1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

622 HDMI worse than component

Discussion in 'DISH™ High Definition Discussion' started by knealy, Apr 25, 2006.

Thread Status:
Not open for further replies.
  1. harsh

    harsh Beware the Attack Basset

    21,192
    183
    Jun 14, 2003
    Salem, OR
    I fully understand the term duty cycle, but I didn't bring it up because of possible confusion by an unfamiliar term. That exercise is left to those who might use those terms to shock and awe (or don't understand the audience).
    Why would you want to convert from digital to analog (except for CRT)? The kinds of transistors found in LSI and VLSI chips are typically not amplifiers -- they are binary switches. If they need to vary something, they do it using duty cycle.
    The signal is modulated into sinus radio frequencies. This is an analog process. Computer people might recognize this as something very much like a MODEM only at a much higher frequency.

    As for your understanding of how LCD works, here's an explanation from the leader in the use of the technology for TVs:

    http://sharp-world.com/sc/library/lcd_e/s2_1_1e.htm

    The key is not the twisting, but rather the spinning of the rods to an endwise orientation which allows the light to pass.
     
  2. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004

    Okay knealy, I believe that you see what you see! I apologise if I gave you the impression that you were crazy or anything.

    Let me ask you a couple of questions.

    You are using a DVI output on the 921.
    When you connect the 622, are you going HDMI to HDMI, or are you using an HDMI to DVI converter on your TV? Are you plugging the thing in the exact same plug on the back of the TV, that being a DVI socket on your Sony in both cases?
     
  3. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004
    We aren't talking about chips here. We are talking about an LCD TV panel that is huge! We are not talking LSI, or VLSI. We are talking three transistors per pixel on a 1080i LCD panel. If anything it is VSSI (Very Small Scale Integration) because you have only 6 million transistors (not even a gate) on a panel that is at least 45 inches in diagonal!

    I think I mis-read what you said there the first time I glanced at it. I think I get what you are trying to say. The transistors on the display are switches that are driven with a duty cycle to modulate intenisty. There has to be one more element to make that work right. If you are varying the switching by modulating duty cycle, then there needs to be something to do the averaging. That would suggest that there either needs to be either a capacitor in there, or the LCD elements that twist need to not be able to react fast enough to the individual switches to do the averaging. If this is the case, then the D/A conversion is done by a process of switching via duty cycle and averaging either by a capacitor or the physical limiations of the liquid crystal twisting.

    In your link, I don't see the word "spin", or "rod", or "endwise", although your explantion sounds okay. I did find "twist" though.

    Here are a few quotes from your link:

    "Light passes through liquid crystals, following the direction in which the molecules are arranged. When the molecule arrangement is twisted 90 degrees as shown in the figure, the light also twists 90 degrees as it passes through the liquid crystals."

    "This figure depicts the principle behind typical twisted nematic (TN) liquid crystal displays"

    "The molecules along the upper plate point in direction 'a' and those along the lower plate in direction 'b,' thus forcing the liquid crystals into a twisted structural arrangement."

    Light bends 90 degrees as it follows the twist of the molecules

    When voltage is applied to a combination of two polarizing filters and twisted liquid crystal, it becomes a LCD display

    "A combination of polarizing filters and twisted liquid crystal creates a liquid crystal display."
     
  4. knealy

    knealy Legend

    120
    0
    Jul 5, 2002
    As I mentioned, I am plugging into the same input on the TV. It is DVI. So I'm using a DVI male to HDMI male cable which goes from the TV to the 622. In the case of the 921 I used a DVI male to DVI male (same input on the TV, -DVI). So for the 622 I had to use a different cable, but no adapter.

    I've tried two different DVI to HDMI cables, one from Gefen, who specializes in such things, and the other from Monster who also made the DVI to DVI cable I was using with the 921. Both brands produced the same results on the 622, -dull and not as clear as the component connection.
     
  5. harsh

    harsh Beware the Attack Basset

    21,192
    183
    Jun 14, 2003
    Salem, OR
    The "twisting" is what gives an LCD increased opacity when not charged (flaccid?).

    The key sentence from the link was the last one:
    What your suggesting is that the voltage is varied to force a behavior more like an iris. While LCD's can be grayed by tinkering with the voltage, it is somewhat difficult to control uniformity. A few pages later, Sharp explains that a TFT (Thin Film Transistor) is a switching transistor (diode).

    http://sharp-world.com/sc/library/lcd_e/s2_4_3e.htm

    (the active matrix diagram)
     
  6. harsh

    harsh Beware the Attack Basset

    21,192
    183
    Jun 14, 2003
    Salem, OR
    I would be asking Dish for an advance replacement of your ViP622.
     
