VideoHelp Forum
+ Reply to Thread
Results 1 to 18 of 18
Thread
  1. Hi all- I am trying to capture analog video (specifically 3/4" Umatic, Composite, NTSC) using a MOTU HDX-SDI, but I'm unable to get the full range of IRE. For testing purposes I have sent color bars from a signal generator straight to the MOTU and then back out to an analog waveform/vectorscope. I'm using the CVBS inputs and outputs on the MOTU.

    The most notable problem is in the bottom range- I am unable to capture values below 7.5 IRE. Has anyone else encountered this?

    The SDI ins/outs work fine. I can see the full range of IRE with the same setup and using a digital waveform/vectorscope. I suppose I could route my analog signals through a DigiBeta deck or something else with A-to-D capabilities before going into the MOTU so I would only have to rely on it's SDI ins/outs, but I'm not sure that I want to go that route.

    I'm attaching screenshots for reference. If you can help confirm or clarify what I'm seeing, I would appreciate it! Also, if you have suggestions for work-arounds I would like to hear them. Thank you!!
    Attached Images      
    Quote Quote  
  2. Member
    Join Date: Sep 2007
    Location: Canada
    Search Comp PM
    The most notable problem is in the bottom range- I am unable to capture values below 7.5 IRE. Has anyone else encountered this?
    Do you see Analog Output Format: 480i setup => USA (7.5%) ? Try changing that to 0
    Quote Quote  
  3. Thanks for that. I'm just confused as to why they would give me the 7.5 IRE option if the box wasn't going to handle it correctly. What do you think?
    Quote Quote  
  4. Broadcaster bigass's Avatar
    Join Date: Dec 2006
    Location: Halifax, NS Canada
    Search Comp PM
    When you have the 7.5 IRE option checked, the box DOES handle it correctly. It's an option. You're telling it to do something specific when you.... wait a minute -- you have some pretty awesome scopes and monitors there. I'm guessing you could explain this better than I could. You're no doubt familiar with 'pedestal' and 'setup' and other peculiarities of NTSC. Just an option.
    Quote Quote  
  5. Member
    Join Date: Sep 2007
    Location: Canada
    Search Comp PM
    7.5 IRE is standard for analog - it's "black". When it's checked, it's actually handling it INcorrectly because you're outputting an analog signal. All analog NTSC equipment is supposed to have that 7.5 IRE setup (excect NTSC-J or Japan)

    But digital can have "blacker than black". 7.5IRE "maps" to digital Y'=16 on a Y' 0-255 digital scale when you go from analog => digital . Y' 0-15 is "blacker than black"
    Quote Quote  
  6. I think I understand it better now...

    I do have one additional question- what would be the use case for the setup at 7.5 IRE option?
    Quote Quote  
  7. Member
    Join Date: Sep 2007
    Location: Canada
    Search Comp PM
    Apologies for the confusion in my last post - It's supposed to say when it's "0" is checked it's Incorrect. Again, 7.5 is correct for NTSC (USA) - When it's set to 0 it's for NTSC-J (japan) equipment .

    Again, the usage scenario is NTSC analog=> analog. If you're staying analog, that 7.5 IRE step up is correct .
    Last edited by poisondeathray; 24th Apr 2014 at 16:53.
    Quote Quote  
  8. The deck I'm working with is a regular old NTSC (USA) Umatic deck and I do have my MOTU box set to setup at 7.5IRE.

    However, when I send SMPTE color bars in and out of the CVBS jacks, it cuts off the black level just below 7.5IRE. (see first photo in my initial post)

    I'm attaching diagrams of what my SMPTE color bars should look like on the waveform.
    Attached Thumbnails Click image for larger version

Name:	smpte bars on waveform.jpg
Views:	56
Size:	33.6 KB
ID:	24791  

    Quote Quote  
  9. Name:  smpte bars on waveform 2.JPG
Views: 127
Size:  29.7 KB

    Attached is a better view of the diagram.
    Quote Quote  
  10. Member Cornucopia's Avatar
    Join Date: Oct 2001
    Location: Deep in the Heart of Texas
    Search Comp PM
    I guess it is a matter of "how would the capture device be correctly doing it's job"? Would it be correct if it capped only 7.5-100 IRE and mapping that to Digital Y 16-235 (clipping all else outside that)? or would it be correct if it capped 0-Over100% and mapped 7.5->16, 100->235 and 0->0 and Over 100%->up to 255?

    Because if it is supposed to do the former, my guess is that IT IS. And so when you make the round trip back out to analog, THERE IS NOTHING LEFT that would (ought to?) be below 7.5. If it is supposed to do the latter, well then you do have a problem.

    Of course, well done analog (US NTSC) video SHOULDN'T have anything below 7.5 anyway, so I don't quite see why you're nitpicking about this.

    Scott
    "When will the rhetorical questions end?!" - George Carlin
    Quote Quote  
  11. Thanks for that. That's a very helpful way of framing the issue. I think I would prefer if it do the latter- mapping 7.5->16, 100->235 and 0->0 and Over 100%->up to 255.

