Converting is supposed to simplify things, so a bigger file feels like a mistake. It is not. It is the price of a format that everything can open, and it is worth understanding before you start lowering quality to chase a number.
Why the same photo weighs more as a JPG
HEIC and JPEG both discard information to save space, but HEIC uses a far more recent method. At the same visible quality it needs roughly half the bytes. Converting does not remove that difference, it gives it up.
So the growth is not damage. The image is not worse than the HEIC in any way you would notice at a high quality setting; it is simply stored less efficiently, in a format that every piece of software on earth can read.
What each preset actually costs
The quality value controls how aggressively detail is thrown away. It does not control the resulting file size directly, which is why the same setting produces very different sizes on a plain wall and on a tree full of leaves.
Rather than take that on trust, you can measure it on your own photo. The quality preview tool encodes the same image at all four presets and reports the resulting sizes side by side, so you can see what your picture costs at each one.
- Maximum, 1.0 - largest files, nothing discarded by choice.
- High, 0.85 - visually indistinguishable in normal use.
- Balanced, 0.7 - clearly smaller, holds up on screen.
- Small, 0.5 - visible artefacts on detailed photos.
What a “make it under 2 MB” setting actually costs
This is the most common thing people want, and it is worth understanding before using it. Hitting an exact byte figure means encoding the image over and over, lowering quality until the number is reached - there is no other way to do it. The browser converter offers it, and reports the quality it ended on so the cost is visible rather than hidden.
What no tool can do is promise a size *and* the quality you started with. If a converter offers a target size and never mentions what it spent, it spent it anyway. The iPhone app takes the other route entirely: it sizes by quality and dimension, and never by bytes.
What to change when the file really is too large
Reduce the longest edge first. It is by far the biggest lever, because file size scales with the number of pixels: halving the longest edge removes about three quarters of them. A photo 4032 pixels wide is more than almost any upload form needs, and 2000 pixels usually lands well under a limit while staying perfectly readable.
Only after that is worth touching quality, and not below 0.7 if the picture contains text. On a photographed document the artefacts appear around the letters, which is the one part that has to survive.
Common questions
Is the converted JPG lower quality than the HEIC?
At a high quality setting the difference is not visible in normal use, but information is discarded, because JPEG is a lossy format. What changes noticeably is the file size, not the appearance.
How do I get a photo under a specific size?
The browser converter has a max file size setting that re-encodes until the file fits and tells you the quality it landed on. Reducing the longest edge first costs less detail, so it is worth trying before handing the decision to a byte target.
Does converting more than once make it worse?
Yes. Each re-encode discards a little more. Converting once from the original HEIC always gives a better result than converting a copy that has already been through a messaging app or an editor.
Which setting should I use for printing?
0.85 or above, with the dimensions left alone. Print uses far more detail per centimetre than a screen, so compression artefacts and missing pixels show up on paper before they show up on a phone.
Why do two photos at the same quality have different sizes?
Because compression depends on the content. Fine detail like foliage or fabric needs many more bytes than a smooth sky or a plain wall, even at an identical quality setting.