Diy Monitors - to build or not to build

[mhelin]

[quote author="Marik"]Yes, in fact I start reconsidering things and thinking more in Seas Excel direction.
Arguably, both are the very best speakers of their size, in existance, and both are... very pretty.

Now, only decide W15 (5'5") or W18 (7"). Both have their very strong points.

The main question--are 7 inchers too big for near field?

DISCUSS!!![/quote]

There are many good nearfield monitors with 7" woofers (and bigger) like the Dynaudios, so they are not too big. Also the the W18EX Excels are very difficult to use, some interesting comments here (in chapter "Designer's Comments"):
http://www.northcreekmusic.com/Pegasus/Pegasus.htm
(about the problem with the low pass & notch filter which creates a second high Q resonance). Another problem may be that metal cones generally have more directivity than paper cones (due to the pistonic behaviour), so at crossover frequency there will be a peak in off-axis response. However, when designing a pure near-field monitor the preferences are different (I propably haven't considered that enough), and the off-axis response is not as important as the on-axis response and the coherence of the sound field. So you want to have the woofer and the tweeter as close to each other as possible (so the possible waveguide must be very small like with Genelec 8020's, or no waveguides are used used at all). Anyway, I would prefer 5" woofers the for near field use except if I want to hear lots of bass frequencies. Even then a subwoofer is a better option. Also the W15 is much easier to use. There seems to be a new paper cone 5" Exel ( http://www.seas.no/Product%20bilder/e0041.htm ), looks like the CSS full-range.

 

[Moby]

I am not familiar with Supravoxes

I still believe that this babies can do something. Here's the link about them and nice measuring graphs...
http://www.supravox.fr/
Sorry about French, but PRR will understand for sure :wink:

 

[Marik]

[quote author="mhelin"]

Also the the W18EX Excels are very difficult to use, some interesting comments here (in chapter "Designer's Comments"):
http://www.northcreekmusic.com/Pegasus/Pegasus.htm
[/quote]

He crosses them too high--at 1600Hz. They "want" to be at 1200Hz.
Not very many tweeters can handle that.

BTW, another way of time alignment (Earthwork Sigma):

The tweeter seems like Vifa XT25TG.

 

[mhelin]

That kind of time-alignment creates heavy diffractions. Instead the Behringer DCX2496 could handle the time-alignment easily just by delaying the tweeter signal. It is also possible to build an electronic circuit (all-pass filter) which delays the signal (I think there are some circuits on the Linkwitzlab pages). Digital delay is another option (if you can find one with good enough converters and digital input you could avoid the successive A/D and D/A, though with analog sources you then had to digitize the signal). Tony Gee often just tilts back the front panel. Maybe if you raise the speaker high enough you get the same effect (the distance from tweeter to your ears is bigger than from the woofer which is closer). Then the off-axis response again becomes an issue.

Coaxial drivers (like the old Tannoys or newer Seas drivers) is one option, too. Though the mixes I've tried to make with Tannoys haven't ever transferred well to other speakers, the Seas might be another story.
Like this one:
http://www.seas.no/Basic%20coax%20PDF/T18RE_XFCTVH1333.pdf

The 30 degree off-axis response looks smoother than the on-axis one. This might be good one, though there must be some FM/IM type distortion with bass-heavy music.

 

[Giganova]

I guess one of the criteria for studio monitors would be accurate and flat frequency response. By looking at the frequency charts of the woofers/tweeters mentioned here I am not sure if that's the case. Could anyone comment on that?

I intend to build near-field monitors for electronic/dance music production. Clearly, I need a good bass response. That's why I'd like to go with an 8-inch woofer (or maybe 10-inch? Would that make sense?). The monitors don't need to sound better than my current ones (M-Audio BX8) since I just want a different "flavor" for checking my mixes. Budget is $500.-/pair excluding cabinets.

Any recommendations what I should look into?
Thanks!

 

[Marik]

[quote author="Giganova"]I guess one of the criteria for studio monitors would be accurate and flat frequency response. By looking at the frequency charts of the woofers/tweeters mentioned here I am not sure if that's the case. Could anyone comment on that?

I intend to build near-field monitors for electronic/dance music production. Clearly, I need a good bass response. That's why I'd like to go with an 8-inch woofer (or maybe 10-inch? Would that make sense?). The monitors don't need to sound better than my current ones (M-Audio BX8) since I just want a different "flavor" for checking my mixes. Budget is $500.-/pair excluding cabinets.

Any recommendations what I should look into?
Thanks![/quote]

The published response (very accurate, BTW) is for unmounted driver in anechoic chamber. On a baffle Seas are very flat.

Personally, I would not make a 2-way with 10" woofer. Even 8" is little too big.
A good 6.5"-7" in a near field could give you all the bass you need.

There are a lot of good woofers on the market. Peerless HDS, or Vifa PL lines are considered as very good budget ones, and are much easier to work with than Kevlar, Aluminum/Magnesium cones. Nomex also seems to have HF breakups.

