Welcome to the new Film-Tech Forums!
The forum you are looking at is entirely new software. Because there was no good way to import all of the old archived data from the last 20 years on the old software, everyone will need to register for a new account to participate.
To access the original forums from 1999-2019 which are now a "read only" status, click on the "FORUM ARCHIVE" link above.
Please remember registering with your first and last REAL name is mandatory. This forum is for professionals and fake names are not permitted. To get to the registration page click here.
Once the registration has been approved, you will be able to login via the link in the upper right corner of this page.
Also, please remember while it is highly encouraged to upload an avatar image to your profile, is not a requirement. If you choose to upload an avatar image, please remember that it IS a requirement that the image must be a clear photo of your face.
Thank you!
In very layman terms, the new method uses a new input signal (m-noise) and looks to find what the maximum SPL of a loudspeaker is before it starts misbehaving and sounding bad. (Indicated by a degraded transfer function input vs output). In contrast, the older AES-1984 method only really tells you how much power a loudspeaker can withstand for 2HRs until it is permanently destroyed. One is a destruction test and the other is a performance test.
Using a TF to compare input and output sounds like a very good idea.
I suspect we would observe a much different ‘real world’ performance for certain brands and models!
I’d be curious to see if that would justify the 3dB headroom often recommended - and also questioned, Steve! ?
thay said the M stands for Meyer Sound. Cinemas are a different beast and I wonder if using a noise with increased crest factor at HF would be the right choice for cinemas.
Is that going to be a new Meyer slogan? When you think of noise, think of Meyer Sound?
I, personally, would welcome a uniform means of specification for speakers, amps and such (that captions as much real-world information as possible). I still want the frequency response plots, and such. M-Noise would help with maximum SPL/power handling, which is but one aspect, though a vital one. One thing about cinema though...we know what our maximum SPL is on any channel...it is much more controlled than live.
I guess the chance of broad industry adaptation of something like "M-Noise" would be better if something like Dolby would've sponsored it, instead of Meyer sound. It's to be seen if the likes of Samsung/JBL and QSC are going to compare their speakers to the competition with a test procedure sponsored by the competition.
M-Noise doesn't stand for Meyer noise, it stands for "music" noise since the test signal better emulates the rising crest factor with frequency which is found in music. The signal itself and the accompanied procedure started as an internal criteria for loudspeaker specification. It was then presented and given to the industry as a whole, as a way to create a more relevant way to specify loudspeaker performance. From there, it was evaluated by a large standards body (comprised many people who work at competing manufacturers), and eventually adopted as the AES75-2022 standard. So at this point it's available for any manufacturer to start adopting. A published number on a specification sheet can simply point to AES75-2022 in the footnote instead of AES2-1985, no mention of Meyer is required.
If this is news to anyone thats because the standard is less than a month old. I would hope, with time, it becomes a new benchmark for loudspeaker specification and removes much of the "specsmanship" thats been problematic in the industry for a long time.
One nice detail of the procedure is that it's relatively straight forward to reproduce without a laboratory and is also non-destructive to the device under test. One only needs a loudspeaker, a microphone, SMAART(+understanding of TF mode), the m-noise wav file, and a neighbor that is tolerant of some SPL.
Thanks for clarifying. I agree that publishing a power figure which is not just an indication of when the driver will mechanically and permanently fail is definitely a step forward.
For me, there is nothing like a @85dBC@ on the level meter at the listening position. This is an approximation for coarse field tests. Remembering the theory, the 7.0 fader value is chosen for that. It should be 70 dB at 7.0 fader within the individual 1/3rd octave bands. There's a knee, where the hf response drops, depending on room size and characteristics. Can be from 2 kC to 5...6 kC, depending on many factors, but is easily visible. When adjusting the sound, for a full range capable speaker system, the average static response measured at the reference position approximates 85 dB C. If your frequency response is not full range, specificly if lacking low end capability, like an old 1940's system, you will end up too high, adjusting for that 85 dBC.
With LFE cabinets, something like "has to read 91 dBC" or whatever is also wrong. This depends on many factors, frequency range covered, room characteristics, standing wave point found, ... For LFE the goal is, to be 10 dB above the 1/3rd octave wideband reading of the main channels, within the usable range of the LFE cabinet. Then, we apply steep filtering below 80 Hertz, to avoid bleeding of noise into the main speaker response.
Efficiency was a key point in the 1940s, where amplification power was limited in the range of 10 to 20 watts RMS at acceptable distortion levels. The frequency range in comparison was limited. Even though analog optical was the best possible recorded format of the time, the limitation of an optical track wasn't much more than a Bell telephone conversation. For a pleasant impression in hearing, the product of lower and upper frequency to be in the range of 360.000 to 650.000, with an optimum around 400k. 70 Hz x 6000 Hz is 420000, a very pleasing result.
