VideoHelp Forum

+ Reply to Thread
Page 3 of 3
FirstFirst 1 2 3
Results 61 to 89 of 89
Thread
  1. Member Cornucopia's Avatar
    Join Date
    Oct 2001
    Location
    Deep in the Heart of Texas
    Search PM
    No, analog does NOT have samples. It is continuous sinusoidal signal.

    Scott
    Quote Quote  
  2. Originally Posted by Cornucopia View Post
    No, analog does NOT have samples. It is continuous sinusoidal signal.
    Sure, it is continuous signal with individual periods representing elements. More elements - more frequency.
    Quote Quote  
  3. Member Cornucopia's Avatar
    Join Date
    Oct 2001
    Location
    Deep in the Heart of Texas
    Search PM
    Individual periods?
    Quote Quote  
  4. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    Originally Posted by SF01 View Post
    I will test it on Saturday.
    Any chance doing some tests?
    Quote Quote  
  5. Originally Posted by dellsam34 View Post
    Originally Posted by SF01 View Post
    I will test it on Saturday.
    Any chance doing some tests?
    You are late, guys. The proof has been provided here by Alwyn:
    https://forum.videohelp.com/threads/407467-Vhs-capture-advice/page12#post2682241

    So yes, the 720x576 is 4:3 i.e. SAR=16:15 in that case of his PAL HDV Camcorder, means ignoring the DVCAM/Rec.601 specs.
    Quote Quote  
  6. I'm a Super Moderator johns0's Avatar
    Join Date
    Jun 2002
    Location
    canada
    Search Comp PM
    vhs-d supports wide screen.
    I think,therefore i am a hamster.
    Quote Quote  
  7. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    Correct, Since Luma, chroma and HiFi audio are modulated into frequencies way higher than them, However when the signal gets demodulate and filtered out by connecting all those periods tips together the resulting signal is a continuous smooth representation of chroma, luma and audio.
    Quote Quote  
  8. Originally Posted by Bwaak View Post
    Analog video has samples, each wave period represents two samples.... ..
    Not quite (or I misread).
    Analog video does not "have samples".
    Analog video is a continuous waveform over time and amplitude. No quantization, no samples.
    When converting an analog video signal to digital one has to sample the analog signal with a sampling frequency which is at least 2x its spectral bandwidth (frequency domain) of the analog signal. For video the sampling frequency for the luma has been standardized to 13.5 MHz (Rec.601), and the spectrum of the analog video signal has to be lowpass- or bandpass filtered (i.e. bandwidth limited) accordingly before being sampled in order to prevent aliasing. (The sampling frequency for the chroma can be "subsampled"=sampled at a lower frequency or decimated according to the different standards.)
    From the sampling process we get time-discrete analog samples, quantized in time, spaced in time 1/13.5MHz=~74ns apart, which can now be quantized in amplitude and converted to bits by means of an analog-to-digital converter. From here onwards we have the bitstream which can be further processed in digital form.

    Anyway, it's going off topic.
    Last edited by Sharc; 8th Mar 2023 at 02:02.
    Quote Quote  
  9. Originally Posted by Sharc View Post
    Originally Posted by dellsam34 View Post
    Originally Posted by SF01 View Post
    I will test it on Saturday.
    Any chance doing some tests?
    You are late, guys. The proof has been provided here by Alwyn:
    https://forum.videohelp.com/threads/407467-Vhs-capture-advice/page12#post2682241

    So yes, the 720x576 is 4:3 i.e. SAR=16:15 in that case of his PAL HDV Camcorder, means ignoring the DVCAM/Rec.601 specs.
    He was shooting DV mode, DV shoots all 720 pixels while still being REC.601 compliant, so while transmitting the image it will not out of the blue crop the 16 additional pixels on sides and substitute them with black when outputting analog video and converter will convert them back into digital. Looking at the wheel it is clearly visible that it's not a perfect circle.

