Studio Monitors (nearfield) - around the NeoPro5i

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm no speaker builder, but I've been interested in the field for some time. I've built from other people's designs.
I'm a new sound engineer and I wish to have new nearfield monitors.
I've heard the lower ADAM (ANF-10, A7), Dynaudio BM5, PMC (DB1+,TB2+), which are the more regarded nearfiled monitors in my price range.

I felt that the ADAMs had the best transient reponse. They have a ribbon tweeter (some custom Eton, as far as I know, not available to us) and are crossed at about 1,8k. nice. A bit colored but they were the most alive and revealed the shape and texture of the soundstage best. I believe that this kind of what we call "revealing" is at least important as the details themselves - this I believe what maxes a mix that sound correct on this kind of speaker translate well on others.

The Dynaudios were more analytical, natural, yet not as dynamic and brutal in the way the translate the music I know. Very nice to listen to, but didn't have depth in the midrange.

The PMC sounds the best as a fun speaker, but they didn't differ different recordings very well, the soundstage tends to be made of their sonic signature everytime, to a certain degree. However, I was impressed by the bass of these things, as it was clearly tighter then the ADAMs, and I demonstrated these when set far away from the walls. They are line transmission design, and the port is on the back. I wonder how they will function when they are close to a rear wall.
Look at the design:
http://www.pmcloudspeaker.com/transmission.html

I can afford the lowest ADAM monitors, which use that ribbon tweet. They are very dynamic, and I think that the lower you go with the XO point, the better it is, if you don't reach driver stress or breakup. A studio monitor should be able to take high levels, but however, in my case, this is for the time when you got a pop, or a wrong gain setting- in other words, just for brief moments. I try not to monitor at high levels as this is not profesional.

So, we have a low XO point. I can't remember how different XOs designs affect off axis response, but I assume that we should aspire to even off axis response as much as possible. I still didn't get how wide should the response of a studio monitor be - of coarse you want to be able to move around the desk, however, too wide probably meens too much reflections from walls and ceiling. I also remember that MTMs have narrow horizontal dispersion - I don't know if that's good, because this only applies about the highs. If the ceiling is acousticaly treated, it meens that the first thing to die are the highs - so the reflected sound is very non linear. Probably better to go MT - in that regard at least. If the tweet is crossed low, maybe an MTM is more of an option because less the chance for comb filtering. Just an idea.

Anyway, wheather it's an MT or MTM, the woofer has to complement the tweeter with a very fast response. Smaller woofers are faster sounding (how is this called, damping factor or something?), would probably wise to define the size of the driver according to the room size, and requirements for the type of music mixed. Actually I'm fine with 40-45Hz, which is quite the low edge of most systems. I care more about dynamics and accuracy, as the woofers must be up to the tweeter, otherwise it was all for nothing, right? (-:

The only reasonably costing, commercial option I can see is the Fountek NeoPro5i. XO point recomended is 1500Hz, lower then the ADAM. Maybe it can go lower.
It's also very efficient at a 102dB/SPL average.
It is not a shielded tweeter and they actually warn about the high flux of it, but it might be worth it, and by the way, I use 2 screens but one obove the other, so the monitors will not be that close. It is a disadvantage though..

There's also the Mark & Daniel Maximus Monitor, which I haven't heard, and is highly regarded. Look:
http://www.mark-daniel.com/english_m/Proucts/P1_2.htm
It uses a big custom made ribbon which is crossed at 800Hz! wow, covering that 1K range with such a fast driver! Well, dream on, dream on...

Woofers - seems to me that good options for woofers that has been used in sucessful studio monitors are Scan-Speak (ProAc Studio 100, which I intend to listen to soon), Dynaudio and Vifa (used by Genelecs, some at least). But I heard that Fostex are revealing drivers, and are very efficient, which is very desirable, and opens up more options for amplifiers. I also seen them used in transmission line designs, IIRC..

I was impressed by the NTC transmission line and wondered how they made it sound right with such a small box like the DB1+ (it's a 5" woofer!). However, if such a designed would have been used, probably would be better to have the port on the front for a studio monitor application, for less reflecting bass from the wall, less dependancy. I base my last sentance on logic and not science, and actually many thing I've said in here, please correct me with my mistakes.

