To Buy or to DIY, here is my question


If I buy the speakers that appear closest to what I desire

they are $11k new and $8k used.

If I buy the raw speakers and build it 

the speakers alone are $2.2k.

That is a 3 way system.

Still must add costs of XOs and cabs.

 

Assume the total custom build cost would be about $3k.

The $8k speakers used are Proac D40Rs.

The raw components would be from ScanSpeak and SB acoustics

and include 10" woofer, 4.5" Mid and a planar ribbon tweeter.

MadiSound provides XO advice. 

 

Comments???

 

chorus

BTW, I do encourage you to check out Troel's website, a builder after my own heart, but he has a lot of pictures about construction and assembly that can really help the new builder see what's involved.

 

http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm

I work at a high end retailer.  I can buy name brand speakers for 50% off or less.  Nevertheless, I built my own.  2 reasons...none of our brands make what I wanted my speakers to be in terms of size, form factor, etc.  AND, even at half price I was not willing to pay for what we sell...all too expensive.  I am not a big one for luxury goods.

My concept was a 7" woofer in a 1.5 cubic foot vented box, with a 2" dome MR and ribbon tweeter.  So I ordered a set of drivers and prefab cabinets from Parts express and paid a cabinet shop in town to rout the front baffles for them.  I bought prefab crossovers from PE, and used in-line L pads to match driver efficiency to get the best blend.  Version 1 did not satisfy me.  Next I swapped out the Aluminum dome MR for Soft dome.  Still not what I wanted.  Then I recreated the crossovers with higher quality components and somewhat different points.  Still not good enough.

At this point I scrapped my original design for one using a completely different MR-TW concept.  I sent the driver selection to Tony G. at humblehomemadehifi.com in Holland, and had Madisound fabricate my crossovers using premium parts, bought another pair of front baffles from PE, and eliminated the L-pads.  These are what I have now.  I wound up spending something like $2.7K on all the parts, labor, design fee for Tony, etc. including all iterations.  My speakers aren't perfect, but for well made symphonic recordings, they are quite satisfyingly true sounding.  Their low end surpasses that of the 6.5" 2-ways that go for around $10K at the shop, and without a robust low end, symphonic material just sounds unbalanced.

So I am firmly in favor of DIY unless....if you are a perfectionist and cannot stop fiddling until you get it just right, you may never get there.  OR, if you can easily afford the speaker of your dreams, then do so and you can always tinker in your spare time for fun!

Cost / Benefit thought experiment: You receive the Proac D40Rs
as a gift? What then? 

Price and cost are not the same.

@chorus , by all means! If you are looking for point source speakers DIY is the best way to go with a few caveats. I would stick with drivers that have the same radiation characteristics. Putting a ribbon tweeter (line source) in a point source loudspeaker is going to significantly change the frequency balance with even small movements in the room. Keeping the drivers phase and time correct is not easy and frequently results in some very complicated crossovers which are impossible for the novice to design. You can stick a basic crossover in there but you will not get the best imaging. You can not do this well without a calibrated microphone system. Depending on luck does not work well with speaker design. If you do not want to get involves with DSP you will have to get very inventive. But, that is the fun in it!

If I were to do this without DSP I would keep the drivers in separate enclosures as small as possible even open (except for the woofer). This will allow you to move them relative to each other. Using the calibrated microphone you can measure the group delays and move the speakers to get them all time aligned which if you are lucky will also get them phase aligned. This should be done outside to avoid reflections that can confuse things. Once you get the locations figured out you can mount them in a framework maintaining those relative locations ala Wilson.