|
Subject:another normalization question
Posted by: Tempest
Date:5/10/2000 1:02:00 PM
OK, so I recorded some DJ sets onto MD. The MD, when simple recorded to WAV through my sound card in really quiet. The peaks on the source file are about -3.5dB. I tried normalizing using SF to -6.0dB which someone had mentioned is what commerical CD's are mastered to. Of course, this seems to have made everything a hair quieter, exactly what I didn't want. My undestanding is that normalizing applies gain to lower the peak levels to whatever you set as the max. I should mention that the peaks in the source aren't particularly louder when listened through speakers, so I'm not sure why they're there. Anyway, what's the best way to increase the perceived loudness of my recording? -Jay |
|
Subject:Re: another normalization question
Reply by: RickZ
Date:5/11/2000 7:15:00 AM
Hi Jay, Try normalizing to -0.5 dB, that should do more of what you wanted, but if you're interested in pushing up the 'perceived' volume, you may want to look into www.sonictimeworks.com Mastering Compressor plug-in. I used it recently with SF 4.5 on an Theatre Organ concert recording, and it sounded great. You could also experiment with the Dynamics processing (compression) that is on the Forge menu. Rick Z JR wrote: >>OK, so I recorded some DJ sets onto MD. The MD, when simple >>recorded to WAV through my sound card in really quiet. The >>peaks on the source file are about -3.5dB. >> >>I tried normalizing using SF to -6.0dB which someone had >>mentioned is what commerical CD's are mastered to. >> >>Of course, this seems to have made everything a hair >>quieter, exactly what I didn't want. My undestanding is >>that normalizing applies gain to lower the peak levels to >>whatever you set as the max. >> >>I should mention that the peaks in the source aren't >>particularly louder when listened through speakers, so I'm >>not sure why they're there. >> >>Anyway, what's the best way to increase the perceived >>loudness of my recording? >> >>-Jay |
|
Subject:Re: another normalization question
Reply by: Jeff_Lowes
Date:5/16/2000 11:38:00 PM
Another method that you might try in order to increase the perceived loudness of your sound file is to use the RMS Normalizing function. Do a Scan levels on the file, set the "if clipping occurs" box to "Apply dynamic compression", then click the Average (RMS) option button and set the fader to a level a little higher than your scanned value. By the way, every major label CD that I have ever digitally loaded into my DAW has come in with peak levels at 0.0 dB (or VERY close to that) which tells me that the big boys master their CDs to zero. Makes sense since everyone wants their CD to be as hot or hotter than the rest. Jeff Lowes On-Track Recording JR wrote: >>OK, so I recorded some DJ sets onto MD. The MD, when simple >>recorded to WAV through my sound card in really quiet. The >>peaks on the source file are about -3.5dB. >> >>I tried normalizing using SF to -6.0dB which someone had >>mentioned is what commerical CD's are mastered to. >> >>Of course, this seems to have made everything a hair >>quieter, exactly what I didn't want. My undestanding is >>that normalizing applies gain to lower the peak levels to >>whatever you set as the max. >> >>I should mention that the peaks in the source aren't >>particularly louder when listened through speakers, so I'm >>not sure why they're there. >> >>Anyway, what's the best way to increase the perceived >>loudness of my recording? >> >>-Jay |
|
Subject:Re: another normalization question...I created a monster! LOL
Reply by: O_G_Killa
Date:5/22/2000 12:15:00 PM