By the way, I think that biamping would be a good idea. maybe with a passive line level crossover. I just have to get the phase correct between my two amps..

MY GOAL? I can't design such a thing myself, honestly. I was hoping that since the DIY community is realy lacking in a realy good studio monitor, it would actually interest someone to design and have something that is available for many times the price, and maybe even make it better. Again, I don't know science, but having a low crossed ribbon (with its ultra low mass) and a small, fast driver with a very low mass and a narrow operating band which is complimented by transmission line to get the lows seems like the best of all worlds.

Adam
 
As a sub-subject of its own, how did PMC make such a small transmission line? Doesn't the length of it suppose to correlate with the wavelength of the frequency range missing?


In both the TB2+ and of the DB1+ the length of that path clearly does not relate to what's missing from the driver... So how does this work?
 
As I said, the woofer probably should be very fast, up to the tweeter.
I don't know much, but looking at Zaph's measurement of the W18EX001 it looks very good, considering the low crossover point for the NeoPro5i:
An externally hosted image should be here but it was not working when we last tested it.

Zaph also states that this woofer would enjoy a very low XO point - 1600Hz, which seems quite up to the tweeter!
He doesn't like ribbons, but it seems that my experience and so-called design goals might be worth a shot with the combination.
He recommends LR 4th order.
The recommended XO type for the NeoPro5i is 2nd order. Why is that? What would happen if crossed at 4th order?
(reminder: I'm thinking of active XO)

However, the W18EX001 is not a high sensitivity driver and that's too bad. Can high sensitivity drivers deliver equal or better timing? Zaph tested some Fostex drivers, but found the non linear distortion figures not optimal. However, considering that the speakers will not be used at high level, maybe meens improvement in that area?
 
neopro5i_waterfall.jpg

Here it's clear why the recommended XO frequency is 1500Hz.

Thi measurement is different then Zaphs, can anybody evaluate the timing difference between the 2 drivers in say, 1500 or 1600Hz?
 
What's the formula for TL by which you can tell? Will it be the same if the port is on the front?

Is the Seas driver suitable for TL loading? (it's supposed to have low enough Qts, IIRC?)

I read different things about TLs: that they have peaks that add coloration, on the other hand offer extension of ported while sounding tight like sealed enclosures. Is there a method to control it, other then acoustic materials and a rigid box?
 

GM

Member
Joined 2003
Greets!

Yes. This doesn't include any near boundry gain that will lower it further:
Fp = (13560"/4)/(path-length+(terminus radius*0.613))

When properly done, the classic TL mimics an IB in a much smaller, though still large cab, ergo whatever the driver's IB response is, the TL will be effectively identical down to Fs, where it will roll off at a slightly faster rate than the IB. So we can say that whatever the driver's Qts is, so will the IB's Qtc and the TL's Qp be for all intent and purpose. TLs can also be tuned for different applications that sound/perform considerably different than the classic alignment, hence the conflicting reviews. Driver position along the line and/or taper ratio has a major effect on the pipe's smoothness also.

Contrary to what some folks seem to believe, any driver can be TL loaded with good results, it's just a matter of choosing the right driver for the app. From what you've posted, it seems to me you need to just buy what you like since without complete tech data there's no way for me or anyone else to 'clone' it. Anyway, having been 'weaned' on large, dynamic Altec, etc., horn loaded studio monitors I'm the wrong person to ask about using a dinky little bookshelf speaker for this app.

That said, if I were going to make a relatively small two way TL nearfield monitor with a ribbon, Jim Griffin's Jordan/Aurum Cantus would be a good starting point.

GM
 
GM,

THANKS for saying it loud and clear - the idea "you need very specialized drivers for a TL" is as stupid as "any driver will perfectly drive any horn". :wave2:

That said - I can't log into the FRD forum at the moment, but I'd very much like to get our misunderstanding out of our lives.

Pit
 
GM
In that formula, there's no importance to the radius of the path itself, and if I relate these correctly, indeed in the PMC the "terminus radius" is as wide as the path itself.
An externally hosted image should be here but it was not working when we last tested it.


However, You can see that the drivers takes considerable volume from the TL path, and the cross section width is higher to allow space. This corelates to the formula, in which it doesn't matter.
Am I right about this?

