Told you Sony's 24mp sensor was noisy

Discussion in 'Digital Cameras' started by Rich, Oct 26, 2011.

  1. Rich

    Rich Guest

    Rich, Oct 26, 2011
    #1
    1. Advertisements

  2. Rich

    Me Guest

    With the Pentax K5 it looks like a /shitload/ of NR goes on between
    reading the sensor and so-called "raw file".
     
    Me, Oct 26, 2011
    #2
    1. Advertisements

  3. Rich

    Bruce Guest


    The level of high ISO noise is quite frightening. We had some images
    from both an NEX-7 and an A77 at a Sony event. The noise levels at
    high ISOs were extremely high. We were reassured that the cameras
    were pre-production samples and that the production cameras would be
    much better. It seems that they aren't.

    Sony Alpha has a long history of disappointments and this is just one
    more to add to the list. You would think that Sony would learn from
    their mistakes, but they just keep on making the same ones, over and
    over again.


    Sony makes some excellent sensors that perform exceptionally well in
    other brands of camera. But put the same (or similar) sensor in a
    Sony camera, and the results are very disappointing.

    The 16 MP sensor in the Pentax K-5 is also used in the Nikon D7000 and
    several Sony models. In the Pentax and Nikon it delivers excellent
    image quality with very good noise control - the Pentax just shades
    the Nikon in this respect. But put the same sensor in a Sony A35 or
    NEX-C3, and it becomes one of the noisiest sensors on the market.

    The same was true of the 24 MP full frame A900 (and later A850) whose
    high ISO noise was desperately bad. Yet the Sony-made 24 MP sensor in
    the full frame Nikon D3X performed extremely well. However the
    sensors are not the same, so the conclusions here are less clear,
    except that if you want low noise at high ISOs, don't buy Sony.

    I had hoped that the A77 would be good enough to win over Sony A700
    users and encourage them to stay with the brand (and buy some of our
    stocks of unsold Sony and Minolta lenses!) but they will be switching
    brands now. Most leaving Sony seem to go to Nikon.

    There could be no excuse for Sony getting the A77 wrong, but ...

    Next, the A99, or A9X as Sony UK prefers to call it. Another profound
    disappointment in the making. :-(
     
    Bruce, Oct 26, 2011
    #3
  4. Rich

    RichA Guest

    If it results in loss of detail or quality, that's not good.
     
    RichA, Oct 26, 2011
    #4
  5. Rich

    RichA Guest

    On a pixel size to pixel size basis, the Panasonic GH2 sensor beats
    all the Sony cameras. Again, if they ever scale up the technology to
    APS or (dare I say it?) FF size, it will be amazing. The Sony A700
    body was unique and remarkably well-suited to the hand, IMO. The new
    77 is really nothing like it.
     
    RichA, Oct 26, 2011
    #5
  6. Rich

    Bowser Guest

    But the NEX looks better than the A77 since it receives more light. I
    expected higher noise levels from the more densely packed sensor, but
    what I'd like to see is prints from these cams. Viewing these images
    at 100% on screen tells me little.

    My NEX 7 is still on pre-order.
     
    Bowser, Oct 26, 2011
    #6
  7. Rich

    Bruce Guest


    That is very unlikely ever to happen, though.


    Sony makes soooo many mistakes in product design.

    I agree, the A700 handled well. It was the best of the Alpha DSLR
    range, and the de facto top-of-the-range model, given the sheer
    incompetence of the full frame A900 and A850. The A700 had/still has
    a lot of loyal followers.

    Now, Sony is trying to replace a very good DSLR with a thoroughly
    mediocre SLT, with all the problems that the pellicle mirror and EVF
    bring. Our Minolta/Sony stalwarts are outraged, and we have far fewer
    A77 pre-orders than we expected. We have had several cancellations of
    pre-orders, the most recent today. It is getting to the point where
    Alpha cameras will be impossible to sell, but we have to offer them to
    get the best terms on the NEX range.

    Two customers also changed their A77 pre-orders to the NEX-7 on the
    basis that it has the same sensor and the same EVF. Their logic was
    that there is no point in having a heavy, bulky SLT when the same
    image quality and viewfinder are available in a smaller package.

