Why do we only have rectangular images on computers?

Replies

• 

  RajdeepCE

  The_Big_K

  why cant we have images of any shape?

  As per my thought the reason behind this is our eyes. Our eyes have viewing window of rectangle size (about 4:3), thats why almost every pictures and videos are recorded in this format. So all the image formats are built according to this ratio.
  

  Also, how can we create images of any shape?

  I am also trying to make images of any shape but I dont know how to make it.
• 

  Kaustubh Katdare

  To the best of my knowledge, you can't create round images. It's more of a technical reason, I suppose.
• 

  shalini_goel14

  Very good idea popped up but they have squared images also 😉
  
  Well I guess the reason is images stored in computers are digital images consisting of rows and columns of pixels. Can anyone tell me how many rows and columns a specific oval or circle or any other shape(other than rectangular and square) has? If we can get the answer of my this question, I guess in future we can save digital images in ovals or circular or any shape other than rectangular and squared.
  
  PS: Feel free to correct me if wrong.
• 

  Corpse-Thrust

  From what I think, having square images just saves us from storing more information about the image. I could be wrong though.
  
  For square/rect images, we'll have to store only length, breadth and coordinates for the top-left corner of the image. That's 4 values.
  
  For circular/oval images, we'll have to store the minimum radius, maximum radius, the coordinates for the centre, the value of that one variable used to define curves, I forgot its name...hypersomethinghoopla?? This makes a total of 5, one being a float.
  
  For polygonal images the best way will be to store the coordinates of each of its corners, which means a minimum of 3x2=6 values. Or possibly storing values of angles and coordinates for one point, and also length of each edge. That'll be a minimum of around 8 values.
  
  Clearly having square images is more economical memory wise. Again, I could be wrong, I'm just hypothesising. (does that word exist btw?)
  
  
  Edit - Having random shaped images in square boxes is possible. You'll have to cut the image for that using Photoshop.
• 

  Kaustubh Katdare

  @ Corpse - You can cut the images in Photoshop , but while saving them, the shape will be enclosed in a rectangular box with defined length & breadth.
• 

  Corpse-Thrust

  ^^^ but those unwanted pixels don't show up at all, so technically you still get the shape you want. btw i'm talking about png's and gif's, not jpg's or bmp's.
  
  
  a square/rectangle is the easiest way to represent an area on a screen which is made up of small squares itself. simple logic, really. it also seemingly comes out to be less resource heavy (see my post above for a small and fail hypothetical explanation 😛 )
• 

  Kaustubh Katdare

  PNGs & GIFs have ability to create a transparent background (which can be defined).
  
  Yes, Rectanglular (PS: Shalini: Square is a special rectangle! 😛 ) form could be the easier way of saving image. Not 100% sure though.
• 

  Ashraf HZ

  Biggie, what kind of issues do you have that you'd have to edit the images?
  
  You can technically achieve a sort of odd shape by slicing the image up, and aligning the different parts. That way you can do things like text wrapping around the images, etc.
• 

  Kaustubh Katdare

  Let me explain by taking an example. Consider following image -
  
  
  Now, the only part of the image, that really matters to me is the part included in yellow circle. However, if you see, the image is included in white rectangular box.
  
  This part is extra and should be removed from the image. It'll save lot of space.
  
  Say what?
• 

  silverscorpion

  I think that's what we are all discussing until now.
  
  Removing the white areas from the picture will minimize space in the web page. It'll give us more space to work on in the page.
  
  But the memory used to represent that image will increase. So, if you want to represent a circular image, then you have to be ready for the extra overhead used to store it.
  
  Anyhow, in this particular image, or any similar image, I don't think you can do much with the space that you got off the corners..
• 

  MaRo

  Interesting..
  
  Corpse I don't think it could has significant resource eating the GPU draws the inner shape anyway..I think it may be relative to the software : a game, photoshop, html engine...etc the shapes already drawn in arbitatry ways like in games you don't play with a character in rectangle that's why I think it's about the receiver.
  
