Resolution Does NOT EQUAL File Size
And similarly file size does not equal resolution 
I’ve come across enough examples of these two very different traits being confused for each other over the last few months, that I thought it deserved a little mention in my blog.
In response to my post relating to Snapseed on the iPad, people sent me lots of queries relating to it producing low “resolution” images and they were wondering how I managed to get high resolution images out of it. And similarly, I have recently had some people that were printing my work also ask for high resolution images, which in fact were what I sent them. In all cases, they were confusing resolution with file size.
Resolution
There is only one thing and one thing only that effects resolution; and that’s the number of pixels. Nothing else! If a file is 1024×768 pixels, then regardless of DPI, PPI, bits per channel, colour space, file format (or host of other factors), the resolution of the image is still 1024×768.
That is the resolution. It is the number of pixels in the x (horizontal) dimension by the number of pixels in the y (vertical) dimension. Multiplying both figures together will give you a total pixel count, or “lumped” resolution value. For instance in the case of a 1024×768 image, there are 786,432 pixels – not quite a million, so not quite a 1 megapixel image.
The higher the resolution image we have, the greater flexibility we have for cropping (throwing away pixels) or for producing “useable” prints. The word useable is really important here, because any image can be printed at any size. But the larger you print, to retain quality, you need higher resolution images. If quality isn’t of concern to you, then low resolution images can be printed as large as you want.
File Size
The file size (size on disc) of an image is certainly influenced by resolution, but it is by no means the only factor, nor is it necessarily the dominant factor.
- Resolution (number of pixels):
The total number of pixels in an image certainly impacts the size of the image on disc. By and large, more pixels equals larger file size. So generally large resolution images will take up more space than lower resolution images (but not always)
- Number of Bits:
The JPG file format is an 8 bit (per channel) format, requiring 8 bits to describe the red channel, 8 bits to describe the green channel and finally 8 bits to describe the blue channel. So each and every pixel requires 24 bits to describe its colour (called 24bit “true” colour). But other formats require or can handle more information. For instance TIFF or PSD can handle 16 or indeed 32 bits per channel. A 16 bit image would take up twice the amount of space as an 8 bit image and indeed a 32 bit image would take up 4 times the amount of space. Taking an 8 bit versus a 16 bit image as an example, assuming both are captured correctly and not heavily processed, to the human eye, both on screen and on print, should be indiscernible from the other. So you could have two “visually identical” images, with identical resolution, but where one is 100% larger on disc than the other.
- Compression:
A lot of graphical file formats offer compression (often “lossy”), which allows you save your images as smaller images on disc. This compression does not effect the number of pixels in the image, hence does not effect resolution. But it can dramatically effect file size. When I say “lossy” it means that data is lost or thrown away during the compression. The algorithm makes decisions about what data it can keep and what it can throw away whilst still maintaining the visual integrity of the file. The higher the compression applied, the greater the loss in quality – often indicated by what are known as “JPEG artefacts” within the image (clear straight divides between pixels of similar colours). The quality factor/compression strength is user controlled.
Compression is probably one of the biggest factors at play when it comes to file size. My standard flow is to save my high resolution and indeed web sized JPG’s at a quality factor of 8. On average this results in web sized images being around 100k~150k and full sized/high resolution images being around 400k~500k. If I were to save these at the highest quality (12) using the least amount of compression, the full resolution images would typically be in and around 3MB~4MB. That’s a 10x increase! And for me, there is no visual discernible difference. I’m happy to print from these JPGs and often do, but what I never do is use them as a source for editing. Any future tweaks and/or resizing, I refer back to the original uncompressed 16bit PSD file.
- Other Factors
Apart from file format itself, I believe the two factors above (bit depth and compression/quality ratio) are the two biggest factors that influence the file size of an image. But other factors can come into play. If you choose to save the EXIF data with an image, it can add to the size of the file on disc. The type of data in an image can also influence file size. For instance an image with lots of small detail (as an example noise) would not lend itself to compression as much as image with large areas of smooth transitions of colour tones (sky, sea, skin etc)
Can I Have A 2MB File Please?
This sort of request does my head in! But yet it’s often a common request from stock libraries or print providers. For what ever insane reason, they request an image based on its file size, not its resolution??? This is madness! They would be better off specifying a minimum resolution and requesting an uncompressed version. Sadly they don’t 
I have often provided files, where the 2MB has been made up by padding the EXIF with junk. The image itself remains the same.
Examples
As a proof of point, here are 5 different files, all representing the exact same image (original image). Two are PNG files (portable network graphics, which is a lossless bitmap image) one 16 bit and one 8 bit. The other 3 are JPGs, one with no compression, one with quality factor set to 8 (my default setting) and the fifth is a higher resolution version than all the others, yet is saved using higher compression, so is smaller on disc. All have completely different file sizes. Can you visually spot the difference?
16 bit PNG (no compression) 600×849: 2,23MB
8 bit PNG (no compression) 600×849: 578k
8 bit JPG (highest quality/low compression) 600×849: 294k
8 bit JPG (medium quality& compression) 600×849: 86.4k
8 bit JPG (low quality/high compression) 800×1132: 82.5k
Top 2, I can’t see the difference.
3 to 5 I see some degrading color rendition in gradation area and the “JPEG artifacts” are getting stronger (worse) to the bottom.
what would you recommend using to save to a disc/ usb to create a high resolution picture to be able to have good print quality?
I save as JPG (quality setting 8) with maximum resolution in the X of 1200 and/or in the Y of 1100.
That usually ends up with a file in and around 200k in size and reasonable quality