  7. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004
    Nuts! I tried to edit my post when I got home from work, but it was already replied to. You can see my edited post there. It does say that it is a switching transistor. I still do think that changing the voltage acts like an iris, one of the previous pages seemed to hint at this. I think that because it is a switching transistor that varying a duty cycle will do the same thing as changing the voltage. If you vary the duty cycle fast enough, the liquid crystal twisting may not be able to twist back all the way between pulses. It takes time to move mass. This could be a natural way of doing the averaging via a duty cycle. I don't see any capacitors in the diagram, however, there may be some sort of capacitance in there too. It sounds like doing high speed duty cycle type switching and averaging with the effects of the liquid crystals twisting and untwisting are doing the digital to analog conversion.
     
  8. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004
    Are you watching the same channel on both the 622 and 921? For instance, if you are watching an MPEG-4 encoded channel on the 622 that would not be good because the bits coming out of the unit would be different from the different encoding/decoding schemes.

    Is there a way that you could temporarily take the component pictures out of your test and just compare the DVI on the 622 against the DVI on the 921? It is entirely possible that the component outputs are biased differently on the two units.

    Have you tried viewing a test pattern on the two units? HDNet has a test pattern that runs at 5am on Tuesday mornings that lasts for ten minutes. I don't know if that is still the same time for it. The tests weren't as good as video essentials, but it might give you a more controlled basis for comparison.
     
  9. knealy

    knealy Legend

    120
    0
    Jul 5, 2002
    Well the 921 was sent back for my $200 credit. I had also returned a 501 for a replacement. Unfortunately they gave me a $25 credit for the 501 in exchange for the 622. So it took quite a while to straighten out that mess. They claimed they hadn't received a 921. But I had the UPS tracking info showing that they signed for it two days before the 501 arrived. Then they said it only indicated that they'd got a package, but not what was in it. So I asked them to tell me what was in it. Finally they relented and told me I'd be getting the credit.

    Now to answer your question. Yes, I used the same channels, HD and SD to check the results. I also paused the HD picture so I'd be sure to have the same image.

    Don't know where to go from here except just to use the component video.
     
  10. Mr. Crowley

    Mr. Crowley New Member

    1
    0
    Apr 27, 2006
    You guys need to realize DVI is 8-bit and HDMI is 10-bit, so there will be a difference when using the DVI input. If you want to read further read this (just add www at the beginning):

    avsforum.com/avs-vb/showthread.php?t=556134
     
  11. Jeff McClellan

    Jeff McClellan Hall Of Fame

    1,266
    0
    Apr 22, 2002
    On a Sony LCD-RP, component does have a better picture with the 622 then HDMI. For now at least.:)
     
  12. LtMunst

    LtMunst Hall Of Fame

    1,267
    3
    Aug 24, 2005
    This is nuts. You are trying to say that software drivers have nothing to do with the video output? :rolleyes: Wonder how software updates have been able to fix video issues in the past (like the 942 smearing issue)? Guess the updates somehow changed out the chipset. :lol:


    All 3 of these receivers are specialized computers running on the Linux operating system. Just like your home PC, anything displayed on your monitor is the result of both hardware and software processing. There is no straight pipe from the MPEG decoder chipset to the HDMI output.
     
  13. LtMunst

    LtMunst Hall Of Fame

    1,267
    3
    Aug 24, 2005

    Ok...lets play semantics a little more. There is NO SUCH THING as a digital broadcast signal. Last I checked, all radio transmissions are in the form of an analog sine wave. Can that analog wave carry digital info?....sure. But that requires an A/D conversion.

    Carrying this even further....Every process within the receiver or TV is completed using an electrical current based on a 60hz alternating sine wave. In this respect, ALL functions are analog.

    Bottom line is that you are mis-applying the term DAC. A DAC, as used in audio/visual components performs one of two functions.

    #1 Converts digital instructions into a single audio sine wave which is directly amplified by a speaker.

    #2 Converts digital instructions into a single sine wave used by a cathode ray to progressively scan across a phosphor matrix.

    That's it.

