Hi Pano, the thread about 'are speakers linear' had me thinking a bit more about peak vs rms, and how it relates to this test.
When you told me that Goldwave can calculate/measure the rms of a track of music, an idea hit me how to maybe make this great test even more precise and more telling.
Because I've always had a problem with how well the rms voltage of a -12dBFS sine wave served as proxy for the average rms of a music track.
I think the accuracy of the test rests on this...
It seems to me that when we find the loudest level we think we will ever listen at, that we should take the track used and let Goldwave calculate the rms....
And that value is what we then use for the sine wave test, instead of -12dBFS.
I guess if we used pink noise for the track to find our maximum listening level, the -12dBFS works automatically (assuming the pink has a 12dB crest factor).
But if we use music, it seems to me that using a calculated average rms from the track itself (-x DBFS from Goldwave) as the (-x dBFS) value for the sine wave we then use to measure voltage, .... will give us readings that more accurately reflect both the average rms level and the peak levels we need to have headroom for.
What do you think?
It's undoubtedly making a simple enlightening test more complicated...perhaps unnecessarily so..but i hope to get to Goldwave and give it a try...
Sorry if all this has previously appeared in the thread somewhere...i figure most likely it has...just couldn't wade through all the pages..
And sorry for not having even tried Goldwave yet...
When you told me that Goldwave can calculate/measure the rms of a track of music, an idea hit me how to maybe make this great test even more precise and more telling.
Because I've always had a problem with how well the rms voltage of a -12dBFS sine wave served as proxy for the average rms of a music track.
I think the accuracy of the test rests on this...
It seems to me that when we find the loudest level we think we will ever listen at, that we should take the track used and let Goldwave calculate the rms....
And that value is what we then use for the sine wave test, instead of -12dBFS.
I guess if we used pink noise for the track to find our maximum listening level, the -12dBFS works automatically (assuming the pink has a 12dB crest factor).
But if we use music, it seems to me that using a calculated average rms from the track itself (-x DBFS from Goldwave) as the (-x dBFS) value for the sine wave we then use to measure voltage, .... will give us readings that more accurately reflect both the average rms level and the peak levels we need to have headroom for.
What do you think?
It's undoubtedly making a simple enlightening test more complicated...perhaps unnecessarily so..but i hope to get to Goldwave and give it a try...
Sorry if all this has previously appeared in the thread somewhere...i figure most likely it has...just couldn't wade through all the pages..
And sorry for not having even tried Goldwave yet...
The test track doesn’t serve as a proxy for music
It is purely a way of injecting a known signal to the input, then the user can measure the output and determine the system gain. Since the input is digital and has a hard upper limit, there is a hard limit to the output and this is the maximum voltage that the system can provide at that gain setting
-12dBFS on the test track was cleverly chosen to limit the power to the speakers during the test and simply calculating the maximum power required, it has nothing to do with the average RMS of music
Brian
Hi Brian, yep I get all that...and agree the -12dBFS was cleverly chosen.
If we are using the test to get a good picture of the rms and peak levels our amps are pulling, I think it fullfills its purpose very well.
But I think the closer the -12dBFS is a proxy to the average rms of music, the better simply adding 12 dB will be at determining the peak maximum power required.
And that the hard limit imposed is really just an arbitrarily limit (again rather cleverly chosen) based on the validity of the -12dBFS sine representing the rms of the music.
If we are using the test to get a good picture of the rms and peak levels our amps are pulling, I think it fullfills its purpose very well.
But I think the closer the -12dBFS is a proxy to the average rms of music, the better simply adding 12 dB will be at determining the peak maximum power required.
And that the hard limit imposed is really just an arbitrarily limit (again rather cleverly chosen) based on the validity of the -12dBFS sine representing the rms of the music.
I understand where you're coming from but there isn't any need to know the RMS of the music for this test. In this test we are just trying to calculate the highest voltage level you'll normally use.
Now, if you want to know what RMS and peak voltages are going to your speakers on a particular song, you could figure that out using a voltmeter and GoldWave. That would be useful and I can show you how to do it.
FYI an average level of -18dBFS was a common CD mastering level. Classical has a lower average level, pop/rock higher. Almost all tracks hit 0dB. A few quiet classic tracks don't. -12dB would be loud, but a lot of recent modern mastering is there, or worse.
Now, if you want to know what RMS and peak voltages are going to your speakers on a particular song, you could figure that out using a voltmeter and GoldWave. That would be useful and I can show you how to do it.
FYI an average level of -18dBFS was a common CD mastering level. Classical has a lower average level, pop/rock higher. Almost all tracks hit 0dB. A few quiet classic tracks don't. -12dB would be loud, but a lot of recent modern mastering is there, or worse.
Hi Pano, sorry for the slow return reply and thx for the continued offer to help.
