Faster...

Examples might be:

- Advanced use of CAT tools (ultra-large TMs, MultiTerm, glossaries converted into TMs, etc.)
- Find and replace macros (run on Unclean Trados files, for example)
- Glossary macros
- Templates
- Collaborative solutions
Powerful PC with 2 large screens (each 24")
- Electronic dictionaries, including my own specialised glossaries
- Medium-sized TMs, pruned from time to time of useless terminology. I find that ultra-large TMs give too many matches, so it is hard to find good ones. As we say in English, you can't see the wood for the trees.
- Speech recognition. Perhaps my single most effective technique. Translating a well-written text in a subject I am familiar with with a well-trained microphone can be highly productive.
3. Fast typing.
4. 9-10 hours of continuous work, no distractions
What I like to do is create an Unclean Word file (translate it first using a blank TM), hide the Source text and do lots of find and replace on terms/expressions. Only the Target text is replaced, so the source is intact. Obviously, you have to be careful, as with any find and replace!
* My CAT tool (Wordfast)
* My CD dictionaries (I have four)
* My fast internet (well, fast enough)
* My browser, built for speed (Opera)
* My word processor's spell-checker (underlines)
* My headphones and such musicians as Vaakevandring, Slechtvalk etc
* My high chair
* Mugs and mugs of liquid chicory (or coffee, if chicory is unavailable)
I also set myself targets of what to get through in 50 minute sessions, using the rest of the hour to answer emails, get up to stretch, make a cup of tea, whatever.
'We've got a specialist for legal stuff who often posts in the forums who claims a typical daily output of somewhere around 10,000 words a day. As far as I know she does most of her work with voice input and sticks strictly to the subject she knows best and practiced as an attorney.'


Voice input: I use remote-location digital audio-typists - a service now offered on the Internet - at up to 110 words a minute who have a strong legal and medical background. Advantages over VRS/voice recognition systems like Dragon:

1. far less installation space required for Olympus, Grundig and Philips transcription kits;

2. having worked together for over 20 years, we can turn around about 2,000 words an hour: I dictate Romance faster than Germanic.

3. a built-in editor: a VRS asks the 'dictator' neither if the output makes sense, nor picks up on source-language words, clauses, paras. or even entire pages omitted in the translation.

4. no need to dictate lists of figures and numbers in tables, such as annual reports and accounts.

5. disadvantage: erosion of profit margin through typists' charges.

At work I use www.oxfordreference.com, which is quite good for English reference material.

Regarding voice recognition software, how easy it is to switch across languages?

I believe that with some of them (like Dragon), you have to close the application and start it again to switch languages, which is really a pain if you want to use it to dictate in your target languages and, at the same time, to make searches in your source language.
Yes. A portable digital dictaphone like a Grundig Digta - no need to be chained to the PC with e.g. a Philips Speechmike microphone - can input the dictation to a PC which serves as a dictation station. Sound files are stored and retrieved at will.

The huge, albeit compressed, voice file as a .wav is then emailed - by ISDN or pref. by ultra-fast Broadband - to the audio-typist's PC at a remote location. Normally, on most transcription kits, 1 hour of dictation generates a 6MB voice file.

The translation files emailed back can still be forwarded to an outside editor - but further eroding the profit margin.

There are Internet audio-typing transcription agencies advertising themselves. The target-audience seems to be mainly legal and secondarily medical and accounting, the latter being more of a number-crunching profession.

It has long been proven that, unless the translator is an extremely fast touch-typist, it's possible to dictate/speak 7 times faster than typing/keyboarding. Output must at least double, incrementally again over the years if you have no dictation recording practice, on the assumption that the work is there in the first place.