Do you have any experience with speaker building? If not I'd highly recommend going for existing one.
This one actually looks very good:

http://www.zaphaudio.com/audio-speaker17.html

If you don't mind and have space for floorstanders, this one looks good, as well:

http://www.troelsgravesen.dk/PL18XT.htm

Both designes are pretty well documented.

P.S. I had for a few days M-Audio BX5A. To me they sounded quite dreadful. I am not sure how much BX8 are different, but if you take a good design and carefully follow instructions, for $500 your speakers can blow socks off M-Audio.

 

[Giganova]

[quote author="Marik"]
A good 6.5"-7" in a near field could give you all the bass you need.
[/quote]
I guess you are right. I will go with a 6.5"-7" near-field setup.

Do you have any experience with speaker building? If not I'd highly recommend going for existing one.
This one actually looks very good:

http://www.zaphaudio.com/audio-speaker17.html

No, I don't have any experience, but I can read circuit diagrams and solder. Its a start, huh? Hey, this looks pretty much exactly what I had in mind! (even though its on the cheap side and would use only 1/2 of my budget).

P.S. I had for a few days M-Audio BX5A. To me they sounded quite dreadful. I am not sure how much BX8 are different

the BX8 sound better, but not much. Do you think a setup like the one above would provide significantly better results?

Also, can the suggested woofers simply be replaced by others (e.g., the L18 for the Excel series woofers) while leaving the crossover? I'm asking because I could spend significanly more than in the setup above and could upgrade the drivers.

 

[Marik]

[quote author="Giganova"][quote author="Marik"]
Do you have any experience with speaker building? If not I'd highly recommend going for existing one.
This one actually looks very good:

http://www.zaphaudio.com/audio-speaker17.html[/quote]
No, I don't have any experience, but I can read circuit diagrams and solder. Its a start, huh? Hey, this looks pretty much exactly what I had in mind! (even though its on the cheap side and would use only 1/2 of my budget).
[/quote]

Let's see roughly. Drivers around $200. Premade box from partsexpress--$150. Crossover parts around $100. Odds and ends like staffing material, binding posts, wiring, shipping, etc., will easily bring you to total $500.

Also, can the suggested woofers simply be replaced by others (e.g., the L18 for the Excel series woofers) while leaving the crossover? I'm asking because I could spend significanly more than in the setup above and could upgrade the drivers.

Don't do it!!! The Excel W18 needs much lower crossover point, and is extremely hard to work with.

If you want to make good speakers, precisely follow all the instructions, use cabinets of the given size, don't change drivers, mount them exactly as in the original, and so on, (of course, unless you know exactly what you are doing).

 

[Giganova]

Thanks for your help, Marik! I think I'll build these speakers following the instructions (which couldn't be more precise). Should be fun! :green:

Cheers,
Gig

 

[PRR]

> can the suggested woofers simply be replaced by others

They're your speakers. Do what you want.

I learned by sorting through surplus drivers, mixing and matching. Got some idea what works and does not work. Which is why I'm sour on the idea of DIY reference speakers. You can knock-out a fine home speaker by guesswork and a little trial/error. Getting a reference-quality speaker is a BIG project. The guys who do it well do it as their main job/hobby.

Later a friend asked for a $100/pair design. I carefully worked it out and gave him Radio Shack part numbers. He actually built them. They sounded awful enough to notice: big dip in the voice range. I pulled off the grilles.... "What are these???" I knew what I had picked and knew this was something else. Same size, similar price, not the same drivers. "Oh, the salesman said these were better." "OK, ask the salesman to design and tune your speakers." In a huff, I bought the exact part-numbers I had picked on paper, and built a pair. It was sure a $100/pair speaker but a lot of fun to listen to, not a hole in the middle kinda sound.

There are a dozen $10-$40 "full" range speakers that make fine monitors; if you aren't mixing loud, if you aren't mixing deep bass, if you learn where to sit for the top octave, if you have other speakers (including multi-way speakers) to cross-check on. Heck, I've polished choir on a $30 pair of JBL PC speakers (with the lovely Fostex 3-inch), but I sure would not mix danceboom on them.

 

[Guest]

Interesting stuff here from Harbeth


Also more interesting a SoS forum thread currently going on where Mr Harbeth (Alan Shaw) is giving out some of his ideas (and an exhange of ideas is going on)

 

[Moby]

I have simple (maybe stupid) question. :roll:
Why designers choose so critical crossover points like 1.5 , 2 or 3.5 khz :?: :?: From side of linearity (electronics) that can be perfect monitor, but when you have to fight with that in "real life" mixing, mastering (music), thing can be real pain. Something like multi-band compressor...
I think human brain can accept some freq. nonlinearities, but phase changes are really unpleasant brain pain :wink:

 

[mhelin]

Crossover point desicion is based on used elements and their attributes, like with the W18EX001 driver discussed which must be crossed over two octaves below the first breakup mode. Another reason is the that the on-axis and off-axis responses of the two drivers should match at the cross-over frequency to avoid any peaks or dips in (usually off-axis) response. It is true that 2-way speakers are problematic - ideally you would like to cross them over at ~1 kHz, but there aren't many tweeters that can be used. Also then the crossover has to be very steep (4th order or more) disturbing the phase. However, there are linear phase crossovers (analog and digital) that could solve some of these phase problems. Digital FIR filters could even correct the acoustic phase problems, but their problem is the latency (signal delay between the input and the output). However, the digital filter can't change drivers acoustic properties so the drivers must still be chosen so that the on-axis and off-axis responses match.

