TIF: Difference between DevonThink and Preview

I’m not Preview does, but it renders tiffs in such a way that makes them readable, whereas in DevonThink they are mostly unreadable. It’s as if Preview adds gradiant or texturing that is just not there in DevonThink.

Any ideas why, or how to get DevonThink to process these correctly?

Could you send me an example? I’ve never encountered such differences and the display/rendering should actually be identical.


three images, one is the tiff as rendered by preview, the other by DT, and the original TIFF file.

I will say that there is a moment in Preview when it opens each tiff that it looks exactly similar to DevonThink. Then it seems to “think” for a second and then it appears with more gradient (and therefore readability).

I get these TIFFS from microfiche machines for my research. typically the text from newspapers is nearly unreadable in preview anyway, rendered accurately. in devonthink it’s often completely unreadable.

Thanks for the assistance on this. =)

The problem is with the way microfilm images were transferred to computer images, I think. I would guess that the film used to photograph the original documents had very high contrast and when the film was later transferred as computer images the objective was to make the images only fairly readable on a microfilm reader. Although the original image’s resolution is shown as 400 x 400 dpi, an OCR conversion program might or might not be able to accurately convert to a PDF with PDF+text.

If you open the original photo it looks like the outfit that captured the microfilm images to computer made no effort to capture true grayscale images (probably the film used for making the microfilms was a very high contrast, as is common). The photo shows it’s a dithered 4-bit grayscale image, but it’s not much different in image quality from a 2-bit dithered B&W image. It’s really all dots with very little gradient smoothing.

DT Pro is rendering the picture fairly accurately, I’m afraid. I noticed the way Preview tries to handle the picture. It initially seems to present it as a 2-bit image and then seems to try to do some interpolation to grayscale. Strange.

Could you post a primarily text image and let us know when you do that? I’d like to see if the quality is good enough to try OCR.

It should be possible to get the same results in DEVONthink, I’ll check this.


Could you please post one of your text-containing images at the same URL?

I’d like to see if it could be OCRd reasonably well.

The rendering will be equal to the one of Preview in V1.1.1.

Looking much better now. Thanks.


Devonthink apparently can’t handle multipage tiffs like Preview can. DT only shows the first page.

Another wrinkle.

I have no idea how many images you are working with. But their value in DT Pro would be enhanced if the text were captured, following OCR of the images.

I’ve done OCR of multi-page TIFF images in the past. That produces searchable text. The accuracy, of course, depends on the quality of the original images.

As noted, i have no idea how you are using the images, but if you need them as continuing references (rather than one-time reading) you might consider using Acrobat Pro 7 or ReadIRIS 11 to process your images to computer-readable form.

I agree. But while some of these newspaper scans may convert easily into OCR, most probably won’t. Since DT contains all my research, it is easier to just have them in the same place and read them from the screen than to go through and individually OCR each one.

Maybe I can find a TIFF-to-PDF convertor. ?

Yes. Graphics Converter will convert TIFFs to PDFs. And it can batch process a large number of images.

Not entirely sure about multi-page TIFFs, as I don’t have one lying around at the moment.

If you don’t have Graphic Converter, it can be demoed, so you can give it a try.

Support for multi-page TIFFs will come in a future release.

Any progress on this?

Thanks Devonpeople!

Maybe v1.1.2 will support this but no promises yet.