    An LCD does not use #2. To create a picture on an LCD, the processor creates a digital map of the pixel matrix for each frame identifying which subpixels need to be turned on and at what intensity. These instructions are then sent to the matrix at one time. The matrix is not refreshed line by line. There is no single analog wave driving the pixel refresh. Are individual analog signals sent to specific electrical components to make this happen?...Sure. It is not DAC, though.
     
  14. SaltiDawg

    SaltiDawg Active Member

    2,384
    0
    Aug 30, 2004
    After your numerous pompous and arrogant posts, suggesting a lack of understanding of the issue, you have the cajones to admit this?:rolleyes:

    Ploink!:nono2:
     
  15. olgeezer

    olgeezer Guest

    1,833
    0
    Dec 5, 2003
    Let's see if we can make any sense. I'm not an engineer, which may be good, but I am a salesman which is 2 steps above pond scum. This is what I think. Your eyes are always the best judge. Nothing has changed in your set up except the receiver. The only possibilities that are left are: 921 better HMDI/DVI than 622, unlikely to be different. 622 better D/A conversion than 921 possible. Display and cables have remained constant so that D/D or A/D conversion changes on display are unlikely. If your DVI/HMDI cabling are greater than 7 meters, there is a possibility of working on one but not the other, but that normally is a cliff not a gradual decline. The 921 was DVI the 622 is HMDI. Another possibility is the HMDI cable. The only thing you can change is the cable. I have a 622 on a analog rear projection. I did notice an improvement on component although not enough to disconnect my DVI/HMDI. At work on 7 digital displays from a 411 I've seen an improvement on component. My best guess would be that the 622 has better D/A conversion than the 921.:D
     
  16. TechnoCat

    TechnoCat Legend

    156
    1
    Sep 4, 2005
    True but completely irrelevant. Data loss doesn't tend to be a one-time thing, and HDMI signals are far more fragile than Component ones.

    Yes, that's one of the points I made, but then you added...
    Name 5 in the top 10 of plasma sales. The typical 42" plasma panel is 1024 x 768. 720p is 1280x720.

    Did you even READ my post, or did you just decide to argue with every sentence? That's pretty much what I said - that they have to rescale on the monitor side.

    I've worked in that field, even developing codecs. From your desire to argue rather than read already demonstrated, I don't think you have a real good concept of what I said. And I was simplifying dramatically for comprehension; fully describing what it really does requires a LOT of VERY complex math.

    Decent scaling takes care of blotching/etc., as would antialiasing or a number of other techniques that are usually built into the scaling. Not that they can fix over-compression, but our cable company was far worse, and so was the DirecTV images I saw last year.
     
  17. dave1234

    dave1234 Icon

    511
    0
    Oct 9, 2005
    But all Mpeg2 is 8 bit so in the end it doesn't matter if HMDI is 10bit and DVI is 8 bit.
     
  18. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004

    You're right LtMunst, software does write to control registers for the decoders and scalers and such, and that can have a huge effect on what is viewed. That code, while it is probably part of a driver doesn't have to be in a driver. As far as chipsets being changed on the fly may sound laughable, it can happen (I'm not suggesting the 622 is this way though)! If you recall, however, the 921 has programmable hardware in it. They got software updates in it that programmed FPGAs. Those chips are generally expensive and used mainly for development, but there are a fair number of shipping products with them.

    As far as the there being no straight pipe from the decoder chip to the digital output goes, you might want to think about that a bit. You are right, these receivers are computers running Echostar DVR software. These computers do not have the processing power to to be in the critical path between the decoder chipset and the video output. The software may put menus and programming guides on the screen by writing to video memory, that software does not process all of the bits going from the decoder to the digital output.
     
  19. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004
    The LCD doesn't use #1, or #2. There are more kinds of DACs though.

    How about #3,
    The Pulse Width Modulator, the simplest DAC type. A stable current (electricity) or voltage is switched into a low pass analog filter with a duration determined by the digital input code.

    In this case, the pulse width modulator (PWM) is what is happening in your LCD. The low pass filter is accomplished via the physical properties of the twisting liquid crystals in the display's pixel.
     
  20. jsanders

    jsanders Hall Of Fame

    1,191
    0
    Jan 21, 2004
    Cool! Maybe you can enlighten us a bit then. What codecs have you developed? I'm aware that the DCT based compression is complicated. Could you take a couple of paragraphs to give us a bit more details? I'm interested to know. Why do they do Discrete Cosine Transforms instead of Discrete Fourier Transforms? It seems that they are similar in what they do.
     
Thread Status:
Not open for further replies.

Share This Page