Phil Leigh;632145 Wrote: > The gain reduction can be set in inguzdsp.exe.config. > > Mine is set at -13dB which seems to minimise/eliminate clipping on > almost all of my albums I have tried. > > I doubt you have many albums with an average level of < -20dB.. Do you > use replaygain (I do). My quietest album is nowhere near that quiet ON > AVERAGE, looking at the calculated albumgain. There are some passages > at -30dB... on analogue they would be very noisy due to the lousy SNR. > On digital they sound fine even with 20-30dB of cut on top... > Obviously you need quiet pre and power amps... :-) > > By the way, a very small amount of sporadic clipping is most likely to > pass undetected, even in critical listening. It generally takes a few > successive samples to show it up at all. > > The point is that the -16dB (or whatever) is a flat gain reduction > across the board that is then mostly recovered by boosting certain > frequencies in the DRC. So, the net gain will tend back to zero at some > frequencies. Obviosuly there is a slight deterioration in the SNR > because of this, hence why I feel it makes sense to do all of this in > 24-bit and hence why we are using SOX to convert everything to 24-bit > before gain manipulation. Note also if you ARE using replaygain to > (generally) reduce playback level, that this is applied AFTER INGUZ... > > How much gain reduction you need is solely determined by how much boost > you dial in with EQ AND how much boost DRC needs to apply to flatten the > response of the frequency response dips and how loud your loudest albums > are (generally you will find this is 0dB). > RE the average level of -20dB I don't know if we are talking about the same thing here but i am referring to the RMS levels measured by the foobar DR plugin (which I think you pointed out on this forum). I attach a doc with a number of measurements which show lots of albums on which most ttracks have RMS figure in the -20-30dB range.[and no the Glee cast album is not mine] I appricate that one can drag the target curve in audiolense up and down, but I was tryign to work out what the required overall gain reduction was for a given correection filter. I have a feeliong that audiolense will show you the applied correction curve as well as the target and the mearement. Am I right in nthinking that the maximum overall gain reduction required is equal to the maximum boost in the correction curve? I have a strong suspicion that no more than 6 Db or so of boost is applied in my correction curve
+-------------------------------------------------------------------+ |Filename: foobar dr analysis.doc | |Download: http://forums.slimdevices.com/attachment.php?attachmentid=11874| +-------------------------------------------------------------------+ -- adamdea ------------------------------------------------------------------------ adamdea's Profile: http://forums.slimdevices.com/member.php?userid=37603 View this thread: http://forums.slimdevices.com/showthread.php?t=77084 _______________________________________________ audiophiles mailing list audiophiles@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/audiophiles