<P> Due to higher accuracy and the ability to couple it with other measuring techniques (% CO2 and% alcohol), most soft drink companies and breweries use an oscillating U-tube density meter . Refractometers are still commonly used for fruit juice . </P> <P> When a sugar solution is measured by refractometer or density meter, the ° Bx or ° P value obtained by entry into the appropriate table only represents the amount of dry solids dissolved in the sample if the dry solids are exclusively sucrose . This is seldom the case . Grape juice (must), for example, contains little sucrose but does contain glucose, fructose, acids and other substances . In such cases, the ° Bx value clearly cannot be equated with the sucrose content, but it may represent a good approximation to the total sugar content . For example, an 11.0% by mass D - Glucose ("grape sugar") solution measured 10.9 ° Bx using a hand held instrument . For these reasons, the sugar content of a solution obtained by use of refractometry with the ICUMSA table is often reported as "Refractometric Dry Substance" (RDS) which could be thought of as an equivalent sucrose content . Where it is desirable to know the actual dry solids content, empirical correction formulas can be developed based on calibrations with solutions similar to those being tested . For example, in sugar refining, dissolved solids can be accurately estimated from refractive index measurement corrected by an optical rotation (polarization) measurement . </P> <P> Alcohol has a higher refractive index (1.361) than water (1.333). As a consequence, a refractometer measurement made on a sugar solution once fermentation has begun will result in a reading substantially higher than the actual solids content . Thus, an operator must be certain that the sample he is testing has not begun to ferment . Brix or Plato measurements based on specific gravity are also affected by fermentation, but in the opposite direction; as ethanol is less dense than water, an ethanol / sugar / water solution gives a Brix or Plato reading which is artificially low . </P>

Why are the brix readings of fermented solutions usually higher than unfermented beverages
find me the text answering this question