26-10-2015, 12:08 PM
Hmm... I don't think the problem here is loudness. I think there are a lot of good or even great sounding records or songs that are pushing -6 LUFS. Although in my opinion the frequency balance is a bit harsh and ear drilling in the "original" mix, soft and loud master.
LRA is a good way of measuring the dynamics of the overall program material, or the musical arrangement. Radio talk show might have a LRA of 0-1 dBs although it's produced at -23 LUFS. There are different ways to measure the micro dynamics of audio, but in my opionion a good rule of thumb is that when the crest factor average goes over 10 dBs the record starts to sound just weak and not dynamic when considering modern rock.
Almost all consumer products are designed and tested for -10 dBV. This in itself doesn't tell much, but almost every -10 consumer level DAC product is tested and calibrated for 10 - 12 dBs of headroom, 16-bit or 24-bit. This is usually reported in the tech specs, where the nominal operating level is at -10 dBV and the max output level for peaks is "always" 0 or +2 dBV. Basically this just means that when the digital signal level is averaging at around -10 dBFS (usually in some RMS time window) the DAC is working at it's nominal level, and peak levels can go as high 10 dB over (closer to 0 dBFS) the average level, also known as headroom. Professional production gear can have as much as 24 dBs of headroom above the nominal level, which in pro gear is usually +4 dBu, also meaning the pro gear signal operates at higher voltage level in analog domain.
When producing material for listening on consumer level gear the average level unfortunately still has to be fairly high, and of course avoid intersampled peaks. From purely technical point of view. It's actually kind of funny how loud masters are criticized for being just loud, when indeed the main purpose of (pre)-mastering is to prepare the original material for the delivery medium and playback gear, and this "always" means increasing the average level of relatively low level 24-bit mixdown closer to -10 dBFS and dithering to 16-bits. Of course there are some records which are loud for just being loud.
LRA is a good way of measuring the dynamics of the overall program material, or the musical arrangement. Radio talk show might have a LRA of 0-1 dBs although it's produced at -23 LUFS. There are different ways to measure the micro dynamics of audio, but in my opionion a good rule of thumb is that when the crest factor average goes over 10 dBs the record starts to sound just weak and not dynamic when considering modern rock.
Almost all consumer products are designed and tested for -10 dBV. This in itself doesn't tell much, but almost every -10 consumer level DAC product is tested and calibrated for 10 - 12 dBs of headroom, 16-bit or 24-bit. This is usually reported in the tech specs, where the nominal operating level is at -10 dBV and the max output level for peaks is "always" 0 or +2 dBV. Basically this just means that when the digital signal level is averaging at around -10 dBFS (usually in some RMS time window) the DAC is working at it's nominal level, and peak levels can go as high 10 dB over (closer to 0 dBFS) the average level, also known as headroom. Professional production gear can have as much as 24 dBs of headroom above the nominal level, which in pro gear is usually +4 dBu, also meaning the pro gear signal operates at higher voltage level in analog domain.
When producing material for listening on consumer level gear the average level unfortunately still has to be fairly high, and of course avoid intersampled peaks. From purely technical point of view. It's actually kind of funny how loud masters are criticized for being just loud, when indeed the main purpose of (pre)-mastering is to prepare the original material for the delivery medium and playback gear, and this "always" means increasing the average level of relatively low level 24-bit mixdown closer to -10 dBFS and dithering to 16-bits. Of course there are some records which are loud for just being loud.