    This way would help avoid clipping even though it shouldn't be happening if everything is configured properly. Also, I could continue to use my SMPTE color bars test signal.
    Quote Quote  
  12. Many capture devices and analog processing devices clip Y<16 and Y>235.
    Quote Quote  
  13. video should be 100IRE (16 - 235) always, 7.5IRE is generated by separate circuitry inside video encoder,
    NTSC encoder will squeeze 100IRE signal to 92.5IRE automatically when 7.5IRE pedestal is active, NTSC decoder should remove 7.5IRE pedestal and restore from 92.5IRE video 100IRE signal (all those operations are performed in analog signal domain thus no loss in quantization range, theoretically SNR is slightly worse).
    Quote Quote  
  14. Member 2Bdecided's Avatar
    Join Date: Nov 2007
    Location: United Kingdom
    Search Comp PM
    Originally Posted by jagabo View Post
    Many capture devices and analog processing devices clip Y<16 and Y>235.
    True, but many don't. If I was capturing video that may be out of range, I'd either want decent monitoring and proc amp before capturing to by 100% sure of getting into range, and/or the ability to capture out of range just fine.

    Even "legal" video (according to most broadcast standards) is allowed to have short transients outside the normal range (16-235 digital, 0-700mV PAL, that weird range NTSC) so it's useful to be able to capture outside that range.

    EDIT: Are you sure the waveform graphs you posted are both showing clipping? What are they showing? The first one appears to show luma clipping (blacker-than-black has gone) but not composite clipping (chroma signals below 7.5IRE are intact - it would be catastrophic if they weren't), the second one does seem to have the blacker-than-black pulse intact, but I don't know whether everything else is right or not.

    Cheers,
    David.
    Last edited by 2Bdecided; 28th Apr 2014 at 06:11.
    Quote Quote  
  15. For those looking for background on how to read a waveform/vectorscope, I posted a few articles that came with my DPS Personal V-Scope over on DFaq: http://www.digitalfaq.com/forum/video-restore/4577-how-calibrate-ntsc.html

    Generally NTSC equipment will output below 7.5IRE blacks, even if its "professional". The NewTek Video Toaster was notorious for this. In order to get a true 7.5IRE black matte, you had to use a black frame store as opposed to its matte generator. Capture devices usually don't have a problem with it because they have to take into account the previously mentioned NTSC-J standard.
    Quote Quote  
  16. Originally Posted by 2Bdecided View Post
    EDIT: Are you sure the waveform graphs you posted are both showing clipping? What are they showing? The first one appears to show luma clipping (blacker-than-black has gone) but not composite clipping (chroma signals below 7.5IRE are intact - it would be catastrophic if they weren't), the second one does seem to have the blacker-than-black pulse intact, but I don't know whether everything else is right or not.
    Right, the chroma isn't clipped on the luminance levels. And the second waveform is digital showing the SDI signal path into and out of the MOTU.

    While it's good that only the luminance levels are being clipped in the analog path, I'm still a little concerned. The videos I'm working with were not professionally shot and tend to be all over the place- too bright, too dark, strange hue problems, etc. I am using some proc amps (specifically a TBC) before the signal gets to the MOTU box and I am trying to adjust as best as possible and I've got monitors set up on both the input and output.

    I'm not sure if that's good enough though. Any advice on what gear would best suited for this type of source material? What capture devices and software do capture the full range??? Thanks!!
    Quote Quote  
  17. Originally Posted by 2Bdecided View Post
    Are you sure the waveform graphs you posted are both showing clipping? What are they showing? The first one appears to show luma clipping (blacker-than-black has gone) but not composite clipping (chroma signals below 7.5IRE are intact - it would be catastrophic if they weren't), the second one does seem to have the blacker-than-black pulse intact, but I don't know whether everything else is right or not.
    Right, the chroma isn't clipped only the luminance levels. And the second waveform is digital showing the SDI signal path into and out of the MOTU.

    While it's good that only the luminance levels are being clipped in the analog path, I'm still a little concerned. The videos I'm working with were not professionally shot and tend to be all over the place- too bright, too dark, strange hue problems, etc. I am using some proc amps (specifically a TBC) before the signal gets to the MOTU box and I am trying to adjust as best as possible and I've got monitors set up on both the input and output.

    I'm not sure if that's good enough though. Any advice on what gear would be best suited for this type of source material? What capture devices and software do capture the full range??? Thanks!!
    Quote Quote  
  18. Member 2Bdecided's Avatar
    Join Date: Nov 2007
    Location: United Kingdom
    Search Comp PM
    I think most people here would get the material in-range so it didn't clip during conversion, and then fix the remaining luma and chroma level problems in software. Don't need to get it right in real time, or keep spooling tapes to re-capture problematic sections that way. We're amateurs though.

    No idea about pro-level capture equipment.

    Cheers,
    David.
    Quote Quote