Previous Thread
Next Thread
Print Thread
Rate Thread
Page 8 of 9 1 2 3 4 5 6 7 8 9
Re: INSANELY BRIGHT?
#16872 08/12/03 04:41 PM
Joined: Apr 2003
Posts: 16,441
shareholder in the making
Offline
shareholder in the making
Joined: Apr 2003
Posts: 16,441
I have a semi-related question that I'm sure anyone with the slightest knowledge of EE can help me with.

When reading specs for some higher-end amps, I often see the term "slew rate," given in volts per microsecond. What is this spec and what does it tell me about the amp?

Re: INSANELY BRIGHT?
#16873 08/12/03 08:49 PM
Joined: May 2002
Posts: 10,654
shareholder in the making
Offline
shareholder in the making
Joined: May 2002
Posts: 10,654
Peter, since semi or sushi haven't had a chance to jump in on this and you did say "slightest knowledge", let me say that slew rate is the rate of change of its output voltage(in volts per microsecond,as you indicate)that an amplifier is capable of. The significance is that the maximum frequency which an amplifier can reproduce with negligible distortion is proportional to the slew rate divided by the peak output voltage which the amp has to supply. So, slew rate is an important engineering consideration, but once an amp can reproduce at least 20,000hz with low distortion at its designed peak output voltage(which any properly designed amp can do), the slew rate is necessarily shown to be high enough for audio purposes. In particular, a higher slew rate doesn't mean that an amp is "quicker" in reproducing music and isn't of any particular significance in the real world.


-----------------------------------

Enjoy the music, not the equipment.


Re: INSANELY BRIGHT?
#16874 08/12/03 11:43 PM
Joined: Apr 2003
Posts: 16,441
shareholder in the making
Offline
shareholder in the making
Joined: Apr 2003
Posts: 16,441
In reply to:

a higher slew rate doesn't mean that an amp is "quicker" in reproducing music and isn't of any particular significance in the real world.



That is what was bugging me. Thanks for the explanation.

Re: INSANELY BRIGHT?
#16875 08/13/03 01:30 AM
Joined: Jul 2003
Posts: 80
old hand
Offline
old hand
Joined: Jul 2003
Posts: 80
In reply to:

As an example, my Adcom power amp (rated 325Wpc into 4 ohms) has nifty instantaneous distortion indicators; I have NEVER managed to turn them on even when listening to highy dynamic music/soundtrack materials at insane (basically intolerable) volume levels in my 5000 cu. ft. room.



I put some calculations into a spreadsheet and came up with the following:
Note: this assumes a speaker sensitivity of 88dB/W/m.

8 ohm, 3 meters, 29 dB loudness = .001V, .001amp, and .00001W. This would be a really quiet passage. 85 dB, close to my average loudness level, would be 6V, .75amp, and 4.5W. Say I crank it really really loud and the peaks bring the music to 100dB. This would mean 34V, 4.25amp, and 145W of power! So as you increase power levels, the instantaneous peaks of required power grow exponentially.

Then don't forget that full-range speakers demand a lot of low-impedance current to power the bass. The "8ohm compatible" Paradigm Reference Studio 100 v.2 has an impedance hovering around 4 ohms for most of the frequencies below 80 Hz. When dealing with very low bass, it becomes feasible to hit 105, maybe 110dB in total power for a moment. Now you're well into the hundreds of watts. Don't forget how many subwoofers have their own power source of 100, 300, even 1000 watts of power.

Boiling frogs
In reply to:

...There are many ways to do this, including the classical frequency sweeps, pink noises, square waves, and more fashionable impulse and step responses, but these are in fact 100% equivalent to each other when measuring amps



100% equivalent?! Absolutely not. Manufacturers choose very different ultrasonic frequencies at which to begin rolloff; some begin the rolloff below 20kHz, even with solid state amps. The resulting square waves can be virtually square with super-high bandwidth amps @ 10kHz, whereas the lower the rolloff, the more rounded the square wave.

At this point we have established that all amps measure differently, though typically the resulting frequency response should be within a decibel. 1.5dB or so is considered close to the threshold of just noticeable difference (JND), right? Well you inspired me to delve into a psychology book. This threshold is usually determined by providing one stimulus, hiding it, then incrementing or decreasing the intensity and showing it again. It can also done by slowly but continuously altering the intensity; this is how you boil a frog without it noticing.

Moving pictures
But what happens if you are staring at a light bulb and it instantly increases in brightness by 1%? Even if its below your "JND" threshold, you can easily notice an abrubt change. Consider color depths on your computer screen. 24-bit is typical, which is 16 million colors. Some people go even higher to 32-bit if they do lots of image work; that is over 4 billion colors. Yet my psychology book says that humans can discriminate about 7 million different colors. Why the difference? Our senses are tuned to use contrast whenever possible; in the real world, colors are not broken into 16 million discrete colors. They are continuous. On the computer screen, an artificial contrast or step is created between each quantum, and our visual system accentuates this difference, so we see a line if we look closely.