Today, power is no problem, but speakers still can't handle high power long term. An 18" woofer with a 4" motor coil, probably aluminum. High power heats the coil, thermal compression as a result.
Whatever manufacturers state, 1.2 kW per driver, not a single VC would handle that long term, it's just a safe point, where it doesn't burn. A realistic value is 250 to 300 W per driver, and then use the factory given sensitivity for your calculations. Remember, for EQing you need headroom. I always suggest 6dB, but other designers feel it's over the top. You cannot discuss the safety margin, it is needed, but we may discuss the figure.
So if in theory at 1500 W capability and 2800 W max power your single dual 18 in cabinet might reach the required SPL, in practice it won't. Another fact is, direct radiators do not really couple well to the surrounding air. They produce fast air movement in a small piston area, whereas the surrounding air is infinite area. This is like pushing a block of concrete with a fast pulse. The pulse will reflect. Slowly pushing on a large surface, same energy, it will move finally.
So it need acoustical transformation, horns, transmission lines, large radiating area by using multiple cabinets and so on. Matching radiator impedance to transmission impedance. A single cabinet won't.
For a room mentioned, 50 ft deep, 4 dual 18" work pretty fine, if placed closely together. With the result of bass, you can fell. And probably, you're down from 2000 W RMS amplifier power to 6 dB lower, 500 Watts. The drivers are never even forced to play in the range of thermal heating, they just have to do little excursion, resulting in low distortion and high usable dynamic range. That's probably another 2 to 2.5 grand extra, which in total is no factor in building and equipping a screen. Keep in mind, the usable time for low excited drivers is extremely long, and the failure rate extremely small. In the end the extra cost does not matter at all.
A more major problem I find is the use of surrounds. They're often placed in positions given by the internal designer, not on a position required by the sound field. As they do not really look nice, small cabinets are preferred, normally 8" bass speaker cabinets have to do an impossible job. For cinema halls larger than a preview studio, a 12" bass is a must. And if possible the mid and hf drivers to be used are identical to the front ones. Then, people use minimum amplifier numbers, make all kind of weird parallel - series connection to maintain a 2 Ohm load. Different cable lengthes, production tolerances, all ends up in an array far from good. Surrounds, bi-amping can be argued, should be connected to an individual power amp per speaker. These don't have to be powerful amps, 100 to 200 Watts per speakers is OK. The industry has a bunch of 6/8/16/...32 channels per 2/3/4 HU cabinet available, with normally 150 to 400 W RMS per channel.
If there's no at least 10" cabinets intended, 6.5" and 8" can be used, if the number is raised accordingly with close spacing. It even gets worse, trying to get a 7.1 rear field was a small number of 8 in cabinets. In that case skip 7.1. Added surround bass management can help, then, where to place these even more obvious and ugly speakers?
You've made quite a few generalizations and who hasn't. I have no doubt that your rules-of-thumb have served you well. However, you present ideas as facts when they are merely your perceptions. 6dB of headroom may be your idea of proper, others have stated 3dB and I'm less worried about headroom in a cinema because I know what the maximum SPL will be. The only real reason to worry about headroom is if you also think the space or equipment is so miserable that a lot of EQ that will require a non-flat response to take away from speaker sensitivity.
BTW...it is 72dB, per ⅓-octave band, in the pass band (100Hz< Passband < 2KHz). The upper hinge point is 2KHz - 10KHz where it goes down 3dB/oct, if following ISO 2969 (aka curve X from SMPTE 202). You are correct that for 85dBc to have meaning, the response has to meet the X-curve and that it is more important for the individual bands to be at the nominal level for the response than what the single number for the wide-band response.
How well a driver can handle higher power levels depends on how the driver cools itself to pull the heat off the voice coil. I can tell you that a JBL 2269 can take 2000-watts about all day...it is a scary thing! I've never seen a JBL 2242 v.c. cook except when it gets DC (thanks Ashley). When a 2242 has come in for recone, 99 times out of 100, it was physically damaged (Screwdriver or something through the cone). It handles an honest 800-watt.
Remember too, the stimulus (pink noise) factors into these measurements too. Pink Noise is not this 1-size fits all stimulus. The crest factor plays into it as well. In cinema, it should have a 12dB crest factor. I believe it was Meyer that has a white paper that shows the crest factors of various pink noise sources and depending on which one an installer uses, they'll end up with different levels.
When it comes to lower frequencies, particularly in large rooms, like cinemas, the room itself can have a huge effect on the overall level (and response). You will get a room-gain that will lower the demands a bit too. In typical cinemas, that is going to be knocking around 3dB (so I guess you get some headroom there).
