not a good solution as I will be losing information.

    Also the VCR player included in the combo DVD/VCR Is clearly more limited than the SVHS JVCīs as it doesnīt have TBC, DNR and a dropout compensation circuit as Iīm getting dropouts ( with bad or old tapes ) something that I never had till now with the JVCīs even with bad tapes.

    I tried also to capture directly in component converting RGB to component YUV but the Lindy converter only outputs component when the DVD section works not the VCR. The Panasonics DVD recorder as two RGB settings RGB itself and RGB video the VCR only works if RGB video is selected, however it works well with an RGB Tv set.

    It seems that capturing VHS using RGB/component output will be probably limited to the privileged that do have devices with direct component output (something very rare) but eventually a different variant workflow of this type with different or better devices could eventually work.

    The workflow detailed in graph edit allows me to do captures directly without using Virtualdub and to use other compression codecs not only for example, Utvideo VCM (vfw) codecs recognized by VirtualDub as for example the DMO Utvideo versions and others supported by Magewell. Is there any advantage in using DMO compressors?
    Last edited by FLP437; 13th Nov 2016 at 22:28.
    Quote Quote  
  20. Sorry, on the first line inside the function I had originally called the result of Histogram() "scope". When I posted the script I decided the name "waveform" was more descriptive. But I forgot the change the name later in the function. So on line 5 change "scope" to "waveform".
    Code:
    function WaveAndHist(clip vid)
    {
      waveform = vid.TurnRight().Histogram().TurnLeft().Crop(0,0,-0,256) # just the waveform monitor
      hist = vid.Histogram(mode="levels").Crop(vid.width,0,-0,256) # just the histogram
      both = StackHorizontal(waveform, hist) #stack waveform and histogram horizontally
      vid = vid.AddBorders(0,0,both.width-vid.width,0) # add a black border to side of the video to match the width of "both"
      StackVertical(both,vid) # stack the analytics over the video
    }
    And yes, Histogram() doesn't work with YUY2 video. Converting to YV12 was the correct thing to do. If you're ever unsure what colorspace you have you can use Info(). That will print information onto the video about it's current state, colorspace, frame size, frame rate, etc.
    Quote Quote  
  21. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    In my search for information that can eventually give additional clues for better and more efficient analog video captures ( without breaking the bank ) I have been reading some sites related to historic video preservation most of them related to governmental services.

    On this site I found some recommendations that do not seem to fit with the usual recommendations for analog SD video capture

    http://videopreservation.conservation-us.org/dig_mig/index.html

    “The recommend option are standard NTSC US resolution (525 x 480) at ITU-BT 709 (10-bit RGB with no black or white set points; no CGR) allowing full scale capture, at a signal bandwidth greater than the original. Use of the CGR options, configures incoming signal to be limited to the "broadcast legal" range between (a) the 7.5 IRE setup point (16 RGB) and (b) the 100 IRE maximum white (235 RGB), this applies compression to the historic video signal, which is not desirable.”

    8 bit vs 10 bit
    While both 8- and 10-bit uncompressed video are capable of providing excellent quality broadcast video, 10-bit represents a significantly higher quality and is preferable in many situations. Because 10-bit video has four times the numerical precision when compared to 8- bit, it has a signal-to-noise ratio 12 db higher than 8-bit video.
    advantage of using 10-bit capture of the analog video signal. 8-bit RGB has 256 steps for each of the red, blue and green channels, while 10-bit scaling has 1024 steps. This allows significantly more RGB values (steps) for the incoming signal to be divided into by the ADC
    10-bit video is often used when the source and output video (or “master”) is also 10- bit. Even if the input and/or output video is 8-bit, a 10-bit “project” will still maintain a higher quality when there is a significant amount of effects rendering involved.

    YUV vs RGB
    Selecting the color values to be used in capture is critical to maintain the visual integrity of its appearance. The concept is to capture the image stream in a format that is wider than the original. Secondarily, is to apply no limitation to the signal. The most common limitation is to make the signal "broadcast legal." This means that the signal is compressed to conform to standards for playback via broadcast.
    The problem described is due to the natural compression of ITU-BT 601 (standard) video where the full 256 steps available in 8-bit digital scaling are reduced to 219 steps due to the definition of the ITU 601 standard. This is because of the blackest black is defined as RGB-step 235, rather than 255, and the whitest white is plateaued at RGB 16, rather than RGB 0; is is a range of (235-16) 219 RGB-steps, a significant compression. White the loss of RGB step in the whites is problematic, the loss of 19 steps in the dark is profound.

    Reading these texts and if we do have the necessary hardware/software it seems the better will be to capture
    10 bit RGB 4:4:4 BT.709 (0-255) for analog SD video however if not completely possible are for instance a 8 bit RGB Bt.709 (0-255) better them the usually recommended 8 bit YUV2 4:2:2 (16-235) or are these recommendation not applicable for any reason to standard analog SD captures ?
    Quote Quote  
  22. 10 bit caps won''t be helpful with VHS. The tape only has about 5 bits worth of signal to noise ratio. Capturing with 10 bits will only give you more precise noise.