Have you ever seen LEDs that seem to flash when you walk by them? When we stare at a fixed spot of pulsing light, it begins to look solid I think in the hundreds of hertz. However, when that light source moves, our threshold jumps to the kilohertz. So in our quantified world of man-made pulses and lines, we need to increase the numbers several-fold in order to properly fool us into sensing continuity, be it vision, hearing, or other senses. I believe this is why DVD-audio and SACD are noticeably better than CD: they up the numbers almost to the point where it sounds no different than the continuity of an LP, but without the low resolution.

Impossibly conclusive
SOOOO, for those of you still here, I will hypothesize that with constantly changing music, our sensitivity to frequency response is higher than with fixed tones. If we can hear a symphonic peak clearly, at 11 million times the power level of a whisper, and if we can distinguish both within a second of each other, then perhaps we can hear all those subtleties that many engineers would love to dismiss. For those of you who have heard the differences (I'm pretty low on the audiophile scale), then this will resonate well with you.

-Cooper

Re: INSANELY BRIGHT?
#16876 08/13/03 05:22 AM
Joined: Feb 2003
Posts: 1,490
connoisseur
Offline
connoisseur
Joined: Feb 2003
Posts: 1,490
Peter, as John has nicely explained, slew rate is essentially another way of describing the top-end frequency response of an amp. As long as the amp's frequency response is quasi-flat up to 20kHz, you don't have to worry about the slew rate.

In reply to:

100% equivalent?! Absolutely not. Manufacturers choose very different ultrasonic frequencies at which to begin rolloff; some begin the rolloff below 20kHz, even with solid state amps.



Cooper, the third sentence above is, of course, entirely true. Accordingly, I never stated that all amp designs are 100% equivalent at all. What I stated was, various methods of deriving the formal "transfer function" of an amp are 100% equivalent from each other. Some of these methods use steady-state test signals; others use transient/dynamic signals. But in the end, they derive an identical transfer function for a given amp. This is mathematically guaranteed. My whole point is, as far as we are discussing the amps, their dynamic/transient behavior can be accurately predicted from the measurements that use steady-state test signals. This is almost counter-intuitive but true, AS LONG AS we do not go beyond the maximum continuously-available power.

In reply to:

So as you increase power levels, the instantaneous peaks of required power grow exponentially.



Of course. That's the very definition of the dB unit, which is based on the logarithmic raw of human auditory perception. Your calculations are accurate, except for two rather unrealistic assumptions: (1) An in-room sensitivity of 88dB/W/m is really close to the bottom end (you are assuming something like the Maggies). For instance, the Axiom M22 is rated at 93dB/W/m in-room. This alone cuts the power demand down to one third. (2) The sound level does NOT decrease with distance in accordance with the inverse square law in a typical listening room and position (since it is not strictly "near-field"). So, at 3m distance, the sound level will NOT drop by -9.5dB -- it will realistically be more like -6dB (exact figure depends on the room size, reflectivity, speaker placement, etc). This cuts the required power output further by half. So, the realistic power demand at 100dB musical peaks would be more like 25W (rather than 145W). And then, all this assumes that you are driving only one channel. If you are talking about a 2-channel stereo system, the power demand per channel will further decrease by 2-3dB, down to 15W/ch or so. Therefore, if you have a good-quality 100W/ch amp, you will still have a healthy ~8dB headroom. In fact, this is exactly why some astute audiophiles are perfectly comfortable with a 30W/ch "pure Class A" power amp, when combined with relatively efficient speakers.

In reply to:

At this point we have established that all amps measure differently, though typically the resulting frequency response should be within a decibel.



I agree, EXCEPT that the resulting frequency response (with a real-world load) would be, in the worst case, within 0.1-0.2dB for today's SS amps. And this is indeed an important fine point: I don't know what your psychology textbook has to say, but people with well-trained ears can discern a 0.5dB hump on the frequency response (and, perhaps, a 1dB dip), if the hump/dip is broad enough to encompass 1-2 octaves.

In reply to:

SOOOO, for those of you still here, I will hypothesize that with constantly changing music, our sensitivity to frequency response is higher than with fixed tones.



Your analogy with the human vision is nice, but unfortunately, it is a wrong analogy. The human visual and auditory systems work very differently in many aspects [one example is the "intra-scene" concurrent dynamic range: the visual system easily handles ~1,000,000:1, whereas the auditory system can do only ~1,000:1 at best]. In fact, on the contrary to your hypothesis, our sensitivity to deviation of frequency response is BETTER when we are listening to a steady-state, wide-spectrum tone. Alan will tell you that a pink noise is a perfect example; try it if you are in doubt. Trust me, I do systems neuroscience as my day-job, when I am not loitering in this forum.

