Is scanning to DT3 possible right now with M1 Macs? ABBYY? (Apple Silicon)

[emptybits]

Hi. I have an M1 Mac Mini on order and I intend to gradually transition to it as my document/knowledge desktop so I’m wondering if it’s realistic to expect my simple workflow to run right away.

Reading comments here, I understand DT3 will work via Rosetta 2 until there’s a Universal Binary. Cool. But on the blog and discourse, DT suggests ABBYY may not be Big Sur and/or Apple Silicon ready right away so does that hoop my usual scan workflow?

1. Input: paper doc
2. Fujitsu ScanSnap iX500
3. Software by Fujitsu/ABBYY/DT
4. Output: text-behind PDF doc now in DT3 database

Fujitsu’s website suggests ScanSnap Home will run fine on BigSur (via Rosetta 2 until they go Universal). But what’s up with ABBYY? I’m actually not sure “where” that piece lives in my current setup, but I assume it’s there because I’m getting OCR. Anyways, is it working under Rosetta 2?

Thanks for any facts or experience with this on M1 hardware.

[rmschne]

Probably good to put your search skills to work to see this already discussed here:

Big Sur's a coming

OCR works fine on Big Sur but not yet on ARM-based Macs due to an issue of the Abbyy engine. The next release will also fix some minor glitches and update the toolbars for Big Sur. And the Mail plug-in hasn’t been updated yet but there are several alternatives (e.g. drag & drop, scripts and View > Sidebar > Import). But as usual I would highly discourage to update to a .0 release of macOS in a production environment, at least as long as you’re not a developer and need it for your work …

[emptybits]

Thank you for the fast response. I did read that. In hindsight, I should have cut to the weak link and put “ABBYY” in the name of this thread to distinguish it from the general Big Sur thread. I just did that. Anyways, I’m left wondering:

1. A lot of extremely complex Intel macOS binaries “just work” on M1 by design of Rosetta 2. So what’s the specific issue with ABBYY’s binary(s)?
2. Is there another component or method to OCR a scan into DT3 that I should consider? I will stick with DT3 as the destination for my docs but I care less about how they get there. Just wondering. I’m not attached to the existence of apps or temporary files that hang around due to Fujitsu and/or ABBYY right now (ScanSnap Home and whatever else seems involved).

[cgrunenberg]

@emptybits

A lot of extremely complex Intel macOS binaries “just work” on M1 by design of Rosetta 2. So what’s the specific issue with ABBYY’s binary(s)?

A good question but actually one for Abbyy.

[pablo]

Well, Sir, thank you for your statement about question number 1.
What about Question number 2.?

@emptybits

2. Is there another component or method to OCR a scan into DT3 that I should consider?

I use the same workflow as our worthy colleague Emptybits an I quess we are not alone in the big DT3-Universe.

[BLUEFROG]

Welcome @pablo
You can use other third-party OCR applications then import them after the OCR is finished. Many scanners already include some OCR software or capabilities.

[APC]

@emptybits

A lot of extremely complex Intel macOS binaries “just work” on M1 by design of Rosetta 2. So what’s the specific issue with ABBYY’s binary(s)?

Rosetta 2 does have specific things it cannot do compatibility wise including kernel extensions and drivers.

However, ABBYY’s own OCR Mac products are listed as not compatible with Big Sur (even Intel) on their website’s small print. The fact it might run on Big Sur Intel is not official support.

[APC]

@BLUEFROG

You can use other third-party OCR applications then import them after the OCR is finished. Many scanners already include some OCR software or capabilities.

I used to use Abby Finereader so I could tweak and get optimal OCR results specially with diagrams etc.

I winder if Tesseract complies on Apple Silicon natively?

an open source one called Tesseract it might not be great but can be good enough to OCR for search capability. Its command line so not the easiest to set up but you might find a GUI wrapper for it online or perhaps someone could wright a devonthink script for temporary usage?

[APC]

is it me or is this working now?

it completed OCR’ing some files today

[cgrunenberg]

It didn’t work on the development machines (DTK) but it’s working on M1 Macs.

[BLUEFROG]

I can second @cgrunenberg’s comments. On an M1 MacBook Air, OCR is working as expected.

[APC]

@cgrunenberg

It didn’t work on the development machines (DTK) but it’s working on M1 Macs.

It didn’t work on my M1 at first but does now

[littlesunjill]

What did you do to make it work?

It didn’t work on my new Mac M1. Did you install Rosetta 2? This is what I saw from another thread.

[APC]

I didn’t really do anything there was an update. I do have Rosetta 2 installed since day one so could be, although the OCR plugin is showing as a Mach-O bundle supporting both x86_64 and arm64. There was an issue with Abby’s OCR SDK used by DEVONthink OCR add-on because Abby’s Finereader, a separate software package, also didn’t work until they updated it. I have used OCR under DevonThink3 today and its working fine, Make sure you have the OCR addon installed from DevonThink3->Install Add-ons . Also make sure you’re uptodate i.e DEVONthink3 3.8. Try uninstalling and reinstalling the add-on.