Page 2 of 2 FirstFirst 12
Results 21 to 36 of 36

Thread: Small size pictures utilize all sensor pixels?

  1. #21

    Join Date
    May 2014
    Location
    amsterdam, netherlands
    Posts
    2,237
    Real Name
    George

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by pnodrog View Post
    The file would be in PPI (if defined) the printing DPI.

    Personally I am far more interested in concepts rather than arguing about semantics.
    You don't have your best day.
    The file is just pixels. The monitor has a resolution of x ppi.

    George

  2. #22
    MrB's Avatar
    Join Date
    Feb 2011
    Location
    Hertfordshire, England
    Posts
    1,125
    Real Name
    Philip

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by pnodrog View Post
    The file would be in PPI (if defined) the printing DPI.

    Personally I am far more interested in concepts rather than arguing about semantics.
    On the assumption that dpi had been typed in error, ppi was posted as a correction, not as an argument.

    Cheers
    Philip

  3. #23
    pnodrog's Avatar
    Join Date
    Sep 2012
    Location
    Waipu, Northand, New Zealand
    Posts
    3,582
    Real Name
    Paul

    Re: Small size pictures utilize all sensor pixels?

    George, I am having a great day - be made better if people could read properly... .. PPi is a value often stored in the file. I even used the word "if" in the comment to make it reasonably clear that it was not a property of the file.

    I admire your determination to try and find fault with postings..

    P.S. A file is not just pixels it is just bits - data in a digital form... not hard to play your game.
    Last edited by pnodrog; 9th November 2017 at 07:31 PM.

  4. #24
    Administrator Manfred M's Avatar
    Join Date
    Mar 2012
    Location
    Ottawa, Canada
    Posts
    13,614
    Real Name
    Manfred Mueller

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by george013 View Post
    No, the file is not printed, just resized, in the question.
    And all the sensor pixels are used for the image.
    For a JPEG file, this is not correct and for RAW data even less correct. A jpeg is an image file and has an R, G and B value for every pixel. Raw data for all cameras using a Bayer array (all cameras except for Sigma cameras and the Leica Monochrom series), the data is R, G, G, B and has to be turned into image data with R, G and B values.

    The data has to be scaled to fit the output device requirements unless we are outputting to exactly fits the height and width of the file. This rarely happens. On a computer screen, the image data is downsampled and new data is created and displayed; some of the original data is "thrown away". This is because most computer displays only use somewhere between 2MP of data and 8MP of data; depending on the specific screen.

    For a small print, the data will be downsampled as well, but for a large one, the data will up upsampled, i.e. new pixel data will be created to fill the paper. For Canon and HP printers, the native resolution is 300 ppi and for Epson it is 360 ppi, but the printer drivers effectively hide this from the end user.
    Last edited by Manfred M; 9th November 2017 at 07:52 PM.

  5. #25

    Join Date
    May 2014
    Location
    amsterdam, netherlands
    Posts
    2,237
    Real Name
    George

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by pnodrog View Post
    George, I am having a great day - be made better if people could read properly... .. PPi is a value often stored in the file. I even used the word "if" in the comment to make it reasonably clear that it was not a property of the file.

    I admire your determination to try and find fault with postings..

    P.S. A file is not just pixels it is just bits - data in a digital form... not hard to play your game.
    I've seen better days of you.
    Phillip mislead you with PPI, it's DPI as you stated at first, the wanted printer resolution. And that's a request to the printer driver to correct your image so that it looks like being printed on the wanted print resolution if printed full size.
    An image is a queue of pixels, every pixel contains 3 values for the 3 colors R,G and B. And those values can be a byte or a word or just an amount of pixels as in a RAW file. The logical unit is how they're addressed: in pixels.

    Back to topic. I just wondered how that magnification came in in this thread.

    George

  6. #26

    Join Date
    May 2014
    Location
    amsterdam, netherlands
    Posts
    2,237
    Real Name
    George

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by Manfred M View Post
    For a JPEG file, this is not correct and for RAW data even less correct. A jpeg is an image file and has an R, G and B value for every pixel. Raw data for all cameras using a Bayer array (all cameras except for Sigma cameras and the Leica Monochrom series), the data is R, G, G, B and has to be turned into image data with R, G and B values.