Re: INSANELY BRIGHT?
#16877 11/08/03 10:53 AM
Joined: Oct 2003
Posts: 604
aficionado
Offline
aficionado
Joined: Oct 2003
Posts: 604
In reply to:

They are continuous. On the computer screen, an artificial contrast or step is created between each quantum, and our visual system accentuates this difference, so we see a line if we look closely


But this is because of lateral inhibition / the whole center surround receptive field of the retinal ganglion and visual cortex (and arguably also the horizontal cells and the amacrine cells, as they begin to sort data in between the different cone and rod cells before the ganglion cells touch it). In other words, contrast is REALLY perceptible in the visual system; it's meant to be that way. Our brains rely on contrast for vision, not so much the actual image; so much so that the magnocellular cortical layers are much thinner than the parvocellular layers. (Magnocellular is basically motion, wheras parvocellular is color and form, both of which get a huge deal of information specifically from contrast.)

In reply to:

SOOOO, for those of you still here, I will hypothesize that with constantly changing music, our sensitivity to frequency response is higher than with fixed tones. If we can hear a symphonic peak clearly, at 11 million times the power level of a whisper, and if we can distinguish both within a second of each other, then perhaps we can hear all those subtleties that many engineers would love to dismiss.


Well, true, but not like we'll be able to do this much; as after that second, there will be data sent from your Medial Superior Olive to the outer hair cells that cause them to lenghten (and comically produce a funny otoacoustic emission of that same frequency in the process, but this is besides the point) and thus make their selective frequency quieter (by bracing the basilar membrane at the corresponding frequency) and decrease the loudness. Plus, I wouldn't imagine it to be too clear until this is happening, and even if it was, you'd be damaging your stereocilia. Even if you don't damage the inner hair cells, damaging the outer hair cells is bad for day to day life, as the noise control is needed for discerning every day sounds from important ones (ie. detecting and understanding voice in lyrics over the loud compuer I have in front of me.)


Re: INSANELY BRIGHT?
#16878 11/08/03 10:55 AM
Joined: Oct 2003
Posts: 604
aficionado
Offline
aficionado
Joined: Oct 2003
Posts: 604
In reply to:

Trust me, I do systems neuroscience as my day-job, when I am not loitering in this forum.


YEAH! for systems! It's such a growing area (engulfing so many other fields - lol). Who'd wanna be cogsci or cell/molec anyhow? They just seem so boring in contrast.


Re: INSANELY BRIGHT?
#16879 11/08/03 03:17 PM
Joined: Feb 2003
Posts: 1,490
connoisseur
Offline
connoisseur
Joined: Feb 2003
Posts: 1,490
Well, DJ... I strive to integrate the systems approach with molecular genetic strategy. I don't think either one alone would bring truly powerful advances in neuroscience.

If you are interested, here is an HHMI bulletin featuring some of my recent work (although it is in lay language).

Re: INSANELY BRIGHT?
#16880 11/08/03 03:31 PM
Joined: Oct 2003
Posts: 604
aficionado
Offline
aficionado
Joined: Oct 2003
Posts: 604
(from article)
In reply to:

...Masashi Yanagisawa had discovered the gene for a neuropeptide that binds to this receptor...


Heh - wow.

And you're right about just strictly systems alone won't be as powerful as a combination of that with something a bit more fine tuned. That's why systems people generally do more than strictly circuits, which systems is by defenition.

Alternatively, even cognitive can integrate to a certain degree with systems and be powerful, don't you think?

Edit: LOL - I'm no longer a vet, and am now a local - lol.


Re: INSANELY BRIGHT?
#16881 11/08/03 04:29 PM
Joined: Feb 2003
Posts: 1,490
connoisseur
Offline
connoisseur
Joined: Feb 2003
Posts: 1,490
In reply to:

Alternatively, even cognitive can integrate to a certain degree with systems and be powerful, don't you think?



If you mean by "systems neuroscience" the circuit- and neuroanatomically-based approaches, then the cognitive people have certainly been using the systems approach for a long time, especially those working with primates.

The fundamental problem is, many meaningful experiments in cognitive neuroscience require humans or primates, in which the techniques for genetic manipulations are next to nonexistent right now. We need a big methodological breakthrough here...

Page 8 of 9 1 2 3 4 5 6 7 8 9

Moderated by  alan, Amie, Andrew, axiomadmin, Brent, Debbie, Ian, Jc 

Link Copied to Clipboard

Need Help Graphic

Forum Statistics
Forums16
Topics24,940
Posts442,457
Members15,616
Most Online2,082
Jan 22nd, 2020
Top Posters
Ken.C 18,044
pmbuko 16,441
SirQuack 13,840
CV 12,077
MarkSJohnson 11,458
Who's Online Now
0 members (), 563 guests, and 1 robot.
Key: Admin, Global Mod, Mod
Newsletter Signup
Powered by UBB.threads™ PHP Forum Software 7.7.4