If you go to Weyermann's web site you'll find this statement:
In order to raise the color of 1 hl beer though 1 EBC, 14 gr. of SINAMAR® is required.
. Thus 140 mg/l gives rise to 1.9685 SRM corresponding to 71.12 mg/L/SRM. Using this number gives 6*19*0.0711 = 8.107 grams for 6 SRM in 19L which is quite differnt from an ounce (28 g). ???
If Beer's law holds the increase
should be linear. We didn't talk about Beer's Law on the show (or if we did I don't remember it). There are theories as to why it might not and there may even be literature out there with exceptions documented. In my experience, however, I have never seen it violated for beer and I do check for it whenever I use dilution as a means of getting dark beer into the accurate range of the spectrophotometer (for color measurement). So while I have never used Sinamar (or even seen it) my guess would be that you should be able to darken a blond to amber and amber to porter color in equal steps by making equal additions of Sinamar. I would expect the results to be very close to linear or linear within the bounds of reasonable measurment error. If there are deviations from linear it is probably going to be found in beers with SRMs of 30 or 40 and up where doesn't make much difference if a beer is 40 or 45 SRM.
The fact that Weyermann reports the coloring capabilities of the stuff as if it were linear (if it were not 14 grams per EBC per hl at any color depth they would have to give us a table or a curve) suggests to me that it is linear or quite close to it.
