AdamZuf said:
http://www.zaphaudio.com/ZD5.html
note the
Unfortunately I don't think it is. the TL builder seemed to more interested
in his own pet theories on line length than what actually measures best.
Some useful related info here though :
http://www.humblehomemadehifi.com/Optimo.html
FWIW I think short TL's are a waste of time, they have no low bass.
/sreten.Yeah I've seen it. I want to try achieving something better.sreten said:
Hey, it works. I heard it and was very impressed, more then any other monitor I heard. I want that type of bass in my monitor. So it's a question of how it works, now if it works.
I never heard sealed monitors but I assume that it's closer to TL then vented.
I read that the dynamics are not right with vented designs, and it now makes sense with the sound I know from these boxes.
Well it's not that short, if used in PMC's way, as noted by another memeber in here. I don't aim for anything lower then flat down to 40Hz. PMC did it with the TB2+, and I have no problem with this size of a box.
I don't know the math but the way they do it is the most efficient, and my ears approved it. Is the cross section of a PMC TL to small, to your opinion?
Thanks!
I didn't know this guy, he looks like a real pro, very organized and esthetic work, I wish I could hear his stuff! (well I can build it but I don't have the time for that at the moment).
Perhaps doing a TL like I'm talking about will be a lot of trial and error to work correctly. OTOH, it seems that the conclusions from Tony Gee's work is that cross section is best, when the woofer is at the begining of the TL. It is also the most predictable.
Thanks
Adam
I think Al was referring to popular studio monitor being a reference in more than just name. For example some people mix so much on Genelecs, ATC's etc. that they instantly feel comfortable with the sound and know how it will translate to an average speaker. This experience is brought about through a familiarity with an accurate studio reference speaker but at the same time knowing that no speaker is truely accurate and, through the aforementioned familiarity, compensating for it in a correct way.
A large range of things, from the acoustics of the room, to design specifics such crossover points, phase tracking, cone material, enclosure material and construction quality, placement, driver configuration, baffle size, bass loading, polar response and so on. Describing a loudspeakers character is hugely more involved than just frequency response.
Depends on the mixing engineer. If they're good then the answer is no. If the recording sounds accurate on accurate speakers then the job is done, when its sounds bad in other situation then its obviously due to an inaccurate system.
What about something like this.
http://www.troelsgravesen.dk/FE126E.htm
With a TL or Fonken style loading it would probably provide the quality of bass you are looking for - though it might not produce the quantity.
Sean
http://www.troelsgravesen.dk/FE126E.htm
With a TL or Fonken style loading it would probably provide the quality of bass you are looking for - though it might not produce the quantity.
Sean
ShinOBIWAN said:
The point is that I have trouble to describe what's accurate. There are very good systems that sound very good and different very good systems that sound very good, some translates better.
However, I have this immediate sense of how it will translate. For example, I head the ADAM ANF-10 when it was very annonimus in Israel, without anyone trying to sell them to me or to convince mw that they are good. Immediately I knew: This monitor is going to translate very very well. That was before I became a mixing engineer. On the one hand, I feel like buing this monitor, on the other hand, although the risk, the likely possibility to go a league up in monitoring prescision is very charming...
I was also toying with athe idea of a high efficiency driver. The CSD looks good for integrating but I expect that the Fostex will get much worse down there?seanzozo said:
Frequency response is bad, as the XO point. It's very logical to persue 2kHz XO point. I'll try to go Raal instead of Fountek.
The design needs a certain quantity and certain resemblance to other consumer audio stereo, as noted before in the thread, and I accept. The fact is that I'm chossing the Scan Speaks woofers over the Excels, even thugh the Excels have better figures in many areas, I know SS are used for killer monitors (ProAc, Dickason's and others).
So this project will very likely to use one of the top of the line Scan Speaks and Raal Ribbon. It turns out to be expensive but hey, you get what you pay for! (By the way, my school might be willing to pay the designer for his work if he's a qualified, very experienced person, and I might have found a very good designer...)
Adam
AdamZuf said:
Greets!
Today, 'reference' is a variable for the type of music to be mastered, though originally it meant that the ~100 - 6 kHz BW be nominally flat with a minor rising on axis response and the two ends rolled off at ~12 dB octave to keep them from unduly 'coloring' the singer's output and IMO should still be the standard today for all music. Once this BW is properly mixed, then the rest of the BW can be EQ'd up to get the desired fill since they fall outside our acute hearing BW.
