|
Post by scumbum on Dec 28, 2016 10:57:47 GMT -6
To be played back on all systems and converted to MP3 without distorting , what level should peaks be hitting on a Mastered Track ?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 28, 2016 14:55:53 GMT -6
To be played back on all systems and converted to MP3 without distorting , what level should peaks be hitting on a Mastered Track ? I'll assume you mean RMS, because if you're peaking over 0dB you've got bigger issues than distortion . There's nothing set in stone, I try to avoid the loundess wars and focus more on quality if at all possible and dynamic music can have a heavier impact on how far you can push. Usually I aim for around -12 RMS and a decent lookahead brick wall usually keeps it around that, with maximum peak limiting being around -0.3 to take into account intersample peaks. At that level it translates fine with with little to no distortion.
|
|
|
Post by mrholmes on Dec 28, 2016 15:00:57 GMT -6
For peak level I go minus 9 dbfs and I don't care what other people do.
|
|
|
Post by scumbum on Dec 28, 2016 20:20:16 GMT -6
I didn't mean RMS , but the peaks .
The output of a limiter can be adjusted so you don't hit right at zero , but instead something like -0.1 . I read that you should aim for -0.6 if your gonna convert to MP3 . Is that true ?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 28, 2016 20:23:12 GMT -6
I didn't mean RMS , but the peaks . The output of a limiter can be adjusted so you don't hit right at zero , but instead something like -0.1 . I read that you should aim for -0.6 if your gonna convert to MP3 . Is that true ? Well I did already mention it just in case you asked , as said in my original post -0.3 is usually fine it gives enough for the limiter to catch intersample peaks correctly (usually). I've not used every limiter out there (just a fair chunk of em) and I've never had an issue. AFAIK one of the mastering from Nirvana's albums in the 90's was -0.3 peak limit. Doubt it was ITB (as in slapping ozone or something across the master), but I've seen many albums usually mastered around that limit.. Besides from some major lables pushing it for the hell of it, I think one of the Radiohead albums was -6.5 RMS and -0.09 (peak). I wouldn't recommend it personally..
|
|
|
Post by popmann on Dec 28, 2016 21:55:49 GMT -6
It's been years since I've bought a major label record that the peak wasn't 0dbfs....and "true peak" +.8-2....that said, it also sounds like ass. It's why digital attenuation pre DAC (or on the DAC chip itself) is important to modern masters sounding OK.
I always use -.5dbfs for my peak normalize for full scale export. I do that because I did -.4 for years but at -.5 I've never been able to have an encoder or anything go over 0 for ISP or "true peak" (reconstruction estimation)....-.4 occassionally can register one--at least in whatever reconstruction modeling they do to detect ISPs....
|
|
|
Post by EmRR on Dec 28, 2016 23:25:44 GMT -6
The couple large facility mastering guys around here tend to deliver -0.2dBFS peaks.
|
|
|
Post by scumbum on Dec 29, 2016 0:05:35 GMT -6
Well as it turns out the Shadow really does know ............
A lot of Pro CD's like popmann has said are hitting 0 .
I'll be safe and go for -0.5 .
|
|
|
Post by Bob Olhsson on Dec 29, 2016 18:47:25 GMT -6
-.3 or .5 keeps you out of trouble in a DAC but -1 is best for lossy-encoding. While lots of stuff is clipped, it's silly to do that with material that will be streamed.
|
|
|
Post by ChaseUTB on Dec 29, 2016 19:29:14 GMT -6
Ok for my genres of rap trap rnb pop top 40 edm I use ozone 5/ FF proL / limiter no.6 with InterSample peak protection enabled, and usually aim for -.5 for MFiT / aac / MP3 encoding. Sometimes -.8dbfs is necessary. I find that 320 Kbp/s constant bit rate sounds better to me than Variable bit rate at 320, which also sounds better than MFiT which 256 Kbp/s VBR AAC. My rms levels may be absurd to some but -9 rms ( dbfs ) to -5 rms. Last mix and master I posted via the de esser thread the rough mix was DR3, with a clipped beat of DR4 and had peaks +3-+4 over O dbfs...
|
|
|
Post by Bob Olhsson on Dec 29, 2016 19:33:50 GMT -6
The problem with lossy encoding is that the first thing the audio hits is a 1/3 octave filter bank. That clips very easily. There's just no reason to push the level up another half dB.
|
|
|
Post by ChaseUTB on Dec 29, 2016 19:57:55 GMT -6
When the client pays the client gets what the client wants... When it comes to my music, productions, mixes and masters, I am more conservative for the genres, DR9-DR6. Mixes peak around -8 to - 6dbfs. I can normalize to peak level, say -.5 dbfs and use a limiter set to prevent ISP if that print or bounce is for a lossy destination, and I think that keeps dynamics and usually is super loud. I also prepare for that loudness level from the beginning of the production. If I feel or client wants louder after normalizing to -.5 dbfs, I can use Pro L add 2-3 db ( - 1 to - 2db GR ) and the next thing we are at DR5 and less, which on most consumers DAC past 80-85% volume will cause distortion...
|
|
|
Post by Ward on Dec 30, 2016 7:57:04 GMT -6
Many many mix tracks are 'pre-mastered' anyhow, and many are mixed at unholy high levels. My main mastering engineer has to routinely knock a lot of things down because they are 'blasting'.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 30, 2016 8:24:44 GMT -6
The irony of spending tens of thousands on equipment with the lowest noise floors and highest dynamic ranges to squish the end product into a distorted brick.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 30, 2016 16:01:12 GMT -6
Andrew Scheps on "that" Metallica record - " I won the loudness war" He also said he had death threats because of it ... serious business this ....
|
|
|
Post by swurveman on Dec 30, 2016 16:15:18 GMT -6
Andrew Scheps on "that" Metallica record - " I won the loudness war" He also said he had death threats because of it ... serious business this .... I saw a video of him talking about how he puts limiter's at .01 on individual channels. The guy seemed proud of being known as the loudness guy, which if you think about it, makes sense from the perspective of wanting to get work from label guys with research saying louder is always better.
|
|