Condense 1.5

Support Page

Frequently Asked Questions:

#1 What does QuickFix do and how does it work?

QuickFix will display a Popover. This popover is designed to be able to be used without moving your hands from the keyboard at all. You can use the mouse / touchpad, but that's optional.
It will show you one misspelled word at a time with 3 words of context. The misspelled word is colored.
You will receive up to 4 different corrections on the Arrow Buttons. You may click them, but you could also use your arrow keys.
The most common suggestions will be on “Arrow Up”. If Condense is >90% sure it is the right word, it will also be colored.
Pressing the “Ask Me Later” Button (Hotkey: Return), will continue with the correction. When you are done with all the other words, Condense will ask you to type in the correct version of the word. So you don’t lose your “flow”.
If you want to ignore a word, press Spacebar.
If a mistakes occurs multiple times, you will only be asked once.
Don’t forget to save when you are done with QuickFix. Saving will automatically copy the current text into your Pasteboard. Just paste the text wherever you need it.

#2 Why don't you offer an autostart option?

Condense is a very complex and thus memory intensive application. I recommend using it on demand. I don't want unexperienced users to run into memory problems. You can enable autostart anytime using this little tutorial.

#3 Why do I receive a whole lot of gibberish or no text at all sometimes?

Optical Character Recognition has been developed with books and print media in mind. Characters are a lot bigger (DPI >600) in size there, so small fonts (<14px) are very difficult to recognize for OCR software. Exquisit Fonts are also very hard to extract.

Here are some tipps to increase accuracy:

• Zoom in. Bigger letters do help, even if they get pixelized a bit
• Using a Retina display increases accuracy immensely
• Try to minimize the number of elements which do not belong to your text: UserInterface, Images, etc. by selecting the smallest area possible
• When working with angled text, try to hit the angle as close as possible. 1° does make a huge difference! Try to rotate until the drawn lines are parallel to horizontal text elements, like the bar on top of the letter 'T'.
• OCR works best on backgrounds with consistent colour
-> If you encounter a gradient, use the medium contrast settings
-> If you try to extract text from a photo (eg. Internet Meme), use high contrast settings
-> Text using borders (typically used on Memes) is horrible to OCR.
• If you have a long text consisting of multiple paragraphs, it does make sense to scan a single paragraph at a time.

#4 Why is the application 450MB in size?

I started Condense with the English language in mind. I was asked by many users to implement more languages and that's what I did. Those languages take up a lot of space unfortunately.

#5 Why can't I download the languages from within the application?

The language files are quite big and I don't have the bandwidth to support the traffic they would create. I don't want to rely on a 3rd party mirror either. I also can't edit my application bundle from within my application (which is a security feature), but you can. If you want to get rid of multiple languages, do the following:
Keep in mind: You have to reinstall Condense to get the languages back.

1. Open your 'Applications' folder in Finder
2. Alternate click on 'Condense.app' and click 'Show Package Contents'
3. Navigate through 'Contents' -> 'Ressources' -> 'Tessdata'
4. Remove the languages you don't need

#6 Why can't I drag and drop whole PDF documents?

I considered and even tested this feature. But I had to realize that this is not the usecase Condense excells at. I do use a lot of algorithms on the captured image to improve its recognition. I can only do so, because I know that the captured size does not reach more than a few million pixels. Scanning documents took either a) way, way too long or b) was too unreliable.
I consider Condense a helper tool you can count on if you need a relatively small amount of text. It is not a flagship OCR Suite to scan whole books.

#7 Which icon pack did you use?

I literally did spend several weeks searching for an appropriate icon package. I was looking for consistency and a modern look. I didn’t find a single one with more than 5-10 good icons out there. I then discovered some free icon fonts: Font Awesome and two iOS 7 style Fonts from www.pixeden.com. Then I used the awesome Ultra Character Map application to export the desired icons as vectors into my graphic application.

If you came here, chances are high I did something wrong or suboptimal. Tell me how to improve:

 

geeky.gent