Folks marketing speakers are interested in making a living, so obviously they are each going for a different sound in the hope that there's will become a 'reference'.
No, it doesn't make sense that reference monitors should somehow resemble the poor specs of much of consumer audio because then you're compounding how bad they are. What you want is a neutral monitor so that all it does is reproduce the signal as accurately as practical, so that each manufacturer's offerings sound as they intended, for better or worse. For this you ideally need a flat amplitude and phase response over the entire audible spectrum, but no speaker can currently do the latter, so at least the ~300 - 3 kHz phone BW should be IMO, ergo no XOs in the ~212 - 4.24 kHz BW if 4th order and even further away for lower orders.
Even this is considered too large a monitor for other than mono, so the major manufacturers kept trying to find how much they could compromise to save space and finally settled on 12 - 15" two ways with ~500 - 1.6 kHz second order XOs and good directivity control above ~2 kHz, increasing them to relatively compact three or four ways with the switch to the wider BW CD format. With the advent of high power SS amps, downsizing consumer speakers became a priority, so the studio monitors followed suit for the most part to stay in step with using similar cone/dome drivers.
As PCs/powerful software became available, downsizing began anew, so the long out of favor ~fullrange driver came back in vogue for awhile since they come closest to meeting the phone BW requirements without multiple drivers and/or compression horn loading. With ever increasing processing power though, otherwise abysmally performing speakers can be digitally corrected to something folks find pleasing, so now waveguide loaded dome tweeters with small, wide BW woofers in tiny cabs and mass quantities of digital processing is the order of the day.
Anyway, so much for answering your Qs in short.
GM
AdamZuf said:
Greets!
You're welcome!
If by 'straight' you mean a constant taper (zero expansion), then yes.
Because the vast majority of consumer speakers still use passives, which affect the speaker's 'character' and you can bi-wire them so that active can be inserted for tuning 'on the fly'.
You won't know how audible or important phasing is by just reading opinions.
?? This makes no sense to me.
You should mix for the highest average SPL most folks listen at and let them use loudness controls for low SPL and tone controls for high. Like I said previously, you want your mix to be neutral so that it will sound ~the way each manufacturer envisioned on its equipment, which is what I assume you mean by 'translate well'. Obviously, your monitors need to be able to play at high average SPLs relative to your listening distance to make sure the mix doesn't turn to $%@# when 'cranked' up and the mix should sound ~balanced when either analog or digital loudness EQ is used .
GM
AdamZuf said:
Greets!
Well, I don't know that they 'got it right', only that you think so.
'Rounding' the corners in a TL only improves its HF response typically above its passband, which we don't want, so squaring them off is actually beneficial to its overall performance.
Getting 'flat' to 40 Hz (half space) with a TL requires a high Q driver, so if you're sure the PMC does, then either they aren't as 'fast' as you think they are or the ones you auditioned benefited from considerable room gain and/or EQ. Anyway, for the SS 15W-853K01 in a 'straight' TL:
Leff = ~92.66"
area = ~25.41"^2
Fold it up however you like as long as you ~preserve the area through the bends and don't exceed ~a 9:1 aspect ratio and stuff it to suit (usually much trial and error as PMC noted). Half space sim with 1 lb/ft^3 of polyfil:
GM
Attachments
Hi GM,
if I may add what's missing in your posting - it should read:
"You should mix for the highest average SPL *their lousy equipment allows folks to* listen at."
"Plastic is phantastic" stereos, €89.99 5.1 sets...oh boy! The only way to get a minuscule amount of SPL out of this c**p is dynamic compression by means of a steamroller - and exactly that is being done to current pop recordings.
Pit
if I may add what's missing in your posting
- it should read:"You should mix for the highest average SPL *their lousy equipment allows folks to* listen at."
"Plastic is phantastic" stereos, €89.99 5.1 sets...oh boy! The only way to get a minuscule amount of SPL out of this c**p is dynamic compression by means of a steamroller - and exactly that is being done to current pop recordings.
Pit
Greets!
