Re: Sony tells DSLR shooters they're idiots

Discussion in 'Digital Photography' started by Mort, Nov 25, 2012.

  1. Mort

    Trevor Guest

    <> wrote in message
    news:...
    >>>>> The purpose of the 24/96 system is to aid in mixing multi-channel
    >>>>> sound.
    >>>>> If you
    >>>>> want to mix down 2 tracks, for example, in 16/44 you would have to
    >>>>> drop
    >>>>> the
    >>>>> levels by 3db. If you want more tracks, you have to cut even more, and
    >>>>> since
    >>>>> each bit represents 3db,

    >>
    >>6dB actually. But don't forget each track you add also gives you 6dB more
    >>level, so you lose no S/N by adding tracks *unless* they contain no
    >>signal.
    >>The idea is to mute or "duck" the tracks where there is no signal.

    >
    > 6db in volts, but db (electricity) is always related to power, which is
    > 3db.


    6dB *SPL* actually, which is what audio is about. Elswehere you quoted 96dB
    dynamic range yourself, which is 16bits by 6dB, the voltage or electrical
    power is irrelevant.

    As for editing, most editors work in 32 or 64 bit internal processing these
    days regardless of the original file bit depth. Just like editing Jpeg, not
    continually saving back to a 16 bit file is a must. Of course starting with
    the highest quality always ensures a better result, just that the advantage
    of recording 16bit Vs 24bit audio is far less than 8 bit Vs 16 bit photo's.

    Trevor.
     
    Trevor, Dec 8, 2012
    1. Advertisements

  2. Mort

    Trevor Guest

    <> wrote in message
    news:...
    >>>>>>> The purpose of the 24/96 system is to aid in mixing multi-channel
    >>>>>>> sound.
    >>>>>>> If you
    >>>>>>> want to mix down 2 tracks, for example, in 16/44 you would have to
    >>>>>>> drop
    >>>>>>> the
    >>>>>>> levels by 3db. If you want more tracks, you have to cut even more,
    >>>>>>> and
    >>>>>>> since
    >>>>>>> each bit represents 3db,
    >>>>
    >>>>6dB actually. But don't forget each track you add also gives you 6dB
    >>>>more
    >>>>level, so you lose no S/N by adding tracks *unless* they contain no
    >>>>signal.
    >>>>The idea is to mute or "duck" the tracks where there is no signal.
    >>>
    >>> 6db in volts, but db (electricity) is always related to power, which is
    >>> 3db.

    >>
    >>6dB *SPL* actually, which is what audio is about.

    >
    > Not when it's inside an electrical device...


    But an electrical power ratio of 3dB *IS* irrelevant when you specify 96DB
    DNR for 16 bit audio! Each bit does NOT "represent 3dB" as you claimed.


    >>Elswehere you quoted 96dB
    >>dynamic range yourself, which is 16bits by 6dB, the voltage or electrical
    >>power is irrelevant.

    >
    > I'm a technician, I design and build audio equipment, I deal with db in
    > power by
    > habit.


    So what, it's still WRONG when used in conjunction with what you said ie. 16
    bit/96dB audio which DOES NOT "represent 3dB each bit".
    Fortunately engineers understand that if some technicians apparently do not
    :)

    Trevor
     
    Trevor, Dec 9, 2012
    1. Advertisements

  3. Mort

    David Taylor Guest

    On 09/12/2012 06:45, Trevor wrote:
    []
    > So what, it's still WRONG when used in conjunction with what you said ie. 16
    > bit/96dB audio which DOES NOT "represent 3dB each bit".
    > Fortunately engineers understand that if some technicians apparently do not
    > :)
    >
    > Trevor


    dB are power ratios. If you double the power, that's a 3 dB increase.
    If you double the voltage, that's four times the power (into the same
    resistance), so it's a 6 dB change. 96 dB would be ~4 x 10^9 in power,
    ~63000 times in voltage. Each bit contributes 6 dB when measured either
    in voltage or power terms.
    --
    Cheers,
    David
    Web: http://www.satsignal.eu
     
