Building a standard curve you can quantitate a DNA or RNA sequence based on a known quantity. First you create a standard curve by diluting a known quantity of target DNA. Basically, you give the machine some "reference" samples having a specific amount of DNA copies (eg. standard 1 = 10 copy of DNA; standard 2 = 100 copy of DNA etc.). Then the machine compares unknowns to the standard curve and extrapolate a value for the unknown samples. It is advised to use a plasmid because:
1. It is highly concentrated (so you can have easily give a sample with a lot of DNA copies)
2. It is relatively pure and clean after miniprep
3. The DNA insert should be in single copy and there should be few "unwanted" DNA sequences.
Plasmids tend be used in qPCR assays when the copy number of the gene is so low that if a standard curve is attempted with cDNA, even when neat the Ct value is > 30 cycles
Thus, even with a 1:2 serial dilution of the cDNA the Ct values are >>30 and in consequence, triplicate tend to be separated by > o.5 cycles and thus the correlation coefficient for the putative standard curve is < 0.9 R2
In effect that means trying to create a standard curve with nascent cDNA with a paucity of the GOI cDNA species is not fit for purpose
This is the function of a plasmid: Because the efficiency of your PCR reaction is principally determined by the primary sequence of your target and thus your primer sequence, if another construct contains that species, e.g. an insert in a plasmid, then it to can be determined to determine the efficincy of your PCR by constructing a standard curve based ona dilution series of that GOI insert/Plasmid
The supercoiled form of plasmid DNA has a reputation for stability, and reproducible results. Like Laurence mentioned above the difference is the threshold cycle numbers. Do you know if you plan on using circular or linearized plasmids? Circular standards range = 16.79 to 36.72 , linearized = 12.89 to 33.59. Here is a good article from NCBI that may help explain more details ...
Article Serious Overestimation in Quantitative PCR by Circular (Supe...
I hadn't thought about that but know for example that it is sometimes easier to sequence and PCR linear plasmids by virtue of emancipation from topological constraints so it makes sense