You can use a reverse taper (TQWT) to shorten it, but the bottom line is that it takes 'X' amount of Vb for a given gain BW, so the box size doesn't change, only its internal baffle layout. Here it is shortened to the PMC's 1.5 m, same damping, terminus area and 2x it at the closed end for ~the same net Vb, which shows marginally more gain, 3rd harmonic dip shifted higher with a slightly higher Q LF roll off:
GM
You can use a reverse taper (TQWT) to shorten it, but the bottom line is that it takes 'X' amount of Vb for a given gain BW, so the box size doesn't change, only its internal baffle layout. Here it is shortened to the PMC's 1.5 m, same damping, terminus area and 2x it at the closed end for ~the same net Vb, which shows marginally more gain, 3rd harmonic dip shifted higher with a slightly higher Q LF roll off:
GM
Attachments
So it is possible...
Would you recommend trying it? Do you think that it will require a lot of experimentation to tune right?
Is it possible to plan everything in the speaker beside the box's volume for a TL, in a case someone plans it for me, but the TL needs tweaking of the depth and height? (or to ask, can one make a good design out of given baffle with crossed drivers?)
I was already in the thought of a sealed MTM but it will cost much more then a bit more wood and also be bigger... How much depends on the extension I choose..
Going from an MTM back to a TL will allow using a waveguide (for a dome), as used by Zaph, which is probably too big for an MTM.
By the way, PMC didn't position the woofer ay the begining of the TL. If I will do that, will the tweeter being in the way of the channel cause a problem? (it's a narrow path anyway)
Adam
Would you recommend trying it? Do you think that it will require a lot of experimentation to tune right?
Is it possible to plan everything in the speaker beside the box's volume for a TL, in a case someone plans it for me, but the TL needs tweaking of the depth and height? (or to ask, can one make a good design out of given baffle with crossed drivers?)
I was already in the thought of a sealed MTM but it will cost much more then a bit more wood and also be bigger... How much depends on the extension I choose..
Going from an MTM back to a TL will allow using a waveguide (for a dome), as used by Zaph, which is probably too big for an MTM.
By the way, PMC didn't position the woofer ay the begining of the TL. If I will do that, will the tweeter being in the way of the channel cause a problem? (it's a narrow path anyway)
Adam
Greets!
I'm probably the last person to ask whether to build them or not since I have no interest in the 'small' size/sound format for critical listening, ergo not very knowledgeable on the criteria for that style of mixing beyond the high compression required. Tuning a TL can be as easy or time consuming as your ears/personal preferences dictate. Most folks get them 'good enough' pretty quick, while apparently PMC spent considerable time/effort getting theirs right and IIRC Ken Kantor mentioned tweaking one of his consumer monitor designs for ~a year before he was satisfied, so as always YMMV.
Not sure I follow, but an adjustable folded pipe that seals well enough for accurate testing would probably be an expensive one. You can make a MTM TL or TQWT, just double the closed and open end areas assuming both woofers are the same........
Neither did I in the sims, I assumed a typical size tweeter was above it. If MTM, put one woofer at the extreme top front and I doubt the sim or actual performance would be significantly different from the sims except for being more efficient.
GM
I'm probably the last person to ask whether to build them or not since I have no interest in the 'small' size/sound format for critical listening, ergo not very knowledgeable on the criteria for that style of mixing beyond the high compression required. Tuning a TL can be as easy or time consuming as your ears/personal preferences dictate. Most folks get them 'good enough' pretty quick, while apparently PMC spent considerable time/effort getting theirs right and IIRC Ken Kantor mentioned tweaking one of his consumer monitor designs for ~a year before he was satisfied, so as always YMMV.
Not sure I follow, but an adjustable folded pipe that seals well enough for accurate testing would probably be an expensive one. You can make a MTM TL or TQWT, just double the closed and open end areas assuming both woofers are the same........
Neither did I in the sims, I assumed a typical size tweeter was above it. If MTM, put one woofer at the extreme top front and I doubt the sim or actual performance would be significantly different from the sims except for being more efficient.
GM
I got my idea of what makes a monitor translate well:
1)How well the engineer knows it.
2)How it reproduces the midrange.
I think everybody will agree on the 1st point, so I will proceed to the second: you know the Fletcher–Munson curves and what they mean (for who does not know, they tell that we hear the midrange louder and more accurately than bass and treble).
