Gene thanks for starting this discussion. Need alignment across manufacturers so consumers can effectively use ratings when making decisions on amps etc. “Measure at the knee” is a great rallying call. Bumper stickers should be issued!
Good video. The problem with what you are asking is that it falls into the same trap we already have: Most consumers just want ONE number; They want to know the watts. However anyone reading this realizes that the reality is FAR more multidimensional than a single number. A single number is one data point from hundreds or thousands. Like most things in life, a reasonable understanding of something involves a lot more than one fact, one number, or one data point. One needs a spectrum of data to see a fuller picture. Also, like most things, the numbers are never linear. That's why graphs are so helpful, and of course a graph is a spectrum of data points. So back to the original point: Getting everyone to agree on one methodology to generate one number (watts) could be helpful in a small way, but it's not going to give enthusiasts the full picture.
Rated at low distortion with 3dB of dynamic headroom. The difference in performance is crazy. I have a power amp rated 100w into 8 ohms at 0.01 percent distortion. Actual peak power is nearly 3 times that. No need to overdrive the amp to get fantastic uncompressed music. I am very sensitive to amplifiers starting to compression as the volume is turned up.
I think "at the knee" would be appropriate and suspect most Audioholics fans would too, which is to say the sort of person who knows what dynamic headroom is and why it's good. That distinction is probably why you see an ultra high end speciality brand like Classé rating that way. It's customers know already or are buying from a dealer who's ready to educate. A more mass market brand like Marantz can't as readily assume a well informed customer or that the buyer is talking to anyone but a check box on a website, so they go with the "big numbers" method that is more widely understood.
Back in the day (40+ years ago) the shop i worked for looked at the "power bandwidth", 20 - 20,000) at 8, 4 and even 2 ohms on 4:55 separate power amps. There was also the issues of how the amp handled a reactive load not just a resistor. (One 40 watt rated amp outperformed a popular 400 watt rated amp at low impedance and a reactive load. The protection circuit would shut down the 400 watt amp rather quickly when driven hard into a low impedance speaker. There were other measurements with regard to distortion, etc. at various frequencies ( a bit too complex for me to understand back then) at the manufactures rated power as well. The customers received a printed sheet with actually measured specs for their product. Today I rarely see a 20-20000 rating especially at 4 ohms. You may get an 8 ohm rating from. 20 - 20,000 but rarely any a 4 ohm rating beyond a 1,000 spec Many receivers will not handle anything below 6 ohms! (We would have described any product like that as lowfy junk!) Other manufacturers actually show reduced power at 4 ohms telling you that thr amp is running out of power. I have a 30 watt commercial amp that out performed a component that the manufacturer rated at 1000 watts! Good grief! My new Class D commercial amp is rated a 650 watts per channel into 4 ohms both channels driven at 1000. At 8 ohms you get a 20 -20000 rating of 400+ as I recall. The amp delivers plenty of power to get music into upper 90s on my rather inefficient speakers nominally rated at 4 ohms. However, What is the "real" power of my amp at 4 ohms? I have no clue! The manufacturer refused to give me a 20-20000 power rating into 4 ohms. I've seen 7 channel amps that offer a rating with "one channel driven." How so you even do that with a 7 channel home theater system? All this serves to mislead and confuse the average customer. What you have is "snake oil" ratings with no reference to real world application. Very sad! 😢
The reason for sharp knee is that output stage transistor get saturated and it is not coming out of saturation, the soft knee version is coming out off saturation easly, you can read about it in Bob Cordell book, you can run test with signal generator with white noise output , limit bandwidth in generator to 24khz, I use Keithley Instruments signal generator for this test, then if you increase the signal level in some point, it will hit the rail limitation voltage, you can use oscilloscope with delay trigger then use stop to study the signal , you will see in some part off signal that the out put transistor get stuck in saturation and it is not follow signal even the signal go down. With soft knee version is not the case , the transistor will not saturated. Bob Cordell explains this missbehave in some off his audio conference. With this test you will see that hard knee is not desirable, this test will not show with sinus signal , I run my test with white noise test and it shows it very easily, so the answer is soft knee amplifier is not misbehaving, and power rating should be when THD hit the 1% rating. And it should be continues average power rating per channel , usually I do it with 20 Hz to see if power supply sagging with all channel active to lowest load (2 ohm). This so called misbehaving are known by some manufacture and some other they don't about it ( or they don't care). they just test it with sinus signal which it doesn't show.
I generally rate amplifier power at 1% below clipping, with all channels driven. It's the end of it's clean range and most reflective of the electronics' real world capabilities.
Sadly power isn't across all channels as it use to be, Dsp modes are effecting the original sound the directors intended and don't benefit the end user , full rated power across all channels should be the main focus as it once was.
I believe the best system for deterring power for a solid state amplifier is what I call "10% below the knee". Most solid state amplifiers will have a point in it's power sweep spectrum where it will "knee" (change very quickly from a mostly horizontal increase in THD to a mostly vertical increase in THD). This often happens almost instantly at a very specific point. What I would do is take that power rating point, and multiply that number by 0.9 and there you have it, the rated power (hence 10% below the knee). If an amplifier has a somewhat gradual increase (some pure class A amplifiers have this as well as most tube amplifiers) then I'd take the point that most closely resembles the knee (even most tube amps will have at least something close to a "knee" even if it's very rounded) then multiply that by 0.9 to get my "10% below the knee" rating. I do not like to use a "THD target" because that sometimes will result in a rated power far below the knee or well up into the knee that gives a very inaccurate statement on an amplifiers real rated power. Now comparing the two amplifiers, I'd say the one of the left has a "knee" rating of around 140 watts meaning I'd only rate that as a 125 watt amplifier. The amplifier on the right does not hit the knee till 344 watts, so that has in my book a rated power of closer to 310 watts but for rounding purposes, 300 is a good round number. I think in the case of the right side amplifier, they had a "target" rating and they wanted to clearly surpass it. The one on the left, they had a target, didn't quite make it, so they decided to tweak the results to call it a 200 watt amplifier using a "THD" target even though clearly it's well past the knee and if you were to push that any further you'd clip that amp pretty hard and quickly.