    We don't have a release date for the A77/NEX-7.
     
    Bruce, Oct 26, 2011
    #7
  8. Rich

    Robert Coe Guest

    : > On 26/10/2011 4:04 p.m., Rich wrote:> I noticed it in the NEX-7 results and now, it seems that the SLT-A77 has
    : > > the issue.  However, the camera does show really high resolution.
    : > > The Pentax K5 remains the noise control king of APS.
    : >
    : > >http://www.dpreview.com/reviews/sonyslta77/page14.asp
    : >
    : > With the Pentax K5 it looks like a /shitload/ of NR goes on between
    : > reading the sensor and so-called "raw file".
    :
    : If it results in loss of detail or quality, that's not good.

    Well, is it that or just that the default presentation of the RAW image is set
    to a high NR level? If the latter (as I think it would be in the Canon world),
    then you could back it off with no overall loss of information.

    Bob
     
    Robert Coe, Oct 26, 2011
    #8
  9. Rich

    RichA Guest

    People need to adapt a bit. I remember them complaining when Minolta
    was sold to Sony.
    Getting away from the optical viewfinder will take some people a long
    time to get used to. I hope that by buying a NEX-7, they will get
    used to it and buy into the SLT's, but I personally don't like the
    idea of the light loss with the pellicle. Then again maybe the less
    expensive A65 will be a good choice, as it has the same EVF and 24M
    sensor?
     
    RichA, Oct 27, 2011
    #9
  10. Rich

    RichA Guest

    I'm still wondering about a choice. A new "pro" Panasonic m4/3rds is
    due and Sony's production is compromised.
     
    RichA, Oct 27, 2011
    #10
  11. Rich

    Bruce Guest


    When a camera manufacturer with a long history of innovation and a
    reputation for high quality products is sold to an electronics firm
    whose reputation has taken a hit, there are good reasons to complain.
    Given the mess Sony has made of the opportunity presented by Minolta,
    the pessimists have been proved right. The Alpha range is a disaster.
    The highly successful NEX has no Minolta DNA.


    The pellicle has been tried before, and rejected. There are too many
    fundamental problems for it to work; light loss is just one of them.


    The same sensor/EVF, yes. A slower processor and a plastic body.
     
    Bruce, Oct 27, 2011
    #11
  12. Rich

    Bowser Guest

    I think the best we can hope for in the GX1 is the sensor in the G3.
    Good, not in Sony's class. I think that the sensor in the NEX 7, when
    not handicapped by a pellicle mirror will be about a stop better. Just
    guessing. For me, if it's good up to ISO 3200 I'm OK with it.
     
    Bowser, Oct 27, 2011
    #12
  13. In other words there's nothing wrong with the sensor. Which suggests
    the problem must be in the processing, i.e. other camera makers have
    better in-camera jpeg noise reduction, and possibly put a bit of NR in
    between the sensor and the supposedly unprocessed RAW image file. In
    recent years we've seen a number of new much improved noise reduction
    methods available as stand-alone programs or editor plug-ins which not
    surprisingly give better results than in-camera noise reduction.

    So the interesting question is whether when these improved noise
    reducers are applied to Sony RAW images you get noise reduction at
    least comparable to that offered by other camera makers. You certainly
    get an amazing improvement over what Sony's own in-camera jpeg noise
    reduction offers.

    But how does it then compare to what you can get from Nikon and Canon
    with similar processing?
     
    Chris Malcolm, Oct 27, 2011
    #13
  14. Plus at least one major advantage of the pellicle mirror: no mirror
    vibration, and the electronic "shutter" opening means no shutter
    vibration either.
     
    Chris Malcolm, Oct 27, 2011
    #14
  15. Rich

    Bruce Guest


    The G3 16 MP sensor competes with Sony's 16 MP sensor when the latter
    is fitted to a Sony camera. The G3's sensor cannot compete with
    Sony's 16 MP sensor when the latter is fitted to a Pentax or Nikon
    camera. Sony has still not mastered the problem of high ISO noise,
    when Pentax and Nikon both have.