So most of the "data" that our earing system passes to our brain is midrange information.
That's why (IMHO) NS10s translate well: they got no bass, inaccurate and somewhat lacking trebles, leaving only half-decent midrange.
That makes an engineer make two things: create a balance that works well in the midrange because It's all he's hearing well, and boost lows and highs to hear them at least a little, creating a FFT plot that resembles an inverted Fletcher–Munson curve (less extreme, of course: just a few dB of mid attenuation).
Except Lowther lovers ( ), most humans seem to like that curve, hence the success of those crappy speakers since the late 70's; that, and some guy mixing The Dark Side of the Moon and then saying he used NS10s as monitors.
In conclusion, if I'd like to suggest:
-go to this url http://sound.westhost.com/articles.htm and read about biamping, that guy's got very good points (in favor).
-to biamp, you need an active XO and I heard that the Behringer DCX2496 is very good for it's price. You said that it's not phase linear and it's true but remember that there are not analog phase linear filters, only phase coherent (not quite the same). Linear phase filtering is possible in the digital domain only (and pretty expensive).
-for the amp section, do not rule out class D and class T designs, their quality is now on the level of Class AB pushpull and very cose to Class A design but with greater efficency and much (MUCH!) lower prices.
But you can start a topic in the appropriate sections of this forums. I can give some URLs of online shops if you ask me.
-If you do not need very high SPL levels, favor drivers with a light cone and good spectral decay and distortion figures to flat responses and high excursions.
The 5 inches Scan-Speak looks like an excellent choice.
Sorry for the long one and good luck!
1)How well the engineer knows it.
2)How it reproduces the midrange.
I think everybody will agree on the 1st point, so I will proceed to the second: you know the Fletcher–Munson curves and what they mean (for who does not know, they tell that we hear the midrange louder and more accurately than bass and treble).
So most of the "data" that our earing system passes to our brain is midrange information.
That's why (IMHO) NS10s translate well: they got no bass, inaccurate and somewhat lacking trebles, leaving only half-decent midrange.
That makes an engineer make two things: create a balance that works well in the midrange because It's all he's hearing well, and boost lows and highs to hear them at least a little, creating a FFT plot that resembles an inverted Fletcher–Munson curve (less extreme, of course: just a few dB of mid attenuation).
Except Lowther lovers (
), most humans seem to like that curve, hence the success of those crappy speakers since the late 70's; that, and some guy mixing The Dark Side of the Moon and then saying he used NS10s as monitors. In conclusion, if I'd like to suggest:
-go to this url http://sound.westhost.com/articles.htm and read about biamping, that guy's got very good points (in favor).
-to biamp, you need an active XO and I heard that the Behringer DCX2496 is very good for it's price. You said that it's not phase linear and it's true but remember that there are not analog phase linear filters, only phase coherent (not quite the same). Linear phase filtering is possible in the digital domain only (and pretty expensive).
-for the amp section, do not rule out class D and class T designs, their quality is now on the level of Class AB pushpull and very cose to Class A design but with greater efficency and much (MUCH!) lower prices.
But you can start a topic in the appropriate sections of this forums. I can give some URLs of online shops if you ask me.
-If you do not need very high SPL levels, favor drivers with a light cone and good spectral decay and distortion figures to flat responses and high excursions.
The 5 inches Scan-Speak looks like an excellent choice.
Sorry for the long one and good luck!
Phobos said:
Greets!
Well, if that's the case then building a max intelligibility speaker makes sense to me:
Maximum speech intelligibility requires that the response above 2kHz rolls off 3dB for each octave, i.e. -3dB at 4kHz; -6dB at 8kHz; -9dB at 16kHz. The response from 125Hz to 2kHz is flat with a window of ±2dB. Frequencies below 125Hz are rolled off at 6dB/octave to minimize rumble/boom. This curve is based on human hearing and perception research.
GM
Yeah but such a roll off is really extreme and will hide what happens up high and down low, wich is clearly dangerous. All that talk of the NS10s was just an example, meaning that I hate them but i use them as a tool: if the whole mix is decent there, it will probably sound decent at home, in the car, in the PC, etc..