Interesting watch. In ye olde days in radio communications systems when we used tube amplifiers - TWTA and Klystron - those were rated at saturation. But as more solid state amplifiers were introduce we started using the 1 dB gain compression point as the metric. That was something like 40ish years ago. There are a number of other metrics we look at as well for amplifiers and other active components such as 3rd order intercept, gain and flatness across the passband and group delay across to name a few. Looking at the THD knee is an interesting approach though.
Just would like to say thanks for talking about equipment for the blue-collar buyer. If we all had a jeff bus budget the audio world would be different. Also audio stores would make a big difference too. I purchased 2 HSU for myself at Xmas and I love them. I run PSB Equipment for years and I enjoy their stuff, but the HSU subs are much better. But thanks guy for doing some of the not so popular Equipment, if you get what I'm saying. Darrylk
When I first began to build a single stage amplifier (early transistor experience) we would do what's called a family of curves. Any higher quality amplifier needs to operate in its linear region for the full frequency response and below it's peak output. It's disingenuous to create an amplifier that reaches its rated peak power at or just above the peak voltage swing of the power supply. The amplifier may show a bit more current, however it has literally run out of the power supplies capacity. If a transistor completely turns on or off, it's now operating in switching mode and more distortion will be present. Power amplifier ratings are also subject to operating temperature and duration. It's quite possible to momentarily achieve maximum power out, but it isn't sustainable. The heat buildup will cause the output as well as many driver transistors to fail. Therefore, there needs to be a period of time factor in the power rating. Most people with hearing accustomed to listening for audio details found in good reproduction systems, will be able to hear the 1% to 3% harmonic distortion. I consider that unacceptable for amplifiers. Transducer distortion is an entirely different matter and a different discussion.
I want to see THD graphs as a function of power but across at least 5 or more frequencies and at multiple loads - 2 (car audio), 4, 6(optional), & 8 ohms. Like: 50Hz 100Hz 500Hz (optional) 1KHz 5KHz 10KHz 15KHz (optional) Lastly, they should also provide the details about power over full bandwidth pink noise with: 1ch 2ch 4ch All channels driven, or at least the 70/30 for mains and surrounds. I realize I'm asking for the moon in specs.
I would be satisfied if the amplifier was measured from 20hz to 20,000hz and an average from that was taken. BUT, the bigger question is how much I really need. I remember the days when 35 watts per channel was a very popular power point for many listeners such as myself back then.
20/25 years ago comment per Harman Kardon rep when they were selling receivers rated much lower watts at higher prices than consumer brands at the time like Sony, Kenwood, Technics "It is not wattifier it is amplifier". They used to specify how much amp/current their amps/receivers were . And their watt rating were all channels driven with full range . Of course this is before Samsung take over and killing the HK amp business
Thank you, Gene! Agreed that these ratings are mostly useless and need to be researched quite a bit to make sense of. Same goes for power at 8, 4 and 2 ohms. Curious - could you do a similar video about speaker power ratings? I’ve always been a little perplexed by how to interpret those as well. For example - my speakers are rated for a max. 500W and have a sensitivity of 90db at 1W. I can’t imagine pumping 500W into them. I feel like I’d be deaf way before or they would go up in smoke. I don’t even have 500W available. What is that relationship between max power handling of a speaker and how much you actually feed it?
For car audio, I just go by the fuse rating times 10. it's usually close enough for me to match to the appropriate speaker. Some of these amplifiers are not coming with fuses making it harder to follow that rule. As for home amps I look at the power consumption its rated for. They all lie, but this don't.
I think we should have an amplifier warbling about 20 Hz while we have an inter-modulated tone you know at various frequencies to see how it can all hold together. Then measure power, etc.. by frequency band.
this, and AVR manufacturers should rate for 5 channels driven I think. It's just silly for a 9 or 11 channel amp to specify 2 channel only power ratings.
@@Jon-nz3dm that could be because at a given time in any movie scene, there could be as many as 2 or max 3 channels running at full levels. But yes as you say they should also mention values when all channels are driven at different loads(8 & 4ohms).
When you think about actual listening, amps will spend the vast majority of time in the low-power regions of their curves with louder dynamic content being mostly sporadic if you're not really driving the whole system really hard. So most of the time when in the high-power parts of their curves, it's going to be loud highly dynamic sound which should mask a fair amount of the distortion. I guess what I seem to be getting at is amps should be designed with more moderate continuous power but with considerable dynamic power. The question you asked was about how to rate amps, though, and that's a different question. Can you build amps with such a output profile, meaning highly dynamic capability even if more modest continuous? And it would need to be clean dynamic power.
I agree with this ideology completely. No matter the case, it should be standardized and mandatory. EVERY home-purposed A/V receiver and Amplifier should be mandated to report this way. That way we can ALL begin to understand the rating with no "tricks". Shouldn't CEDIA and other similar organizations be working to standardize these types of specifications? I'm all for the more conservative Rating before the Knee. Anthem and NAD have always produced amplification that is "understated". And their performance is certainly exceptional.