    Apologies for not doing the tests, I am occupied with a priority project that I need to finish quickly, I will try to run some tests on this weekend with analog, digital and test board signals.
    Quote Quote  
  10. Originally Posted by SF01 View Post
    He was shooting DV mode, DV shoots all 720 pixels while still being REC.601 compliant ....
    No, I would have to disagree, it's not Rec.601. And it's not Rec.601 plus filling the picture to 720.
    Proof 1 is given in post#357
    The wheel becomes an exact circle only when playing the full 720x576 frame as 4:3, which means it is not Rec.601.

    Looking at the wheel it is clearly visible that it's not a perfect circle.
    Of course it is not a circle when you view it as stored. But this doesn't mean it is Rec.601 compliant.
    Proof 2: For Rec.601 the SAR would have to be 12:11 (in very close approximation of Rec.601) for PAL. But in fact it is 16:15 (sometimes called "generic") as has been shown by doing the circle test in post#358. So again it is not Rec.601 compliant.
    Last edited by Sharc; 8th Mar 2023 at 12:04. Reason: typos
    Quote Quote  
  11. Originally Posted by Sharc View Post
    Originally Posted by SF01 View Post
    He was shooting DV mode, DV shoots all 720 pixels while still being REC.601 compliant ....
    No, I would have to disagree, it's not Rec.601. And it's not Rec.601 plus filling the picture to 720.
    Proof 1 is given in post#357
    The wheel becomes an exact circle only when playing the full 720x576 frame as 4:3, which means it is not Rec.601.

    Looking at the wheel it is clearly visible that it's not a perfect circle.
    Of course it is not a circle when you view it as stored. But this doesn't mean it is Rec.601 compliant.
    Proof 2: For Rec.601 the SAR would have to be 12:11 (in very close approximation of Rec.601) for PAL. But in fact it is 16:15 (sometimes called "generic") as has been shown by doing the circle test in post#358. So again it is not Rec.601 compliant.
    Right, it is not expanded. However DV itself as produced by native DV camcorder is REC.601 compliant with actually extra image outside regular 4:3/16:9 frame, so that it is REC.470 compatible when output through analog outputs, especially that DV/DVCAM was used by TV stations.
    I will have to further test.

    If that is indeed the case, then I have unnecessarily cropped edges of video digitized using HDV deck. And yet Vegas and other NLE interpret such video with proper REC.601-compliant SAR...
    Quote Quote  
  12. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    I believed that some newer DV and D8 camcorders did not follow the rec.601 SAR, but some diagreed with me. I would still like to see a test from a strictly DV/D8 imaging sensor and not a one that was designed for anamorphic HD 1440x1080, That would confirm my suspecion about those camcorders.
    Quote Quote  
  13. Originally Posted by SF01 View Post
    However DV itself as produced by native DV camcorder is REC.601 compliant with actually extra image outside regular 4:3/16:9 frame, ....
    Yes, adding "extra" image is what a DVCAM/Rec.601 compliant videocam can do. But apparently the HDV Canon of the poster does NOT do it accordingly as the circle test cleary indicated. The HDV Canon ignores Rec.601. It is native 720 rather than 704+filled with extra image. Please do the test with Alywyn's clip.
    And yes do the tests with your camara. Make a shot with a wheel, a circular clock or similar. Maybe YOUR camera is Rec.601/DVCAM compliant. We will see.
    Last edited by Sharc; 8th Mar 2023 at 12:49.
    Quote Quote  
  14. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    Here is a demonstration done by this guy on youtube to record anamorphic analog video into uMatic video tape machine using a pro Sony FS700 camera as the anamorphic digital source and converted the digital output to analog in the 4:3 anamorphic state, Neither the digital to analog converter nor the uMatic machine know what's in the 4:3 frame but when playing back the recording in a 16:9 canvas it produced a perfect video frame without distorsion, albeit with 16 black pixels included in the frame. This can be done on any analog tape format including VHS.
    Quote Quote  
  15. Member Skiller's Avatar
    Join Date
    Oct 2013
    Location
    Germany
    Search PM
    Originally Posted by SF01 View Post
    If that is indeed the case, then I have unnecessarily cropped edges of video digitized using HDV deck.
    Digitized as in the source was analog video converted to DV by an HDV deck? Then it's not affected and follows Rec.601. Simply because it was analog and there is only one standard of how to sample analog video.
    What's affected is video filmed by the HDV camcorder itself in DV mode and when an HDV tape is played back but in DV mode. In other words whenever there is a conversion from HD to SD inside the machine.


