Hi Pano
There is still a niggling little issue with the title of this thread and it has to do with the difference between what is required to reproduce a signal, what is present on an oscilloscope display and what “a meter” reads.
At least at the signal requirement end, this can be explored without test equipment by using “a meter” which shows what is actually present as well as integrated over a short time like a Volt or sound level meter does.
ORBAN Loudness Meter
This free signal monitoring software displays several different integration's of the audio signal as well as the instantaneous peak values, all of which are needed to reproduce the signal in it’s entirety.
You can see for whatever kind of music you choose (and it’s highly variable) how large the short peaks are compared to the short average level like a “fast” sound level meter would read or as your ears judge as “subjective loudness”. One can reach the same average loudness without the short peaks (when present) and it sounds fine, but when you A vs B compare, the version with the peaks sounds more dynamic than without. It’s funny how often one can find instantaneous clipping on an amplifier output if you look with an oscilloscope as any problem in gain structure shows up at the end of the chain.
What’s new out your way? I have moved since we talked last, in Georgia now nearer the shop since things started taking off.
Best,
Tom
There is still a niggling little issue with the title of this thread and it has to do with the difference between what is required to reproduce a signal, what is present on an oscilloscope display and what “a meter” reads.
At least at the signal requirement end, this can be explored without test equipment by using “a meter” which shows what is actually present as well as integrated over a short time like a Volt or sound level meter does.
ORBAN Loudness Meter
This free signal monitoring software displays several different integration's of the audio signal as well as the instantaneous peak values, all of which are needed to reproduce the signal in it’s entirety.
You can see for whatever kind of music you choose (and it’s highly variable) how large the short peaks are compared to the short average level like a “fast” sound level meter would read or as your ears judge as “subjective loudness”. One can reach the same average loudness without the short peaks (when present) and it sounds fine, but when you A vs B compare, the version with the peaks sounds more dynamic than without. It’s funny how often one can find instantaneous clipping on an amplifier output if you look with an oscilloscope as any problem in gain structure shows up at the end of the chain.
What’s new out your way? I have moved since we talked last, in Georgia now nearer the shop since things started taking off.
Best,
Tom
Hey Tom, I've moved again too. Now much closer to Georgia. In fact I'll be flying over Georgia a dozen or so times this month. 
Some day I'll get by the shop. Meanwhile I follow DSL on Facebook.
About the test. There is no problem with measuring and calculating. Since a digital signal has an easy to determine maximum level, every other level must be below that. My test signal is at 12dB RMS below the maximum of full scale. If you measure the voltage of the test signal, you will know - that as long as you don't twist the volume knob - the peak voltage will never be more than 12dB above what you measured. That's as high as the signal can ever go.
Where the only difficulty lies is in picking that volume setting. You might run your system "As loud as it ever goes", but be wrong. For example you might pick a highly compressed music track that sounds super loud with the volume knob at 9:00. But if you played a classical track you might turn it up to 1:30 (past 11!) Your measurement of the test signal would be valid for compressed tracks that you play at 9 o'clock on the dial, but not for the classical that you set higher.
That's why there has been a good bit of discussion on picking a track to set the max level. That said, most people know where the volume knob is generally set, and go a little beyond that for the test. As long as the volume is not pushed past that point, you can be sure that the amp peak voltage will never be more than 12dB (~4X) higher than the test signal. There is no higher level on digital.
Some day I'll get by the shop. Meanwhile I follow DSL on Facebook.About the test. There is no problem with measuring and calculating. Since a digital signal has an easy to determine maximum level, every other level must be below that. My test signal is at 12dB RMS below the maximum of full scale. If you measure the voltage of the test signal, you will know - that as long as you don't twist the volume knob - the peak voltage will never be more than 12dB above what you measured. That's as high as the signal can ever go.
Where the only difficulty lies is in picking that volume setting. You might run your system "As loud as it ever goes", but be wrong. For example you might pick a highly compressed music track that sounds super loud with the volume knob at 9:00. But if you played a classical track you might turn it up to 1:30 (past 11!) Your measurement of the test signal would be valid for compressed tracks that you play at 9 o'clock on the dial, but not for the classical that you set higher.
That's why there has been a good bit of discussion on picking a track to set the max level. That said, most people know where the volume knob is generally set, and go a little beyond that for the test. As long as the volume is not pushed past that point, you can be sure that the amp peak voltage will never be more than 12dB (~4X) higher than the test signal. There is no higher level on digital.
100mVac is roughly 29dB below 1W into 8ohms.
That makes your speaker output ~ (96-29) = 67dB @ 1m
are you listening at 500mm or at 2500mm?
How quiet can you tolerate?
The distance is about 3 meters, as shown here (Dallas II with FE06En).
https://www.youtube.com/watch?v=uYogOGWotvk
I'll redo the measurement with a sinus if you wish (play at normal volumes, then replace the signal with a 1000 Hz sinus, then measure the voltage ). Because of our apartment situation, I don't listen extremely loud (for me 1W with these speakers is about the limit). The Dallas II is incredibly detailed even at these low volumes. BTW, at 1000 Hz, they are 12 Ohm .
https://www.youtube.com/watch?v=uYogOGWotvk
I'll redo the measurement with a sinus if you wish (play at normal volumes, then replace the signal with a 1000 Hz sinus, then measure the voltage
). Because of our apartment situation, I don't listen extremely loud (for me 1W with these speakers is about the limit). The Dallas II is incredibly detailed even at these low volumes. BTW, at 1000 Hz, they are 12 Ohm .
I tend to listen at low levels like you, I like listening at night when there is less ambient noise and I can reduce the volume even further. Low volume reveals more detail especially as your hearing adjusts. I measure around 150mV when listening like this and an added bonus is a class AB amp is in class A most of the time at these levels
The absolute level of the playback file doesn't matter, it will vary between all users anyway depending what you play it on. Its the ratio of the files level to your chosen volume setting that makes this work so well.
Just pick your favourite music tracks and play them at the loudest you would ever normally use.
Now without touching the volume control play the test file and note the voltage levels you measure.
You must use the same device to play back the test file and the music file though.
(the zipped files just need unzipping in Windows. Right click the file and select 'extract')
Just pick your favourite music tracks and play them at the loudest you would ever normally use.
Now without touching the volume control play the test file and note the voltage levels you measure.
You must use the same device to play back the test file and the music file though.
(the zipped files just need unzipping in Windows. Right click the file and select 'extract')
Attachments
You have to set the initial level by playing music using the same source component as the wav file. Its no good playing a CD on a CD player and using a mobile phone or laptop for the test wav file (or vice versa). Both have to be played on the same device so that the ratio of the levels is consistent.
Yes, you set the level by playing a loud and dynamic piece of music at as high a volume as you would ever dare to use. Then without altering the volume control play the wav file and record the AC voltage across the speaker.
Yes, you set the level by playing a loud and dynamic piece of music at as high a volume as you would ever dare to use. Then without altering the volume control play the wav file and record the AC voltage across the speaker.
I'm not sure on that. It really needs to be a music track played using the same configuration as will be used for the test file. Streaming something and then playing a media file (the wav file) might not give correct results... the playback gains (I can't think what else to call it) could be different streaming compared to media playback.
You can certainly try it but the results might be a little off although probably not massively.
You can certainly try it but the results might be a little off although probably not massively.