But of course it has to sound BETTER on a good monitor or hifi system. If something sounds good on the Yamahas, it usually is some ubercompressed track that will make you sick in 2 mins on a good monitor.
That's why engineers alternate between monitors and other systems such as the NS10s.
Just my 2 cents.
But of course it has to sound BETTER on a good monitor or hifi system. If something sounds good on the Yamahas, it usually is some ubercompressed track that will make you sick in 2 mins on a good monitor.
That's why engineers alternate between monitors and other systems such as the NS10s.
Just my 2 cents.
It is an interesting idea to low cut the speaker and work on part of the mix like that. With active crossover using a DCX2496 it will be a breeze.
ok. MTM of two Revelators. Nearfiled listening. Vertical lobing diminishing is important to some extent, sometimes customer sits in different hight, and has to be impressed.
Would you space the woofers according to D'Appolito's formula of 1 wavelength of XO freq? Or would you just make them as near as possible? I still don't know the XO frequency, I'll have to experiment with the DCX2496.
What about driver offset? How does that affect horizontal response? (I only know it affects lobbing). I was thinking big champfers, I attached a very quick general sletch. How would that affect baffle size calculations, and the descistion wether to offset drivers?
Thanks
Adam
ok. MTM of two Revelators. Nearfiled listening. Vertical lobing diminishing is important to some extent, sometimes customer sits in different hight, and has to be impressed.
Would you space the woofers according to D'Appolito's formula of 1 wavelength of XO freq? Or would you just make them as near as possible? I still don't know the XO frequency, I'll have to experiment with the DCX2496.
What about driver offset? How does that affect horizontal response? (I only know it affects lobbing). I was thinking big champfers, I attached a very quick general sletch. How would that affect baffle size calculations, and the descistion wether to offset drivers?
Thanks
Adam
Attachments
Tweeters:
Seas 27TBFCG - good distortion figures, great CSD, good off axis response.
Peerless 810921 - great distortion figures (especially on the lows), good CSD, a bit lesser off axis response.
Zaph's tweeter comparison Mark K's tweeter comparison
I aim for low XO point, in the idea of good step response and staying away from the ear's most sensitive band. I can see Linkwitz crossed his orion at 1440Hz with the Millenium, which is a higher distortion tweeter then any of the above (at the lows).
I'm curious about the Peerless - if it would result a similar XO point to the Seas (although it can reach lower), then I'm staying with better distortion figures, but a bit less good off axis res. and CSD. But in case I will use its ability to cross lower then the Seas, then of coarse it's a winner.
Please help, or maybe suggest another.
Thanks
Adam
Seas 27TBFCG - good distortion figures, great CSD, good off axis response.
Peerless 810921 - great distortion figures (especially on the lows), good CSD, a bit lesser off axis response.
Zaph's tweeter comparison Mark K's tweeter comparison
I aim for low XO point, in the idea of good step response and staying away from the ear's most sensitive band. I can see Linkwitz crossed his orion at 1440Hz with the Millenium, which is a higher distortion tweeter then any of the above (at the lows).
I'm curious about the Peerless - if it would result a similar XO point to the Seas (although it can reach lower), then I'm staying with better distortion figures, but a bit less good off axis res. and CSD. But in case I will use its ability to cross lower then the Seas, then of coarse it's a winner.
Please help, or maybe suggest another.
Thanks
Adam
I wouldn't cross a tweeter that low in a studio monitor, unless it was one of the few that the manufacturer states that it can be done i.e. Accutons can be crossed with 1st order crossovers from 1.7/1.8 kHz: maybe with a 4th order they could be crossed lower, since Fs is 350 Hz (thx to a tuned rear chamber?). Some ribbon manufacturers state that they can be crossed lower than 2000 but it's true for hi-fi use only.
A tweeter that has to be used in a studio monitor will be subject to waaaay more abuse than a hifi one.
Just think of a check to find the right settings for the mic pres... sometimes the signal is too low, sometimes is so high that the red light on the monitors tells you that the internal limiter just saved your tweeter. And it doesn't matter how careful you are, sooner or later you will make an error and if you don't have limiters or strong tweeters your monitors will be ****ed up for good.
That's why ribbons are used in three way monitors as super tweeters (Genelec 1022 cross at 4 kHz).
