Translation from English to Spanish options

[Sarah32263431mzw1]

Hello! I work at a small community college and we need to convert our existing materials from English to Spanish and then going forward, all materials will need to be produced in both English and Spanish.

I have done some light research trying to find the best option for our small college. I tend to lean towards purchasing a subscription to Redukon or something similar where I can upload, translate, download, and then tweak the file knowing there will be a considerable amount of extra characters.

But the control freak in me likes the idea of a plug-in such as ID Extra's Translate where I can translate and tweak the file as I go. I am concerned about the pay-as-you-go model since we will have a lot of materials that will need translation on an ongoing basis.

My question is for those of you who have translated multiple documents, what would you recommend and why? Am I missing any other alternative way to translate our materials?

Thank you for your help!

[James Gifford—NitroPress]

Not directly addressing your translation issue, but it's worth considering that subscription services (like Adobe's model, for example) usually mean continuous update, better customer support and problem resolution, never having to budget for irregular updates, etc. Since your need/workflow is continuous, I can't see a downside to a continuously "current" tool for it.

(It's much different when a need is one-shot or very infrequent, and then subscription tools can become a budgetary burden, especially for solo/small shops.)

[Sarah32263431mzw1]

Thank you! I don't have a problem with a subscription, but I want to be sure that I am getting the most for my money. I will need to do a little digging to see how much a project would cost buying credits as with ID Translate vs a monthly fee, but really - I am wondering if there are any pros and cons between those 2 types of platforms. Thank you!

[TᴀW]

Well, as the Id-Extras Translate creator, obviously I second @James Gifford—NitroPress , but in any case, the entry price for Translate is very low ($15, which includes a $5 translation credit). So I would recommend giving it a test drive, and seeing if you're happy with the results.

You can include a custom dictionary and/or a glossary (which is like a dictionary, only smarter as it takes into account different grammatical constructions, whereas a dictionary is a strict replacement).

The big advantage with our Translate add-on is that all the formatting is kept intact in your InDesign file. It's like "Google Translate" for a web page, just for an InDesign document! I'm not sure if other systems make it quite as easy...

Coming soon, too, is the ability to add a paragraph or two of "context," to assist the translation. This is a new DeepL feature, and I hope to include it in our Translate add-on in the nearish future...

[Sarah32263431mzw1]

Thank you! I wish there was a free trial option. I realize the initial cost is very low, but when working within an organization such as a college, I need to have expenses approved before I can make purchases. It would be nice to be able to try it out and let my supervisors know why the purchase I am looking to make will be so helpful.

Redokun has a free trial option, so I have taken advantage of that to translate one project from English to Spanish. It was all done online and then I was able to download the indd file. There were a couple of formatting tweaks that I had to make since the Spanish language uses so many more characters than English, but it kept all my formatting and styles, which was nice. 

How is ID Translate different than Redokun?

I appreciate your help!

[TᴀW]

ID Translate offers an extremely simple workflow: You select some text in InDesign, or choose an entire story or document from the Translate window, click Translate, and a few moments later the selected text, or story, or document, is translated. The result is a document that looks identical, but is in a different language. The translation uses DeepL's engine to machine translate the text.

All paragraph styles and character styles, as well as any local overrides (that is, text that is formatted differently but does not have a character style or paragraph style applied to it) are preserved.

You don't need to fiddle with an external browser-based UI. It's a simple one-click operation.

Of course, you can provide a comprehensive glossary for many language pairs, and DeepL will use that during machine translation to ensure accuracy and consistency to match your requirements.

Redokun is geared towards human translation, although it does also have a machine translation option. But it is much more complex, and requires exporting the text from InDesign, fiddling around in a browser-based environment, and then update the file.

I haven't tested how well it works with local overrides, but I wonder if it does as good a job in that department as Translate?

And of course, there's the price. The cheapest Redokun plan that includes machine translation is $325 per month. Translate is a simple pay-as-you add-on, with an initial translation key costing only $14.95.

[Joel Cherney]

My question is for those of you who have translated multiple documents, what would you recommend and why? Am I missing any other alternative way to translate our materials?

Well, there are lots of ways that involve commercial services, and either human translators, or human editors correcting machine translation, but these are typically more expensive than DIYing the formatting on AI-driven translation. For some kinds of translation - especially relating to health, PHI, or regulatory compliance - there are really strong arguments against relying on AI. But I don't imagine that your small community college has these kinds of concerns.

One way you can estimate the scale of such concerns is this: you can ask yourself "What are the consequences of an incorrect translation?" Because the consequences scale, to a degree; "I didn't make the sale" isn't as bad as "The reader filed an accessibility lawsuit over it." Lawsuite are better, in the grand scheme, than "the reader was injured as a result of believing the incorrect translation." I just saw a Critical Error a few months ago, where the poorly trained AI confused the Chinese glyphs for weeks, months, and days. That doesn't sound that bad, but it was a childcare handbook describing the ages at which children could be left alone, and for how long. So, that particular incorrect translation could have resulted in the death of a child. The moral of my little story here is that you shouldn't skip the part where you hire an editor who already knows how to do post-MT review. 

So, I can absolutely make some suggestions if you're interested (20+ years in the language industry, have formatted billions of words in dozens of languages over those years), but my advice is likely to be right out of your budget. (For example, if I were given your job, I would absolutely get the best translations I could for the English documents I already had in my archive, and then I would use the new translations to build bilingual corpora. Then I would use that data to roll my own machine translation tool based off of the vetted corpora, and could apply that to all new projects henceforth. ) But how much volume are you looking at, here? How many words in your archives currently, and what is your projected volume of new translations? Those numbers will let you estimate your total cost per year, right? That will help you figure out which tool will likely be the most cost-effective.