    Sony claims that the pellicle mirror robs only about one third of a
    stop of light. Without any means of verifying that claim, we have no
    option but to take it at face value.
     
    Bruce, Oct 27, 2011
    #15
  16. Rich

    Robert Coe Guest

    : >>I noticed it in the NEX-7 results and now, it seems that the SLT-A77 has
    : >>the issue. However, the camera does show really high resolution.
    : >>
    : >>http://www.dpreview.com/reviews/sonyslta77/page14.asp
    :
    : > The level of high ISO noise is quite frightening. We had some images
    : > from both an NEX-7 and an A77 at a Sony event. The noise levels at
    : > high ISOs were extremely high. We were reassured that the cameras
    : > were pre-production samples and that the production cameras would be
    : > much better. It seems that they aren't.
    :
    : > Sony Alpha has a long history of disappointments and this is just one
    : > more to add to the list. You would think that Sony would learn from
    : > their mistakes, but they just keep on making the same ones, over and
    : > over again.
    :
    : >>The Pentax K5 remains the noise control king of APS.
    :
    : > Sony makes some excellent sensors that perform exceptionally well in
    : > other brands of camera. But put the same (or similar) sensor in a
    : > Sony camera, and the results are very disappointing.
    :
    : > The 16 MP sensor in the Pentax K-5 is also used in the Nikon D7000 and
    : > several Sony models. In the Pentax and Nikon it delivers excellent
    : > image quality with very good noise control - the Pentax just shades
    : > the Nikon in this respect. But put the same sensor in a Sony A35 or
    : > NEX-C3, and it becomes one of the noisiest sensors on the market.
    :
    : > The same was true of the 24 MP full frame A900 (and later A850) whose
    : > high ISO noise was desperately bad. Yet the Sony-made 24 MP sensor in
    : > the full frame Nikon D3X performed extremely well. However the
    : > sensors are not the same, so the conclusions here are less clear,
    : > except that if you want low noise at high ISOs, don't buy Sony.
    :
    : In other words there's nothing wrong with the sensor. Which suggests
    : the problem must be in the processing, i.e. other camera makers have
    : better in-camera jpeg noise reduction, and possibly put a bit of NR in
    : between the sensor and the supposedly unprocessed RAW image file. In
    : recent years we've seen a number of new much improved noise reduction
    : methods available as stand-alone programs or editor plug-ins which not
    : surprisingly give better results than in-camera noise reduction.
    :
    : So the interesting question is whether when these improved noise
    : reducers are applied to Sony RAW images you get noise reduction at
    : least comparable to that offered by other camera makers. You certainly
    : get an amazing improvement over what Sony's own in-camera jpeg noise
    : reduction offers.
    :
    : But how does it then compare to what you can get from Nikon and Canon
    : with similar processing?

    By default, Canon estimates the amount of noise reduction you need and applies
    it to the RAW file. But no information is lost; and if you don't want the NR,
    you can easily reverse it in the photo editor. Occasionally I'll decide that
    an image needs more NR and tweak it upwards in PP. It makes the image a bit
    blurrier, but usually the overall effect is improved.

    I'm not sure this is an answer to the question you posed, but maybe it helps
    define the context.

    Bob
     
    Robert Coe, Oct 28, 2011
    #16
  17. Those who have compared NEX-7 with A77 exposures have found about 1/3
    stop difference.
     
    Chris Malcolm, Oct 28, 2011
    #17
  18. Rich

    RichA Guest

    True, its DR isn't up to the APC. But, up to around 3200 ISO, the GH2
    sensor is nearly on-par with the D7000, I compared them both. The G3
    sensor which some had claimed was a stop less noise controlled than
    the GH2, is really almost the same. So, the G3 should hold its own
    with the Nikon, and is clearly less noisy than the A77 and the NEX
    7.
     
    RichA, Oct 29, 2011
    #18
  19. Rich

    PeterN Guest


    AFAIK NR is usually accomplished by color blurring.
     
    PeterN, Oct 29, 2011
    #19
  20. Is it or is he comparing apples with oranges?
     
    R. Mark Clayton, Oct 29, 2011
    #20
    1. Advertisements

Ask a Question

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

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.