Which format is more legible for a scanned document, JPG or PNG?

It depends on the content of the document. If you are dealing with a clean scan containing only black text and lines, PNG offers a clear advantage, because it works losslessly and the sharp edges of the letters are preserved without any degradation. By contrast, with scans containing photographs, greyscale shading or paper texture, the PNG file becomes very large and in that case a high-quality JPG is a more balanced choice. If you plan to convert the file to text with OCR, preferring PNG is safer in terms of accuracy.

If PNG is lossless, why does everyone use JPG?

Because being lossless is not always an advantage. PNG works very inefficiently with images containing millions of colour tones and continuous transitions, such as photographs, and the file ends up many times larger than the JPG. Where web page speed matters, that difference becomes unacceptable. Because JPG's lossy compression was designed specifically for photographic content, it produces a far smaller file at a quality most viewers cannot tell apart. So the choice is not about quality but about the type of content.

My TIFF has transparent areas — which should I choose?

You should choose PNG, because JPG does not support the alpha channel, that is transparency information, at all. If there are transparent regions in a TIFF and you convert the file to JPG, those areas get filled with a solid colour, usually white or black, and that cannot be undone. If you are working with a logo, a product image or a cut-out with the background removed, PNG is the only correct option. If transparency does not need to be preserved and the content is photographic, JPG can be reconsidered.

Does it make sense to produce and keep both?

In some cases yes, provided it does not create needless clutter. For example, keeping a product image both as a JPG that will load fast on the website and as a PNG to be used with transparency in design work is a reasonable workflow. But what always needs to be kept is the original TIFF file; you can regenerate both formats from it whenever you like. Keeping the master copy safe is a healthier archival arrangement than multiplying derivative copies.

Should You Convert TIFF to JPG or PNG? A Comparative Decision Guide

The right question is not "which is better"

You have a TIFF file and you need to use it on the web, send it by email or put it into a presentation. TIFF is unsuitable for all of these, so you have to convert it. But to which format?

There is no single right answer to this question, because JPG and PNG were designed to solve different problems. The wrong choice results either in a needlessly large file or in a degraded image. The decision depends on what is inside your file.

The fundamental difference between the two formats

JPG is lossy. During compression it permanently discards part of the image data. It does this by targeting the weak points of the human eye: it reduces colour information more aggressively than brightness and eliminates fine frequency detail. In return it produces very small files.

PNG is lossless. It preserves the exact value of every pixel and merely encodes the data more efficiently. It also has alpha channel, that is transparency, support. In return, file size grows considerably with photographic content.

Because TIFF is also lossless, converting to PNG creates no loss in terms of pixel data. Converting to JPG, on the other hand, is an irreversible discarding operation. That is the most critical axis of the decision.

Deciding by content type

Photographs and photographic content → JPG

A landscape, a portrait, a product photograph, a digitised old photograph. This kind of content contains continuous colour transitions and natural texture. JPG was designed for exactly this.

Using PNG for this content bloats the file many times over for no reason and gives no visible quality gain in return. At a high quality setting a JPG is indistinguishable from the original for most viewers.

Sharp-edged graphics, drawings, logos, screenshots → PNG

Images with flat colour areas, sharp boundaries and few colours. JPG behaves badly with this kind of content: it produces grey shadowing known as "ringing" and block artefacts around sharp edges. That dirtiness appearing at the edge of a logo is unacceptable in professional output.

PNG gives both a cleaner result with this content and usually produces a smaller file, because lossless compression works very efficiently on flat areas containing few colours. So there is not even a trade-off here — PNG wins on both counts.

Scanned text documents → it depends

This is the category that causes the most hesitation.

A clean scan containing only black text and lines is logically close to graphic content: PNG is more suitable. Letter edges are not degraded, OCR accuracy is preserved, and the file usually stays reasonable.

But if the scan also includes paper texture, greyscale shading or photographs on the page, the PNG file swells rapidly. In that case a high-quality JPG is a more balanced choice. Do not keep the quality setting low; a low-quality JPG of scanned text both looks ugly and makes text recognition harder.

Images containing transparency → PNG, no argument

JPG does not support transparency. Full stop. If your TIFF has an alpha channel and you convert to JPG, the transparent areas fill with a solid colour and that cannot be undone. If you are working with a product image with the background removed or a logo, converting the TIFF to PNG is the only right route.

Does file size actually matter?

It depends on the context. On a website, page load time directly affects user experience; there JPG's size advantage is a serious, real benefit. Email attachments have size limits too.

But if you are putting the image into a presentation, using it in an internal document or producing an archival copy, a difference of a few megabytes is practically irrelevant. In those situations choosing PNG to preserve quality is reasonable.

Do not forget what you are losing

Whichever format you choose, some of TIFF's features are lost in both conversions:

Multi-page support. TIFF can carry more than one page; neither JPG nor PNG can. With a multi-page scan, either only the first page is taken or each page becomes a separate file. If the pages need to stay together, think about PDF.
CMYK colour space. Both work with RGB. When a CMYK TIFF prepared for print is converted, a colour shift occurs.
16-bit colour depth. JPG is limited to 8 bits; PNG technically supports 16 but most workflows drop to 8.

That is why, whatever the conversion, the original TIFF must be kept. The JPG or PNG you produce is a delivery copy, not the master copy.

Quick decision summary

Photograph, speed matters on the web → JPG, at a high quality setting
Logo, graphic, screenshot, drawing → PNG
There is transparency → PNG
Clean black-and-white text scan, OCR to be run → PNG
Mixed scan with paper texture and greyscale → high-quality JPG
Going to print → neither; give them the original TIFF
Multi-page document must stay whole → PDF

In summary

The choice between JPG and PNG is not a quality contest but a matter of matching the content. With photographic content JPG's compression works intelligently and the size advantage is real. With sharp-edged content and content containing transparency, PNG is both cleaner and usually smaller. Looking at your file and asking "is this a photograph or a graphic?" answers almost the whole question. For the rest, keeping the original TIFF is enough — if you realise you made the wrong choice, you can convert again from the start.