Frequency response of video switchers in preamp

Discussion in 'Archived Threads 2001-2004' started by Sam D, Aug 27, 2002.

  1. Sam D

    Sam D Auditioning

    Joined:
    Aug 26, 2002
    Messages:
    8
    Likes Received:
    0
    I want to understand video bandwidth, and frequency response of video switchers.

    While investigating surround preamps, I've noticed several that some have a video frequency response of dc-10mhz, while some are dc-100mhz.

    For instance, the Rotel 1066 is dc-10mhz, the Anthem SF20 is dc-100mhz.

    I want to know how much response I really need, how much only applies in specialized situations.

    For video components, I own a Sony 32hv600 tv and a Sony ns715p dvd. The dvd player has a 108mhz dac, which means that the maximum bandwidth is 54mhz, if the dac doesn't oversample (I don't know if it does or not).

    So, what I think I care about, is how much bandwidth is necessary to provide a good connection between those two components. I run the dvd player in progressive mode. Also, I probably care about near-future technologies which will require more bandwidth.

    I scanned through the faq's but I might have missed the discussion if it exists there.

    Thanks!
     
  2. Bob McElfresh

    Bob McElfresh Producer

    Joined:
    May 22, 1999
    Messages:
    5,182
    Likes Received:
    0
    Hi Sam. Welcome to HTF! [​IMG]
    The Video Bandwidth spec on a receiver is a fairly new thing. Ordinary, interlace video signals top out at about 4 Mhz. So to correctly pass video, you want the receiver to have a bandwidth higher than this.
    The 3 types of video signals in your system have upper frequencies of:
    Ordinary Video: 4 Mhz
    Progressive Video: 7 or 13 Mhz (I'm having trouble finding a definitive number for this)
    HD Video (1080p): 35 Mhz
    So the Rotel unit is designed only for ordinary/interlace video.
    The Anthem unit is designed to handle Progressive/HD video.
    Note: this is only an issue if you only have 1 HD input on your TV and more than 1 HD/Progressive source. We often advise people to wire the video from their DVD player, or HD Box directly to the TV.
    On the other hand, I can see a system that has the following sources:
    - Progressive DVD player
    - STB (Set Top Box for over-the-air HD signals)
    - XBox with the Advanced/Progressive video output
    For this system, having a receiver that can switch HD signals with 100 Mhz bandwidth is very nice.
    Hope this helps.
     
  3. Sam D

    Sam D Auditioning

    Joined:
    Aug 26, 2002
    Messages:
    8
    Likes Received:
    0
    Thanks that all makes sense. I did some more google-ing and I think the 480p is 13.5 mhz, which means the video dac on my dvd player is 8x oversampling. And, also means that a 10mhz switcher probably looks pretty cruddy with a 480p signal. I can understand why you recommend going straight to the tv and not through the switcher.
     
  4. David Guill

    David Guill Extra

    Joined:
    Nov 9, 1999
    Messages:
    16
    Likes Received:
    0
    The MHz rating for A DAC refers to the sampling rate of the digital signal to "accurately" rebuild the analog signal. Nyquists Theorem tells us to get the "general" characteristics of the waveform we need to sample the signal at twice its frequency (hence the 108 MHz, twice the 54 MHz signal).
     
  5. Bob McElfresh

    Bob McElfresh Producer

    Joined:
    May 22, 1999
    Messages:
    5,182
    Likes Received:
    0
    David is correct. Be very careful to NOT try and calculate video frequency issues by looking at the CPU or DAC speed on the DVD player. Yes, these chips bridge the digital & analog domain, but the relationship between clock-speed and video-frequency handling is not quite as simple.

    Example: Does a video game run twice as fast on a 800 Mhz computer than a 400 Mhz? Of course not. But it likely runs a bit smoother.
     
  6. Sam D

    Sam D Auditioning

    Joined:
    Aug 26, 2002
    Messages:
    8
    Likes Received:
    0
    David's post seems to imply that you can derive the frequency repsonse of a dac by halving its clock frequency, which is not correct in the presence of oversampling.

    But I do agree that my methodology of figuring out video bandwidth from a dac clock frequency is flawed, which is why I posted the question.

    Thanks!
     

Share This Page