Why are masters so different?
Sep 16, 2016 at 10:06 PM Post #31 of 132
i think its a marketing trick, as i have found out here not many people know about mastering quality.
 
so instead of companies releasing just 24/96 versions of the CD master and hoping placebo sells the "Hi-Res" albums they also (most of the time) secretly use a higher dynamic range master to go alongside its 24/96 label, then its not inaudible its massively different.but because people dont know about mastering differences they instantly assume it sound so good because its "Hi-Res" not because its got a dynamic range rating of something like 12 instead of the usual 5/6/7 
wink.gif

 
Sep 17, 2016 at 2:15 AM Post #32 of 132
I don't know how about trying to get a good sound without turning so quickly to compression. Nevermind is a good example.  Some level of compression is integral to its sound.  Yet, in my opinion, it would have been better with less than it has.  Not none, but less. ...

 
If you're referring to the original version, I disagree. It has "the sound" (that defined the genre) without actually being heavily compressed. If all the music produced since then stuck to those sorts of levels, I'd be a happy customer.
 
You know something's wrong when a heavy metal album ends up less compressed than a semi-acoustic soul/blues album. (Iron Maiden, "The Final Frontier" versus Tom Jones, "Praise And Blame". Both mastered in 2010 by Bob Ludwig.)
 
Sep 17, 2016 at 3:00 AM Post #33 of 132
 
[1] I don't know how about trying to get a good sound without turning so quickly to compression.
[2] Nevermind is a good example.  Some level of compression is integral to its sound.  Yet, in my opinion, it would have been better with less than it has.  Not none, but less. 
[3] EDM has to be highly compressed to work in its intended environment which is a noisy dance club.
[4] So were I picking levels of compression I would stop for the tremendous majority of music with levels adequate to enjoy it in a car environment.
[5] I'll grant you the opinion that well done even heavy compression adds to the quality of the mix.  Please grant me the opinion that just going by what I hear way too much compression is being applied routinely.

 
1. I've been trying to explain that compression is one of the earliest and most fundamental audio processing tools invented. It's effects cannot be replicated any other way. It's similar to saying; how about trying to get a good sound without using any EQ.
 
2. Maybe a small amount less but without compression entirely or with very minimal compression the mixes on Nevermind would have to sound very substantially different and that would be a sacrilege because it's a great album (from a production point of view, regardless of whether one likes the genre).
 
3. I was just using EDM as an obvious example. Techno, hardcore, drum and bass, modern R&B and many others completely depend on fairly substantial use of compression. Even most forms of rock music does, a kick drum does not sound that way in real life. Again, compression is a fundamental production tool; popular genres, even some fairly old ones, have evolved with it's use in mind. Without it, those genres wouldn't exist or would be substantially different. What would metal sound like with just the real sound of a kick drum?
 
4. That would require a moderate to heavy use of compression then, not minimal or none at all! Are you changing your original assertion?
 
5. As I've also tried to already explain; not only do I grant you that opinion but I entirely agree with it! Too much compression is very common, the loudness war definitely exists. I'm not arguing against the idea that in many/most cases the use of compression should be reduced, quite the contrary, I've been arguing for a reduction for over 20 years! However, my argument is for a reduction from the massive over application of compression to an appropriate level of compression. In practise this would generally mean anywhere from minimal to heavy compression (rather than massive over compression), rather than the minimal or no compression you originally asserted. Minimal or none at all is even worse than massive over compression IMO because over compression merely damages mixes, whereas no compression at all impacts artistic decisions to the point of genres being very substantially different or not even existing.
 
Quote:
  i think its a marketing trick ...

 
That's also a bit of a grey area, rather than entirely black and white. For example, there is a valid argument that a standard 16/44.1 version is likely to be converted to an MP3 or similar for portable use where a high dynamic range is undesirable. A higher sample rate version is more likely to be listened to on a better quality system in a quieter/better environment, where a higher dynamic range is more desirable. So, while there is some valid justification, there's also no doubt that this fact is sorely abused by marketing depts. It's turned around and used to imply/demonstrate that higher sample rates/bit depths are intrinsically better and the price hike justified.
 
Long term this is a problem for the consumer and an opportunity for distributors. What happens when the so called Hi-res formats become standard on portable devices? Technologies such as MQA are attempting to make "hi-res" formats practical for portable storage/streaming. At that point 24/96 or 24/192 could effectively become the new standard res and even higher rates required for "hi-res", with the potential justification that 24/96 or 24/192 need less dynamic range for portable use.
 
G
 
Sep 20, 2016 at 10:23 AM Post #34 of 132
I hear the word mastering mentioned so often but in reality a lot of the compression is performed at the mixing stage on individual stems. Vocals, drums, snares, guitars, bass, you name it. A lot of EDM such as trance additionally relies on the pumping effect of a sidechain compressor going into a bass line, for instance.That's just the nature of the beast. The kick and bass carries the track, and without heavy compression (and generous amounts of reverb!) there would probably be nothing to dance to. At the other end of the spectrum you may have a classical music piece, where one typically ends up with very little compression to preserve the dynamics of the sound. One just has to look at recordings such as the Goldberg Variations or the Hammerklavier to see the inherent dynamic range even on one instrument such as solo piano.
 
