I have a four way active horn system (three way horns with bass reflex subwoofers) that I use DSP for crossover and processing. Was the smartest thing I ever did. There are still plenty of things to get wrong, but once you dial everything in, it's a beautiful thing.
Allen
following i was never able to overcome :
- beaming. Specially in my midrange horn
- agressiveness in the upper midrange region. Tried many compression drivers, but that was always a issue.
- horn coloration
All these short comings came suddenly to an end with my actual system. And it took me not much tweaking to get a final result.
I use DSP to run my 4 way system (currently 2 horns for the highs, 2 DR's for the lows). I found linear phase crossovers and / or steep slopes screw up the directivity and thus the listening window and power responses quite a bit around the crossover points, at least with my system. It is not a cure-all. I ended up with shallower slopes cutting to higher orders away from the crossover in some cases. If you can take good measurements and then post-process them while designing your crossover, you can get better results. If you have large driver spacings you will probably get some lobing around crossover but I would imagine this would be better while keeping a fairly constant DI than having a very sharp change in DI with no lobing.
I don't really want to hijack the OPs thread, so I'll just post an image. I haven't started my own thread yet as I've only really just completed the build. I'm still fine tuning settings etc. In fact even these pics are a little out of date as I've made modifications to the horn stands since this was taken;
All those problems go away (mostly) with an OS horn
But then you inherit a few others that are more benign
Don't forget that some people LIKE the colouration of traditional horns. Just as you like the dipole colouration from your AMT. Nothing wrong with that
Not everyone will find nirvana with a AMT crossed to a 12P80Nd. You need to realise that and stop ramming it down everyones throat. Its becoming annoying
Yes, that could be a problem. You have found something that works, I believe that it works well for you. Midrange is usually where the difficult compromises come out in all their glory.
A baffle doesn't have to be flat. To me there is much in having a smooth transition. There is a middle ground between a flat baffle and a loading horn. If you go from narrow to wide, you have to cross this ground somewhere and some way.
Yes, exactly. I have a 3-way horn loaded system - mid and high frequency with direct radiator 15". I use steep linear phase digital xo, along with driver linearization and time alignment of the drivers. I wrote an article on it:
Computer Audiophile - Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
I changed out the high frequency CD and waveguide after the picture was taken to a QSC 90 x 50 constant directivity waveguide, similar in size and polar response to the JBL 90 x 40 constant directivity midrange horn. The small high frequency exponential horn in the pic drove me nuts with it's beaming response, Now it is seamless between all three drivers and I can sit anywhere on a 6' foot couch with smooth response, and does not sound harsh in anyway.
To me, there is a lot of value in matching directivity/polar responses ala Earl Geddes: http://gedlee.azurewebsites.net/Papers/directivity.pdf. Using linear phase digital xo's and time aligning the drivers, especially horns with big z offsets, really made a difference in my DIY system. While this is an older article, the principles of why time alignment remain the same: Phase, Time and Distortion in Loudspeakers
Kind regards, Mitch
How do you guess that, it may be working very well
I assume you know the profiles and have an educated guess on their dispersion by the photo. Still without exact knowledge of listening distance and the crossover points and the crossover orders of these channels, one can not know if they will sound right or will have problems. You can have even greater distance between these two channels and still could have good balance continuous sound across the bandwidth. Lobing does not start and end with the driver/horn positions.
Rewind, I’m sure it won’t surprise you to learn that I don’t necessarily share your opinion. While I’m not saying a larger distance between my midbass and high frequency horn is a necessity, I am saying that in this case it’s not detrimental. I do however, agree with kodomo. If I’ve learned anything over the years of listening to many systems, it’s that you just cannot tell how good or bad a system sounds by simply looking at it. I’ve been surprised on a good number of occasions (for the better and worse).
My listening room is pretty large, with a high ceiling. The listening position is perhaps further away from the speakers than your typical living room setup. I wanted to generate a large, realistic soundstage. I don’t run steep crossover slopes and I spent a good deal of time experimenting with crossover points and overall configuration. It sounds very good. I’m not saying it’s perfect, I think the subwoofer integration could use some work but I’m working on that. This is only my opinion. You are of course entitled to yours. But to even have an opinion based on a single photograph seems pre-emptive.
Nice setup! Is that Electro Voice (HR90) mid horns? How big is the room and what is the relationship of the horns to the walls and where you listen?
Thanks POOH. Top horns are Yuichi 290FL (no fins) machined from laminated baltic birch ply. Room is about 8.5m long and 5m wide. Listening position is around 4.5 to 5 m from horns. Unfortunately horns are too close to back and one side wall to be considered ideal, but as this is a living area I’m somewhat restricted. To answer your question though, back wall is about .5m away and side wall about 1m. Back wall is treated with acoustic panels.