Obviously full-range drivers would be one solution, also well designed three-way speakers, though they have same problems with the phase. Some designers use 1st order filters to minimize the phase distortion, but then the sweet spot will became very small. That may be ok for the mastering speakers but not for general monitoring use. One such speaker with 1.st order (6 db/oct) filter slope is Dynaudio BM5A (http://www.dynaudioacoustics.com/Default.asp?Id=6611&AjrNws=749&AjrNwsPg=1). It still crosses quite low (@1500 Hz), maybe the Esotec tweater they use can be used that low. B&W however crosses their drivers quite high, so you might want to hear some of their products (http://www.bwspeakers.com). I don't think it's any better solution as the midwoofer becames very directive and all the breakup modes distort the sound. B&W for an example uses special Kevlar cone midranges with semi-flexible diaphragm which widenes the dispersion of the driver when compared to stiffer cones. So you need special drivers which are expensive. However, some of these full-range drivers are cheap and are an alternative for usual 2-way speakers. Such design you find for an example here: http://www.zaphaudio.com/audio-speaker18.html.

 

[Moby]

crossover point decision is based on used elements and their attributes

Yes, i know that, that's why I'm talking about some wide range mid-bass drivers. I'm not talking about DIY cheep, pristine monitor, because they are available to buy, from companies like B*e*r*i*n*g*e*r, or some other in that class. (BTW truth's sounds nice, but just nice).
What I'm talking about?
About DIY good, better or maybe say reference monitor for the less price of overpriced P*M*C,s or B*W....
And, again , maybe I'm wrong, but if we DIY-n N*E*V*E, SS*L, A*PI , why don't we go for some "more serious" monitors?
Here's my situation. I payed around 2000 euro for my Tannoy AMS8 (not really happy with them, even flush-mounted..), and must say that all DIY outboard sounds better!
Comments :green:

 

[mhelin]

Moby,
It's possible to clone some "more serious" monitors only if you can get hold of schematics and original drivers, and maybe the original monitors are needed as well to find the contruction details. Also you have to duplicate the speaker box construction and use the same materials and measurements, maybe also use same kind of stuffing material etc. you never know which affects to sound. However, the manufacturers seldom use unmodified OEM drivers, and also some OEM drivers like Dynaudios are not sold to diyers any more. So I think a new design using off-the shelf speaker elements would be much better solution. Also IMHO you don't have to stick with any design as such - just pick up quality components which are likely to perform well together and use a general purpose crossover (either electronic or DSP) and parametric EQ for tuning the system. You may start with a combination of drivers used on commercial speakers, I think some people on this forum know which drivers are used. I *think* for an example that you could use Volt DVC220.1 woofer with Dynaudio D-260 (or MD130 which is pretty much the same driver) to build a Quested VS2108/VH2108 kind of speaker (not exactly same as the Quested woofer is a custom model). However, you could build better monitors using newer technology like the Seas Excel drivers.
~Mikko

 

[tardishead]

anyone out there have any more progress with this thread?
anyone build any kits? what about scan speak reference monitors?

 

[tardishead]

bump

 

[Svart]

bump

 

[bahrens]

> can the suggested woofers simply be replaced by others

They're your speakers. Do what you want.

I learned by sorting through surplus drivers, mixing and matching. Got some idea what works and does not work. Which is why I'm sour on the idea of DIY reference speakers. You can knock-out a fine home speaker by guesswork and a little trial/error. Getting a reference-quality speaker is a BIG project. The guys who do it well do it as their main job/hobby.

Later a friend asked for a $100/pair design. I carefully worked it out and gave him Radio Shack part numbers. He actually built them. They sounded awful enough to notice: big dip in the voice range. I pulled off the grilles.... "What are these???" I knew what I had picked and knew this was something else. Same size, similar price, not the same drivers. "Oh, the salesman said these were better." "OK, ask the salesman to design and tune your speakers." In a huff, I bought the exact part-numbers I had picked on paper, and built a pair. It was sure a $100/pair speaker but a lot of fun to listen to, not a hole in the middle kinda sound.

There are a dozen $10-$40 "full" range speakers that make fine monitors; if you aren't mixing loud, if you aren't mixing deep bass, if you learn where to sit for the top octave, if you have other speakers (including multi-way speakers) to cross-check on. Heck, I've polished choir on a $30 pair of JBL PC speakers (with the lovely Fostex 3-inch), but I sure would not mix danceboom on them.

PRR,
What DO you consider to be a good reference monitor?