A question came up about the certificate that accompanies each Q47-D capsule system. Though it may appear that the stack of certificates in the image below are non-unique, bulk printed cards, they definitely are not.
Each certificate is printed with the actual, measured frequency response of the capsule in question. Many companies who have serialized inventory and certificates use either a hand-written or stamped method for marking serial numbers and other product characteristics that may be unique to each unit (such as frequency response or rated power or weight, etc.). Although the cost to us is greater, we decided to have each certificate custom-printed, which reduces the chances of human-error in our process and looks nicer to us.
Here are two cards I pulled at random from each “deck” above…capsules with serial numbers EQ110043 and EQ110207. You can see that even though the curves look very similar, they are unique. The image on the far right is a chart export from EquiLog of these two capsules’ responses overlapped (EQ110043 = blue, EQ110207 = purple). Interestingly, these two capsules are still not a tight enough match for our matched pair tolerance guidelines even though the responses are really quite similar.
Just for kicks, here’s another random pair of certificates. EQ110028 versus EQ110199.Again, you can see that each certificate is unique.
UPDATE: Here’s a chart export of one of our matched pairs (EQ110172 & EQ110187). There are actually two curves here, though it only looks like one since they’re so closely matched!
Oct 4th, 2011 @ 11:15 pm:David Pickett
Do you have any idea what causes these small diffences?
Oct 5th, 2011 @ 12:05 am:skylar
I don’t mean to sound rude, but, in short, physics. There will always be slight differences in any products that are manufactured. Even the measurement system may be responsible for certain fluctuations. It gets to a point where you typically accept such fluctuations as negligible. In this case, the differences you see are fractions of a decibel, which most would agree is inaudible.
I looked up data points for the upper example above. At 100Hz, they are 0.19dB different. At 1kHz, 0.37dB different, and at 5kHz, 0.93dB different. Incredibly tight if you ask me…and our matched pairs are even tighter!
Even with these small differences, I haven’t seen any capsules on the market that approach the consistency that we have been able to achieve. I would be very interested to know if any other capsule manufacturers have better tolerances than ours.
Oct 17th, 2011 @ 10:02 pm:Milton
We assume the freq. response cards are for cardioid only for only the front side of the capsule… though, there should be 2 cards showing each side of these capsules, since they are dual diaphragm, unless the front is matched exactly as the back.
Oct 18th, 2011 @ 9:51 am:skylar
The front and back membranes are matched within a tolerance limit that we have set at 0.2dB. But the majority of our capsules’ front & rear membranes are matched much closer than that. In fact, it’s pretty typical for our measurement system to report the front & rear responses as indistinguishable (0.00dB difference in other words). The frequency response plot on the certificate is a composite of both sides in cardioid operation. So you can expect both sides’ frequency response characteristics to be as reported on the certificate.