Based on some of my expereinces, I would say that not only are there 
quite a number of misaligned systems, but there are also a number of 
processors which were just plain defective. Unless you have the Dolby 
tester for checking the cards, there is no way to know if there is an 
internal electrical problem or not.

So, many users could have indeed set up the cal levels properly, and 
aligned their recorders properly, but would still end up with a bad 
recording due to an internal card problem.While we have generally found 
the cards to be stable, we have run across problem cards in some of the 
used NR units we have bought over the years (especially ones from large 
studios, with many multi-channel  frames to maintain).

As you state, when moving a DBX tape from one machine to another, you 
need to pay careful attention to both the frequency response, and to 
some degree, the level of the recording. For example, if a recording is 
done on one machine at a level of 250 nW/m, and that same tape is played 
back on another machine calibrated at 185 nW/m, and the DBX processor is 
located in the machine (ahead of the reproduce level control), you could 
run the risk that the procesor would be driven into the unity gain 
portion of the expander slope, thereby causing the louder portions of 
the recording to sound compressed. I have seen situations exactly like 
this occur when levels were not properly matched between machines.

Also, quite a few recorders do not allow you to adjust the LF response 
properly, which, when coupled with head bump issues, could cause easily 
result in a response error in excess of 2 DB. This then would be 
magnified into a 4 db error by a DBX Type I processor, which would 
indeed be quite audible!

However, as long as the respose is flat, and both the recording and 
reproducer processor are kept within their linear region, I haven't had 
too many issues. (Again, this also assumes that there isn't a problem 
with the linearity of the processer itself, which I have run into, 
especially on older units).

Unfortunately, there is scant information available as to how to make 
any calibration adjustments on the boards. We have blueline drawings for 
some of the processors, but they don't have much calibration info or 
component criteria on them. Guess that DBX just wanted the units to be 
returned to the factory. (To be fair, I think that there are some hand 
selected components in the compander circuit which probably require some 
special test equipment, and known selection parameters, which would be 
outside of the abilities of most studio techs. This was also the case 
with some of the UREI equipment we've had over the years.).

Bob Olhsson wrote:

>Scott D. Smith writes:
>>As long as the system isn't driven so hard that you exceed the range of 
>>the level detector, then it *should* track properly at any level.
>Where level calibration got challenging was in applications where you needed to move tapes between different machines. It wasn't nearly as bad as Dolby in that respect however it was far more critical than the manufacturer would have people believe. For that matter Dolby grossly understated how critical level alignment was. I've encountered a huge percentage of  sloppily aligned Dolby A and SR systems.