Images created by a camera or by a scanner can get a very large file size. this huge size has downsides:

  • Can't send over mail.
  • Can't publish on a web site
  • Any action abuses the computer CPU and takes a long time
  • Storage on a cloud or even locally may trigger some issues

Reducing the file size may solve these issues but can distort the image so badly that efficient as it gets, the file becomes useless.

So the question is: what is, if even possible, the best approach to reduce the file size while not losing too much quality? Change file type? Resize (if so, how)? Drop picture information (what and how)?

  • 2
    You should be able to get a huge saving in file size by using .png with compression. But for a photo you can generally afford a lossy format like .jpg. Are your images actually photos though? .bmp is an unusual format.
    – Chris H
    Mar 27, 2016 at 21:22
  • Downvote because no specific question is stated (not even a single question mark either). Mar 28, 2016 at 11:07
  • @Pinhollow: my question is.... is the place where my question begins :) anyway... its not really a question about photography... its about the expirence of photograper to manipulate thier files afterwards... so you might be right that it is not clear enough
    – Asaf
    Mar 28, 2016 at 13:10
  • What are the images? Are they photos, screenshots, vector images? Why 100kb? Are you uploading them to an image service with a hard and fast limit? If so, what is the service and what are its file type requirements? Are you just storing them locally, and you can use whatever format is required? Do you have photoshop or another image editor? Do you have so many that you will need to do a batch conversion, or will you be manually shrinking each one?
    – wedstrom
    Mar 28, 2016 at 17:16

2 Answers 2


Here are some statements which you can compile into the answer of the question which you have but we do not know about:

  • downsizing cannot be lossless period
  • downsizing by 1/2, 1/3, 1/4 etc means that no resampling is required and will preserve image apperance somewhat better
  • however, most images (particularly unedited images from cameras with Bayer sensors) do not require the resolution which they have for preserving details which they have, i.e. have surplus information some of which may be wasted without visual effect, and resampling will not be too bad in this case
  • image encoding is abstracted from any visible manipulation in vast majority of cases. Image encoding is most oftenly (have not seen a counter example yet) preserving image pixel dimensions, therefore there is no sense in asking about how resizing is related to lossless
  • image encoding may affect following characteristics directly or indirectly: bit depth, tone distribution, chromatic resolution
  • image encoding may create artifacts (false details) of different kinds (for example, low quality JPEG2000 is quite different from LQ JPEG)
  • resizing can be done with various algorythms from which the worst one is "nearest neightbour", the better one is "bilinear", the one which is fine in all cases is "bicubic", the one which is good for resizing unsharped picture is "bicubic sharper" (increases sharpness along with downsizing, should be applied carefully to sharpened images), and probably the best of all available is "Lanczos" (does not include sharpening).
  • there are numerous image encoding algorythms from which I recommend to try JPEG2000, WebP and FLIF (may be lossless and lossy) or, if you do not have particular interest in quality, you will be fine with JPEG. PNG encoding is good for images with big uniform areas which do not contain noise (this is important).
  • PNG encoding can be lossy depending on implementation but when there are no options in software to choose lossy/lossless (most of the software does not), it is safe to assume that image will be compressed losslessly.

P.S. I am intentionally saying that there may be no resampling in case of said resizing scale because for resampling to occur there should be an interpolant first but it is not required to compute an interpolant if all input pixels are strictly divided between output pixels and cannot influence other pixels. It is of course possible to implement resizing in such a way that interpolant is always computed but it still is NOT required (this is what I said originally, not that resampling is always skipped in this case).

It might be proven that if continious image is sampled with AxB and (N*A)x(N*B) samples then latter may be exactly reduced to former without ANY signal loss. This is what provides somewhat better quality, not the absence of resampling itself.

It is common to name any change in discretisation "resampling" but I ignore that (in same way as I ignore the commond statement saying that long focal length objectives compress perspective). Please create a chat if you want to disprove my statements.

P.P.S. The Nyquist-Shannon theorem is related to this, BTW.

  • 1
    Downsizing by 1/2, 1/3, 1/4, etc. absolutely requires resampling. I'm not sure what you mean by that statement. Mar 29, 2016 at 2:32
  • 1
    @user1118321: no, it can be done both with resampling and without it. One may average each group of NxN pixels and avoid resampling. Resampling means interpolating image and sampling the interpolant. No interpolating is required for averaging pixels. Mar 29, 2016 at 8:57
  • You're mistaken about what resampling means. If you're averaging NxN pixels you are resampling. Mar 29, 2016 at 16:07
  • 1
    @user1118321: please suggest another term which suites better. "Resampling" is not inherited from resizing. If no interpolant is required (in case of simple averaging for example) than what would be sampled? Why call local computations (all input pixels are strictly divided between output pixels and cannot influence pixels adjacent to only output pixel) resampling? I know that it might be common to call any discretion change "resampling" but it is only called so because of most resizing reuires resampling, not because any resizing does. I added more to the answer. Mar 29, 2016 at 16:33
  • It's called resampling because you are sampling the image again to produce a new one, not because there's an interpolant. It has nothing to do with resizing, it has to do with sampling. Mathematically, an NxN average for downsampling is equivalent to a convolution with an NxN box blur and then a nearest-neighbor resampling of that result. I've never seen any literature that makes a special case for "not having an interpolant" and others will be confused by that use of the word. But if you want to ignore the actual meaning of the word, then by all means, go ahead. Mar 30, 2016 at 1:55

This question lacks of information, becouse a 3Mb BMP (uncompressed) file can be 10k or 1Mb compressed depending on what you have in that image. Let us think it is a photo, for example a portrait.

You are not saying either if you can resample it or not. There are two different terms. File size (pixel dimensions) and File weight (Disk space usage).

For a photo, with no resampling, the only way to go is using a lossy compression like JPG.

In theese tests I made http://otake.com.mx/Apuntes/PruebasDeCompresion2/4-CompresionJpg5Porciento.htm I found that you can compress a photo to like 5% ratio without a noticable artifacts. But there is a chance you need to aditionaly resample a bit.

1) In Photoshop resample using bicubic sharper. Try diferent ratios for example 80% the original size.

2) Sharpen a bit.

3) Save for web. Experiment with diferent values untill you are ok with the results in terms of visual quality and size.

For the complete methodology I used see this other page: http://otake.com.mx/Apuntes/PruebasDeCompresion2/1-CompresionJpgProceso.htm

Please use google translate.

  • "For a photo, with no resampling, the only way to go is using JPG compression." - there are miriads of image compression aloorythms, and many of them are lossless. Why JPEG? Mar 28, 2016 at 11:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.