    The data has to be scaled to fit the output device requirements unless we are outputting to exactly fits the height and width of the file. This rarely happens. On a computer screen, the image data is downsampled and new data is created and displayed; some of the original data is "thrown away". This is because most computer displays only use somewhere between 2MP of data and 8MP of data; depending on the specific screen.

    For a small print, the data will be downsampled as well, but for a large one, the data will up upsampled, i.e. new pixel data will be created to fill the paper. For Canon and HP printers, the native resolution is 300 ppi and for Epson it is 360 ppi, but the printer drivers effectively hide this from the end user.
    Manfred, read the original question.
    Are all sensors pixels used regardless of picture size or is only a subset of the sensor pixels used as needed to create a picture of the selected size?
    My first question was what he meant. Was picture size meant in framing or content as he mentioned later, or was meant the file size. It turned out to be the latter. And that smaller file size can be archieved either by heigher compression or by resizing, or both. But in his example it's resizing with the same content. The original raster image of 3648x2736 has changed to 640x424, with the same content. A thumbnail. All sensor info is used.

    George

  7. #27
    pnodrog's Avatar
    Join Date
    Sep 2012
    Location
    Waipu, Northand, New Zealand
    Posts
    3,582
    Real Name
    Paul

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by george013 View Post
    I've seen better days of you.
    Phillip mislead you with PPI, it's DPI as you stated at first, the wanted printer resolution. And that's a request to the printer driver to correct your image so that it looks like being printed on the wanted print resolution if printed full size.
    An image is a queue of pixels, every pixel contains 3 values for the 3 colors R,G and B. And those values can be a byte or a word or just an amount of pixels as in a RAW file. The logical unit is how they're addressed: in pixels.

    Back to topic. I just wondered how that magnification came in in this thread.

    George
    The word would be array not queue...

  8. #28

    Join Date
    May 2014
    Location
    amsterdam, netherlands
    Posts
    2,237
    Real Name
    George

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by pnodrog View Post
    The word would be array not queue...
    No, an array does have an index, a queue not. I first wrote list but changed that afterwards in queue.

    George

  9. #29
    Administrator Manfred M's Avatar
    Join Date
    Mar 2012
    Location
    Ottawa, Canada
    Posts
    13,614
    Real Name
    Manfred Mueller

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by george013 View Post
    Manfred, read the original question.
    My first question was what he meant. Was picture size meant in framing or content as he mentioned later, or was meant the file size. It turned out to be the latter. And that smaller file size can be archieved either by heigher compression or by resizing, or both. But in his example it's resizing with the same content. The original raster image of 3648x2736 has changed to 640x424, with the same content. A thumbnail. All sensor info is used.

    George
    George - I did read the question and agree with your comments with respect to the original question. As per my response #11, as there is one exception to this and that concerns the way still cameras record video data.

    The original question did not get into compression at all, but regardless, the input data is the same in all cases. The usual issues regarding ppi and dpi have crept into the forks in this thread.

    ppi - is an outcome and depends on screen pixel pitch of the screen.

    dpi - is dictated by the printer manufacturer; Canon and HP printers have a native resolution of 300 dpi whereas Epson printers have a native resolution of 360 dpi.

  10. #30
    pnodrog's Avatar
    Join Date
    Sep 2012
    Location
    Waipu, Northand, New Zealand
    Posts
    3,582
    Real Name
    Paul

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by george013 View Post
    No, an array does have an index, a queue not. I first wrote list but changed that afterwards in queue.

    George
    An array has a structure.

  11. #31

    Join Date
    May 2014
    Location
    amsterdam, netherlands
    Posts
    2,237
    Real Name
    George

    Re: Small size pictures utilize all sensor pixels?

    Every variable has a structure. That's defined when creating them. But beside that we're talking about disk files. An array is a memory variable. And yes, in memory one can approach an image as a 2 dimensional array, or as a queue, or list. A matter of programming.

    George

  12. #32
    pnodrog's Avatar
    Join Date
    Sep 2012
    Location
    Waipu, Northand, New Zealand
    Posts
    3,582
    Real Name
    Paul

    Re: Small size pictures utilize all sensor pixels?

