| Author
|
Preferred RMS reading device
|
Elad
Tsabeat/Sattel Battle
Started Topics :
158
Posts :
5306
Posted : Jun 24, 2008 15:00
|
hehe i thought RMS is the way you actualy hear it (loudness) , u cant really messure peaks tho with ears 
My ears can easy tell me if there is overcompression and distortion
i dont say it doesnt sound good sometimes tho , it really matter what the music inside is.
crazy breakcore track can have -7rms and its all good since it should be distorted etc. while chillout track i wouldnt go for over then -11rms or so
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Jun 24, 2008 16:09
|
The difference is probably due to different averaging time as Elad mentions but also whether 0 dB FS is considered a full scale square wave or a full scale sine wave.
In theory, you can't have a full scale square wave because the reconstructed waveform that comes out of the converters would be above 0 dB FS (inter sample peaks). This is technically an illegal signal that does not satisfy the Nyquist requirements for digital sampling. Having said that, you can easily create these signals within the digital domain. (Clipping, limiting, boosting the high-end or by synthesis)
The correct AES-17 meters use sine = FS and will show levels about 3 dB higher than other meters. I think iXL is correct. In Adobe Audition you can select whether sine or square is FS (and change the averaging time).
Oh and MOST audio engineers including mastering engineers don't seem to be aware of this so they might be giving you the wrong numbers. Usually not much of an issue but keep it in mind when you get weird numbers that don't make sense. 
UnderTow
|
|
|
|
psylevation
IsraTrance Full Member
Started Topics :
52
Posts :
841
Posted : Jun 25, 2008 08:43
|
Greg...keep it down..shhhhhh, free G and inspector XL are trying to end the loudness war 
It could be that some of them are trying to be exact. For example to get the RMS a sine wave you can multiply the amplitude by .707 but to be more exact you multiply by the square root of 2. So it could be that some of the plugs are using more exact methods and others are using a general method.
But honestly I don't know.
 ~Airyck~
~Unoccupied Mind ~
Psyowa! |
|
|
|
uiop
Inactive User
Started Topics :
0
Posts :
9
Posted : Jun 25, 2008 09:03
|
|
Audio post production would be a good place to investigate......they are much more specific about levels and equipment spec. |
|
|
|
Elad
Tsabeat/Sattel Battle
Started Topics :
158
Posts :
5306
Posted : Jun 25, 2008 14:51
|
i cant find exacly the link sorry
but read about K-system
defintly the default preset of PAZ is -5DB rms then the RME Digicheck
its all about the speed of reading
btw the last cd i heard i really liked the sound then checked it and saw its freakin square track lol so im not sure what to do .. i guess as always , as long i dont hear any major dynamic change or distortion or loosing stereo width i will maximize as much possible , in 95% of the cases we talk on -10DB Rms averege by wavlab analize check. |
|
|
|
vegetal
Vegetal/Peacespect
Started Topics :
19
Posts :
1055
Posted : Jun 25, 2008 14:57
|
|
piko_bianko
Oxya
Started Topics :
57
Posts :
974
Posted : Jun 26, 2008 02:29
|
Soundforge
i find it very precise
extreme |
|
|
|
makus
Overdream
Started Topics :
82
Posts :
3087
Posted : Jun 26, 2008 17:41
|
i dont check rms until the task that i do requires exact level in result (eg tv or radio clip). in that case i use iXL for real-time check and wavelab for offline one. never had any problems with those two.
www.overdreamstudio.com |
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Jun 26, 2008 19:16
|
Quote:
|
On 2008-06-26 02:29, piko_bianko wrote:
Soundforge
i find it very precise
|
|
Soundforge (at least the version I have) like Waves PAZ is wrong by 3 dB. (FS = Square). Audition, Wavelab, Inspector XL and others are correct.
UnderTow |
|
|
|
Elad
Tsabeat/Sattel Battle
Started Topics :
158
Posts :
5306
Posted : Jun 26, 2008 20:28
|
Quote:
|
On 2008-06-26 19:16, UnderTow wrote:
Quote:
|
On 2008-06-26 02:29, piko_bianko wrote:
Soundforge
i find it very precise
|
|
Soundforge (at least the version I have) like Waves PAZ is wrong by 3 dB. (FS = Square). Audition, Wavelab, Inspector XL and others are correct.
UnderTow
|
|
again PAZ cant be "wrong" , it can act like evry other machine if u change the settings
ever looked on the "load" preset list? or played with the "response" button?
all in all after mastered 2 v/a's i can say sometimes there is complete diffrence from the numbers to what you hear.
some tracks are louder with no extra rms , and some dirty sound tracks cant be fully maximized since the bands blend too much |
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Jun 27, 2008 14:53
|
Quote:
|
On 2008-06-26 20:28, Elad wrote:
Quote:
|
On 2008-06-26 19:16, UnderTow wrote:
Waves PAZ is wrong by 3 dB. (FS = Square).
|
|
again PAZ cant be "wrong" , it can act like evry other machine if u change the settings
ever looked on the "load" preset list? or played with the "response" button?
|
|
PAZ sure can be wrong. The "Response" only affects the averaging time. PAZ does not follow AES specifications that a sine wave which peaks at 0 dB FS is the maximum level in digital audio. PAZ, like many other digital level meters, considers a square wave that peaks at 0 dB FS to be the maximum level in digital audio.
So what is wrong with taking a square wave that peaks at 0 dB FS to be the maximum? To sample a square wave you need infinite bandwidth but we don't have infinite bandwidth (Nyquist theory etc), we have limited bandwidth. We filter out anything above 1/2 the sampling rate at the inputs and outputs with a low-pass filter (anti-aliasing filter at the input and reconstruction filter at the output).
If you take what looks like a square wave peaking at 0 dB FS in a typical DAW, and filter out all the harmonics (reconstruction filter) you end up with a wave that isn't a square and that peaks higher than 0 dB FS. If all the harmonics are above 1/2 Nyquist, you end up with a sine wave that peaks 3 dB above 0 dB FS. This is quite obviously wrong. Hence the AES specifications which Wave PAZ doesn't follow.
Here is an example of an "illegal" digital signal (Not an exact square wave): http://puretone.nl/Illegal_signal.jpg
All the sample values are at 0 dB FS yet what comes out of your converters peaks at close to +7 dB FS!
Quote:
|
all in all after mastered 2 v/a's i can say sometimes there is complete diffrence from the numbers to what you hear.
some tracks are louder with no extra rms , and some dirty sound tracks cant be fully maximized since the bands blend too much
|
|
Absolutely. RMS is closer to our hearing than peak level but it still isn't an exact indicator of perceived loudness.
UnderTow |
|
|
|
Aker
Inactive User
Started Topics :
0
Posts :
111
Posted : Jul 1, 2008 10:02
|
|
What is this AES specification? |
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Jul 1, 2008 20:26
|
Quote:
|
On 2008-07-01 10:02, Aker wrote:
What is this AES specification?
|
|
AES-17. I can mail you the PDF if you are interested.
UnderTow |
|
|
|
trollup
Inactive User
Started Topics :
0
Posts :
55
Posted : Jul 2, 2008 06:08
|
|
hey undertow where can i find your music? |
|
|
|
trollup
Inactive User
Started Topics :
0
Posts :
55
Posted : Jul 2, 2008 08:53
|
|
IS that the AES-17 1998 specification? |
|
|
|
