I'm sorry. It's still not clear if the treble unit has it's own amp or if it shares the same amp as the midrange unit.
How many channels of amplification are you using for each speaker system? Is it 3 each side?
When you quoted your voltage figure for Pano's test, was it on the midrange amp?
Why are the bass units of different size? Is that Left & Right different? What sensitivity?
Krglee, I think to further respond to your repeated questions about my system will not make the debate to progress. I suggest you submit your own data about Power amps, dynamics and sound levels at home to the readers.
forr, the reason for my interest in your system is that I've a lot of experience in designing complex multi-way, multi-amped systems with active analogue & DSP EQ.
If designed properly, these can have substantial advantage in peak handling compared to 'conventional' systems. Rod Elliot has a page with some important stuff that bears on this.
But even without this information about your system, we can conclude certain things.
- It is VERY unlikely that YOU will clip YOUR system cos you play at quite soft levels.
- Pano's little survey test-how-much-voltage-power-do-your-speakers-need.html has 60% of people playing louder than you and 28% MUCH louder such that they would probably find your system inadequate .. probably more as your speakers are quite insensitive. My own survey suggests even more people.
I'm not sure this makes you 'typical' .. especially in this august forum 🙂
As you say, cos the known levels of digital stuff bla bla .. you can tell from this if the user's system will overload. Pano's test evaluates this precisely.
My own info is that I often clip a 50W 8R amp playing commercial Classical CDs to a degree that I find audible & irritating.
When playing commercial Classical CDs, mastertapes and my own recordings, I clip a 200W 8R amp into 90dB/2.83V speakers but not to an annoying extent.
When doing Double Blind Listening Test bla bla on speakers, I clip a 1000W 4R amp sometimes but not usually noticeably.
There are important exceptions to this. 'The Garage Door' will clip the 1000W 4R amp (and State of the Art microphones & 16b recorders of its day) .. and even if the speakers don't burst into flames, will probably generate unacceptable 'distortion'.
You could play 'The Garage Door' with your Volume Control at 7 o'clock and it probably won't clip even your system but then what's the point?
______________
My original comment was that a 50W 8R amp needs to be carefully designed to ensure good overload & recovery performance. This is a view shared by Bob Cordell .. ie this is probably the major audible difference between 50W 8R amps.
Your view was that you NEVER clip your amps so this is irrelevant. I can see that is true for YOU.
But other people, including more than 60% of those in Pano's survey WILL clip a 50W amp.
So if I'm designing a 50W 8R amp, especially for commercial sale, I'd pay close attention to this.
I can see that your answer is to write in the Instruction Manual
- DO NOT CLIP THIS AMP
- USE ONLY > 96dB/2.83V @ 1m speakers
- Use Cordell's Klever clipper
- other impractical stuff that the user will ignore
If that's what rocks your boat and makes you happy .. but please don't criticise other people who haven't got your restraint and may be using their amps for monitoring live recordings.
Richard, these are extraordinarily high levels.
I have a 180 watt amp and 280 watt amp and I'd hate to think what would happen to my ears at 3 o'clock volume. Totally unpleasant listening experience because it would be too loud.
A good test of classic music dynamic range is Mikhail Pletnev's Swan Lake recording.
Speakers are B&W 703's claimed 89 dB/ watt and located in my medium sized listening room.
I have a 180 watt amp and 280 watt amp and I'd hate to think what would happen to my ears at 3 o'clock volume. Totally unpleasant listening experience because it would be too loud.
A good test of classic music dynamic range is Mikhail Pletnev's Swan Lake recording.
Speakers are B&W 703's claimed 89 dB/ watt and located in my medium sized listening room.
Last edited:
Well 50 watts is certainly not good enough for serious listening on the B&W's in my current larger room, but the 15 W sx-Amp still makes a nice sound and in my smaller room in the old house sounded very fine indeed.
Sorry to say that a person saying to have lot of experience is not making me to be much convinced on what he/she can assert. I prefer to rely on facts and data.