Great vid, 100% headroom is best not peak number. Consumer cheap audio overstates capability, true Hifi gear underates. On a seperate note; NAD M23 listening review!!! Please make a vid on its sound 😁 im a proud NAD M3 owner, love to hear your thoughts on the M23
I think we should abandon the use of a nominal value. It’s like neatly packaged for consumers as a marketing tool, rather than really conveying performance to the consumer. Amplifier Watts/distortion, output resistance, and speaker impedance. Are all stories of undulating dynamic change, not singular specs. So I feel graphs are what we should use. It’s not the 1970’s where graphs were hand drawn, and magazine ink increased cost. There are no barriers or limitations to using graphs in the expression of amplifier performance. Lets be real, if someone is such a simpleton that they prefer a singular spec, they also don’t have the nerdy curiosity to understand the singular spec or the context of it’s meaning, and would be better served buying from brand identity or model number tiers. High fidelity sound is strongly influenced by the distortion, so shopping for really low distortion isn’t served by a singular spec, but by observing the graph. How low the distortion rides through that the whole span before the knee, and it’s consistency is not captured at all in traditional printed ftc specs. Take this example given in the video, the first graph has elevated distortion in the first few watts relative to later on when it becomes cleaner. The other graph is just low from the start and is basically a flat line all the way out to the knee. So which one has lower distortion at your expected listening levels? The listening levels may be behind the knee, making the power comparison irrelevant, leaving only the distortion levels to consider. Or the listening levels may draw power beyond the knee, in which case the steep change in a sharp knee may have some audible character that a rolling off knee would not. So for me, I really want to see the graph, and I feel for others not into the nerdy details, they aren’t even served by seeing a nominal spec, and actually confused by being bothered to think about them.
Actually, there needs to be two standards set. One at the old standard of 20 to 20k +-1 db with .09% distortion or less, two channels driven (Standard). A second one at peak 20 to 20k +-1 db with 1% or less distortion, two channels driven (Dynamic). If the manufacturer would produce a chart that shows the power using those standards with each additional channel /channels added would be a great selling point.
I’m an old school broadcast engineer, used to measuring harmonic distortion in 5,000 and 50,000 watt AM broadcast transmitters. Sheesh… talk about problems with power supplies giving out at high levels! But it’s instructive in that the critical performance characteristics are (1) the max output from the power supply before it gives out (causing clipping) and the RISE TIME of the power supply at any output level. It surprises me how little attention is devoted to the INTERMODULATION DISTORTION figures for high powered amplifiers. With everything else in good shape, if the PS can’t restore the DC voltage level to the filter capacitors after a transient, you’ll get INTERMOD-the mixture of two or more frequencies in the program material to produce sum and difference products. This is WORSE than harmonic distortion because at least the harmonics are related (harmonically) to the original sound. Intermod products are not. And what’s more, they tend to fall in the middle of the bandwidth, not the upper reaches, so they mask original frequencies that give the program definition and clarity. General rule: harmonic distortion makes the music sound raspy. Intermod distortion makes it sound muddy.
Thanks for rasing the debate. I think it's simple, specify output power at the rated distortion, so if manufacturer says THD= 0.03% than one should look at max power within that number. Another thought; dynamic power is dynamic; some amps can sustain very high power levels well beyond the FTC 20ms rating; which is short, should be at least be 10x times longer so that a sloppy orchestra can get all the notes in that window for a crescendo. Expressing dynamic power as a linear function by defining desired dynamic power in dB's and then subtract that to arrive a continuous power rating doesn't make sense in my book. True dynamic power (being able to produce short power bursts for say, 300ms) is what makes music sound alive and fast from an amp with headroom.
This has been discussed since back in the 1970's. I don't care whether you have 400 watts per channel or 75 Watts per channel. The quality of sound at the volume you yourself require is more of what I'm interested in. Who runs them at wide open throttle says wow that sounds great.
Absolutely amps should be rated just before the knee. It is an important issue, as it speaks to long term reliability. If you run amps to the clipping range or even close, tube amps excepted, then they will have a short life. I advise keeping 3 db. of heard room. That means if you think you need 100 watts of power, then you need a 200 watt amp. That is the cheapest approach on the long run.
Amplifier power should be rated highly, very highly. There's no such thing as to much, only playing to loud, for your ears, or for your speakers. I know, I'm being difficult here. But really, with modern Class D amps hardly anyone needs to make any kind of compromise when it comes to amps. We're living in an age where a larger part of your budget than ever before can be directed towards the speakers. And that's a wonderful thing!
I've had this thought before and I think they should state both 2db headroom rating and peak watt rating. I don't think the audio companies would want their customers to get too smart though. They like it when customers that use
I would say, make is the same as the speaker measurment. If a speaker is rated 200w RMS, then the standard is to rate the peak as 2x that, or 400w. So if an amp is rated at 200w, it should do a mininum of 400w dynamic, if not, then they shouldn't rate it.Or even better, give the dynamic power, and the rated becomes the /2 of that. Standards is all we need, and to force companies to give enough data.
Great video Gene, Question I bought the Marantz SR7015 based on your channel, Klipsch Reference R-26FA 5.1 Home Theater Pack, the Klipsch R-41SA as my rear height atmos speakers and Two Subwoofers by Klipsch, and am about to add the M5100X to the Marantz SR7015 . Should I connect the M5100X to the two front tower speakers for when I listen to stereo music two channels, or connect the two front, center, and side surround speakers, and let the Marantz SR7015 drive the Klipsch Reference R-26FA atmos and rear height behind me? Thanks in advance.