  Well, I'm not technically sure but I believe it's all in software.
• 

  Kaustubh Katdare

  I'm yet to come across any software that lets me save a file limited by the boundary of the image.
• 

  MaRo

  At last all images limited by the view of the window OSes which all in rectangles that's why you won't see it arbitrary shapes, but as I said in games (doesn't take GUI control from OSes) you can find any shape.
• 

  oldboy1985

  my view would be
  
  1. reason why we are storing an image is not only to just view it or to store
  it
  2.we use images for various manipulations like
  enlargement,rotation,contraction,printing and etc
  
  all these operations can be done more easily if it is a rectangle
  
  if image is of any other shape then the image processing will be more difficult
  
  
  and we would be requiring more complex technology and more sophisticated algorithm to scan it
  
  so in order to ease our processing and manipulation we make a trade off with space
  we dont have any problem with space
  
  so they are stored in rectangle
  this is my hypothesis
• 

  Kaustubh Katdare

  *bump*
  
  Anyone with ideas?
• 

  skipper

  I once wrote a graphics program for a storage display (this is coincidental, all it meant was that changes and edits to the on-screen image required redrawing the whole image).
  It also used a plotter and transparencies to make photo-ready exposure masks for 2-sided PC copper laminates.
  
  I used flat arrays, that distributed algorithmic shapes over the flat screen. However when the plotter was working it moved in 3 directions. The electron beam does the same thing, it writes like a pen on the storage display (now redundant). A modern LCD is maintained continuously by refreshing the display memory of an image periodically instead of when needed.
  Holographics might be the next step, you can already create 3d images in polymers, by carving them with a laser.
• 

  Kaustubh Katdare

  bump!
  
  Why is this thread dead? No one wants to discuss this?
• 

  Tim1

  First, with respect to storage requirements, the actual image shape doesn't matter. It takes very little space to store a white/transparent background, since it is uniform and any compression algorithm will shrink it to almost nothing.
  
  Second, since it is easy to have any image shape in practice by using a transparent background, the "physical" image shape is completely unrelated to the shape of the object in the image, so this question is in effect null and void: computers can already represent non-rectangular images.
  
  Can you suggest any benefits of a non-rectangular image storage format? Having them rectangular has many benefits, such as:
  
  * Supporting a simple universal model when displaying the image. Consider that computer monitors, web browsers, TVs, HTML markup, digital picture frames, LCDs on cameras, iPhones, camera film, 4x6 camera prints, polaroids, video footage, billboards, Photoshop windows, your profile pic area on Facebook, etc. etc. etc. are all rectangles.
  
  * Simple code when doing image manipulation that can also be easily parallelized and/or put into an ASIC for HW acceleration.
  
  And many more. There are many drawbacks to using non-rectangular image storage formats, and I fail to see any benefits.
• 

  debu

  @big_k: The real reason for images being rectangular is; All image formats that have been successful thus far, involve the use of Cartesian Coordinate System (clickie) to represent pixels.
  
  So, if I wanted to reach any particular pixel or reference it elsewhere I would use a Cartesian set to enumerate it. eg. (2,10) or (43, 17). This is an unfortunate side effect of the Raster Graphics based systems that have deeply penetrated all forms of computing now days.
  
  Some of us (at least me) still remember the Macintosh Ii and the old Acorn Bbc Micro, both of which used a non traditional vector graphics system which used a brilliant Curvilinear Coordinates to represent images. A few image formats still use curvilinear coordinates (like the famous CDR and SVG formats), but they are all blocked by various patents and restrictions and are not allowed to grow.
  
  Regards,
  
  Debu 😀
• 

  Kaustubh Katdare

  Wow, that's something 😁 !
  
  Thanks a lot for the information Debu & Tim1
• 

  sookie

  Sorry but I am not able to agree completely with Cartesian coordinates system things. A related question - "Why do graphics system allow us to create rounded rectangles or circles" in the same computer?Which kind of Cartesian system it follows at that time? Can anyone explain? 😕
  
  Regarding first question: I am not able to recall exactly related thing I read in our Graphics subject during engineering but "image cropping"[or something like that] is a technique used to convert rectangular images into non-rectangular. It was based on discarding less than half-filled squares[A set of small squares into which rectangular image is divided into]
  
  Check the work a guy had done #-Link-Snipped-#
  
  Correct me if anything wrong. Thanks !
• 

  Sahithi Pallavi

  Nice and as well as interesting question Biggie..! And answers too....
  
  
  
  
  
  WINNERS DONT DO DIFFERENT THINGS....THEY DO THINGS DIFFERENTLY....