    4:4:4 chroma subsampling (ie, 720 pixels for the luma and chroma of each line) is overkill for VHS which only has a resolution of only about 350 pixels for the luma and 40 pixels for the chroma across each line.

    https://forum.videohelp.com/threads/319420-Who-uses-a-DVD-recorder-as-a-line-TBC-and-wh...=1#post1981589
    Last edited by jagabo; 17th Nov 2016 at 17:06.
    Quote Quote  
  23. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Itīs not only for VHS they propose this is for SD analog video in general so also S-VHS,video8, HI8 , HI8 XR, etc

    However I do agree with you, but what seems curious is why it is recommended ?

    However is not the 10bit/ 8bit and Chroma subsampling 4:4:4/4:2:2 that eventually interests me as I donīt see also a benefit and even if I do I would be obliged to invest heavy to have a 10 bit functional workflow ( hardware and software).

    My curiosity goes more for the capture between 0-255 and in BT.709 and YUV2 /RGB why ?

    I detected with the DVD/VCR outputting RGB to the Panasonics DVD recorder and capturing over hdmi that I have problems to adjust contrast and brightness with the magewell capture card ( probably with others cards to I will have to confirm ). If I try to capture between 16-235 itīs complicated or impossible to adjust the histogram as I canīt compress the frequencies to fit in between the limits without getting an histogram full of spikes or interruptions/cuts . To get a normal uniform histogram without disturbances I will be getting some sort of clipping on the super blacks or in the superwhites or both that could be very residual or not depending on the tape.

    If I do use contrast to compress to fit to the boundaries I get an abnormal histogram if I do try to get a normal histogram I canīt fit in between the boundaries and I get clipping, losing some information more or less depending on the tape

    However if I choose to record a BT.609 stream( 576i) over HDMI in BT.709 itīs easier to use contrast to compress to fit in between the boundaries 16-235 ( or 0-255 ) maintaining a uniform histogram .

    This as not to do with the DVD/VCR but with the DVD recorder probably because Iīm using Hdmi. It seems that the key that allows me to adjust the histogram without clipping and having a normal histogram is to record in BT.709. This is true for both situation using the RGB input or the svideo input on the Panasonic.

    Related to capturing in 0-255 vs 16-235 can the recommendation be useful as we do have a less compressed recorded stream ?

    also capturing in RGB as most of the times the firt option is to convert it to RGB32 by using virtualdub filters and the capture card can eventually do a better work?
    Last edited by FLP437; 17th Nov 2016 at 22:02.
    Quote Quote  
  24. If you are planning on filtering later it doesn't matter whether you capture full range (0-255) or limited range (16-235). Use whatever works. But most "viewing" formats will expect limited range -- so you'll need to convert to limited range in the end anyway. And for SD video they'll expect rec.601, not rec.709. Both are limited range but the way colors are handled is slightly different.
    Quote Quote  
  25. Member vhelp's Avatar
    Join Date
    Mar 2001
    Location
    New York
    Search Comp PM
    I can't recall ever seeing a vcr put out proper color levels. They seem to all have blown out brights to some degree while the black levels aren't as bad and easier to fix later. The capture card is also to consider in this. They can make what might be slightly below specs even lower. I've been blaming the vcr's for the blown out bright mainly. But I do recall my old Sharp VC-H960 not having this issue. It had good color levels if I recall, but i'll have to double-check that to be sure. If that turns out better, then it might just boil down to the capture card, assuming the process is vcr-to-capture_card. I have a lot of vcr's
    Quote Quote  
  26. Member vhelp's Avatar
    Join Date
    Mar 2001
    Location
    New York
    Search Comp PM
    Can we see some captures of commercial vhs tape from the Magewell ?
    Quote Quote  
  27. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    It seems the magewell card doesnīt have problems with macrovison at least I tried several original commercial VHS tittles and it worked fine. HDCP is another story but an hdmi splitter that doesnīt pass along hdcp works.

    It curious but if the VHS tapes are in good / medium condition the best results are obtained with the VCR tbc/dnr on but if the tape is not in good condition the best results are obtained with the VCR tbc/dnr off.

    Also if a tape have been recorded under medium/poor tv reception tbc on the svhs deck has to be disconnected if not the results are very poor and itīs not only the visual quality that suffers also drops could occur , if the tbc on the deck is disconnected and only the tbc on the card remains we get the best results in this situation and without any dropped or inserted frames..
    Iīm putting a sample from a cartoon film, the sample has been recorded in utvideo YUV2 16-235 .It seems to me but itīs difficult to prove it that capturing full range 0-255 gives marginally better results.
    The results in general are better with direct captures to the magewell than through the Panasonic DVD recorder and after via hdmi to the card.Also brightness and contrast adjust in direct captures are not a problem only when the dvd recorder is in between.
    I have not tried with original tapes with the combo VCR/DVD using the RGB output the results seem to be cleaner but I have problems adjusting contrast and brightness
    Image Attached Files
    Last edited by FLP437; 21st Nov 2016 at 00:54.
    Quote Quote  
  28. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    Iīm trying to develop some scripts to restore VHS this is the result of one of these scripts applied to the raw sample
    However is not optimized an even less for cartoons , for cartoons is possible to do better
    Image Attached Files
    Last edited by FLP437; 21st Nov 2016 at 00:57.
    Quote Quote  
  29. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    here you are a second sample not cartoons this time
    Quote Quote  
  30. Member
    Join Date
    Mar 2015
    Location
    Europe
    Search Comp PM
    I forgot to attach the sample
    Image Attached Files
    Last edited by FLP437; 22nd Nov 2016 at 08:36.
    Quote Quote  



Similar Threads

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