Adams uses folded ribbons for his monitors, which are more resistant due to the increased surface (=less motion required for same SPL) and still they cross their tweeter at 1800 Hz.
In the end my point is that you need a lot more headroom in a studio monitor than in a hifi cause if you put in a cd you know what to expect:1,225 volts for a balanced output, if i remeber correctly.
In the studio since you need some headroom to avoid clipping/saturationg the input of your ADC/tape, and due to the fact that you will be listening to untreated material (at least at the start of a project) your volume pot will be pretty high to get a decent (average) SPL, but that could expose your loudspeakers to higher peaks that may rip a ribbon (except a few models maybe, like the Fountek's flagship model).
A tweeter that has to be used in a studio monitor will be subject to waaaay more abuse than a hifi one.
Just think of a check to find the right settings for the mic pres... sometimes the signal is too low, sometimes is so high that the red light on the monitors tells you that the internal limiter just saved your tweeter. And it doesn't matter how careful you are, sooner or later you will make an error and if you don't have limiters or strong tweeters your monitors will be ****ed up for good.
That's why ribbons are used in three way monitors as super tweeters (Genelec 1022 cross at 4 kHz).
Adams uses folded ribbons for his monitors, which are more resistant due to the increased surface (=less motion required for same SPL) and still they cross their tweeter at 1800 Hz.
In the end my point is that you need a lot more headroom in a studio monitor than in a hifi cause if you put in a cd you know what to expect:1,225 volts for a balanced output, if i remeber correctly.
In the studio since you need some headroom to avoid clipping/saturationg the input of your ADC/tape, and due to the fact that you will be listening to untreated material (at least at the start of a project) your volume pot will be pretty high to get a decent (average) SPL, but that could expose your loudspeakers to higher peaks that may rip a ribbon (except a few models maybe, like the Fountek's flagship model).
I see your point with the power thing of monitors vs. hi-fi.
It is most likely that I cross 4th order. I'll also put a cap on the tweeter against DC.
I don't know the distortion figures of the Accuton, but I read that the 810921 is high Xmax, and got great distortion figures.
That compares to the 27TBFC/G specs - it's recommended XO frequency starts at 1.5kHz, and the 810921 is lower distortion.That compares to the 27TBFC/G specs - it's recommended XO frequency starts at 1.5kHz, and the 810921 is lower distortion.
I use Linkwitz usage of the Millenium (lesser distortion specs) as a reference because he cares so much about the issue I'm quite sure he is far away from streching the tweeter, even at high levels. The Millenium is 88dB/spl avg. vs. 810921 which is 93dB/spl, with roughly equal power rating. On the other hand, the Millenium reommended range starts at 2kHz (higher then the others)... ok I'm starting to get lost, I need more educated view of this. Of coarse I really respect yours but I want to hear what others say as well.
I don't say I'll cross as low as 1.5k but it is useful to have the option, I never tuned a crossover so I don't know what to expect.
How loud are the peaks recieved at the 1.5kHz point that the driver should handle? (studio usage)
(PS there's a thread discussing the very same thing here )
It is most likely that I cross 4th order. I'll also put a cap on the tweeter against DC.
I don't know the distortion figures of the Accuton, but I read that the 810921 is high Xmax, and got great distortion figures.
That compares to the 27TBFC/G specs - it's recommended XO frequency starts at 1.5kHz, and the 810921 is lower distortion.That compares to the 27TBFC/G specs - it's recommended XO frequency starts at 1.5kHz, and the 810921 is lower distortion.
I use Linkwitz usage of the Millenium (lesser distortion specs) as a reference because he cares so much about the issue I'm quite sure he is far away from streching the tweeter, even at high levels. The Millenium is 88dB/spl avg. vs. 810921 which is 93dB/spl, with roughly equal power rating. On the other hand, the Millenium reommended range starts at 2kHz (higher then the others)... ok I'm starting to get lost, I need more educated view of this. Of coarse I really respect yours but I want to hear what others say as well.
I don't say I'll cross as low as 1.5k but it is useful to have the option, I never tuned a crossover so I don't know what to expect.
How loud are the peaks recieved at the 1.5kHz point that the driver should handle? (studio usage)
(PS there's a thread discussing the very same thing here )
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.