    David Taylor, Dec 9, 2012
  4. Mort

    Trevor Guest

    "David Taylor" <> wrote in message
    news:ka1k5p$6t6$...
    >> So what, it's still WRONG when used in conjunction with what you said ie.
    >> 16
    >> bit/96dB audio which DOES NOT "represent 3dB each bit".
    >> Fortunately engineers understand that if some technicians apparently do
    >> not
    >> :)

    >
    > dB are power ratios. If you double the power, that's a 3 dB increase. If
    > you double the voltage, that's four times the power (into the same
    > resistance), so it's a 6 dB change.


    So far so good.


    >96 dB would be ~4 x 10^9 in power, ~63000 times in voltage. Each bit
    >contributes 6 dB when measured either in voltage or power terms.


    And you originally wrote : "each *bit* represents 3db" which is all I
    objected to.
    Your continual blustering does not alter that simple fact.

    Trevor.
     
    Trevor, Dec 9, 2012
  5. Mort

    David Taylor Guest

    On 09/12/2012 09:15, Trevor wrote:
    []
    > And you originally wrote : "each *bit* represents 3db" which is all I
    > objected to.
    > Your continual blustering does not alter that simple fact.
    >
    > Trevor.


    No, you are mistaking me for someone else, this is my first contribution
    on this topic.
    --
    Cheers,
    David
    Web: http://www.satsignal.eu
     
    David Taylor, Dec 9, 2012
  6. Mort

    Trevor Guest

    <> wrote in message
    news:...
    > On Sun, 9 Dec 2012 17:45:53 +1100, "Trevor" <> wrote:
    >><> wrote in message
    >>news:...
    >>>>>>>>> The purpose of the 24/96 system is to aid in mixing multi-channel
    >>>>>>>>> sound. If you
    >>>>>>>>> want to mix down 2 tracks, for example, in 16/44 you would have to
    >>>>>>>>> drop the
    >>>>>>>>> levels by 3db. If you want more tracks, you have to cut even more,
    >>>>>>>>> and since each bit represents 3db,
    >>>>>>
    >>>>>>6dB actually. But don't forget each track you add also gives you 6dB
    >>>>>>more
    >>>>>>level, so you lose no S/N by adding tracks *unless* they contain no
    >>>>>>signal.
    >>>>>>The idea is to mute or "duck" the tracks where there is no signal.
    >>>>>
    >>>>> 6db in volts, but db (electricity) is always related to power, which
    >>>>> is
    >>>>> 3db.
    >>>>
    >>>>6dB *SPL* actually, which is what audio is about.
    >>>
    >>> Not when it's inside an electrical device...

    >>
    >>But an electrical power ratio of 3dB *IS* irrelevant when you specify 96DB
    >>DNR for 16 bit audio! Each bit does NOT "represent 3dB" as you claimed.

    >
    > The 2 measurements are equal as I said, but I see what you mean, yes I
    > should
    > have said 6db, but I'm so used to dealing with 3db it makes no difference
    > to
    > me...
    >
    >>>>Elswehere you quoted 96dB
    >>>>dynamic range yourself, which is 16bits by 6dB, the voltage or
    >>>>electrical
    >>>>power is irrelevant.
    >>>
    >>> I'm a technician, I design and build audio equipment, I deal with db in
    >>> power by
    >>> habit.

    >>
    >>So what, it's still WRONG when used in conjunction with what you said ie.
    >>16
    >>bit/96dB audio which DOES NOT "represent 3dB each bit".
    >> Fortunately engineers understand that if some technicians apparently do
    >> not
    >>:)

    >
    > I understand db quite well, having spent 50 years as a communications
    > tech,
    > repairing the crap engineers come up with... :)



    You should hang your head in shame for getting it so publicly wrong then!
    :)
    (your claim that "each bit represents 3db' still included above!)
    Not to mention the whole point of your post that you lose significant S/N
    ratio with every added track is mostly wrong in any case, with any modern
    editing software at least.
    Perhaps you are 50 years behind the times? :)

    Trevor.
     