All manufacturers: RELEASE those THD vs power graphs for different frequencies to the public. People who don't understand them, or don't want to learn, will always believe whatever you tell them anyway.
RMS watts @1khz 0.1%THD. Dynamic power is just a bonus. I really shouldn't need to use more than 75% of the amplifiers rated power anyway. If I do, then I need an amplifier with more power.
I've been casually looking at amps for months and get more confused as I learn more. I have no idea how to get a good budget friendly amp for a klh model 5.
Creek 4240 or 4230 used. It should run you about $200. And it sounds delightful with plenty of power. It’s hard to beat for the money since vintage Marantz/Pioneer is overpriced.
What should be or what it will remain? Rhetorical question. It should be max output before the knee at the lowest distortion with max dynamic headroom ALL CHANNELS DRIVEN. That's what it should be, but as long as there is no forced industry standard for providing specs, this is not going to change. Companies will continue to market to the average 80%, which look at watts/channel vs watts/channel. Engineering will tell marketing that it should be spec'd at lower wattage with more dynamic headroom, while marketing will tell Enginering, thanks we will take it from here...advertising more wattage sells more. That's just how it is whether it's right or wrong.
we should not assume dynamic headroom, a rating for the continuous power output at 0.1% give me the information that is also the dynamic max. i can then get a the amp that has the watts to get me to the dynamic headroom i need. if we also need a dynamic headroom rating that could be more confusing. some of the bass i listen to are so close to a sine wave anyways.
Back in the day I used JBL L300 speakers with Carver TFM 55 amps. Mono Bridge 1K per channel at 8 ohms. Head room was about 4 dB. Sounded great. When the mass ring hit the pole piece I knew to turn the volume down. Sounded like someone hit a welding table with a sledge hammer. Speaker was rated 50-100 watts RMS. But 350-400 watts did sound great. Don't think that did the alnico magnet's or my ears any good. I'm 67 now ears stop at about 13K.
My $0.02, I would prefer a Conservative, medium, and aggressive rating. The Marantz is Agressive while the Classe is Conservative. This could be an industry standard.
If the amp is a Multi-Channel the rating should only list ALL-Channels Driven rating. If it a Stereo ( 2- Channels it should be rated both channels drive simultaneously from bottom (HZ) frequencies to the highest (KHz) frequency, and has of total harmonic distortion they should list 2 or 3 different points. ( Example: audiophile acceptable THD what ever % is set, and a acceptable average / normal THD, and of course that 1 % THD all channel driven... They should stop that 1 channel driven crap rating....
Idk, something like the A11 tribute or the audio labs 6000a only get you 50wpc into 8 ohms into 2 channels for 800-1200$. A monolith 1600$ 200wpc x5 channels seems like an okay deal. If you actually need the power that is, and 95% of people don’t for their home theater.
Clean power cost money, for $1200 you most of them only get 100 watts per channel and many times less than that, but what most important for an amplifier is not what it can deliver in 8 ohms, i rather have numbers for 4 and 2 ohms, and here most amplifiers below $3000 cant deliver
@@andreasmoller9798 I run the Polk Audio Reserve line (R700 mains). They're all 8 ohm speakers...so 8 ohm power was very important to me. ...and I forked out the $$$ for solid gear, too. Solid in my opinion anyway. All of my electronic components are Marantz. ...and my amps were overpriced as well. I have about 30 years of experience in car audio....so I'm not used to spending $1.50 to $2.00 per watt anymore. That died out in 12v back in the early 2000s. Sure, there are a handful of amps that cost a fortune still...but they are made by brands with amazing reputations who've been around for decades...and who are basically "flexing". It's very much a "Hey, look what we can do" type of thing. ...but those same brands have "take down" models that don't cost quite as much. Helix have their Thesis lineup that costs many thousands of dollars...but they also have amps that cost $350-$500. In the HiFi/Home Theater side of audio, the only budget friendly stuff is either super sketchy ChiFi....or used....or refurb/open box. ...and speaking on the impedance thing one more time in this novel. I am eventually going to set up a separate 2.1 music system in one of my spare bedrooms. That room is getting Arendal 1723 gear, so that'll need 4 ohm power...
Gene thanks for starting this discussion. Need alignment across manufacturers so consumers can effectively use ratings when making decisions on amps etc. “Measure at the knee” is a great rallying call. Bumper stickers should be issued!
Good video. The problem with what you are asking is that it falls into the same trap we already have: Most consumers just want ONE number; They want to know the watts. However anyone reading this realizes that the reality is FAR more multidimensional than a single number. A single number is one data point from hundreds or thousands. Like most things in life, a reasonable understanding of something involves a lot more than one fact, one number, or one data point. One needs a spectrum of data to see a fuller picture. Also, like most things, the numbers are never linear. That's why graphs are so helpful, and of course a graph is a spectrum of data points. So back to the original point: Getting everyone to agree on one methodology to generate one number (watts) could be helpful in a small way, but it's not going to give enthusiasts the full picture.
Rated at low distortion with 3dB of dynamic headroom. The difference in performance is crazy. I have a power amp rated 100w into 8 ohms at 0.01 percent distortion. Actual peak power is nearly 3 times that. No need to overdrive the amp to get fantastic uncompressed music. I am very sensitive to amplifiers starting to compression as the volume is turned up.