    And I'm not too surprised Alwyn's HDV camcorder does not stick to Rec.601 in DV mode. DV mode is just an afterthought on these machines and a dumb downscale from their native HDV resolution.
    I bet it's not so easy to find a native DV camcorder that does not stick to Rec.601.
    Last edited by Skiller; 26th Mar 2023 at 20:22. Reason: typo
    Quote Quote  
  16. Originally Posted by Skiller View Post
    Originally Posted by SF01 View Post
    If that is indeed the case, then I have unnecessarily cropped edges of video digitized using HDV deck.
    Digitized as in the source was analog video converted to DV by an HDV deck? Then it's not affected and follows Rec.601. Simply because it was analog and there is only one standard of how to sample analog video.
    What's affected is video filmed by the HDV camcorder itself in DV mode and when an HDV tape is played back but in DV mode. In other words whenever there is a conversion from HD to SD inside the machine.


    And I'm not too surprised Alwyn's HDV camcorder does not stick to Rec.601 in DV mode. DV mode is just an afterthought on these machines and a dumb downscale from their native HDV resolution.
    I bet it's not so easy to find a native DV camcorder that does not stick to Rec.601.
    Not analog, HDV downsampled to DV, that's the issue we want to check. However, I ws converting NTSC tapes with HDV deck from analog to DV and it was disturbing, because when I was trimming to 704 there was image in the 16 samples.
    In the video about U-matic, the converter clearly does not comply with 470 and 601, as the image is not padded with blanking.
    Quote Quote  
  17. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    Originally Posted by SF01 View Post
    In the video about U-matic, the converter clearly does not comply with 470 and 601, as the image is not padded with blanking.
    It does, Don't you see the black borders? The source camcorder however doesn't but it did provide a perfect anamorphic frame.
    Quote Quote  
  18. There are black borders, but the 16:9 should be entirely between the black borders, not cropped. Unless it's the issue of capture dievice not capturing the borders at all and the ctual blanking was cut off in post and the remaining black borders are still withing 702/704 samples expanded o 16:9 frame. I need to motivate myself to run tests finally...
    Quote Quote  
  19. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    This is the reverse process, so the converter converts the entire scan line by blanking the sides, I'm not entirely sure how this supposed to work because rec.601 defines converting analog to digital not the other way around. Maybe the converter has a limited timing for every scan line so it cannot add blanking over the limited scan line length.
    Quote Quote  
  20. The person says it outputs SD signal over SDI, so it might actually be the camera's fault, not the conevrter, now that I rewatched the video.
    Anyway, I've done test recordings and captures from VHS to DV with DV converter and HDV deck in DV mode, next step is to use HDV mode, output analog and capture, the final step will be to record HDV tape and output DV, I will also be taking photos of the screen.
    Quote Quote  
  21. Attached are the recordings made from test paterns in SD 4:3, SD 16:9, BBC full frame 16:9 with image stretching to full 720 samples, HD 16:9.
    Test patterns were exported as DV files to D8 camcorder and from there output via S-Video to D-VHS recording on standard VHS tape (I suspect this tape is no longer suitable for recording, pity, as I was about to record a cnocert from my collection there).
    The resulting VHS tape was then captured using D8 camcorder acting as a converter to DV, as well as using HDV deck in DV mode.
    Finally HDV video test pattern was recorded on HDV tape and played back analog to D8 camcorder capturing in DV as well as downconverted DV signal straight thruogh FireWire.