    Trevor, Dec 10, 2012
  7. Mort

    Trevor Guest

    <> wrote in message
    news:...
    > On Mon, 10 Dec 2012 20:42:01 +1100, "Trevor" <> wrote:
    >>You should hang your head in shame for getting it so publicly wrong then!
    >>:)
    >>(your claim that "each bit represents 3db' still included above!)
    >>Not to mention the whole point of your post that you lose significant S/N
    >>ratio with every added track is mostly wrong in any case, with any modern
    >>editing software at least.
    >>Perhaps you are 50 years behind the times? :)

    >
    > We were discussing straight 16 bit audio mixing and I pointed out the
    > flaws,
    > that's all.


    And incorrectly since when you add tracks the signal is usually correlated
    and the noise is uncorrelated, thus the S/N ratio often increases, NOT
    reduces as you claimed. Unless you add tracks with no signal, which as I
    pointed out is not necessary.


    >I don't know why the industry felt the need to go to 24 bit since,
    > as you know, software uses 32bit floating point and converts to 16 anyway.


    Nothing wrong with 24bit, there is simply no need to transfer 4 bytes for
    every sample when 3 bytes is already more than any real world converter will
    ever need.


    > Perhaps it was a marketing scheme?


    Nope, 24 bits is the optimum for digital recording. Editing is better, and
    easy to do in 32 or 64 bits.


    > Hey, go back 50 years and check out the available audio mixers!


    Why, I'm glad I don't have to use them any more, others moan the loss of
    valves and transformers they used.

    Trevor.
     
    Trevor, Dec 11, 2012
  8. Eric Stevens <> wrote:
    > On Thu, 6 Dec 2012 17:03:01 +0100, Wolfgang Weisselberg
    >>Eric Stevens <> wrote:


    >>> control colour balance and that the F801s could only control exposure.
    >>> It did this on the basis of a data base of thousands of images. It
    >>> must have had color sensitivity of some kind as, for example, it could
    >>> tell the difference between a large white-walled building and a snow
    >>> scene.


    >>So you're basically saying that there's no way to detect the
    >>difference between a large white-walled building and a snow
    >>scene without colour sensitivity, not even by seeing that the
    >>snow scene was much brighter than the building?


    > But was it much brighter than the building?


    So why do you think snow is a different colour than a white
    walled building?

    -Wolfgang
     
    Wolfgang Weisselberg, Dec 16, 2012
  9. Eric Stevens <> wrote:
    > On Sun, 16 Dec 2012 18:25:21 +0100, Wolfgang Weisselberg
    >>Eric Stevens <> wrote:
    >>> On Thu, 6 Dec 2012 17:03:01 +0100, Wolfgang Weisselberg
    >>>>Eric Stevens <> wrote:


    >>>>> control colour balance and that the F801s could only control exposure.
    >>>>> It did this on the basis of a data base of thousands of images. It
    >>>>> must have had color sensitivity of some kind as, for example, it could
    >>>>> tell the difference between a large white-walled building and a snow
    >>>>> scene.


    >>>>So you're basically saying that there's no way to detect the
    >>>>difference between a large white-walled building and a snow
    >>>>scene without colour sensitivity, not even by seeing that the
    >>>>snow scene was much brighter than the building?


    >>> But was it much brighter than the building?


    >>So why do you think snow is a different colour than a white
    >>walled building?


    > http://skaremedia.com/sites/default/files/imagecache/resize_to_page/flickr/MyEye.jpg


    Using an instrument with an amazingly variable automatic
    white balance (amongst other really wild stiff) --- which is
    also fed and strongly influenced by "how it should look" ---
    is a very good method to measure.

    You can't even tell what thing is brighter when they're only
    a couple centimeters away from each other --- see optical
    illusions.

