How I loathe Apple’s eco system. Not available as a VST.
I use this meter in demo mode and as far as I tested this meter and the iTunes plugin, -1 dBTP avoids any red light. It gave me a ceiling of about -1.5 dB on the last limiter.
Do home recording guys regularly do this??
I dunno, I don’t know much but I thought -1 or -0.5 as you say, was more or less an accepted norm/standard.
I have used -0.3dB for a long time. This discussion has me re-thinking that though, when planning to bounce to 320kbps MP3, which I do a lot. It does seem safer in that case to go with -0.5dB.
1 Like
If you press the “MFIT” (Mastered for iTunes) button on Boz’s “The Wall” limiter, it reduces the ceiling to -1dB.
2 Likes
I “master” to -1db…
as -0.3 or even -0.5db stand the chance to get “turned down” by streaming services’ auto level matching and such… But I have stopped crushing my mixes as well. So maybe I could get away with more.
All the commercial tracks I’ve ever analysed were mastered to 0db, so I’ve always replicated that. What is the reason for mastering lower than 0db?
1 Like
Converting to mp3/AAC creates inter-sample peaks that cause 0dB to go over. Cheap converters (like the one in my iPod) don’t like this, and it comes out pretty distorted. Eg. I have a Gomez album where this is particularly noticeable.
5 Likes
I think we are mainly concerned with ISP’s in mp3 and that’s what started the convo…
In my case, I am “mastering” inside the same daw and mixing rig. So it’s just a safety net.
have fun
rich
Ok thank you Andrew.
Yeah, I was doing the same as you Stan until I read a few articles on the intersample peaks created by lossy compression. Then I started Mastering to -1db. Most of the genres I mix tend to fit best in the -7 to -6db RMS range and lowering the peak level didn’t affect this enough to be noticeable. I’m sure someone will argue that a -7db rock mix isn’t loud enough but I don’t care, I’ve heard many mixes that were louder and I find they all lack a depth when they are squeezed more. But I digress.
1 Like
Hey Bobby! Good to see and hear from you! Good thoughts on the process. I have been using the LUFS method more now than the DR method, but yeah I’m going more for the 9-10 range or even lower in order to not have stuff pushed so hard. Setting the peak lower shouldn’t be a big deal.
1 Like
Hey Stan, I glad many migrated over. Would be a shame for everyone to be lost in the void of the interweb.
I’ve been leaving everything at -0.5 as well, having seen it suggested in various places. But I have to admit I knew nothing about intersample peaks when I did that… well, I still don’t know much about intersample peaks!! 
I ran some checks on stuff I’d mastered to -0.5. It’s NOT enough.
Falbfilter ProL has a intersample peak meter included (didn’t know that) and it showed that some peaks were still rising to 0.9. .
Check this video out. Guy explains and you can see he drops the limit to 0.5 and still gets ISPs.
I guess it’s a question of how often they happen. I mean, many masters over the last 15 years have deliberately clipped the converters to get more volume, so that’s tens if not hundreds of oversamples every second. On the other hand, if you master something so it still has some crest factor but peaks at -0.5dBfs, you might only get a millisecond of clipping IPS’s at the top of a snare transient a couple of times a bar which will probably not really be noticeable - after all, while an actual clip at 0dBfs is encoding a distortion in the waveform by way of a flat top, at least with ISP’s the actual, undistorted signal is fully described in the file - it’s just a question of how the individual playback system deals with those peaks; maybe the analogue op-amps have the headroom to reproduce them, maybe they don’t, it’s hard to believe they’ll sound as bad as genuine digital clipping.
That said, I read recently on another forum by a guy who was around in the early days of MP3, one of the designers of the format suggested 6dB (IIRCWIDTBF) headroom would give the best results. Obviously he was scoffed at - who’s going to waste 6dB of volume?
I’ve decided going forward I’m going to shoot for the Spotify LUFS normalisation standard, as it’s the loudest of the streaming services. They normalise to -11LUFS and I think limit at -1dBfs to avoid ISP’s. In the context of the loudness war, -11LUFS is pretty reasonable, I was surprised how easy it is to get there with just a couple of DB limiting.
So, if you master to that volume and peak at -1dB, Spotify won’t turn you down or up (if it turns your mix UP it might further limit it) and you’ve got room for some punchy transients. And that’ll work fine for CD too, I figure if someone’s going to buy your CD you don’t really need to compete with anyone else volume wise as they’ll either play your CD in isolation or rip it to their computer… in which case it’ll likely be volume normalised and that -1dB headroom will reduce the chance of ISPs!
I sorely hope that’s the direction we’re going in, anyway.
That’s Spotify. Interesting: I didn’t know that, but I’ve been aiming at -11 LUFS for all tracks on my new album.
I may be kidding myself, but the recordings all sound better (to me) - in terms of having decent dynamic range - than my last album.
I agree that much of this stuff has to do with the capacities of the playback medium.
That’s just a plyback loudness, it doesn’t necessarily mean you should master at -11 LUFS, it means that there is no point in deliberately mastering loud any more because it will be physically turned down to -11 LUFS.
I guess that depends on the LUFS of your last album.
I have just started the new Salem album, I’m mastering to -10 LUFS. Not bad (especially for heavy rock/metal) when you consider that the 2013 album was -6, and the last one was -8. I tend to master most of my studio work at -10 now, I find that any lower can be counter productive because what you gain in dynamics you lose in glue. I reckon as long as you’re into double figures you’re maximising the dynamics of the recording.
1 Like
True. My last album was pre LUFs.