Is info for acoustic centers of drivers as mounted on baffle part of data package? Is shimming drivers to a small degree allowed to adjust time delays? Or do we have to work with drivers as mounted on baffles flush? Or maybe one driver can be reverse mounted on backside?
Yes, the drivers have already been measured in the intended box, on the baffle, flush mounted. You should be able to use the measurement files to determine the AC offset.
The rules state that you should design the crossover using the supplied measurements on the system. Do not try to "adjust" time delays by assuming shimming or alternate mounting. I know that this is somewhat constraining, but we need to apply the same rules to everyone.
From math perspective, all filters are applied by convolution. Passive, active, DSP; it is all the same thing.
All practical IIR filters converge into noise floor of system; and are readily truncated for use with freely available convolution engines.
Linkwitz-Riley crossover is convolution of two Butterworth filters; Linear phase version is merely convolution of time reversed copy of one of the Butterworth filters. Similarly linear phase PEQ may be constructed.
What are judgment criteria? Listening only? At just apex of listening triangle?
SPL capabilities? What kind/size listening space and speaker placement therein?
Much better to include wave files of impulse response measurements, and calibration data for microphone.
It seems to me that best way to make a winner is to purchase drivers, build the box, and design crossover correctly by being able to measure and listen. Doubtless some will do this.
The contest is actually quite simple - don't overthink it! Just design a crossover using the files supplied. The assumption is that the measurements already supply all the information that the designer needs to know about the system and the listening environment. What is being judged is the material that is submitted with the contest entry and nothing else. Have fun.
What you and the crowd have come up with this contest is really clever.
I think I can sacrifice a dozen hours of work time and take a shot.
Entries are rolling in...
Just keeping this on people's radar. Deadline for entries is October 9.
Just keeping this on people's radar. Deadline for entries is October 9.
Hahaha! Yes, but I have no friends 🙁
(now accepting friend requests, for a short time only hahaha)
You don't need to attend or "have a friend in the Bay Area" to enter a design in the contest. You can still "win", you just won't be able to collect the prizes. It's just for fun anyway...
So, yes, you can enter and win even if you are "on the wrong continent" as you say.
Folks:
At the risk of over-stepping bounds here, I'm planning on attending Burning Amp. If anyone who wants to enter the contest is interested, I'd be happy to be your team member. If you win, I'd also be happy to ship everything to you at your cost. I don't want anything in return.
Regards,
Scott
At the risk of over-stepping bounds here, I'm planning on attending Burning Amp. If anyone who wants to enter the contest is interested, I'd be happy to be your team member. If you win, I'd also be happy to ship everything to you at your cost. I don't want anything in return.
Regards,
Scott
My best friend is Scottish. I never said I had anything against Scots or Scotts. It was just funny because I was wondering if Weldon was a Scottish name🙂.
Hi Charlie, I have a couple of questions: about what to submit, when you say in the instructions "the PCD worksheet", you mean the whole Excel file? Wouldn't it be enough to send the (much smaller) session file? Also you mention component tolerances, is there a way to enter those in PCD? Sorry but I've never used it before (normally I use my own Excel spreadsheet), I've just started playing with it and it's pretty intuitive but I probably haven't discovered all the features yet.
Also you mentioned before that there's no limit to the complexity of the filter, that it will be judged only on the results it achieves. Do we not get extra points for "simplicity"? I mean, you can always throw a bunch of notches at it to iron out each and every wrinkle, but isn't it better to aim for something more "realistic" that gets a good response with a reasonable number of components?
Many thanks in advance,
Cabirio
Also you mentioned before that there's no limit to the complexity of the filter, that it will be judged only on the results it achieves. Do we not get extra points for "simplicity"? I mean, you can always throw a bunch of notches at it to iron out each and every wrinkle, but isn't it better to aim for something more "realistic" that gets a good response with a reasonable number of components?
Many thanks in advance,
Cabirio
I don't use PCD myself, but you can reduce the file size using the button "Compress all program tables & calculations" (the button at the top right of the design page - you have to scroll to the right to find it). After compressing, the file size should be something like 5-6MB in size. Just send it over like that.
The judges have complete freedom to determine which crossover design is "best". They might look at an overly complex crossover and have some concerns, but that could totally depend on the design. There are so many different variables involved we are trying to stress the bottom line - achieving relatively flat on-axis response - and not come up with some kind of shopping list of judging criteria. I know that this is not reassuring, but that's what was decided for this year.
Last edited:
- Status
- Not open for further replies.