3 ms·
-If you wanted to, you could master an LP with 0 dynamic range; it would work just fine. (Provided the average signal level was low enough to keep the needle fr
by lb1lf 10y ago
-If you wanted to, you could master an LP with 0 dynamic range; it would work just fine. (Provided the average signal level was low enough to keep the needle from going haywire attempting to follow the LF grooves)
What quickly becomes a problem for LPs is too much energy in the LF/bass region, which will cause the needle to skip and jump all over the place.
The RIAA correction curve is a measure to alleviate this problem, as it strongly emphasizes low frequencies on playback (and, inversely, reduces them before cutting the master disc, leaving much less LF content in the grooves than would otherwise be needed for accurate reproduction.)
I suspect (Mind, this is just a guess) that when someone says an album has 'a separate master for vinyl', what they are really saying is that someone took a copy of the album master before it was sent through the last loudness war compressor - and sent that copy to the LP plant, using the compressed-as-heck master for any other format.
- beat 10y agoThere's a lot more to mastering than "sent through the last loudness war compressor", for cd as well as vinyl. Not a guess - I've sat in the room with a mastering engineer while working and talked to him about it.
- semi-extrinsic 10y agoDo elaborate, please! :)
- beat 10y agoMastering at album-length is largely about coherence... making a bunch of separately recorded things with often wildly varying sounds sound like a single piece of music. It involves a combination of gentle compression and brickwall limiting, broadband equalization and high-Q deep-cut equalization. Some of it is visual metrics (especially looking for frequency problems), and some of it is just ears. A good master will sound good on any playback equipment. This is where I point and laugh at all the people lauding the "accuracy" of their favorite medium. Speakers and amps are the limiting factor, not the medium! I've been dinged by a mastering engineer for accidentally having a 32hz synth tone (a sub-bass sound) on a mix. He eq'd it away. My mixing system couldn't reproduce it - hardly any can. But it wastes tremendous power and distorts amps and speakers, so out it goes. For cd, it's usually brickwall-limited to no more than 10db max dynamic range (often no more than 6db). This makes things sound "better" on typical playback equipment, which is... not good, compared to the sound of live instruments in a room. Vinyl is a different beast. It needs RIAA equalization on the master (which is reverse-eq'd by the playback equipment) to managed the size of bass grooves and keep the needle from jumping out of the track, and to make the highs physically big enough for the needle to pick up well. Dynamics are constrained by the physical limitations of the media. So it's a separate master. That's not a big deal, though. The constraints of vinyl slot in quite well with the constraints of speakers and amps! This is why vinyl, despite its on-paper inferiority, can sound as good or better than digital. But in the end, it's all about the speakers. I mix on a pair of Tannoy System 12 speakers and an old UREI studio amp that would be the envy of many hi-fi nerds - but I don't consider a mix done until it sounds good on my stock iPhone earbuds and shitty car stereo. Mix to the output, not the medium.
- lb1lf 10y agoI believe you; I hardly know the first thing of the process - and should've made it clear that I didn't. (the comment being intended somewhat tongue-in-cheek)