If so, why are they the first ones to imply a path with a cross section as wide as the port (for the most part of it) in a design? What's the advantage of other TL shapes?

From what you've posted, it seems to me you need to just buy what you like since without complete tech data there's no way for me or anyone else to 'clone' it.

I don't wish to clone it since I'll use a different driver. I just want to understand the idea behind it, so there will be no need to buy a speaker I can hardly afford, and even if so, I'll have no money to make a new one.

Thanks
Adam
 
Ex-Moderator
Joined 2002
Reference speakers are such for a reason. As you noted, they may well sound different from each other, but those differences are a known quantity. You may well be able to equal, or even better the quality of commercial monitors for the same money, if you spend a lot of time and research, but the resulting speakers would not be a reference, because they will sound different to widely available commercial monitors.

If you are the only one ever going to use your speakers, and you will not be using other studios, then by all means build your own, but you might find difficulty if other engineers want to use your studio, (paid or otherwise), as they will not be a known quantity, or even if you want to use other facilities for remixing or whatever.
 
pinkmouse said:
Reference speakers are such for a reason. As you noted, they may well sound different from each other, but those differences are a known quantity. You may well be able to equal, or even better the quality of commercial monitors for the same money, if you spend a lot of time and research, but the resulting speakers would not be a reference, because they will sound different to widely available commercial monitors.
All speakers sound different. The difference in professional "reference" monitors maybe is narrower then hi-fi, because the use is very specific, however they are very different. I assume this design won't make an exception. We are still talking about concepts used in professional studio monitors.
If a company sells their monitors which sounds different (much better!) and translates well, for $100,000, only then it's supposed to be called reference?

You can mix on about anything if you know it. But the time spent to compensate for lies of the speakers is, IMO, better spent on aquiring knowledge and understanding what works best for me. Maybe this design will be a keeper for years to come, who knows. Maybe it will save me 5 pairs of monitors until I reach a $xx,xxx monitor. Maybe not. Worst case scenrio - I lost a bit of money and time. Best case scenario - I have much better mixes, much faster then these years until I get a monitor I really "communicate" with.

As I said, I will not design this monitor myself. Actually, I might have found a very profesional person to help me with that. I'm here to gather more opinions on the subject. Be sure that if the project is excuted, it will be a successful one.

Personaly, It's hard for me to accept that I hear flaws in every form of design I hear in my price range, and I'll always feel that I missed something if I won't do a thing about it.


If you are the only one ever going to use your speakers, and you will not be using other studios, then by all means build your own, but you might find difficulty if other engineers want to use your studio, (paid or otherwise), as they will not be a known quantity, or even if you want to use other facilities for remixing or whatever.

You have a point. However, I don't have professional studio - I'm still a student. So for now, it's only for me. I rely on my mixing abilities to bring clients for mixing purposes only for now. I want my mixes to translate well to other systems. What I'm doing here is a risk - However, if I can nail what makes a speaker "transparent" by my definition of it, by my ears - which I mix with, then maybe something great will come out of this project. You can see that i already got a gut feeling which sums to 1 major point:
Best transient response in a given frequency, with minimum number of drivers to achive that. In my aim to cause a uniform dynamic response throughout the spectrum (as much as possible), the bass department asks for more control then offered by reflex boxes, but the extension that closed boxes can't offer. The PMC design proved the solution for me.

My concern is that the speaker will be so precise (I'm aiming towards very fast elements, as you see) that although it will make the best recordings sound good and the worse recordings sound really bad, it still won't translate well - Because on the one hand, consumer audio usually has bad figures of all measurements, and it makes sense that the speakers should somehow resemble.

On the other hand, I don't know how similar these distortions figures are between consumer audio stereos. Maybe they can not be generalized as it's easy to believe, and indeed in every systems different recordings translate differenty to a certain extent. We've all been in the scenario that recording we thought was very good was mediocre on our favourite system and vise versa. But the fact is, that the best studios use speakers that are definetly "different" then most speaker, even low end (I meen, $xxx) monitors, by being 2 things, the first is being ultra presice, and the second, which is the obscure part, they translate well to other speakers, which not what all speakers we regard as very presice, do.