Hi JR, I believe I was the one that posted something about CDs being normalized to -6dB. I guess I should have elaborated on that more. Actually, CDs are constantly being mastered hotter and hotter. The latest prince CD I got (Joy Unto the Rave Fantastic) has a Peak of 0dB and an RMS of -2dB or -3dB!!!! As technology gets better, the levels get hotter!!! I would like to clarify this, EVERYONE PLEASE READ... There are two types of normalization in SF, Peak and average RMS. When I refered to the -6dB, that was in Reference to the average RMS level of the recording, NOT the peak level of the recording... THE PEAK LEVEL OF YOUR RECORDING SHOULD ALWAYS BE 0.0dB!!!!!!!!! That said, you can have a file with a Peak of 0.0dB but a RMS of - 20dB!!!!! The Average RMS is the average volume over time. Try this.... Open your DJ mixes you want to normalize in SF. Go to Process Menu | Normalize. In this window you will see a few things.. First, change the "Normalize Using:" To "Average RMS power". Second, click the "Scan Levels" button in the lower left corner of the window. Let it do it's thing (this will take a little while) Third, Write down the Peak and RMS number it gives you (should display the numbers right below the "Scan Levels" button). You now know what your project's levels are at. Since you are recording DJs and most DJs use samples, you may not really need to normalize that much because most samples are normalized to begin with...and since the DJs are spinning other peoples songs, those songs they are spinning are most likely normalized as well...so when working with DJs you are usually going to have an easier time Normalizing levels....anyway...moving on.... :-) now that you know what your peak and average RMS are...you can now accurately normalize using RMS. you said that when you used a setting of -6dB it got quieter. There are two obvious reasons for this... One, you didn't change the "Normalize Using:" to "Average RMS power". Or Two, your file already has an anverage RMS that is higher than -6dB. since you are working with a recording from a DJ either one is very possible... Now going back to those numbers you wrote down a few minutes ago...the Peak and RMS values of the file. If your peak is -3.5 and your RMS is say -4dB...then you really don't need to Normalize using RMS!!! But if your peak is -3.5dB and your RMS is say -12dB. You may want to normalize it. Basically...When you Normalize using average RMS it is similar to Compressing the file...in fact most professionally mastered CDs use compressor/limiters to "Normalize" the recordings. For Example I know that some mastering studios use Wave's L1 plugin (www.waves.com) for Normalizing. The one thing you have to remember is.................. You normalize to get all of the recordings going on to the CD at the same volume. You don't want track one to be very loud and then track two to be 15dB quieter!!! So if your average RMS is -6dB or -20dB, just as long as all the songs going onto the CD you are making are that same level you shouldn't have a problem. One other thing I would like to add....Sonic Foundry's effects and plugins are great!!!!...but sometimes 3rd party plugins are what you need to get the sound you want. I use Wave's Renaissance Compressor/Limiter to Normalize all of my recordings. It is not that I think the Sound Forge Compressor and Normalizer are inferior, I do use them. But, Every time you do something to a recording it is going to "color" the recording. I like the way the RCL (short for Renaissance Compressor/Limiter) sounds and what it does to my recordings, that's why I use it. Hope this helps. Sorry for the confusion! JR wrote: >>OK, so I recorded some DJ sets onto MD. The MD, when simple >>recorded to WAV through my sound card in really quiet. The >>peaks on the source file are about -3.5dB. >> >>I tried normalizing using SF to -6.0dB which someone had >>mentioned is what commerical CD's are mastered to. >> >>Of course, this seems to have made everything a hair >>quieter, exactly what I didn't want. My undestanding is >>that normalizing applies gain to lower the peak levels to >>whatever you set as the max. >> >>I should mention that the peaks in the source aren't >>particularly louder when listened through speakers, so I'm >>not sure why they're there. >> >>Anyway, what's the best way to increase the perceived >>loudness of my recording? >> >>-Jay |
|
Subject:Re: another normalization question...I created a monster! LOL
Reply by: Hermes
Date:5/23/2000 7:25:00 AM