My professional digital processor can deliver no more about 7.75 Vrms
at its outputs. These outputs are loaded by a 20 dB attenuator, so my amps never see more than 0.775 V at their input. As they can all handle more than 1 Vrms at their input, they never clip in my system.
Everybody can manage this sort of things on his/her system, (with 2 Vrms maximal for a CD output)
If I needed more SPL, I would change my amps or even my drivers, to always benefit from minimum distortion.
Amps with clean clipping behavior is certainly useful in some circumstances (BTW, I need one) but it can't be considered as a quality criterion in a hifi system for which only the clips of the records should exist.
If very powerful amps clip with 90dB/2.83V/1 m loudspeakers , it means that the drivers are inefficient for the demanding listening level and that they work in conditions not far from extreme with a lot of distortion.
Hi Guys
Music can have "apparent" dynamics that do not relate to true changes of power required. This sonic perception is the result of what instruments are being played, or what sounds are present, as this is an effect used quite a lot in sound tracks for movies and TV shows.
Given the averaging incorporated into most SPL meters, it really is average levels that matter with respect to ear fatigue. Clipped peaks will exacerbate the issue where clean peaks won't. On another thread I described how I measured the SPL in my room while also monitoring signal level to the speakers using a scope. For my system, 1W produces 90dB of sound at 1m. This is too loud even for "blasting". Comfortable loudness is in the 70db range, and I measure loudness at the seating position.
I calibrated my clip meters to light at 1W even though each channel can produce a whopping 8W. The lights never come on.
It never surprises me as how low the SPL is that seems loud. I am, however, dumbfounded at how enamoured many people are of the high-end of SPL tolerance. Our hearing evolved to "survive" brief high-SPL sounds - not to endure them continuously. Taking an SPL meter to a concert would be pointless as the level would be pegged or off-scale all the time, at least for rock.
Has anyone heard how loud a saxophone is? a trumpet? drums? in a real room. leave out electric instruments, as their loudness is pretty much unlimited... but why does anyone think that they should reproduce concert or live sound levels in their living room? It seems like madness to me. I would rather listen to music at a medium level, where the room resonances and reflections will not not interfere with musical intellegibility, and where my hearing is "flattest" inasmuch as loudness and ease of recognition are low energy expenditure processes.
Turn it down. Enjoy music for your whole life.
Have fun
Kevin O'Connor
Music can have "apparent" dynamics that do not relate to true changes of power required. This sonic perception is the result of what instruments are being played, or what sounds are present, as this is an effect used quite a lot in sound tracks for movies and TV shows.
Given the averaging incorporated into most SPL meters, it really is average levels that matter with respect to ear fatigue. Clipped peaks will exacerbate the issue where clean peaks won't. On another thread I described how I measured the SPL in my room while also monitoring signal level to the speakers using a scope. For my system, 1W produces 90dB of sound at 1m. This is too loud even for "blasting". Comfortable loudness is in the 70db range, and I measure loudness at the seating position.
I calibrated my clip meters to light at 1W even though each channel can produce a whopping 8W. The lights never come on.
It never surprises me as how low the SPL is that seems loud. I am, however, dumbfounded at how enamoured many people are of the high-end of SPL tolerance. Our hearing evolved to "survive" brief high-SPL sounds - not to endure them continuously. Taking an SPL meter to a concert would be pointless as the level would be pegged or off-scale all the time, at least for rock.
Has anyone heard how loud a saxophone is? a trumpet? drums? in a real room. leave out electric instruments, as their loudness is pretty much unlimited... but why does anyone think that they should reproduce concert or live sound levels in their living room? It seems like madness to me. I would rather listen to music at a medium level, where the room resonances and reflections will not not interfere with musical intellegibility, and where my hearing is "flattest" inasmuch as loudness and ease of recognition are low energy expenditure processes.
Turn it down. Enjoy music for your whole life.
Have fun
Kevin O'Connor
I tried out my new amp tonight at a bit higher level than before using mixed material from my starred-list on spotify (ranging from Neil Young via Antony Johnson to Tangerine Dream).
Definitely loud enough. Clipping was not considered during design so I do not expect the amp to behave, +- 20V supply (regulated, dipping some under heavy load) and 90 dB speakers. I estimate some 20 W RMS max and 103 dBSPL peak.
Definitely loud enough. Clipping was not considered during design so I do not expect the amp to behave, +- 20V supply (regulated, dipping some under heavy load) and 90 dB speakers. I estimate some 20 W RMS max and 103 dBSPL peak.
FACT
"My own info is that I often clip a 50W 8R amp playing commercial Classical CDs to a degree that I find audible & irritating.
When playing commercial Classical CDs, mastertapes and my own recordings, I clip a 200W 8R amp into 90dB/2.83V speakers but not to an annoying extent.
When doing Double Blind Listening Test bla bla on speakers, I clip a 1000W 4R amp sometimes but not usually noticeably."
This is also FACT
- It is VERY unlikely that YOU will clip YOUR system cos you play at quite soft levels.
- Pano's little survey test-how-much-voltage-power-do-your-speakers-need.html has 60% of people playing louder than you and 28% MUCH louder such that they would probably find your system inadequate .. probably more as your speakers are quite insensitive. My own survey suggests even more people.
I'm already TOTALLY CONVINCED that YOU NEVER clip your system. But please excuse me if I don't recommend any 50W amps that you design 🙂
As a commercial speaker designer in most of my previous life, I'm well aware of this. As I've said, most speakers don't sound any better with a bigger amp than about 200W 8R.
But it IS possible to design speakers that will sound better with bigger amps up to about 1000W 4R and still have 90dB/2.83V sens. I can't vouch for bigger amps cos I've never used them. 😀
Andrew, I used to be an amatuer recording engineer though with some rather good gear.Bonsai said:
Uncompressed classical music has very high peak to mean ratios. The levels that forr quotes from ancient HiFiNews etc are average levels measured with an spl meter. If you record eg a brass instrument or even piano, you'll soon discover that these have VERY asymmetric waveforms and MUCH higher peak to mean levels than you'd think. When you have small groups like excellent choirs, this is even worse.
Most commercial classical recordings are compressed but even then, there are notable exceptions. If the Pletnev Swan Lake is the Waterlily one, I know quite a lot about that from the recording engineer on another forum.
Of course most people don't play stuff anywhere near the correct levels .. but recording engineers try to.
And as I've said, most pop is unbearably loud long before your 50W amp clips.
But to get back to the thread ... the setting of your Volume Control is the most accurate indicator of whether you will clip your amp. Pano's survey is a precise handle on this.
If you have your volume control at 3 o'clock cos the recording as a very low average level, it probably means it has very high peaks which will clip even your high powered system.
When a system is working correctly it has very different subjective characteristics from that of a typical hifi - which is one of the key ways of realising, or registering that it's in the "zone" ...
One is, it always remains subjectively dynamic - you can adjust the volume from the highest undistorting levels down to whisper quiet and the sound always remains engaging, "alive" -- I've had situations where I've literally had to mute the sound because I or my wife had to momentarily strongly turn our attention to completely different matters ... and the content of the music still distracted us too much even when whisper quiet. In other words, the sound doesn't become boring, turn into background wallpaper at low levels.
And another: modern, highly compressed pop becomes extremely intense - subjectively it can only be listened to at very moderate volumes, the richness and impact of the sound is overpowering otherwise. If the volume is then raised a bit your ears start ringing within 5 minutes or so, because the average intensity is too much for the hearing system - you're given a warning by your body to back off ...
One is, it always remains subjectively dynamic - you can adjust the volume from the highest undistorting levels down to whisper quiet and the sound always remains engaging, "alive" -- I've had situations where I've literally had to mute the sound because I or my wife had to momentarily strongly turn our attention to completely different matters ... and the content of the music still distracted us too much even when whisper quiet. In other words, the sound doesn't become boring, turn into background wallpaper at low levels.
And another: modern, highly compressed pop becomes extremely intense - subjectively it can only be listened to at very moderate volumes, the richness and impact of the sound is overpowering otherwise. If the volume is then raised a bit your ears start ringing within 5 minutes or so, because the average intensity is too much for the hearing system - you're given a warning by your body to back off ...
My Pletnev recording is DG recording of the Russian National Orchestra. Fantastic.
Yes, huge dynamic range and this will clip if played at anything like realistic levels even on my big amps - agreed.
Yes, huge dynamic range and this will clip if played at anything like realistic levels even on my big amps - agreed.
I don't want to disappoint you but I can't find any reference to your work with that search term. Did you actually try it?
I want to play with FIR filters but most of the commercial DSP products just have pre-canned versions of standard filters. What's the use of that?
So, to answer your email question in a thread where it seems to fit, I haven't decided the implementation of my new system. Hence my interest in your experience.
Best wishes
David
Powered Integrated Super Sub tech. was the R&D Department's project name for our speaker+amp/analogue/DSP/linear/non-linear/EQ stuff.
For some reason, Marketing never liked the name 😱
Oh! The tech. is/was real and I've tried it all. Simple versions appeared in several commercial products. The DSP stuff was too expensive at the time 🙂
The FIR stuff, although novel at the time is old hat today. There's big advantage if you can specify IIRs at coefficient level.
Simple Arbitary IIRs is novel & presently SOA
Kgrlee,
It is VERY unlikely that YOU will clip YOUR system cos you play at quite soft levels.
I can play a lot louder without clipping the sytem.
Clipping happens in amps when they are undersized towards the demanding listening level from the loudspeakers.
Pano's test results are :
2 volts or less 38.14%
Between 2-5 volts 34.29%
The peak voltage across the voice coil of my medium drivers being above 5.5 V peak (3.9 Vrms), I belong to the above category of 72% of people.
You seem to have a problem with the 60% of people who play louder than you.
You misinterpret Pano's results.
When Mike Renardson asked people which amount of power they wished for the Mosfet amp he was setting up, there was a clear majority prefering 30 W / 8 Ohm.
Most people have comfortably lived with BBC LS5/3a or QUAD Esl which have an SPL limited to around 96 dB.
I plainly agree with Struth's statement :
"comfortable loudness is in the 70dB range"
With the rarely dynamic peak/average of 25 dB, peaks are at 95 dB SPL.
And my system is able to deliver more than 105 dB SPL.
All this makes me think I listen music more at the people average level
than you.
But please excuse me if I don't recommend any 50W amps that you design.
I have not published any amp, I use Renardson's, Self's and Yamaha circuits.
Uncompressed classical music has very high peak to mean ratios. The levels that forr quotes from ancient HiFiNews etc are average levels measured with an spl meter. If you record eg a brass instrument or even piano, you'll soon discover that these have VERY asymmetric waveforms and MUCH higher peak to mean levels than you'd think. When you have small groups like excellent choirs, this is even worse.
Where are your numbers ? Your statements ignore the fundamental concept applying to the recording art : Full Scale.
It is VERY unlikely that YOU will clip YOUR system cos you play at quite soft levels.
I can play a lot louder without clipping the sytem.
Clipping happens in amps when they are undersized towards the demanding listening level from the loudspeakers.
Pano's test results are :
2 volts or less 38.14%
Between 2-5 volts 34.29%
The peak voltage across the voice coil of my medium drivers being above 5.5 V peak (3.9 Vrms), I belong to the above category of 72% of people.
You seem to have a problem with the 60% of people who play louder than you.
You misinterpret Pano's results.
When Mike Renardson asked people which amount of power they wished for the Mosfet amp he was setting up, there was a clear majority prefering 30 W / 8 Ohm.
Most people have comfortably lived with BBC LS5/3a or QUAD Esl which have an SPL limited to around 96 dB.
I plainly agree with Struth's statement :
"comfortable loudness is in the 70dB range"
With the rarely dynamic peak/average of 25 dB, peaks are at 95 dB SPL.
And my system is able to deliver more than 105 dB SPL.
All this makes me think I listen music more at the people average level
than you.
But please excuse me if I don't recommend any 50W amps that you design.
I have not published any amp, I use Renardson's, Self's and Yamaha circuits.
Uncompressed classical music has very high peak to mean ratios. The levels that forr quotes from ancient HiFiNews etc are average levels measured with an spl meter. If you record eg a brass instrument or even piano, you'll soon discover that these have VERY asymmetric waveforms and MUCH higher peak to mean levels than you'd think. When you have small groups like excellent choirs, this is even worse.
Where are your numbers ? Your statements ignore the fundamental concept applying to the recording art : Full Scale.
I won't comment on your points except you might want to look at Pano's results again with care and check your arithmetic. All the facts are there. Make what you will of them.
Another obvious conclusion from your posts is that you have probably never tried to record a good musical performance with simple microphone arrangements.
I don't see much point in continuing this discussion .. but I would appreciate further examples of commercial CDs with good dynamic range .. not that I am suggesting yus guys should play them at any decreed level .. either by me or forr.
No need to do analysis. I'm sure we are all capable of deciding where we set our Volume Control without help from anyone else. Then you can do Pano's test to find out if your system is up to scratch.
If its not, forr will advise on how to improve it. 🙂
Last edited:
Richard, in the more fun category for dynamics, is the soundtrack CD for the "Moulin Rouge" movie. This is a system buster of the first order - with huge level and intensity changes, there would be very few systems that could replay this cleanly at realistic volumes.
Sorry I was unclear. I meant - did you actually try the search you posted?
I don't doubt that the tech. was tried. That's why I want to learn more.
But I can't find any information about it.
Thanks for the link to the IIR paper. Looks excellent.
Now I just have to find a copy.😉
Best wishes
David
Last edited:
It has nothing to do with the clipping limits of an amp at home.
Whatever you record, there is a Full Scale limit.
It belongs to the recorders and sound engineers of the whole record process to decide if their products will reach this limit or not.
It should not happen at home.
If you want to make a CD track a good musical performance with a lot of dynamics, you have no other choice than to have a low average level, let' say -30 dBFS or to change (probably manually to avoid the deleterious work of a compressor) the gain between the quite moments and the loud moments. If you do not, a listener in domestic conditions (which present usually a brackground noise of 50 dB SPL) will either not well perceive the details of the music in the quite moments and/or will find the loud moments unbearable.
The problem of dynamics in the sound reproduction process is quite similar to the one encountered in photo. It you take a photo of a scene with a lot of contrast , you can't hope to reproduce it in a "realistic" manner : shadows will be overcast, too dark, and highly illuminated parts too white, like "burnt" . In both cases, details are lost and the photo is not nice to look at.
However, tricks like additional light in the shadows when taking the picture or manipulations during the lab process (photoshop and the like, nowadays) can reduce the initial overall contrast such that the final result will be very detailed everywhere. Nobody will be disappointed despite the reduction of the initial contrast.
The details make the appearance more lifelike than a so-called "realistic rendition" which is not a claim made by photographers. The introduction of this fallacious notion in sound reproduction is very ancient and is of the pure responsibility of marketing people.
The notion of "best rendition" is the right one, is well known in photography and is only obtained by "cheating" with the reality. Human visual and hearing perceptions work in the same manner. We look for enjoyment, not for a strict realism in reproduction which does not exist.
Note that in both processes, photographic and sound reproductions, the dynamics numbers are comparable. Taking in account the background parasitic values of the environment (ambient light, ambient noise) at the reproduction time :
a very high contrast in projected photography is 1:1000 (3 in log)
a very high dynamics for a sound record system is 1:1000 (3 in log, +/-30 dB).
Me too. I do not want to be alone to propose examples.
Freeware tools have already proposed to quantify the dynamic ranges.
- Status
- Not open for further replies.