    Conclusions:
    1. HDV works in full compliance with REC.601 in EE mode, capturing analog input into DV.
    2. The full BBC test pattern is consistently cut on both sides and the analog recording does not utilize extra samples in area normally designated for digital blanking.
    3. Unfortunately HDV decks do not comply with REC.601 when downconverting HDV to DV signal in anamorphic mode, I hae not tested the trimming the edges mode, but I can assume the resuts will be identical, it will not take into the account that only the middle 702 samples are the active image area and will only include center 4:3 and downsample it to 720 samples.
    Image Attached Thumbnails Click image for larger version

Name:	Philips_PM5544.gif
Views:	4
Size:	19.0 KB
ID:	70016  

    Click image for larger version

Name:	EBU Test card G SD 16-9 EDIT2 1024px.jpg
Views:	3
Size:	106.4 KB
ID:	70017  

    Click image for larger version

Name:	16x9.png
Views:	3
Size:	2.33 MB
ID:	70020  

    Click image for larger version

Name:	PM5644.svg.png
Views:	3
Size:	106.6 KB
ID:	70021  

    Image Attached Files
    Quote Quote  
  22. Member Skiller's Avatar
    Join Date
    Oct 2013
    Location
    Germany
    Search PM
    Thanks for proving. All conclusions 1), 2) and 3) are exactly what was to be expected at this point (imo).


    By the way, when it comes to NTSC, it absolutely can utilize more than 702-704 samples per line, thus this is not a reliable indicator of what's happening. Depending on where you look you'll find numbers claiming it can go up to 710.85 samples, in conjuction with 480-485 active scanlines. This seems to affect mostly old recordings made with analog equipment, prior to digital 480 line formats spreading.
    Quote Quote  
  23. Originally Posted by Skiller View Post
    Thanks for proving. All conclusions 1), 2) and 3) are exactly what was to be expected at this point (imo).


    By the way, when it comes to NTSC, it absolutely can utilize more than 702-704 samples per line, thus this is not a reliable indicator of what's happening. Depending on where you look you'll find numbers claiming it can go up to 710.85 samples, in conjuction with 480-485 active scanlines. This seems to affect mostly old recordings made with analog equipment, prior to digital 480 line formats spreading.
    I performed only PAL tests.

    Yes, I've heard about 710, when the resolution was 486 lines that would add up.
    Quote Quote  
  24. Capturing Memories dellsam34's Avatar
    Join Date
    Jan 2016
    Location
    Member Since 2005, Re-joined in 2016
    Search PM
    You got me lost with HDV and D-VHS so not sure what is supposed to be shown there, Anyway my intial inquery was about later models D8 and miniDV camcorders and tested with recording from a physical test circle not a file or from VHS, So until we have one from that generation I guess we can put this to sleep for now.
    Quote Quote  
  25. Originally Posted by SF01 View Post
    Yes, I've heard about 710, when the resolution was 486 lines that would add up.
    Pre ITU sampling schemes are based on multiple chrominance subcarrier frequency so for NTSC 4fsc we have sampling rate 14.318MHz not 13.5MHz.
    Quote Quote  
  26. Originally Posted by dellsam34 View Post
    You got me lost with HDV and D-VHS so not sure what is supposed to be shown there, Anyway my intial inquery was about later models D8 and miniDV camcorders and tested with recording from a physical test circle not a file or from VHS, So until we have one from that generation I guess we can put this to sleep for now.
    I recorded analog input in D-VHS to check how it DV and HDV decks handle analog input, operation in compliance with 601.
    Then I tested id HDV would properly downsample HDV video to DV, it did not.
    I have DCR-TRV828E, is this modern enough? I will do the tests of shooting a perfect circle to confirm. I also have a Samsung miniDV somewhere, ig I find it, I will also test it.
    Quote Quote  
  27. Member Skiller's Avatar
    Join Date
    Oct 2013
    Location
    Germany
    Search PM
    Originally Posted by SF01 View Post
    I have DCR-TRV828E, is this modern enough?
    Model year is 2001. Sony is known for being a stickler when it comes to standards, so I would be surprised if there exist a single DV (MiniDV/D8) camcorder on the planet made by Sony that doesn't stick to Rec.601.
    I have a MiniDV Sony DCR-TRV33E. Model year is 2003, so quite late. It does stick to Rec.601. I did the test many years ago, would have to redo it to prove. Maybe later.

    I presume Panasonic and JVC are much more likely to be candidates if we ever find one.
    Quote Quote  
  28. I will ask my friends who collects DV camcorders.
    Quote Quote  



Similar Threads