GREAT EXPLANATION, O.G. Killa but there is just one thing you forgot to explain: is that Normalizing in peak mode is nice, because you can get the maximum bit accuracy, but if you are using RMS, in fact you can destroy your favourite recording, because you are compresing the audio contents instead optimizing up to the last (but not least) bit. In other words, pay attention at what are you doing... with your EARS !!!. Listen first,... and then apply just the minimum process in order to get right in front of the massive attach of HIPERMAXIMIZED- AND-OVEREQUALIZED-PLUS-PERFECTLEVELED-AND-JUST-SUPEREXCITED-AUDIO Sorry at all, but I'm annoyed, because I't seams that where we are going is to a perfect 192Khz samplig rate........ at 2 bits !!!! Please let the music have DYNAMIC AND EXPRESION AND LIFE !!! Regards Hermes O.G. Killa wrote: >>Hi JR, >> >>I believe I was the one that posted something about CDs being >>normalized to -6dB. I guess I should have elaborated on that more. >> >>Actually, CDs are constantly being mastered hotter and hotter. The >>latest prince CD I got (Joy Unto the Rave Fantastic) has a Peak of >>0dB and an RMS of -2dB or -3dB!!!! As technology gets better, the >>levels get hotter!!! >> >>I would like to clarify this, EVERYONE PLEASE READ... >> >>There are two types of normalization in SF, Peak and average RMS. >>When I refered to the -6dB, that was in Reference to the average RMS >>level of the recording, NOT the peak level of the recording... >> >>THE PEAK LEVEL OF YOUR RECORDING SHOULD ALWAYS BE 0.0dB!!!!!!!!! >> >>That said, you can have a file with a Peak of 0.0dB but a RMS of - >>20dB!!!!! The Average RMS is the average volume over time. >> >>Try this.... >> >>Open your DJ mixes you want to normalize in SF. Go to Process Menu | >>Normalize. >> >>In this window you will see a few things.. >> >>First, change the "Normalize Using:" To "Average RMS power". >> >>Second, click the "Scan Levels" button in the lower left corner of >>the window. Let it do it's thing (this will take a little while) >> >>Third, Write down the Peak and RMS number it gives you (should >>display the numbers right below the "Scan Levels" button). >> >>You now know what your project's levels are at. Since you are >>recording DJs and most DJs use samples, you may not really need to >>normalize that much because most samples are normalized to begin >>with...and since the DJs are spinning other peoples songs, those >>songs they are spinning are most likely normalized as well...so when >>working with DJs you are usually going to have an easier time >>Normalizing levels....anyway...moving on.... :-) >> >>now that you know what your peak and average RMS are...you can now >>accurately normalize using RMS. you said that when you used a >>setting of -6dB it got quieter. There are two obvious reasons for >>this... >> >>One, you didn't change the "Normalize Using:" to "Average RMS power". >> >>Or >> >>Two, your file already has an anverage RMS that is higher than -6dB. >> >>since you are working with a recording from a DJ either one is very >>possible... >> >>Now going back to those numbers you wrote down a few minutes >>ago...the Peak and RMS values of the file. If your peak is -3.5 and >>your RMS is say -4dB...then you really don't need to Normalize using >>RMS!!! But if your peak is -3.5dB and your RMS is say -12dB. You >>may want to normalize it. >> >>Basically...When you Normalize using average RMS it is similar to >>Compressing the file...in fact most professionally mastered CDs use >>compressor/limiters to "Normalize" the recordings. For Example I >>know that some mastering studios use Wave's L1 plugin (www.waves.com) >>for Normalizing. >> >>The one thing you have to remember is.................. You >>normalize to get all of the recordings going on to the CD at the same >>volume. You don't want track one to be very loud and then track two >>to be 15dB quieter!!! So if your average RMS is -6dB or -20dB, just >>as long as all the songs going onto the CD you are making are that >>same level you shouldn't have a problem. >> >>One other thing I would like to add....Sonic Foundry's effects and >>plugins are great!!!!...but sometimes 3rd party plugins are what you >>need to get the sound you want. >> >>I use Wave's Renaissance Compressor/Limiter to Normalize all of my >>recordings. It is not that I think the Sound Forge Compressor and >>Normalizer are inferior, I do use them. But, Every time you do >>something to a recording it is going to "color" the recording. I >>like the way the RCL (short for Renaissance Compressor/Limiter) >>sounds and what it does to my recordings, that's why I use it. >> >>Hope this