I use a digital refractometer now. If I question a reading then I grab a hydrometer and it is always within 1.001-/+. I use the downloadable excel spreadsheet you will find within Silent’s link above to Sean Terrils site. You can use either the instant calc on the link to figure out SG/ABV via Brix pre/post ferment or download the excel sheet. You will only need to account for your personal wort correction factor as explained below.
One thing you will need to do to use these calculators correctly to find SG values before/ after-ferment is to use your refractometer along with a hydrometer for a couple samples until you find out your breweries wort correction factor.
I personally found that I need to have the correction factor set to 0.96 to match Original SG (before ferment) and then my final readings always line up as well. If you look to the calculators the default wort correction factor is 1.040.
Heres an example: If I left my correction factor at 1.040, my most recent APA that read 12.4 Brix would show an OSG of 1.048 in the calculator/spreadsheet. Which is much lower than the hydrometer is showing for 12.4 Brix which is 1.052, Which also matches another sheet I use that has proven pretty accurate for OSG numbers only. (see below for that link also.) Now if I simply rachet down my wort correction factor until it matches my dual reading(hydro&refract on same SG sample), which again in my case was 0.96 instead of 1.040 I always show correct OSG/FSG numbers.
Great OSG reference, This is by far the best conversion table I have ever found:
Zymurgy Jan/Feb 2001 Created by Mike Schwartz
The link will open to a Google docs folder which prompts you to download or open the file as soon as you land on the page this link directs you to.
The file is safe as of today, as I downloaded it and double checked it was the right chart and virus free. This chart is relatively hard to find so I recommend downloading a copy.
http://www.google.com/url?sa=t&rct=j&q= ... Wc&cad=rja
EDit* If you are using a scope style refractometer that reads in Brix AND SG then we may have another issue. Which is to trust your Brix readings only and convert as directed above only. The reason I say this is because I think the scope styles that read in dual units like this were seriously flawed in matching SG to Brix within the scope you would read values on. Others can verify or debunk this factor as I have never used one, only heard the reports when these first came on the market, maybe I think 2-3 years ago.