I did some further reading about PLR and PSR since yesterday's post.
PLR is peak minus integrated loudness. So if I am trying to have all my tracks -14 Integrated LUFS or higher (-13...-12...etc.), and if I am trying to have my peaks <1dBTP, the best PLR I can get is -14- -1 = 13 PLR value. This would be a track with integrated loudness of -14 and peaks not higher than -1.
So my range cannot be -14 to -12 integrated LUFS, because if I want to believe Ian Shepherd and Bob Katz, PLR should be >=12 and that would imply -13 integrated LUFS at the most, with peaks hitting -1 dBTP. But Ian says also to keep your PLR within the target platform's headroom. If a track is too low integrated LUFS, then it will be turned up and can experience limiting or clipping. Back to PLR at the end...
But first, PSR...
PSR is the true peak minus the short-term LUFS. If My short-term LUFS need to be -10 at the most, then this would say that the maximum PSR should be -1- -10= 9 PSR value. And from Ian Shepherd PSR should be >=8.
Short-term LUFS is a 3-second window of time where the music is at that level, but there are four different ways to measure this and the AES recommendation is one that is the 3-second window but at any time, so read about that if you're interested. Klangfreund is where my multimeter comes from and they have a great product. If PSR is 8 (like you're on the limit), then it implies that short-term LUFS is what? 8=-1 - S. S=-9. Short-term LUFS = -9. But don't we want to have S of -10 maximum? -9 is higher than that. A paradox.
If I look at the Dynameter plugin that Ian helped develop, it shows these values for PSR targets:
Wide=PSR of 14 minimum.
Balanced=PSR of 12 minimum.
Competitive=PSR of 10 min.
Limited=PSR of 8 min.
...back to PLR...
It seems that PLR does not need to be >=12. 12 means the music is dynamic overall. Something less means less dynamics, but this could be acceptable. Something more means more dynamics but this could be o.k. too although the integrated level will be below -14. That could be o.k. because look at classical music - it is down around -16 to -20 integrated loudness, I think.
Which brings me to short-term LUFS...
Ian Shepherd says look only at short-term LUFS and make the loudest part a maximum of -10 LUFS. O.k., but that is a maximum. You can have lower like -11, say. That could be o.k., and PSR would be 10 for a -1 dBTP maximum level. If I look at his PSR targets, the corresponding Smax values (short-term LUFS) would be these if the max true peaks were -1 dBTP:
Wide=PSR=14, Smax=-15 LUFS
Balanced=PSR=12, Smax=-13
Competitive=PSR=10, Smax=-11
Limited=PSR=8, Smax=-9
So to get higher PSR levels, reduce your Smax. Short-term loud passages would be less than 10 by his rule, but less than 10 would give PSR of 9 minimum. So it is a good rule to be Smax at -10 max because you'll have a good PSR that is not limited. But to achieve the higher PSRs that indicate competitive, balanced, and wide levels according to the labels in the Dynameter plugin, you'd need to have Smax somewhere less than -11 LUFS. But in actual practice, it doesn't work this way due to the crazy way that PSR is measured.
Sigh.
More later.
PLR is peak minus integrated loudness. So if I am trying to have all my tracks -14 Integrated LUFS or higher (-13...-12...etc.), and if I am trying to have my peaks <1dBTP, the best PLR I can get is -14- -1 = 13 PLR value. This would be a track with integrated loudness of -14 and peaks not higher than -1.
So my range cannot be -14 to -12 integrated LUFS, because if I want to believe Ian Shepherd and Bob Katz, PLR should be >=12 and that would imply -13 integrated LUFS at the most, with peaks hitting -1 dBTP. But Ian says also to keep your PLR within the target platform's headroom. If a track is too low integrated LUFS, then it will be turned up and can experience limiting or clipping. Back to PLR at the end...
But first, PSR...
PSR is the true peak minus the short-term LUFS. If My short-term LUFS need to be -10 at the most, then this would say that the maximum PSR should be -1- -10= 9 PSR value. And from Ian Shepherd PSR should be >=8.
Short-term LUFS is a 3-second window of time where the music is at that level, but there are four different ways to measure this and the AES recommendation is one that is the 3-second window but at any time, so read about that if you're interested. Klangfreund is where my multimeter comes from and they have a great product. If PSR is 8 (like you're on the limit), then it implies that short-term LUFS is what? 8=-1 - S. S=-9. Short-term LUFS = -9. But don't we want to have S of -10 maximum? -9 is higher than that. A paradox.
If I look at the Dynameter plugin that Ian helped develop, it shows these values for PSR targets:
Wide=PSR of 14 minimum.
Balanced=PSR of 12 minimum.
Competitive=PSR of 10 min.
Limited=PSR of 8 min.
...back to PLR...
It seems that PLR does not need to be >=12. 12 means the music is dynamic overall. Something less means less dynamics, but this could be acceptable. Something more means more dynamics but this could be o.k. too although the integrated level will be below -14. That could be o.k. because look at classical music - it is down around -16 to -20 integrated loudness, I think.
Which brings me to short-term LUFS...
Ian Shepherd says look only at short-term LUFS and make the loudest part a maximum of -10 LUFS. O.k., but that is a maximum. You can have lower like -11, say. That could be o.k., and PSR would be 10 for a -1 dBTP maximum level. If I look at his PSR targets, the corresponding Smax values (short-term LUFS) would be these if the max true peaks were -1 dBTP:
Wide=PSR=14, Smax=-15 LUFS
Balanced=PSR=12, Smax=-13
Competitive=PSR=10, Smax=-11
Limited=PSR=8, Smax=-9
So to get higher PSR levels, reduce your Smax. Short-term loud passages would be less than 10 by his rule, but less than 10 would give PSR of 9 minimum. So it is a good rule to be Smax at -10 max because you'll have a good PSR that is not limited. But to achieve the higher PSRs that indicate competitive, balanced, and wide levels according to the labels in the Dynameter plugin, you'd need to have Smax somewhere less than -11 LUFS. But in actual practice, it doesn't work this way due to the crazy way that PSR is measured.
Sigh.
More later.