There are 5 color specification scales in various degrees of use. The earliest was the Lovibond scale obtained by comparing the beer sample to standardized colored glasses. It's weaknesses are obvious. So an instrumental method was devised in 1949 in which the optical absorption of 430 nm light passing through 1/2" of beer was multiplied by 10 and reported as the SRM. In Europe a similar measurement was made but in 1 cm, a different wavelength was used and the result multiplied by 25. Interconversion was not possible (unless one had the absorptions at the EBC wavelength and 430 nm). Eventually EBC switched over to 430 nm but retained the multiplier of 25 and the 1 cm optical path. Thus if a brewer measures A units of absorption at 430 nm in 1 cm the SRM would be 12.7*A and the EBC 25*A.
As the path and scaling for the SRM were chosen to get numbers similar to those obtained with the Lovibond Timtometer the Lovibond and SRM scales are similar at the low end and are, consequently, often interconverted at levels where the Lovibond concept makes no sense (e.g. farbebiers with SRM values of thousands).
The fourth system is the ASBC tristimulus system in which the color of beer viewed in a 1 cm container by light characterized as 'Illuminant C' is computed. It is not a very useful measure as one cannot back out the color seen in a 5 cm wide glass by under illuminant D65, for example (which you can do, at least approximately with just the SRM (or EBC). The fifth method has been proposed but not adopted (at least not to my knowledge). It relies on the fact that all beer spectra are approximately the same shape. The SRM of the beer is measured and the spectrum is then normalized so that it can be compared to the average normalized spectrum of beer. The deviation of the actual spectrum from the average spectrum is encoded in a couple of Spectral Deviation Coefficients. These plus the SRM (or EBC) can be used to calculate the color of beers in any width glass by any illuminant.