    Next up (and with much more justification): Santa Claus really
    exists and brings everyone gifts, I've seen him in dozens of
    copies in the mall.

    -Wolfgang
     
    Wolfgang Weisselberg, Dec 19, 2012
  10. Eric Stevens <> wrote:
    > On Wed, 19 Dec 2012 01:20:17 +0100, Wolfgang Weisselberg
    >>Eric Stevens <> wrote:
    >>> On Sun, 16 Dec 2012 18:25:21 +0100, Wolfgang Weisselberg
    >>>>Eric Stevens <> wrote:
    >>>>> On Thu, 6 Dec 2012 17:03:01 +0100, Wolfgang Weisselberg
    >>>>>>Eric Stevens <> wrote:


    >>>>>>> control colour balance and that the F801s could only control exposure.
    >>>>>>> It did this on the basis of a data base of thousands of images. It
    >>>>>>> must have had color sensitivity of some kind as, for example, it could
    >>>>>>> tell the difference between a large white-walled building and a snow
    >>>>>>> scene.


    >>>>>>So you're basically saying that there's no way to detect the
    >>>>>>difference between a large white-walled building and a snow
    >>>>>>scene without colour sensitivity, not even by seeing that the
    >>>>>>snow scene was much brighter than the building?


    >>>>> But was it much brighter than the building?


    >>>>So why do you think snow is a different colour than a white
    >>>>walled building?


    >>> http://skaremedia.com/sites/default/files/imagecache/resize_to_page/flickr/MyEye.jpg


    >>Using an instrument with an amazingly variable automatic
    >>white balance (amongst other really wild stiff) --- which is
    >>also fed and strongly influenced by "how it should look" ---
    >>is a very good method to measure.


    >>You can't even tell what thing is brighter when they're only
    >>a couple centimeters away from each other --- see optical
    >>illusions.


    >>Next up (and with much more justification): Santa Claus really
    >>exists and brings everyone gifts, I've seen him in dozens of
    >>copies in the mall.


    > So we now have it on record that you think you don't know what you are
    > looking at.


    We have it on record that *you* think your eyes are reliable
    colour measurement instruments --- even when used in different
    circumstances and without being able to see both scenes at the
    same time --- whereas *I* claim they can easily be tricked /and/
    have given you enough pointers to research that on your own.

    Then we have it on record that you believe knowing that the eye
    can be tricked equals "you don't know what you are looking at".

    Maybe you /can/ bring proof that *your* eyes are good, absolute
    colourimeters, but I wouldn't bet a cent to an Earldom on it.


    Oh, BTW, let me congratulate you on embarking on your quest
    to ridicule me since I don't agree with you. But look up what
    Ghandi said about that.

    -Wolfgang
     
    Wolfgang Weisselberg, Dec 20, 2012
  11. Eric Stevens <> wrote:
    > On Thu, 20 Dec 2012 16:58:32 +0100, Wolfgang Weisselberg
    >>Eric Stevens <> wrote:
    >>> On Wed, 19 Dec 2012 01:20:17 +0100, Wolfgang Weisselberg
    >>>>Eric Stevens <> wrote:
    >>>>> On Sun, 16 Dec 2012 18:25:21 +0100, Wolfgang Weisselberg
    >>>>>>Eric Stevens <> wrote:
    >>>>>>> On Thu, 6 Dec 2012 17:03:01 +0100, Wolfgang Weisselberg
    >>>>>>>>Eric Stevens <> wrote:


    >>>>>>>>> control colour balance and that the F801s could only control exposure.
    >>>>>>>>> It did this on the basis of a data base of thousands of images. It
    >>>>>>>>> must have had color sensitivity of some kind as, for example, it could
    >>>>>>>>> tell the difference between a large white-walled building and a snow
    >>>>>>>>> scene.


