AMC does need LaTeX to extract the scoring strategy from the source file, but neither during data capture, nor during scoring (if you did not check the box "update marking scale").

thanks for your support.

-AMC can be installed without latex on mobile phones ?!

-AMC can be installed without latex on PC ?!
i mean with Unified form (for all teachers) like this:
http://www.osterud.name/omr.png

Combining the mobile phone shooting with a server side program could be the architectural solution.
You would need no more then an app to shoot the photo and the server could do the rest.
The work flow would be mainly the same as in the current way of working, with the addition that the server could do the OMR recognition and correction
as soon as a photo arrives through this app.
In theory this is possible with current devices (tables, smart phones) and someone willing to write the app.

Scanning with mobile phone would probably be very impractical.

If you wish to use a camera you would need one with a decent macro lens to avoid distortion at a short distance, plus a tripod or a makeshift stand. I've tried it, and it creates more problems than it solves, I've lost a lot of time post-processing images and having to do same photos several times. It's hard to consistently get reliable images.

Flatbed scanner is more practical by comparison. But in the end, you really want ADF scanner.

Dear Nikola:

I agree with you, I think scanning with mobile phone would be impractical with AMC's current OMR engine. But it's not an intrinsically unreachable development-stage for AMC, I think.

Just consider:

1. ZipGrade (www.zipgrade.com/): Android/iOS app for grading multiple choice tests with you phone camera. Although not finished yet (it's still under active development), it's functional for some limited formats.
2. GradeCam(www.gradecam.org): software for graing tests with your laptop's webcam (or with a document camera). It has a very powreful OMR engine, supporting 3D deformation of the answer sheet (http://www.youtube.com/watch?v=QTIzV78nsNw&feature=c4-overview&list=UU_bZ1W-b7quSSsBdpjYKucA).

The real problem here is the complexity of developing such an OMR engine, although it's not about technical limitations: a single developer will never reach such a target, a lot of help is needed.

Of course, you should consider first several pros and cons before deciding to implement this feature: what kind of users will get fenefited for this feature? What will be the involved costs? These are some of the questions to be answered before.

See also use mobile camera to scan the papers ?: Nikola seems to have found some solution.

Yes, this can actually be done by using CamScan Software (or MD scan) on your tablet or smartphone to post-process the image taken by smart-phone camera. Both softwares work equally well to produce usable images for AMC, but for MD Scan you need to pay if you want to use batch processing option.

With CamScan the trick is to take photos using the standard smartphone camera interface, than go to the picture folder, select the photos and batch import them to CamScan software. Processing might take a few minutes.

We had some issues with pencil marks under direct neon lights, at a certain angle the office neon light will reflect right back at the camera shutter so the mark will appear very light in color and not get recognized. We have this problem as our offices use naked neon lighting without any masks or diffusing screen.

An easy way around it is to use black pens for filling the multiple choice options. At that point lighting stops being an issue. Another way is to use diffuse light source, take pictures outside under natural light, or construct a simple light-box from cardboard box and aluminum wrapping (takes 5-10 minutes) that will bounce around the light regardless of the source.

This was a fun side-project and this is a workable option for places without ADF scanners. However, a solid ADF scanner is cheaper than Galaxy S4 and such, so in the end, if many teachers have such devices, they probably work at a school that can afford to procure a proper ADF scanner. Working with a scanner is still much more preferred for high volume of exams.