Sep 21, 2016 at 3:08 PM Post #35 of 132
One doesn't have to wonder then why no one uses EDM for evaluating any type of equipment, well..... maybe subwoofers. :wink:
 
Sep 24, 2016 at 7:20 AM Post #36 of 132
  i think its a marketing trick, as i have found out here not many people know about mastering quality.
 
so instead of companies releasing just 24/96 versions of the CD master and hoping placebo sells the "Hi-Res" albums they also (most of the time) secretly use a higher dynamic range master to go alongside its 24/96 label, then its not inaudible its massively different.but because people dont know about mastering differences they instantly assume it sound so good because its "Hi-Res" not because its got a dynamic range rating of something like 12 instead of the usual 5/6/7 
wink.gif

There is a lot of truth in that.  Some hybrid SACDs do actually use the same mastering on the CD layer as the SACD.  Most people would play the CD layer and then the SACD layer (or vice versa) and say they can hear a clear difference in sound quality in favour of the SACD layer.
 
However SACD players vary in how intuitive it is to pick which layer is playing.  Many of the same people then later find out after enjoying the higher sound quality of the SACD layer that they were in fact listening to the CD layer - ie a proper double blind test.  Perhaps this part or most of the reason why most hybrid SACDs do use a different master for each layer?
 
The other evidence comes from the fall out with the Myer and Moran study which found under proper controlled conditions, even musicians, audiophiles and music producers could not pick the 16/44 version from the hi res at a rate greater than the flipping of a coin.  What is more telling is that some of the industry producers came out and said that many of the SACDs used in that study actually came from 16/44 or 16/48 masters.   What they were saying, in effect, is that in many years up to 2007 at least, people were being sold SACDs with the same source sample rates and word lengths as CDs yet no audiophile or golden eared reviewer noticed.  In the end it was objective measurements which confirmed it.
 
Sep 24, 2016 at 7:23 AM Post #37 of 132
Got a slightly off topic question.  My Bluesound streamer has options for replay gain or not.  For replay gain the options are track gain, album gain or "smart" gain - whatever that is.  I barely notice any difference which ever one I pick.
 
Which is the better gain to use, if any, and why?  Does this affect the fidelity of playback in addition to normalising volume?
 
Sep 24, 2016 at 9:29 AM Post #38 of 132
  Got a slightly off topic question.  My Bluesound streamer has options for replay gain or not.  For replay gain the options are track gain, album gain or "smart" gain - whatever that is.  I barely notice any difference which ever one I pick.
 
Which is the better gain to use, if any, and why?  Does this affect the fidelity of playback in addition to normalising volume?

http://wiki.hydrogenaud.io/index.php?title=ReplayGain
 
Sep 24, 2016 at 12:27 PM Post #39 of 132
  Got a slightly off topic question.  My Bluesound streamer has options for replay gain or not.  For replay gain the options are track gain, album gain or "smart" gain - whatever that is.  I barely notice any difference which ever one I pick.
 
Which is the better gain to use, if any, and why?  Does this affect the fidelity of playback in addition to normalising volume?

maybe more practical answer here.
I'd imagine smart gain to be some kind of limiter for when your files don't have replay gain metadata? IDK.  for the other options, you first need to scan your songs for the desired setting, then write the gain changes in the metadata of each song for the usual replay gain to work. if the song was never scanned for gain, it will just play as is.
 
the typical use is the per track gain so that you can shuffle your library without feeling the need to change the volume level all the time between each song going from Cradle of Filth, to Rihanna and then Chopin(my kind of shuffle).
 
Sep 24, 2016 at 9:24 PM Post #40 of 132
  maybe more practical answer here.
I'd imagine smart gain to be some kind of limiter for when your files don't have replay gain metadata? IDK.  for the other options, you first need to scan your songs for the desired setting, then write the gain changes in the metadata of each song for the usual replay gain to work. if the song was never scanned for gain, it will just play as is.
 
the typical use is the per track gain so that you can shuffle your library without feeling the need to change the volume level all the time between each song going from Cradle of Filth, to Rihanna and then Chopin(my kind of shuffle).

 
Ah, I see.  I didn't know about the metadata, most of my music are flac files and some mp3a, perhaps that is why I could not hear any difference whether the replay gain was on or off.
 
Such a feature would be very useful for my mixed tracks folder, though I wonder how much work is involved to modify the metadata for 3000 tracks.
 
Sep 24, 2016 at 9:27 PM Post #41 of 132
  http://wiki.hydrogenaud.io/index.php?title=ReplayGain

Thanks for the link.  So is my reading correct that replay gain can result in some loss of fidelity?
 
Is there a program that can write the relevant data to Flac and Mp3 files in batches?  What is the relevant data to use in this case?
 
Sep 24, 2016 at 9:54 PM Post #42 of 132
  Thanks for the link.  So is my reading correct that replay gain can result in some loss of fidelity?
 
Is there a program that can write the relevant data to Flac and Mp3 files in batches?  What is the relevant data to use in this case?

 
If you have an album / track that is below the reference level, then replaygain will say to add gain to that track. If the track already has content at/near full-scale, this could cause clipping if the gain is actually added. If you're a fan of classical this can actually be a pretty common issue.
 
There are lots of tagging programs. To make album gains you need tracks grouped into albums either by metadata or by directories.
 
Sep 24, 2016 at 10:26 PM Post #43 of 132
  Thanks for the link.  So is my reading correct that replay gain can result in some loss of fidelity?
 
Is there a program that can write the relevant data to Flac and Mp3 files in batches?  What is the relevant data to use in this case?

RRod gave you the best answer.
 
To add replaygain in batches I would use Foobar2000. Jus see the Replaygain options.
 
Sep 29, 2016 at 12:14 AM Post #45 of 132
Oooh nice. I'll be attending the AES conference in Los Angeles and they'll be having a panel called "Recording & Production: RP5 - Mastering for Vinyl." Maybe I'll learn a thing or two about the mastering process!


...except it's at the same time as the "Listening Tests—Understanding the Basic Concept" pane. : (
 

Users who are viewing this thread

Back
Top