Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
I think it like entry generally combinational logic, but when I design this block, I put the gamma table on memory, and connect your un-correacted data on memory address bus, output data is gamma corrected data.
to implement it with a table, it's not easy
to change the correction point when needed, it's better to store the turn point and multiply it with different value if the
speed is allowed.
implementing lookup tables for the purpose of gamma correction is quite straight forward with FPGA these days, LCD controllers I have used have tended to have a gamma lookup table to allow the user set the values.
You can approximate whole chip resource and physical application on gamma correation...
Because our data bus width not more than 10 bits, generally. so we can get one smaller memory block. (i.e. at most 1024x10), but if we change design into multiplier and another circuit, please check the gate count and chip size, you can find out it's larger than you entry the same feature by memory.
When low-level/cost LCD T-CON, generally the data bus width about 6 to 8 bits, so only one small ROM / RAM (i.e. 64x6 to 256x8) is needed when you entry gamma correation frature.
Another word, if you want to modify gamma table,you only have to doing table-edit job!
:wink:
in contrast to lut, how about the interpolation between the turn point given, I think this will definitely save area,especially for high definition use,
for example 10bit
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.