I think "at the knee" would be appropriate and suspect most Audioholics fans would too, which is to say the sort of person who knows what dynamic headroom is and why it's good. That distinction is probably why you see an ultra high end speciality brand like Classé rating that way. It's customers know already or are buying from a dealer who's ready to educate. A more mass market brand like Marantz can't as readily assume a well informed customer or that the buyer is talking to anyone but a check box on a website, so they go with the "big numbers" method that is more widely understood.
Back in the day (40+ years ago) the shop i worked for looked at the "power bandwidth", 20 - 20,000) at 8, 4 and even 2 ohms on 4:55 separate power amps. There was also the issues of how the amp handled a reactive load not just a resistor. (One 40 watt rated amp outperformed a popular 400 watt rated amp at low impedance and a reactive load. The protection circuit would shut down the 400 watt amp rather quickly when driven hard into a low impedance speaker. There were other measurements with regard to distortion, etc. at various frequencies ( a bit too complex for me to understand back then) at the manufactures rated power as well. The customers received a printed sheet with actually measured specs for their product.
Today I rarely see a 20-20000 rating especially at 4 ohms. You may get an 8 ohm rating from. 20 - 20,000 but rarely any a 4 ohm rating beyond a 1,000 spec Many receivers will not handle anything below 6 ohms! (We would have described any product like that as lowfy junk!) Other manufacturers actually show reduced power at 4 ohms telling you that thr amp is running out of power. I have a 30 watt commercial amp that out performed a component that the manufacturer rated at 1000 watts! Good grief!
My new Class D commercial amp is rated a 650 watts per channel into 4 ohms both channels driven at 1000. At 8 ohms you get a 20 -20000 rating of 400+ as I recall. The amp delivers plenty of power to get music into upper 90s on my rather inefficient speakers nominally rated at 4 ohms. However, What is the "real" power of my amp at 4 ohms? I have no clue! The manufacturer refused to give me a 20-20000 power rating into 4 ohms. I've seen 7 channel amps that offer a rating with "one channel driven." How so you even do that with a 7 channel home theater system? All this serves to mislead and confuse the average customer. What you have is "snake oil" ratings with no reference to real world application. Very sad! 😢
Looking forward to Audio Advice Live. Will be my second AV show in a months time. I’m excited to get out with the AV community this summer.
The reason for sharp knee is that output stage transistor get saturated and it is not coming out of saturation, the soft knee version is coming out off saturation easly, you can read about it in Bob Cordell book, you can run test with signal generator with white noise output , limit bandwidth in generator to 24khz, I use Keithley Instruments signal generator for this test, then if you increase the signal level in some point, it will hit the rail limitation voltage, you can use oscilloscope with delay trigger then use stop to study the signal , you will see in some part off signal that the out put transistor get stuck in saturation and it is not follow signal even the signal go down. With soft knee version is not the case , the transistor will not saturated. Bob Cordell explains this missbehave in some off his audio conference. With this test you will see that hard knee is not desirable, this test will not show with sinus signal , I run my test with white noise test and it shows it very easily, so the answer is soft knee amplifier is not misbehaving, and power rating should be when THD hit the 1% rating. And it should be continues average power rating per channel , usually I do it with 20 Hz to see if power supply sagging with all channel active to lowest load (2 ohm). This so called misbehaving are known by some manufacture and some other they don't about it ( or they don't care). they just test it with sinus signal which it doesn't show.
They can rate whatever they can as long as you r here to test it for us😉
Thank you🎉
I generally rate amplifier power at 1% below clipping, with all channels driven.
It's the end of it's clean range and most reflective of the electronics' real world capabilities.
Sadly power isn't across all channels as it use to be, Dsp modes are effecting the original sound the directors intended and don't benefit the end user , full rated power across all channels should be the main focus as it once was.
I believe the best system for deterring power for a solid state amplifier is what I call "10% below the knee". Most solid state amplifiers will have a point in it's power sweep spectrum where it will "knee" (change very quickly from a mostly horizontal increase in THD to a mostly vertical increase in THD). This often happens almost instantly at a very specific point. What I would do is take that power rating point, and multiply that number by 0.9 and there you have it, the rated power (hence 10% below the knee). If an amplifier has a somewhat gradual increase (some pure class A amplifiers have this as well as most tube amplifiers) then I'd take the point that most closely resembles the knee (even most tube amps will have at least something close to a "knee" even if it's very rounded) then multiply that by 0.9 to get my "10% below the knee" rating. I do not like to use a "THD target" because that sometimes will result in a rated power far below the knee or well up into the knee that gives a very inaccurate statement on an amplifiers real rated power. Now comparing the two amplifiers, I'd say the one of the left has a "knee" rating of around 140 watts meaning I'd only rate that as a 125 watt amplifier. The amplifier on the right does not hit the knee till 344 watts, so that has in my book a rated power of closer to 310 watts but for rounding purposes, 300 is a good round number. I think in the case of the right side amplifier, they had a "target" rating and they wanted to clearly surpass it. The one on the left, they had a target, didn't quite make it, so they decided to tweak the results to call it a 200 watt amplifier using a "THD" target even though clearly it's well past the knee and if you were to push that any further you'd clip that amp pretty hard and quickly.
Interesting watch. In ye olde days in radio communications systems when we used tube amplifiers - TWTA and Klystron - those were rated at saturation. But as more solid state amplifiers were introduce we started using the 1 dB gain compression point as the metric. That was something like 40ish years ago.
There are a number of other metrics we look at as well for amplifiers and other active components such as 3rd order intercept, gain and flatness across the passband and group delay across to name a few.
Looking at the THD knee is an interesting approach though.
For me amp should be rated with osciloscope. When you see that sine wave started to reach the limit, that is the max voltage you get from amp.
