Is HEIC a lossy or a lossless format?

HEIC is lossy. It uses a still-image adaptation of the HEVC video codec and, just like JPG, shrinks the file by discarding detail the eye struggles to notice. What sets it apart is that it stores the same visual quality with markedly less data than JPG. So HEIC is more efficient, but it is not lossless; some information was permanently gone the moment the photo was taken.

Doesn't converting to a lossless format make up for that loss?

It doesn't, because losslessness does not work retroactively. PNG being lossless means that when you save a file and open it again you get bit-for-bit identical pixels. But the pixel array you hand to PNG has already been through HEIC compression and is already a lossy result. PNG preserves that array faithfully; it does not improve it. In practice it performs a freeze, not a repair.

Why is the PNG output so large?

The two formats compress on fundamentally different principles. HEIC shrinks aggressively by discarding data where the human visual system is weakest. PNG discards no pixel information at all; it only packs repeating patterns losslessly. In a photograph, noise and texture mean almost every pixel differs from its neighbour, so there is very little repetition to pack and the file stays large. That same PNG logic works brilliantly on flat-colour graphics — the problem is not the format, it is the nature of the content.

So what is the point of converting HEIC to PNG?

The point is to stop future losses and gain new capabilities. You can edit and save a PNG over and over without accumulating a single new compression artefact. PNG also carries an alpha channel, which makes transparency-dependent work such as background removal possible. On top of that, virtually every piece of software and every platform opens PNG, whereas HEIC causes trouble especially on Windows and on the web.

What Converting HEIC to PNG Really Buys You, and What It Costs

This conversion is not really about quality

The quickest way to understand the move from HEIC to PNG is to look at the question each format was designed to answer. HEIC asks: "How few bytes can I fit this photo into while keeping it visually reasonable?" PNG asks something entirely different: "How do I store this pixel array without losing any information?"

They are not rivals; they are tools built for different jobs. Whether the conversion makes sense depends entirely on which question you are trying to answer.

What HEIC does, and what it sacrifices

HEIC is the HEVC video codec adapted to still images. Apple's reason for making it the iPhone default is clear: it holds the same perceived quality in noticeably smaller files than JPG. On phone storage that is a serious win.

But the win is not free. HEIC is lossy. The moment the photo is taken, frequency components and colour detail the human visual system is insensitive to are discarded. That is hard to notice by eye — which is precisely why the format succeeds. But mathematically the information is gone, and it does not come back.

So the HEIC file in your hand is a compressed interpretation of the scene the sensor saw. Not a perfect interpretation — a very good one.

What PNG does, and what it doesn't

PNG is lossless. It packs the pixel array you give it with the DEFLATE algorithm and hands back a bit-for-bit identical array when you open it. No colour value changes, no detail is discarded.

Here is the key distinction: PNG's losslessness works forwards only. It preserves the data you give it perfectly. But the data you give it is data that already came out of HEIC compression. The information HEIC discarded never reaches PNG, so PNG has nothing to preserve.

Think of it this way: if you photocopy a document and then throw away the original paper, putting the photocopy in an archival-grade safe does not bring back the lost original. The safe simply prevents the photocopy from degrading further. In this scenario, PNG is the safe.

The cost: why the file size explodes

PNG's compression is pattern-based. Long runs of identically coloured pixels, repeating structures, flat backgrounds — it packs those extremely efficiently. On a logo, a screenshot or an interface design, PNG produces surprisingly small files.

A photograph is the exact opposite. Sensor noise, texture and light transitions mean almost no pixel matches its neighbour. DEFLATE finds no repetition to pack. The result: PNG output is often several times the size of a HEIC original that relies on lossy compression.

This is not an error or a sign of a bad tool. It is a direct consequence of the two formats' design philosophies and is entirely expected behaviour.

So what does the conversion actually gain you?

The gain is not in quality; it is in capability and compatibility.

Generation loss stops. If you open a HEIC, edit it and save it as HEIC again, every save means another round of lossy compression. Flaws stack up and accumulate. Once you move to PNG you can edit and save as often as you like — quality stays fixed at the level of that first PNG save.

Transparency becomes possible. PNG carries an alpha channel capable of storing 256 distinct levels of opacity per pixel. That is what makes the edges of an object look smooth rather than jagged after you remove its background. Every workflow that needs transparency needs a lossless, alpha-capable container.

Compatibility. This is HEIC's biggest practical problem. Windows struggles without an extra codec, many web forms reject it, older editors do not recognise it. PNG, in practice, opens everywhere. Most people fleeing HEIC are really trying to solve this problem.

When is that cost worth paying?

The cost of the conversion is disk space. The scenarios where paying it makes sense are narrow but clear:

You are putting the photo through a multi-stage editing process and do not want quality loss on intermediate saves. You are removing the background to produce a transparent output. Or the system on the receiving end accepts nothing but PNG.

Outside of those — especially for archiving and sharing — PNG is the wrong choice. If you are batch-converting for compatibility, your target should be JPG or WEBP; both solve the compatibility problem HEIC creates while keeping file sizes reasonable. Use the HEIC to PNG converter for the individual files you are actually going to work on.

In summary

HEIC is lossy and very efficient; PNG is lossless and very large. Converting between them does not raise quality, because the loss happened at capture time and cannot be undone. The conversion's real return is that further loss stops, transparency becomes possible, and the file opens everywhere. If you need none of those three things, the size cost you pay buys you nothing.