I think that the big questions I have for you guys, are:
A.What makes a monitor translate well more then anything? (except good frequency response, of coarse)
B.Is it a likely scenario that you can create something so precise that even though great recordings sound really good and bad sounds really bad, it still won't translate well?
 
AdamZuf said:
GM
In that formula, there's no importance to the radius of the path itself, and if I relate these correctly, indeed in the PMC the "terminus radius" is as wide as the path itself.

However, You can see that the drivers takes considerable volume from the TL path, and the cross section width is higher to allow space. This corelates to the formula, in which it doesn't matter.
Am I right about this?

If so, why are they the first ones to imply a path with a cross section as wide as the port (for the most part of it) in a design? What's the advantage of other TL shapes?

I don't wish to clone it.........

Greets!

You're welcome!

No, it's just a basic formula and for it to be accurate requires that the bends preserve the pipe's expansion rate. For all else, more complex pressure wave formulas are required.

Right, they were just explaining it's a simple TL, not MLTL (mass loaded, i.e. has a reduced vent area) or TQWT (reverse tapered) and they adjusted the pipe's expansion to ~account for the loss of Vb due to the drivers. If they hadn't, then the formula would only be good for a first approximation.

Other flare rates (shapes) can smooth out pipe harmonics (horn) or shorten the length required for a given Fp (TQWT). For an in-depth look at pipe/horn design, visit MJK's and Bob Brine's sites: http://www.quarter-wave.com/ http://www.geocities.com/rbrines1/

OK, then even after reading later responses I still think my suggestion complete with passive XO (even if you bi-amp) is a good starting point.

GM
 
A.What makes a monitor translate well more then anything? (except good frequency response, of coarse)
B.Is it a likely scenario that you can create something so precise that even though great recordings sound really good and bad sounds really bad, it still won't translate well? [/B]


Re: "translating", I think that basically you just need to make sure that your reference monitors aren't unusual in any particular way compared to regular speakers that most people have in their homes. Think: ghetto blasters, mini-systems, micro-systems, alarm clock radios, 5.1 HT systems with crappy 1-way satellites...

I'm not suggesting that you should mix recordings with the lowest common denominator in mind, just that the monitors should somehow represent the average.

If I had a mixing studio (and the money), I'd get myself a combination of cheapish 2-way bookshelf monitors (ported; paper/poly cone midwoofers; fabric dome tweeters), a pair of much better hifi loudspeakers, as well as headphones (Sennheiser or Beyerdynamic or something like that). It would be set in a "homely" environment with a sofa, coffee table etc in the background so that the acoustics are more likely to resemble actual listening environments.

Re: B, it depends on you mean by precise. A specification such as "20 to 20kHz +-3dB" could range anywhere between meaning that the speakers have a couple of smooth dips and peaks not exceeding 6db in the FR plot... Or it might mean that the tweeter is obnoxiously harsh and the coarse "sandpaper" pattern in the FR plot is indicative of severe resonances in the spectral decay plot.
 
GM said:
No, it's just a basic formula and for it to be accurate requires that the bends preserve the pipe's expansion rate.
So I understand this is only for a straight pipe. I will read about the TL design in the link you provided. Thanks.

OK, then even after reading later responses I still think my suggestion complete with passive XO (even if you bi-amp) is a good starting point.
Why is that? I prefer using the Behringer DCX2496 system, since it will get me more involved in the process, as I don't know much about electronics. With my instructor I can solve the issues by myself easier and have an option for fine tuning by ear.
The DCX2496 filtering and EQ is not phase linear. I'm still reading the different opinions about phase being audiable or not.
Anyway, in case I'll reach the conclution that I want a phase linear system, there's a software for multiple ouput ASIO driven soundcards (I have a presonus Firepod ah ha!), made by http://www.thuneau.com/ . I happened to be a Beta tester so I already have it (-:

CeramicMan said:
I'm not suggesting that you should mix recordings with the lowest common denominator in mind, just that the monitors should somehow represent the average.
This what appeals to me with the ADAM monitors, at least the low end of ANF-10 and A7 (I also heard P11 and it's crap, funny).
In case I won't excute this project, I'll get the ANF-10, it's a great mid point between fidelity, somewhat cheap and edgy and good translating monitor. I like that sound, I'm a Grado guy :) (well at least I used to be, can't afford to keep the RS-1 with all of the gear I need...)