I can't stand the 1khz power rating it should be done below 100hz.
20Hz-20KHz & 1KHz, like old crowns
Ya Id like to see 100hz, 1khz, & 15khz. To see how it handles the ranges.
I am perfectly happy with my Hegel h390, so much power and control, and it dosen’t give up even when the load is below 4 ohms
Most definitely they should start being rated before the knee giving a lot more Headroom...!!🎅👍👍
Just would like to say thanks for talking about equipment for the blue-collar buyer. If we all had a jeff bus budget the audio world would be different. Also audio stores would make a big difference too. I purchased 2 HSU for myself at Xmas and I love them. I run PSB Equipment for years and I enjoy their stuff, but the HSU subs are much better. But thanks guy for doing some of the not so popular Equipment, if you get what I'm saying. Darrylk
When I first began to build a single stage amplifier (early transistor experience) we would do what's called a family of curves. Any higher quality amplifier needs to operate in its linear region for the full frequency response and below it's peak output.
It's disingenuous to create an amplifier that reaches its rated peak power at or just above the peak voltage swing of the power supply.
The amplifier may show a bit more current, however it has literally run out of the power supplies capacity. If a transistor completely turns on or off, it's now operating in switching mode and more distortion will be present.
Power amplifier ratings are also subject to operating temperature and duration. It's quite possible to momentarily achieve maximum power out, but it isn't sustainable. The heat buildup will cause the output as well as many driver transistors to fail. Therefore, there needs to be a period of time factor in the power rating.
Most people with hearing accustomed to listening for audio details found in good reproduction systems, will be able to hear the 1% to 3% harmonic distortion. I consider that unacceptable for amplifiers. Transducer distortion is an entirely different matter and a different discussion.
I want to see THD graphs as a function of power but across at least 5 or more frequencies and at multiple loads - 2 (car audio), 4, 6(optional), & 8 ohms.
Like:
50Hz
100Hz
500Hz (optional)
1KHz
5KHz
10KHz
15KHz (optional)
Lastly, they should also provide the details about power over full bandwidth pink noise with:
1ch
2ch
4ch
All channels driven, or at least the 70/30 for mains and surrounds.
I realize I'm asking for the moon in specs.
Amplifier specs should be based on highest output with the lowest distortion before the knee allowing for headroom .
Agreed
That's a good point.
Agreed! YES!
I would be satisfied if the amplifier was measured from 20hz to 20,000hz and an average from that was taken. BUT, the bigger question is how much I really need. I remember the days when 35 watts per channel was a very popular power point for many listeners such as myself back then.
20/25 years ago comment per Harman Kardon rep when they were selling receivers rated much lower watts at higher prices than consumer brands at the time like Sony, Kenwood, Technics "It is not wattifier it is amplifier". They used to specify how much amp/current their amps/receivers were . And their watt rating were all channels driven with full range . Of course this is before Samsung take over and killing the HK amp business
Thank you, Gene! Agreed that these ratings are mostly useless and need to be researched quite a bit to make sense of. Same goes for power at 8, 4 and 2 ohms.
Curious - could you do a similar video about speaker power ratings? I’ve always been a little perplexed by how to interpret those as well. For example - my speakers are rated for a max. 500W and have a sensitivity of 90db at 1W. I can’t imagine pumping 500W into them. I feel like I’d be deaf way before or they would go up in smoke. I don’t even have 500W available.
What is that relationship between max power handling of a speaker and how much you actually feed it?
For car audio, I just go by the fuse rating times 10. it's usually close enough for me to match to the appropriate speaker. Some of these amplifiers are not coming with fuses making it harder to follow that rule. As for home amps I look at the power consumption its rated for. They all lie, but this don't.
I think we should have an amplifier warbling about 20 Hz while we have an inter-modulated tone you know at various frequencies to see how it can all hold together. Then measure power, etc.. by frequency band.
Below a specified THD for different loads at different frequencies from 20Hz to 20kHz.
this, and AVR manufacturers should rate for 5 channels driven I think. It's just silly for a 9 or 11 channel amp to specify 2 channel only power ratings.
@@Jon-nz3dm that could be because at a given time in any movie scene, there could be as many as 2 or max 3 channels running at full levels. But yes as you say they should also mention values when all channels are driven at different loads(8 & 4ohms).
Great Video Boss.
When you think about actual listening, amps will spend the vast majority of time in the low-power regions of their curves with louder dynamic content being mostly sporadic if you're not really driving the whole system really hard. So most of the time when in the high-power parts of their curves, it's going to be loud highly dynamic sound which should mask a fair amount of the distortion. I guess what I seem to be getting at is amps should be designed with more moderate continuous power but with considerable dynamic power. The question you asked was about how to rate amps, though, and that's a different question. Can you build amps with such a output profile, meaning highly dynamic capability even if more modest continuous? And it would need to be clean dynamic power.
See you around at Audio Advice Live! Sony room is supposed to be even better this year!?
I agree with this ideology completely. No matter the case, it should be standardized and mandatory. EVERY home-purposed A/V receiver and Amplifier should be mandated to report this way. That way we can ALL begin to understand the rating with no "tricks". Shouldn't CEDIA and other similar organizations be working to standardize these types of specifications?
I'm all for the more conservative Rating before the Knee. Anthem and NAD have always produced amplification that is "understated". And their performance is certainly exceptional.