Thought I'd download Goldwave and do a little homework before asking you to show me how. But aargh...I've never really used any wave editors before and unless it's gonna be real easy, not sure I should ask for your help.
I would be cool though...I've done alot of work on limiters..setting rms, peak, and even instantaneous peak via a scope. That work has made it clear for me it's all about waveform structure.
So if it's something you can show me how do do really easy..yes, pls.
Otherwise, I'll just say thx again for the offer.
Oh, and yep, familiar with -18dBFS. I have a piece of Apogee studio gear, and even the x-32 mixers I have are set for +4dBu output to meter at -18dBFS.
Getting those levels/equivalencies firmly in my mellon was a big part of the limiter work. I think I'll always struggle with levels Lol
Thought I'd download Goldwave and do a little homework before asking you to show me how. But aargh...I've never really used any wave editors before and unless it's gonna be real easy, not sure I should ask for your help.
I would be cool though...I've done alot of work on limiters..setting rms, peak, and even instantaneous peak via a scope. That work has made it clear for me it's all about waveform structure.
So if it's something you can show me how do do really easy..yes, pls.
Otherwise, I'll just say thx again for the offer.
Oh, and yep, familiar with -18dBFS. I have a piece of Apogee studio gear, and even the x-32 mixers I have are set for +4dBu output to meter at -18dBFS.
Getting those levels/equivalencies firmly in my mellon was a big part of the limiter work. I think I'll always struggle with levels Lol
That's a handy one. But for some reason it does not list the RMS value of the track. For that you need to go to EFFECTS|MATCH VOLUME. It will scan the track or selections and give you the RMS value. Do NOT hit apply, you don't want to change anything, just look at it. If you do change the levels, just EDIT|UNDO or Control Z. No harm.
The icon for this tool is a little blue speaker with the = after it.
The icon for this tool is a little blue speaker with the = after it.
Attachments
From here you can calculate how much voltage RMS is going to your speakers in a song.
Once you've done the 120Hz sine test posted in the beginning of the thread, you can calculate voltage of your song. Normally it will be 3-6dB lower than the test tone. And now you have the tools to check that more closely.
Once you've done the 120Hz sine test posted in the beginning of the thread, you can calculate voltage of your song. Normally it will be 3-6dB lower than the test tone. And now you have the tools to check that more closely.
Hi All, I'm a little late to the test seeing as I first started reading this thread in 2017, Lol.
Here are my numbers:
Listening Room - 14'W x 24'L x 8'H
Speakers - JBL L150 circa 1979 8ohm 88db/1W/M
Sitting Position - 10' Away from Speakers
Aleph J DiYAudio Clone
120 Hz Test Signal Reading = 1.9V to 3.5V Max Depending on my mood or song
If the calculations are correct (3.5 * 3.5=12.25W) the Aleph J and my Diy M2 at 25watts/ch are more than sufficient for my listening levels.
Here are my numbers:
Listening Room - 14'W x 24'L x 8'H
Speakers - JBL L150 circa 1979 8ohm 88db/1W/M
Sitting Position - 10' Away from Speakers
Aleph J DiYAudio Clone
120 Hz Test Signal Reading = 1.9V to 3.5V Max Depending on my mood or song
If the calculations are correct (3.5 * 3.5=12.25W) the Aleph J and my Diy M2 at 25watts/ch are more than sufficient for my listening levels.
Hi CdDb, thanks for your reply. Admittedly I haven’t read through the whole thread and there are parts and explanations that I have missed. I just “remembered” this thread because it was brought up in a recent thread somewhere else in the PassLabs Forum. To me, my system doesn’t sound strained or distorted at any point, which obviously doesn’t mean it isn’t, and the 3.5V is not the average at what I listen to, it’s mostly in the 2.2 range.
If you can expand on your reply so I can better understand what you wrote that would be greatly appreciated!
Thanks!
Thanks for your reply. What you’ve now explained falls in line with my observation that I didn’t feel I was running into distortion or clipping.
Overthinking it all, and doing/reading other things had me understanding what you wrote backwards, not sure why. Time for some fresh air!!
Thanks again!
Since low tones files were put in first post. Out of curiosity I checked how much voltage I am getting at speaker terminal. I also checked 120 & 220hz to compare. For 120 hz I am getting 1.3v For 220hz I am getting 1.5v. It is OK I think. No Problem
Now with low tones I am getting 0.1, 0.14, 0.24, 0.38 and 0.54v for five tones of the file. Low frequencies at same volume level I should get more volts. Isn't it? Why I am getting low volts compared to 120hz ?
thanks and regards.
Now with low tones I am getting 0.1, 0.14, 0.24, 0.38 and 0.54v for five tones of the file. Low frequencies at same volume level I should get more volts. Isn't it? Why I am getting low volts compared to 120hz ?
thanks and regards.