    >>>>>>>>So you're basically saying that there's no way to detect the
    >>>>>>>>difference between a large white-walled building and a snow
    >>>>>>>>scene without colour sensitivity, not even by seeing that the
    >>>>>>>>snow scene was much brighter than the building?


    >>>>>>> But was it much brighter than the building?


    >>>>>>So why do you think snow is a different colour than a white
    >>>>>>walled building?


    >>>>> http://skaremedia.com/sites/default/files/imagecache/resize_to_page/flickr/MyEye.jpg


    >>>>Using an instrument with an amazingly variable automatic
    >>>>white balance (amongst other really wild stiff) --- which is
    >>>>also fed and strongly influenced by "how it should look" ---
    >>>>is a very good method to measure.


    >>>>You can't even tell what thing is brighter when they're only
    >>>>a couple centimeters away from each other --- see optical
    >>>>illusions.


    >>>>Next up (and with much more justification): Santa Claus really
    >>>>exists and brings everyone gifts, I've seen him in dozens of
    >>>>copies in the mall.


    >>> So we now have it on record that you think you don't know what you are
    >>> looking at.


    >>We have it on record that *you* think your eyes are reliable
    >>colour measurement instruments --- even when used in different
    >>circumstances and without being able to see both scenes at the
    >>same time --- whereas *I* claim they can easily be tricked /and/
    >>have given you enough pointers to research that on your own.


    > You are reading too much into what I wrote, unless you believe white
    > walls are never found in the presence of snow.


    Oh, sure, there are. So, what did the camera make from that
    scene?


    >>Then we have it on record that you believe knowing that the eye
    >>can be tricked equals "you don't know what you are looking at".


    > Well, doesn't it?


    Say, how do you know you saw white walls in the presence of
    snow, or can't your eyes be tricked? Probably that was a
    yellow wall and smog-stained snow you say ...

    >>Maybe you /can/ bring proof that *your* eyes are good, absolute
    >>colourimeters, but I wouldn't bet a cent to an Earldom on it.


    > This is not just a discussion of colour.


    Well, either colour measuring is necessary, or it's not. In
    the latter case "It must have had color sensitivity of some
    ^^^^
    kind" is proven wrong --- which was my point, in the former
    case it *is* a discussion of colour and whether your eyes can
    report them correctly.

    >>Oh, BTW, let me congratulate you on embarking on your quest
    >>to ridicule me since I don't agree with you. But look up what
    >>Ghandi said about that.


    > Tsk, tsk. You are too sensitive.


    .... for your ploys to work.

    > Maybe you have seen too many Santa
    > Clauses in the mall?


    Are they spitting images of you, or where's the connection?

    -Wolfgang
     
    Wolfgang Weisselberg, Dec 25, 2012
    1. Advertisements

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Pat McGroyn

    Re: Sony tells DSLR shooters they're idiots

    Pat McGroyn, Nov 24, 2012, in forum: Digital Photography
    Replies:
    8
    Views:
    403
    PeterN
    Nov 27, 2012
  2. Anthony Polson

    Re: Sony tells DSLR shooters they're idiots

    Anthony Polson, Nov 24, 2012, in forum: Digital Photography
    Replies:
    1
    Views:
    298
    Trevor
    Nov 26, 2012
  3. Anthony Polson

    Re: Sony tells DSLR shooters they're idiots

    Anthony Polson, Nov 24, 2012, in forum: Digital Photography
    Replies:
    8
    Views:
    360
    Whisky-dave
    Nov 26, 2012
  4. RichA

    Re: Sony tells DSLR shooters they're idiots

    RichA, Nov 25, 2012, in forum: Digital Photography
    Replies:
    1
    Views:
    251
    Anthony Polson
    Nov 25, 2012
  5. Tim Conway

    Re: Sony tells DSLR shooters they're idiots

    Tim Conway, Nov 25, 2012, in forum: Digital Photography
    Replies:
    0
    Views:
    273
    Tim Conway
    Nov 25, 2012
Loading...

Share This Page