Great vid, 100% headroom is best not peak number. Consumer cheap audio overstates capability, true Hifi gear underates. On a seperate note; NAD M23 listening review!!! Please make a vid on its sound 😁 im a proud NAD M3 owner, love to hear your thoughts on the M23
I think we should abandon the use of a nominal value. It’s like neatly packaged for consumers as a marketing tool, rather than really conveying performance to the consumer. Amplifier Watts/distortion, output resistance, and speaker impedance. Are all stories of undulating dynamic change, not singular specs.
So I feel graphs are what we should use. It’s not the 1970’s where graphs were hand drawn, and magazine ink increased cost. There are no barriers or limitations to using graphs in the expression of amplifier performance.
Lets be real, if someone is such a simpleton that they prefer a singular spec, they also don’t have the nerdy curiosity to understand the singular spec or the context of it’s meaning, and would be better served buying from brand identity or model number tiers.
High fidelity sound is strongly influenced by the distortion, so shopping for really low distortion isn’t served by a singular spec, but by observing the graph. How low the distortion rides through that the whole span before the knee, and it’s consistency is not captured at all in traditional printed ftc specs.
Take this example given in the video, the first graph has elevated distortion in the first few watts relative to later on when it becomes cleaner. The other graph is just low from the start and is basically a flat line all the way out to the knee. So which one has lower distortion at your expected listening levels?
The listening levels may be behind the knee, making the power comparison irrelevant, leaving only the distortion levels to consider. Or the listening levels may draw power beyond the knee, in which case the steep change in a sharp knee may have some audible character that a rolling off knee would not.
So for me, I really want to see the graph, and I feel for others not into the nerdy details, they aren’t even served by seeing a nominal spec, and actually confused by being bothered to think about them.
Great video..👌
Actually, there needs to be two standards set. One at the old standard of 20 to 20k +-1 db with .09% distortion or less, two channels driven (Standard). A second one at peak 20 to 20k +-1 db with 1% or less distortion, two channels driven (Dynamic). If the manufacturer would produce a chart that shows the power using those standards with each additional channel /channels added would be a great selling point.
Hi Gene what are the new speakers in the background?
ua-cam.com/users/shorts6HCsboIu0Bo?feature=share3
I’m an old school broadcast engineer, used to measuring harmonic distortion in 5,000 and 50,000 watt AM broadcast transmitters. Sheesh… talk about problems with power supplies giving out at high levels! But it’s instructive in that the critical performance characteristics are (1) the max output from the power supply before it gives out (causing clipping) and the RISE TIME of the power supply at any output level. It surprises me how little attention is devoted to the INTERMODULATION DISTORTION figures for high powered amplifiers. With everything else in good shape, if the PS can’t restore the DC voltage level to the filter capacitors after a transient, you’ll get INTERMOD-the mixture of two or more frequencies in the program material to produce sum and difference products. This is WORSE than harmonic distortion because at least the harmonics are related (harmonically) to the original sound. Intermod products are not. And what’s more, they tend to fall in the middle of the bandwidth, not the upper reaches, so they mask original frequencies that give the program definition and clarity. General rule: harmonic distortion makes the music sound raspy. Intermod distortion makes it sound muddy.
Before the knee absolutely no doubt!
Thanks for rasing the debate. I think it's simple, specify output power at the rated distortion, so if manufacturer says THD= 0.03% than one should look at max power within that number. Another thought; dynamic power is dynamic; some amps can sustain very high power levels well beyond the FTC 20ms rating; which is short, should be at least be 10x times longer so that a sloppy orchestra can get all the notes in that window for a crescendo. Expressing dynamic power as a linear function by defining desired dynamic power in dB's and then subtract that to arrive a continuous power rating doesn't make sense in my book. True dynamic power (being able to produce short power bursts for say, 300ms) is what makes music sound alive and fast from an amp with headroom.
This has been discussed since back in the 1970's. I don't care whether you have 400 watts per channel or 75 Watts per channel. The quality of sound at the volume you yourself require is more of what I'm interested in. Who runs them at wide open throttle says wow that sounds great.
Absolutely amps should be rated just before the knee. It is an important issue, as it speaks to long term reliability. If you run amps to the clipping range or even close, tube amps excepted, then they will have a short life. I advise keeping 3 db. of heard room. That means if you think you need 100 watts of power, then you need a 200 watt amp. That is the cheapest approach on the long run.
Amplifier power should be rated highly, very highly. There's no such thing as to much, only playing to loud, for your ears, or for your speakers.
I know, I'm being difficult here. But really, with modern Class D amps hardly anyone needs to make any kind of compromise when it comes to amps. We're living in an age where a larger part of your budget than ever before can be directed towards the speakers. And that's a wonderful thing!
I've had this thought before and I think they should state both 2db headroom rating and peak watt rating.
I don't think the audio companies would want their customers to get too smart though. They like it when customers that use
I would say, make is the same as the speaker measurment. If a speaker is rated 200w RMS, then the standard is to rate the peak as 2x that, or 400w.
So if an amp is rated at 200w, it should do a mininum of 400w dynamic, if not, then they shouldn't rate it.Or even better, give the dynamic power, and the rated becomes the /2 of that.
Standards is all we need, and to force companies to give enough data.
Great video Gene, Question I bought the Marantz SR7015 based on your channel, Klipsch Reference R-26FA 5.1 Home Theater Pack, the Klipsch R-41SA as my rear height atmos speakers and Two Subwoofers by Klipsch, and am about to add the M5100X to the Marantz SR7015 . Should I connect the M5100X to the two front tower speakers for when I listen to stereo music two channels, or connect the two front, center, and side surround speakers, and let the Marantz SR7015 drive the Klipsch Reference R-26FA atmos and rear height behind me? Thanks in advance.