As for wiring up surrounds. Yeah, in a non-Atmos system, I normally wire up surrounds in parallel and keep it to 2Ω or higher for most amps or 1.6Ω or higher for DCA 1222 because it can drive into 1.6Ω. I'll do series-parallel as well with a low an impedance as practical in the series portion. So, if I have say 6 speakers on a side wall, I'll wire three speakers in parallel, wire another three speakers in parallel and then series the two groups so each speaker only sees 2.67Ω (plus the cable ) in series with it. The amp sees a comfortable 5.33Ω. I've yet to hear ANYONEs system that wires one amp channel per speaker sound better. I do try to minimize series connections as it does decrease damping factor but it is also surrounds. I've heard a LOT of soundtracks. There just isn't all that much non-effects sounds in the surrounds to get all hot and bothered about going to the nth degree about each and every aspect on it. Heck, if you have the time, wire one side of a theatre with discrete amplifiers and the other with a good parallel or series-parallel and I defy you to have anyone come out of the room wondering why the two sides sound so different. There are probably upwards of 1000 things you could do different that would have a more noticeable effect on the overall sound. But hey, with Dolby Atmos, we do have discrete amplifiers per speaker and your wish of 12" or larger drivers is granted (thus far, in cinemas, all of ours have been 12" and 15" LF drivers). But there is a difference there...each speaker can be called upon to carry the entire audio for the channel...not to be merely part of a team. Surround LFE, consequently, makes a degree of sense there too, depending on the signal being fed. I'm not in the camp of those that think that Surround LFE is something we've been missing all of these years because, again, the typical surround audio, just doesn't seem to benefit from it (unless you have really small surround speakers that can't move any bass). Then again, with the Dolby DMA amplifiers, having discrete amplifier channels per speaker isn't such a big thing anymore...and in a tidy 4U rack space.
The JBL 2269 is built like a tank, it has a massive heatsink built right into the design. With sufficient ventilation in the cabinet, it will do 2 kW non-stop if it needs to. But no realistic real-world application will ever come close to that. That's why something like "M-Noise" is a practical figure to have, let's see how well the industry will adapt to it, because in general, the focus is on big numbers that look great on paper but are only "useful" for pure marketing purposes or for pissing contests...
Taking the 2242 as an example, I do not question it can take 800W no problems. However, JBL warns of a 600W power compression at rated power (300W at -3dB) so, unless I am misunderstanding the meaning of "power compression", there is little gain in pushing that driver over half of its power, at least for prolonged periods of time, which is what a cinema LFE does pretty often.
I agree that with Music it can be a different story: power compression happens after a few milliseconds (I cannot find a nice SPL/time graph I stumbled upon years ago) so for music LF kick, power compression figures should be taken into account differently.
No, the 2242 won't break at 800W. But it won't output what one expects. Hence the idea that 3dB headroom on LF in a cinema is a good idea. I'll quote Stefan here:
Keep in mind, the usable time for low excited drivers is extremely long, and the failure rate extremely small. In the end the extra cost does not matter at all.
Hold up there Tex...er Marco. So do you find that you are driving your subwoofers for 5 minutes, continuously at full rated power? Because that is what you'd need to do to get a 3dB power compression on the 2242.
Have you ever tracked how long subwoofers (or any channel) are hit during a typical movie? Far less than you are leading on. Are you pre-increasing your subwoofer output level by 3dB on the off chance that Tenet is going to run, forever, and you want to pre-compensate for power compression? Remember, you are tuning up a -20dB...well below where significant power compression is going to be a factor unless you are severely underspecified...something that a headroom guy, like yourself, would never do.
So, if you are pumping 800W into it for 5 minutes, you'll be down 3.3dB. While tuning up, at -20dB...and that is if you are driving it to 800-watts at 0dB...even if you are slow, you'll have less than ½ dB of difference...not something one should worry about given their microphones and analyzer have more variation (as do the drivers themselves). I'd think most audiences would welcome a 3.3dB compression after sitting through that many continuous explosions!
So the summary (for us dumb dumbs) is that under typical operating situations (i.e. not playing Interstellar at 7), power compression on the subwoofer isn't really something to be too concerned about?
It all depends (don't you love those non-concrete answers)? The JBL 2242 has a very good cooling system designed into the driver to move heat off of the voice coil as well as possible. This is not the case for all drivers; not by a long shot. Power compression is a real thing and, depending on the driver and how it is being used, it could play into things. That said, I wouldn't worry about designing my system with power compression in mind. You are not going to be bumping up the subwoofer level over the course of a movie because the v.c. has heated up.
Then again, this is Tenet and the power I was feeding to just one subwoofer. The room is about 65-feet long, the fader was at 7 and tuned to spec. Wattage is on the left, time is on the bottom and this was measured for the entire feature This is peak power, not rms but still:
As you can see...the subwoofer, in most movies, is, typically, not doing much and when it does do serious work, it is for brief instances. Nothing that would approach minutes of continuous.
Last edited by Steve Guttag; 04-11-2022, 02:09 PM.
I would consider anything from Captain 'burn out your system' Nolan to be outside of a typical situation. BTW I do appreciate all the information, even though I'm sure you've given it out multiple times over the years.
Tweet
Comment