    George, quaint.

    "Arrays are used to implement other data structures, such as lists, heaps, hash tables, deques, queues, stacks, strings"

    It would be unproductive for us to continue this thread.
    Last edited by pnodrog; 10th November 2017 at 08:00 AM.

  13. #33

    Join Date
    Dec 2011
    Location
    Staffordshire UK
    Posts
    77
    Real Name
    Barry

    Re: Small size pictures utilize all sensor pixels?

    This topic takes me back a long, long time. I can remember during the early 1950's hearing the parents of one of my friends discussing the purchase of their first television set. Their options were confined to 9-inch and 12-inch models in those days (14-inch screens were just about to appear on the horizon). My friend's mother stipulated that they must get a 12-inch model because there would be more of the picture on the screen.

  14. #34
    xpatUSA's Avatar
    Join Date
    Feb 2012
    Location
    Texas
    Posts
    3,398
    Real Name
    Ted

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by Dave Humphries View Post
    Now whether it is averaging all the intermediate pixels (probably) or just using every 6th one to do this is largely irrelevant because; with memory not being the cost it was cameras were originally developed with this functionality, it serves no purpose in general photography to be skimping - always shoot at the maximum resolution; your "1)".
    Good advice - especially if the intended output is a print of more than 8x10" or A4.

    As to "averaging all the intermediate pixels", it's not quite that simple these days.

    A quick look here will give us an idea:

    http://entropymine.com/imageworsener/resample/

    http://www.microimages.com/documenta...resampling.pdf

    Enjoy.

  15. #35
    pnodrog's Avatar
    Join Date
    Sep 2012
    Location
    Waipu, Northand, New Zealand
    Posts
    3,582
    Real Name
    Paul

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by xpatUSA View Post
    Good advice - especially if the intended output is a print of more than 8x10" or A4.

    As to "averaging all the intermediate pixels", it's not quite that simple these days.

    A quick look here will give us an idea:

    http://entropymine.com/imageworsener/resample/

    http://www.microimages.com/documenta...resampling.pdf

    Enjoy.
    I was hoping to see a note or comment of the effect on the reduction of noise that is sometimes claimed is a possible benefit when downsizing. The chance of highlighting noise by picking on individual pixels rather than doing some type of averaging seems far too risky to me.
    Last edited by pnodrog; 10th November 2017 at 05:55 PM.

  16. #36
    xpatUSA's Avatar
    Join Date
    Feb 2012
    Location
    Texas
    Posts
    3,398
    Real Name
    Ted

    Re: Small size pictures utilize all sensor pixels?

    Quote Originally Posted by pnodrog View Post
    I was hoping to see a note or comment of the effect on the reduction of noise that is sometimes claimed as a possible benefit when downsizing.
    Indeed, we should also remember that there are many different downsizing methods, all of which affect "noise" differently. Also, "noise" is often mentioned with no definition of the type of noise, making the whole topic subject to misunderstanding.

    Recently I was measuring the actual ISO of one of my cameras.

    This needed shots outside of a gray card, using infinite focus resulting in a fairly homogeneous image - perfect for measuring image noise.

    I downsized a shot three different ways.

    Enter the forum Pedant: for this exercise I used SNR to compare the original and the three downsized ones. I used Signal-to-Noise Ratio (SNR) in order to even out the slight luminance variations of my selections in ImageJ.

    So, image actual SNRs(not dB):

    Original: 94.319
    Maker's raw converter: 78.359
    FastStone Lanczos3 (default): 71.704
    FastStone Lanczos2 (sharper): 66.456

    Since a higher SNR implies less "noise" . . the original appears to be "best".

    My SNR is (mean signal value)/(standard deviation) both easily available in ImageJ.

    No doubt there are many other definitions of "noise" here . . .

    The chance of highlighting noise by picking on individual pixels rather than doing some type of averaging seems far too risky to me.
    Quite so. Especially with the weighted types of averaging as can be seen in the links I posted earlier, e.g. bicubic, Lanczos, et al.
    Last edited by xpatUSA; 10th November 2017 at 08:39 PM.

Page 2 of 2 FirstFirst 12

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •