Nania,
If the stereo pair of speakers are playing the same sound, the SPL's will add or subtract depending on the relative distance from each speaker. Equal-distance between the two speakers, the SPL will add to double which is equal to 3db increase in SPL.
Your example speaker has a sensitivity of 80db/Watt/m. That means if you input 1Watt the speaker will produce 80db SPL at 1m (on-axis). Your example amp is capable of 0.5Watt output. If you feed 0.5Watt to your speaker, it will produce half what it would produce if you put 1Watt in. Half power in (0.5W) is equal to -3db which will produce -3db SPL or 77db at 1m.
This is sound recording for movies. This will be severely compressed so as to eliminate dynamic range problems farther down in the production chain. The recordings I spoke of were special recording that minimized the compression constraints of the average record pressing which were limited to 20 to 30 db of dynamic range. There are recordings where the masters were cut in the studio in real time from the live performance (Direct-to-Disk).
Look at these sites:
http://hometown.aol.com/motionizer/page28.html
http://www.nanophon.com/audio/dynrange.pdf
Rodd Yamashita
If the stereo pair of speakers are playing the same sound, the SPL's will add or subtract depending on the relative distance from each speaker. Equal-distance between the two speakers, the SPL will add to double which is equal to 3db increase in SPL.
Your example speaker has a sensitivity of 80db/Watt/m. That means if you input 1Watt the speaker will produce 80db SPL at 1m (on-axis). Your example amp is capable of 0.5Watt output. If you feed 0.5Watt to your speaker, it will produce half what it would produce if you put 1Watt in. Half power in (0.5W) is equal to -3db which will produce -3db SPL or 77db at 1m.
This is sound recording for movies. This will be severely compressed so as to eliminate dynamic range problems farther down in the production chain. The recordings I spoke of were special recording that minimized the compression constraints of the average record pressing which were limited to 20 to 30 db of dynamic range. There are recordings where the masters were cut in the studio in real time from the live performance (Direct-to-Disk).
Look at these sites:
http://hometown.aol.com/motionizer/page28.html
http://www.nanophon.com/audio/dynrange.pdf
Rodd Yamashita
Roddyama
I checked out those sights and they don't refute what I have heard from sound engineers. Yes, the media is capable of greater dynamics but not the process. Here is the "dope from the dog":
The best microphones can capture 150dB of dynamic range only at the mic. You must subtract a few dB for line loss to the recording device. In an extreme example of a large concert hall with an orchestra playing a huge dynamic burst out of silence, the noise floor will still be steady at 55dB. This leaves you with around 90dB in a best case scenario where the orchestra can create a sound level that is an ear piercing 150dB. Here is where it gets complicated. The "clean" 90dB that remains has nusiance spikes above the floor also known as the "creak factor" that must be mitigated along with other electro-mechanical noise. More importantly, there is the desire to recreate the "palpable presence" by positioning the sound within the space. To create the separation the engineer uses compression. Not using enough compression will favor dynamics by sacrificing presence. The standard is a compression factor of 5 which leaves the mastered recording (no matter what the media is capable of) to 90/5 or 18dB of dynamic range. Some daring engineers get crazy and try to use 4.5 but a compression factor of 4 is unheard of in modern recording because of the sacrifices it inevitably brings about. So even in your example of direct to disk (which I'm told still requires a mixing board which introduces its own comprimises) your numbers for power requirements in a typical listening room seem whack. I would love to hear someone with authority set this one straight because there is definitely a contraversey here.
Video engineers also contend with audio and some would argue that recreating realistic sound is an even greater challenge in video because of the intense scrutiny that can be focussed on a single sound out of space, like a set of keys dropping on a table or a single person talking in a room. Here the dynamic of the sound conveys the emotion just as importantly as with music.
I checked out those sights and they don't refute what I have heard from sound engineers. Yes, the media is capable of greater dynamics but not the process. Here is the "dope from the dog":
The best microphones can capture 150dB of dynamic range only at the mic. You must subtract a few dB for line loss to the recording device. In an extreme example of a large concert hall with an orchestra playing a huge dynamic burst out of silence, the noise floor will still be steady at 55dB. This leaves you with around 90dB in a best case scenario where the orchestra can create a sound level that is an ear piercing 150dB. Here is where it gets complicated. The "clean" 90dB that remains has nusiance spikes above the floor also known as the "creak factor" that must be mitigated along with other electro-mechanical noise. More importantly, there is the desire to recreate the "palpable presence" by positioning the sound within the space. To create the separation the engineer uses compression. Not using enough compression will favor dynamics by sacrificing presence. The standard is a compression factor of 5 which leaves the mastered recording (no matter what the media is capable of) to 90/5 or 18dB of dynamic range. Some daring engineers get crazy and try to use 4.5 but a compression factor of 4 is unheard of in modern recording because of the sacrifices it inevitably brings about. So even in your example of direct to disk (which I'm told still requires a mixing board which introduces its own comprimises) your numbers for power requirements in a typical listening room seem whack. I would love to hear someone with authority set this one straight because there is definitely a contraversey here.
Video engineers also contend with audio and some would argue that recreating realistic sound is an even greater challenge in video because of the intense scrutiny that can be focussed on a single sound out of space, like a set of keys dropping on a table or a single person talking in a room. Here the dynamic of the sound conveys the emotion just as importantly as with music.
Nania,
This is a quote from the following link.
http://georgegraham.com/compress.html
I must be brief as I'm at work.
Rodd Yamashita
This is a quote from the following link.
http://georgegraham.com/compress.html
I must be brief as I'm at work.
Rodd Yamashita
...oh and I forgot to mention that even if you could get 20dB of good dynamic range out of your cd you would still have to contend with the compression the sepaker will create in a listening room at loud levels. This is yet another reason why you can't play too loud and still have musical integrity. Its starting to look like excessive watts are for the pazze. I have been told that means crazy in Italian.
Nania,
Here is another reference from a rather authoritative source (JAES).
http://www.bbc.co.uk/rd/pubs/papers/paper_23/paper_23.html
Rodd Yamashita
Here is another reference from a rather authoritative source (JAES).
http://www.bbc.co.uk/rd/pubs/papers/paper_23/paper_23.html
Rodd Yamashita
roddyama
Once again, I believe this is only a reference to the capability of the media.
How did you derive this calculation? The compression algorithm divides the raw dynamic range by its factor. 90/5=18 and 90/4=22.5 and that is a 4.5dB difference on a fixed noise floor which will exaggerate the percentage of noise to the music signal. The explanation I placed in this thread are confirmed by two music engineers who own successful recording studios that are in operation today. It seems peculiar to negate their experience with your (admitted) approximation. I think you are reading more into the links you have posted than is being said or the links you have provided are theoretical and have no basis describing real audio production. Anyone else with other evidence? I think this is important since many of us want to know how much amp to build.
Nania,
It doesn't appear that anyone is too eager to chime into this thread now that it's starting to address your original question so I'll keep going. I've been at this hobby for 25 years and would be a bit disappointed to find one of my system requirements were based on bad information.
I will, however, concede to your friends experience on the issue of compression ratio, having no first hand experience in that area, but I have been looking for information (obviously). Below is a link to the Rane web site that explains the operation of a signal compressor. You are correct about the ratio and how it is a ratio in db, but the link indicates that the compression does not start to compress the signal until the signal level reaches a "threshold". So you must add what left of the signal after compression to the threshold level (also in db). You may have noticed that the recording industry always references to a 0db level. This reference level is some average recording level above the noise floor. I'm getting the sense that this level is about 20db to 40db above the noise floor, but that needs to be verified. The threshold would then be a level above and below this 0db reference. If, for example, the threshold was 10db and the compression ratio was 5/1 and you were recording music with a dynamic range that extended 30db above 0db and 20db below, the dynamic range of the output would be 10db + 10db + 6db + 4db = 30db (Threshold above + threshold below + compressed above + compressed below = total dynamic range).
Ask your friends how that sounds. here the link:
http://www.rane.com/pdf/note141.pdf
Rodd Yamashita
It doesn't appear that anyone is too eager to chime into this thread now that it's starting to address your original question so I'll keep going. I've been at this hobby for 25 years and would be a bit disappointed to find one of my system requirements were based on bad information.
I will, however, concede to your friends experience on the issue of compression ratio, having no first hand experience in that area, but I have been looking for information (obviously). Below is a link to the Rane web site that explains the operation of a signal compressor. You are correct about the ratio and how it is a ratio in db, but the link indicates that the compression does not start to compress the signal until the signal level reaches a "threshold". So you must add what left of the signal after compression to the threshold level (also in db). You may have noticed that the recording industry always references to a 0db level. This reference level is some average recording level above the noise floor. I'm getting the sense that this level is about 20db to 40db above the noise floor, but that needs to be verified. The threshold would then be a level above and below this 0db reference. If, for example, the threshold was 10db and the compression ratio was 5/1 and you were recording music with a dynamic range that extended 30db above 0db and 20db below, the dynamic range of the output would be 10db + 10db + 6db + 4db = 30db (Threshold above + threshold below + compressed above + compressed below = total dynamic range).
Ask your friends how that sounds. here the link:
http://www.rane.com/pdf/note141.pdf
Rodd Yamashita
roddyama
As with any art, there are several ways to achieve the goal of recreating the event but what seems to be dogma is that with the current state of the technology you cannot record greater than a 20dB sound dynamic and make it sound true. One of the ways that sound engineers manipulate the signal to noise is by using different effects on different mixing channels but even that has its foibles because of the limitations of the mixer so we are back to that 20dB limit.
Nania,
You see from the Rane site that modern mixing boards have much more than 20db of dynamic range.
Below are two links to the Digital Domain web site with articles by Bob Katz. These are very comprehensive discussions on professional recording practices and standards WRT dynamic range, dynamic headroom, level, and other aspects of the recording process.
The 0VU level, as set by THX standards is 83dbSPL. You were correct about the engineers being limited to a 20db window, but not dynamic range. This 20db is the practical limit for the “peak to average” ratio, where the average level is the 0VU level. This means the engineer, in order to record to THX standards, must set their equipment so that 0VU equals 83db from their monitors. They should have the necessary dynamic headroom above 0VU, which the article states as typically being about +14db, but is limited by the 0dbFS (Full Scale, signal clipping point) point. The 0dbFS point is the 20bd window mentioned above.
The dynamic range was never stated in any of some 50 sites I looked at. The question is, if 83db = 0VU = the average level, how far below 0VU does the music extend?
I found the following Stereophile Magazine site with a review of a special System Testing CD where they did a spectral analysis of the CD. They found that a guitar note, recorded on the CD had a peak-to-average ratio of 11.8db (within the “window”), but there were harmonics recorded on the CD at –64db. That equates to a dynamic range of more then 75db!
Here are the links.
http://www.digido.com/integrated.html#anchor3646434
http://www.digido.com/compression.html
http://www.stereophile.com/showarchives.cgi?338
Rodd Yamashita
You see from the Rane site that modern mixing boards have much more than 20db of dynamic range.
Below are two links to the Digital Domain web site with articles by Bob Katz. These are very comprehensive discussions on professional recording practices and standards WRT dynamic range, dynamic headroom, level, and other aspects of the recording process.
The 0VU level, as set by THX standards is 83dbSPL. You were correct about the engineers being limited to a 20db window, but not dynamic range. This 20db is the practical limit for the “peak to average” ratio, where the average level is the 0VU level. This means the engineer, in order to record to THX standards, must set their equipment so that 0VU equals 83db from their monitors. They should have the necessary dynamic headroom above 0VU, which the article states as typically being about +14db, but is limited by the 0dbFS (Full Scale, signal clipping point) point. The 0dbFS point is the 20bd window mentioned above.
The dynamic range was never stated in any of some 50 sites I looked at. The question is, if 83db = 0VU = the average level, how far below 0VU does the music extend?
I found the following Stereophile Magazine site with a review of a special System Testing CD where they did a spectral analysis of the CD. They found that a guitar note, recorded on the CD had a peak-to-average ratio of 11.8db (within the “window”), but there were harmonics recorded on the CD at –64db. That equates to a dynamic range of more then 75db!
Here are the links.
http://www.digido.com/integrated.html#anchor3646434
http://www.digido.com/compression.html
http://www.stereophile.com/showarchives.cgi?338
Rodd Yamashita
Hmmm. Its alot to read but I suspect you have misinterpreted what is being calculated and what is resultant on the media. We agree that for all practical purposes there is a maximum "raw signal" of around 90dB so it is entirely possible that the numbers you are quoting are "pre-production" and not what is on the final recording. I think what is being described is the process of setting a maximum spL of 103dB with an 83dB sound floor set to 0dB. The real question now is: What is actually meant by the "peak to average" reference if not to describe the dynamic range?
Nania,
The article on level has a good explenation of "peak-to-average". Also at the end of the article on compression, there is a list of recordings that Katz considers to be exceptional for their dynamic quality. Each of these recording were recorded with 0VU greater then 75db.
Since we agree on the headroom limit of less then 20db, the question still remains, how far below 0VU does the recorded music extend? My guess is somewhere between you and I.
The link to Stereophile give an indication of this, but I must admit, the recording was not through a mixer and it was only a single note.
I, like you was hoping that this was a more important issue to others. I still believe it is.
Rodd Yamashita
The article on level has a good explenation of "peak-to-average". Also at the end of the article on compression, there is a list of recordings that Katz considers to be exceptional for their dynamic quality. Each of these recording were recorded with 0VU greater then 75db.
