There are normally three places that I would check for character display issues such as this:
Do the characters look correct inside of the Components in the Tridion Content Management Explorer (CME)?
Can you open the published page (assuming that this is not a Dynamic Component Presentation) in Notepad++ or similar on the web server? Do the characters look ok in here?
How do the characters look inside of the browser?
Given that the content is correct on the QA environment, I'd assume that what you're seeing in the CME (Number 1) is correct.
If the characters are not correct when you open up the published page in Notepad++ (Number 2), then it is probably that the
Default Code Page setting is not correct on your Publishing Target.
If the characters are correct on the published page in Notepad++ (Number 2), but are not correct in the browser (Number 3) then I would normally check that the encoding used by web application.
- In the .NET web.config:
<globalization fileEncoding="UTF-8" requestEncoding="UTF-8" responseEncoding="UTF-8"/>.
- If it is Java based then I think that
response.setCharacterEncoding("UTF-8"); will do this.
However, I'd assume that your application code and globalisation configuration is the same between your QA and Production environments, so it is probably not that.