helps. Sorry for the confusion! >> >> >> >> >> >> >> >> >> >> >> >>JR wrote: >>>>OK, so I recorded some DJ sets onto MD. The MD, when simple >>>>recorded to WAV through my sound card in really quiet. The >>>>peaks on the source file are about -3.5dB. >>>> >>>>I tried normalizing using SF to -6.0dB which someone had >>>>mentioned is what commerical CD's are mastered to. >>>> >>>>Of course, this seems to have made everything a hair >>>>quieter, exactly what I didn't want. My undestanding is >>>>that normalizing applies gain to lower the peak levels to >>>>whatever you set as the max. >>>> >>>>I should mention that the peaks in the source aren't >>>>particularly louder when listened through speakers, so I'm >>>>not sure why they're there. >>>> >>>>Anyway, what's the best way to increase the perceived >>>>loudness of my recording? >>>> >>>>-Jay |
|
Subject:Re: another normalization question...I created a monster! LOL
Reply by: O_G_Killa
Date:6/7/2000 11:29:00 AM
Hi Hermes, Thanks for the heads up...I have another post on here (the one in which I originally mentioned -6dB RMS Level) that talks more about using the Peak Normalization Feature. But really, my post was to help this particular person with his particular problem...I was not posting a "Technical Note on Normalization", it was more directed specifically towards RMS normalizing techniques. In a previous post I had explained the difference between Peak normalizing and RMS normalizing. When this person I was replied to attempted to try and normalize his recording he wasn't exactly in the right spot. He followed my suggestions but he was following my suggestions for RMS while he was using Peak in SF. He never Switched to RMS in the Normalize window, so the result was not what I had described. Hence, it did not have the affect he desired. I hear what your saying about dynamics and I agree with you. But fact of the matter is, to the untrained ear, the louder the song is the better the perceived sound is. There is a great interview with Bob Ludwig about mastering on Parsons Audio's website (http://www.paudio.com/Pages/learning_Ludwig.html) and he talks about this. Being a musician I love dynamics, I love the excitment cresendos and diminuendos create. But, since I work for other people and not myself, I have to do what they request. Most of the music today is highly compressed, drastically EQ'd, and drenched in effects. Just because I don't agree with it doesn't mean I shouldn't know how to do it nor should I deny someone else that knowledge. So, I appologize for not mentioning Peak normalization in this post. And I would like to reiterate to anyone who has not learned about peak normalization to do so. Thanks. Hermes wrote: >>GREAT EXPLANATION, O.G. Killa >> >>but there is just one thing you forgot to explain: >>is that Normalizing in peak mode is nice, because you can get the >>maximum bit accuracy, but if you are using RMS, in fact you can >>destroy your favourite recording, because you are compresing the >>audio contents instead optimizing up to the last (but not least) bit. >> >>In other words, pay attention at what are you doing... with your >>EARS !!!. Listen first,... and then apply just the minimum process in >>order to get right in front of the massive attach of HIPERMAXIMIZED- >>AND-OVEREQUALIZED-PLUS-PERFECTLEVELED-AND-JUST-SUPEREXCITED-AUDIO >> >>Sorry at all, but I'm annoyed, because I't seams that where we are >>going is to a perfect 192Khz samplig rate........ at 2 bits !!!! >> >>Please let the music have DYNAMIC AND EXPRESION AND LIFE !!! >>Regards >>Hermes >> >>O.G. Killa wrote: >>>>Hi JR, >>>> >>>>I believe I was the one that posted something about CDs being >>>>normalized to -6dB. I guess I should have elaborated on that more. >>>> >>>>Actually, CDs are constantly being mastered hotter and hotter. The >>>>latest prince CD I got (Joy Unto the Rave Fantastic) has a Peak of >>>>0dB and an RMS of -2dB or -3dB!!!! As technology gets better, the >>>>levels get hotter!!! >>>> >>>>I would like to clarify this, EVERYONE PLEASE READ... >>>> >>>>There are two types of normalization in SF, Peak and average RMS. >>>>When I refered to the -6dB, that was in Reference to the average >>RMS >>>>level of the recording, NOT the peak level of the