Since we agree on the headroom limit of less then 20db, the question still remains, how far below 0VU does the recorded music extend? My guess is somewhere between you and I.
The link to Stereophile give an indication of this, but I must admit, the recording was not through a mixer and it was only a single note.
I, like you was hoping that this was a more important issue to others. I still believe it is.
Rodd Yamashita
Nania,
I believe that the "monitor gain" setting is the 0DBVU setting of the monitor during the recording process. That would be set at a level the same number of db above the noise floor. The actual dynamic range would be the -XXdb (below 0VU = to the monitor gain setting) that they managed to record. They do not give this number. This is the missing piece of information that we need to determine the real dynamic range of the recording.
Rodd Yamashita
I believe that the "monitor gain" setting is the 0DBVU setting of the monitor during the recording process. That would be set at a level the same number of db above the noise floor. The actual dynamic range would be the -XXdb (below 0VU = to the monitor gain setting) that they managed to record. They do not give this number. This is the missing piece of information that we need to determine the real dynamic range of the recording.
Rodd Yamashita
This was an EXCELLENT thread years ago.
I am curious what your opinions are about it after 3 years.
As long as I measured "dynamic headroom", in my system, with all kind of recordings (SACD, DVD-A, CD, DVD-V, LPs) never ever went over 30db.
It seems that there is a confusion between SPL and dynamic range. SPL can go to 120db. But show me a recording with a change in dynamic range of 80db and I will eat it.
I am curious what your opinions are about it after 3 years.
As long as I measured "dynamic headroom", in my system, with all kind of recordings (SACD, DVD-A, CD, DVD-V, LPs) never ever went over 30db.
It seems that there is a confusion between SPL and dynamic range. SPL can go to 120db. But show me a recording with a change in dynamic range of 80db and I will eat it.
djk said:
I was researching possible upgrades/modifications to my NAD 2600A, and came across this gem of a thread!
I was listening to some of my favourite music just now, got a bit carried away with the volume knob, and suddenly no bass! Looked at the amp rack, the red 'protection' light on the NAD was 'ON'. (2600 bridged mono, driving my 8 ohm 15 inch sub). Quickly turned the volume down, and the light went off after a few seconds, and everything was fine. I think i just found the limit of the mighty 2600!
Was reading the earlier posts about the NAD power envelope circuit, with dual voltage rails, and how the PSU sags when loaded. So the 150W RMS for the 2600 is from the lower voltage rails?
How does the protection circuit come in for the 2600? When the PSU voltage rails sags excessively, or due to output current limiting?
Is there any way to mod the 2600 to improve the performance? IE, what is the weakest link now? More capacitance in the power supplies? Larger transformer? Or it is limited by the output transistors?
I'm not looking to up the continuous power rating, more like the dynamic headroom, so i was thinking along PSU capacitors upgrade. Currently, think there are 2x 10,000uF 120V and 2x 10,000uF 80 V capacitors.
Thank for the input and comments!!
This thread was a good discussion, a bit tedious at times but you have to accept that everyone sees things different from you. I haven't posted in years, and I am no longer an active DIY'er. And I know this is probably irrelevant to you now, but maybe some others can get something from my reply. I'll reply as if it is still relevant.
Been there, done that a few times. I used my 2600 for my stereo subs (stereo TAD 1601a), and tripped the protection circuit a few times. You can image the SPL I was getting at the time (150W, 97db/W X 2) The SPL was over 120db.
Yes. This is the advertised continuous rated power capability of the 2600. I think it was somewhere around 360W bridged. (I think someone posted the bridged power spec in an earlier post)
I don't know for sure, but I would suspect the protection circuit is triggered by the a voltage threshold on the upper rail charging a timer cap.
I don't know this either. There may be a way, but when I triggered the protection circuit, once I realized I had over driven the sub amp, but it was still ok, my next concern was my mid and upper range speakers/amps. My upper range were horns, and they were super efficient with a 50W AR D52, so I probably wasn't over driving the horns. But my mids were 2 JBL LE10a's, and are pretty inefficient. Even though I was running them with a 200W/ch Bryston, when I tripped the NAD2600 protection circuit, I was probably clipping the Bryston. My solution at the time was to not play so loud.
I think you need more continuous power (different amp), and/or more efficient speakers, and/or stereo subs. My preference would be stereo efficient subs. Also, for subs, and depending on your current sub placement in the room, you can get a boost in efficiency by corner loading (put them in the corner of the room).
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.