Great video as usual! Y the way the Audioholics Intro has great bass, any idea what hz it hits?
RMS ALL CHANNELS DRIVEN AT LOWEST DISTORTION
All manufacturers: RELEASE those THD vs power graphs for different frequencies to the public.
People who don't understand them, or don't want to learn, will always believe whatever you tell them anyway.
RMS watts @1khz 0.1%THD. Dynamic power is just a bonus. I really shouldn't need to use more than 75% of the amplifiers rated power anyway. If I do, then I need an amplifier with more power.
I've been casually looking at amps for months and get more confused as I learn more. I have no idea how to get a good budget friendly amp for a klh model 5.
Creek 4240 or 4230 used. It should run you about $200. And it sounds delightful with plenty of power. It’s hard to beat for the money since vintage Marantz/Pioneer is overpriced.
definitely rated at 2 channels driven and at all channels driven.
What should be or what it will remain? Rhetorical question. It should be max output before the knee at the lowest distortion with max dynamic headroom ALL CHANNELS DRIVEN. That's what it should be, but as long as there is no forced industry standard for providing specs, this is not going to change. Companies will continue to market to the average 80%, which look at watts/channel vs watts/channel. Engineering will tell marketing that it should be spec'd at lower wattage with more dynamic headroom, while marketing will tell Enginering, thanks we will take it from here...advertising more wattage sells more. That's just how it is whether it's right or wrong.
hi your videos are very good, so upgrade your next videos to 4k
All of our produced videos are in 4K. Not the quick ones at my desk due to limitation of streamyard.
we should not assume dynamic headroom, a rating for the continuous power output at 0.1% give me the information that is also the dynamic max. i can then get a the amp that has the watts to get me to the dynamic headroom i need. if we also need a dynamic headroom rating that could be more confusing.
some of the bass i listen to are so close to a sine wave anyways.
Both numbers - power and headroom. Peak power is all that matters.
No, not peak. RMS ALL CHANNELS DRIVEN AT LOWEST Distortion!
Gene, you need to start lobbying for regulations. Honestly. Stop just pondering and start kicking ass.
lets ask HUGO!
Hugo volunteered to be in the 5-year isolation dome to study future colonization of Mars and is unavailable for comment at this time.
Back in the day I used JBL L300 speakers with Carver TFM 55 amps. Mono Bridge 1K per channel at 8 ohms. Head room was about 4 dB. Sounded great. When the mass ring hit the pole piece I knew to turn the volume down. Sounded like someone hit a welding table with a sledge hammer. Speaker was rated 50-100 watts RMS. But 350-400 watts did sound great. Don't think that did the alnico magnet's or my ears any good. I'm 67 now ears stop at about 13K.
2ch driven, 5ch driven, and all channels driven, 0.1%THD, 1kHz, and report dynamic headroom. That is all.
My $0.02, I would prefer a Conservative, medium, and aggressive rating. The Marantz is Agressive while the Classe is Conservative. This could be an industry standard.
If the amp is a Multi-Channel the rating should only list ALL-Channels Driven rating.
If it a Stereo ( 2- Channels it should be rated both channels drive simultaneously from bottom (HZ) frequencies to the highest (KHz) frequency, and has of total harmonic distortion they should list 2 or 3 different points. ( Example: audiophile acceptable THD what ever % is set, and a acceptable average / normal THD, and of course that 1 % THD all channel driven... They should stop that 1 channel driven crap rating....
Any power spec result above max 0,1 % distortion is BS for me!
telling lies about spec's is too common in the audio game; trading standards should get involved.
HiFi & home theater amplifiers are some of the most overpriced products one can purchase. These brands want $1,200 for a 400 watt amp. It's tarded.
Idk, something like the A11 tribute or the audio labs 6000a only get you 50wpc into 8 ohms into 2 channels for 800-1200$. A monolith 1600$ 200wpc x5 channels seems like an okay deal. If you actually need the power that is, and 95% of people don’t for their home theater.
Clean power cost money, for $1200 you most of them only get 100 watts per channel and many times less than that, but what most important for an amplifier is not what it can deliver in 8 ohms, i rather have numbers for 4 and 2 ohms, and here most amplifiers below $3000 cant deliver
@@andreasmoller9798 I run the Polk Audio Reserve line (R700 mains). They're all 8 ohm speakers...so 8 ohm power was very important to me.
...and I forked out the $$$ for solid gear, too. Solid in my opinion anyway. All of my electronic components are Marantz. ...and my amps were overpriced as well.
I have about 30 years of experience in car audio....so I'm not used to spending $1.50 to $2.00 per watt anymore. That died out in 12v back in the early 2000s. Sure, there are a handful of amps that cost a fortune still...but they are made by brands with amazing reputations who've been around for decades...and who are basically "flexing". It's very much a "Hey, look what we can do" type of thing. ...but those same brands have "take down" models that don't cost quite as much. Helix have their Thesis lineup that costs many thousands of dollars...but they also have amps that cost $350-$500.
In the HiFi/Home Theater side of audio, the only budget friendly stuff is either super sketchy ChiFi....or used....or refurb/open box.
...and speaking on the impedance thing one more time in this novel. I am eventually going to set up a separate 2.1 music system in one of my spare bedrooms. That room is getting Arendal 1723 gear, so that'll need 4 ohm power...
There all crap modern amps..
Lol not sure what that means but amplifiers are better today than ever.
Rms 20-20000 all channels driven 1 % at 8 ohms and at 4 ohms if the power is close to double at 4 ohms you know it is a pretty decent amp power wise.
Powerful video. 😅⚡