recording... >>>> >>>>THE PEAK LEVEL OF YOUR RECORDING SHOULD ALWAYS BE 0.0dB!!!!!!!!! >>>> >>>>That said, you can have a file with a Peak of 0.0dB but a RMS of - >>>>20dB!!!!! The Average RMS is the average volume over time. >>>> >>>>Try this.... >>>> >>>>Open your DJ mixes you want to normalize in SF. Go to Process Menu >>| >>>>Normalize. >>>> >>>>In this window you will see a few things.. >>>> >>>>First, change the "Normalize Using:" To "Average RMS power". >>>> >>>>Second, click the "Scan Levels" button in the lower left corner of >>>>the window. Let it do it's thing (this will take a little while) >>>> >>>>Third, Write down the Peak and RMS number it gives you (should >>>>display the numbers right below the "Scan Levels" button). >>>> >>>>You now know what your project's levels are at. Since you are >>>>recording DJs and most DJs use samples, you may not really need to >>>>normalize that much because most samples are normalized to begin >>>>with...and since the DJs are spinning other peoples songs, those >>>>songs they are spinning are most likely normalized as well...so >>when >>>>working with DJs you are usually going to have an easier time >>>>Normalizing levels....anyway...moving on.... :-) >>>> >>>>now that you know what your peak and average RMS are...you can now >>>>accurately normalize using RMS. you said that when you used a >>>>setting of -6dB it got quieter. There are two obvious reasons for >>>>this... >>>> >>>>One, you didn't change the "Normalize Using:" to "Average RMS >>power". >>>> >>>>Or >>>> >>>>Two, your file already has an anverage RMS that is higher than - 6dB. >>>> >>>>since you are working with a recording from a DJ either one is very >>>>possible... >>>> >>>>Now going back to those numbers you wrote down a few minutes >>>>ago...the Peak and RMS values of the file. If your peak is -3.5 >>and >>>>your RMS is say -4dB...then you really don't need to Normalize >>using >>>>RMS!!! But if your peak is -3.5dB and your RMS is say -12dB. You >>>>may want to normalize it. >>>> >>>>Basically...When you Normalize using average RMS it is similar to >>>>Compressing the file...in fact most professionally mastered CDs use >>>>compressor/limiters to "Normalize" the recordings. For Example I >>>>know that some mastering studios use Wave's L1 plugin >>(www.waves.com) >>>>for Normalizing. >>>> >>>>The one thing you have to remember is.................. You >>>>normalize to get all of the recordings going on to the CD at the >>same >>>>volume. You don't want track one to be very loud and then track >>two >>>>to be 15dB quieter!!! So if your average RMS is -6dB or -20dB, >>just >>>>as long as all the songs going onto the CD you are making are that >>>>same level you shouldn't have a problem. >>>> >>>>One other thing I would like to add....Sonic Foundry's effects and >>>>plugins are great!!!!...but sometimes 3rd party plugins are what >>you >>>>need to get the sound you want. >>>> >>>>I use Wave's Renaissance Compressor/Limiter to Normalize all of my >>>>recordings. It is not that I think the Sound Forge Compressor and >>>>Normalizer are inferior, I do use them. But, Every time you do >>>>something to a recording it is going to "color" the recording. I >>>>like the way the RCL (short for Renaissance Compressor/Limiter) >>>>sounds and what it does to my recordings, that's why I use it. >>>> >>>>Hope this helps. Sorry for the confusion! >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>>JR wrote: >>>>>>OK, so I recorded some DJ sets onto MD. The MD, when simple >>>>>>recorded to WAV through my sound card in really quiet. The >>>>>>peaks on the source file are about -3.5dB. >>>>>> >>>>>>I tried normalizing using SF to -6.0dB which someone had >>>>>>mentioned is what commerical CD's are mastered to. >>>>>> >>>>>>Of course, this seems to have made everything a hair >>>>>>quieter, exactly what I didn't want. My undestanding is >>>>>>that normalizing applies gain to lower the peak levels to >>>>>>whatever you set as the max. >>>>>> >>>>>>I should mention that the peaks in the source aren't >>>>>>particularly louder when listened through speakers, so I'm >>>>>>not sure why they're there. >>>>>> >>>>>>Anyway, what's the best way to increase the perceived >>>>>>loudness of my recording? >>>>>> >>>>>>-Jay |