If I had a mixing studio (and the money), I'd get myself a combination of cheapish 2-way bookshelf monitors (ported; paper/poly cone midwoofers; fabric dome tweeters), a pair of much better hifi loudspeakers, as well as headphones (Sennheiser or Beyerdynamic or something like that). It would be set in a "homely" environment with a sofa, coffee table etc in the background so that the acoustics are more likely to resemble actual listening environments.
I already have all those stuff, it's basic stuff, give me some more credit :)
I have Selah audio Carnelian, Aiwa boombox, Sony MDR-V6 (slightly modded), laptop speakers.

Re: B, it depends on you mean by precise. A specification such as "20 to 20kHz +-3dB" could range anywhere between meaning that the speakers have a couple of smooth dips and peaks not exceeding 6db in the FR plot... Or it might mean that the tweeter is obnoxiously harsh and the coarse "sandpaper" pattern in the FR plot is indicative of severe resonances in the spectral decay plot.

By presice I don't meen frequency response. I meen all of the other stuff that makes a speaker good. It's just that the translation thing is obscure to me (you have some very different designs that translates well, some cheaper and some ultra expensive).

From my experience, frequency response is highly overrated. Almost all response graphs of finished products look very similar and can sound miles different.

For example, you might have a "flat" speaker which has a woofer which is slugish in the bass region and has a long delayed impulse response, it meens you'll have more bass and you'll have better results EQ'ing with a non high end equalizer.
Logic tells me that if you are familiar with the Fletcher Monsun curve and able to omit it from the equation as a professional studio engineer (understanding how things should sound in different levels), for a given speaker you'll have a different EQ setting for different levels of listening, because the speaker behaves differently in different levels. Of coarse the amout of this is to be debated, and is design specific, and it's not done anyway, and these differences are better left to knowing the speaker by ear and intuition.
but maybe it somehow relates to the custom of professional engineers to monitor at a certain dB level. It's a better way to really know your speakers, at least it makes sense to me. I try to do it myself but I still don't have a fixed levels at the master,control room and amplifier :)
 
AdamZuf said:

From my experience, frequency response is highly overrated. Almost all response graphs of finished products look very similar and can sound miles different.

For example, you might have a "flat" speaker which has a woofer which is slugish in the bass region and has a long delayed impulse response, it meens you'll have more bass and you'll have better results EQ'ing with a non high end equalizer.


When I shopped around for speakers for some DIYing, I steered well clear of "smoothed" frequency response plots as they are utterly worthless IMO. Who cares if the speaker has a wide hump or dip in the FR plot? Broad-spectrum imperfections generally have a very low Q and relatively benign effects (eg: the speaker sounds cold/warm/bassy/thin etc), which can be easily fixed with an EQ.

What I wanted to know was: will a speaker have problems with resonant effects such as harshness, fatiguing, sibilance, hash, boominess??? Those things can look like insignificant little ripples on an FR plot, but they have higher Q factors, which are harder to correct. So I started relying on waterfall plots and step response graphs more than anything else.
 
AdamZuf said:

Do you have any practical technique or guidlines you found true on how to use them in speaker design, besides driver selection?


I'm not sure really, back-wave resonances that press against the cone are probably the next biggie.

Intuitively: a larger box = lower back-wave SPL, but partial oscillations (as opposed to pressure loading) generally start at a lower frequency.
 
An externally hosted image should be here but it was not working when we last tested it.


How do you think PMC got it right? (I say right because I've heard it, was really good)
The partition is right behind the woofer, at least behind most of the cone.
Do you think that angeling the corners of the TL will result in less back wave resonance and better performace? (the question is did they cut costs or is their way of doing it is actually better, to avoid oscillations)

I think I'll go for one of the
5" Revelator woofers
Any recommendations for a particular one for the TL use?
How large should the box be in order to reach down to about 40 Hz flat? (PMC monitors style). I would like to have the width as small as possible, I'm more flexible with the height and depth.

I think I'll use sonnex to absorve reflections. Its NRC is the best for low frequencies per volume.

Thanks
Adam
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.