The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Reaper Master Track clipping sounds fabulous!
Old 30th July 2017
  #1
Deleted User
Guest
Reaper Master Track clipping sounds fabulous!

So this is very interesting.

I rendered a track to MP3 with the meter set to RMS. I pushed the fader until it has an RMS of -6. Never bothered to look at peaks.

The track is loud and punchy.

There is something going on here between internal headroom at 64bit and the rendered MP3 track output which my DAC is not reading as "clipped".

I noticed when I did this on the mixer track itself it sounded distorted, but most likely because I had effects that were clipping.

I'm going to test this simple volume increase without any effects later today.

What's going on here with the gain staging?

It's supposed to sound like a mangled horrid mess but it doesn't.
Old 30th July 2017
  #2
Deleted User
Guest
Just in case you guys are wondering why you have low RMS with Reaper mixdowns:

The meters are wrong.


Quote:
Originally Posted by planetnine View Post
I think this is wrong, and -18dBFSD should correspond to about 0VU (+4dBu). It should be user-settable as VU is an analogue reading and not everyone's DACs give out +22 dBu for 0 dBFS.



Answer:
You're not seeing things--the three JS "VU meters" that come with Reaper are wrong and not usable. All they are is a peak meter with a needle. They should be fixed or removed from Reaper.

Here's how I setup the master fader to give VU characteristics based on AES17's definition of 0 dBFS with a sine wave. Reaper's RMS calculation is not correct based on the AES17 standard--a sine wave should read the same value for peak and RMS.



Window size of 600 gives an approximation of 300 ms integration time (I visually compared it to the behavior of PSP Vintage Meter)

Offset gives you your desired headroom (I chose 20 dB here for the SMPTE standard).

Display gain must be 3.0 dB to correct for the normally RMS reading meter offset (this is where Reaper takes a math approach to RMS instead of the audio-correct AES17 definition). With a 600 ms integration time, the meter always read -0.1 dB lower for some reason, so I chose 3.1 to correct it such that -20 dBFS tone would read -20 for peak and RMS values.

Red threshold I put conservatively at 0 VU so that I don't stray far from my 0 VU target. 3 is a better choice probably, so you can use the remaining end of the VU scale without too much worry.

You'll also want to change the project's pan law to be -3 or -4.5, the default setting does not make sense for how most consoles operate.

On another note, Reaper's internal tone generator seems to output -0.1 dB less than its specified value. That should be fixed--it's why I had to use mDa Test Tone instead.

Since we're picking apart meters, the last quibble I have with Reaper is that they call everything a VU meter. The meters in Reaper are peak, not VU, and they should not use the word VU to describe them. Renoise does the same thing, so we can expect future generations of sound engineers to be thoroughly confused by "VU meters" forever, which is exactly why the programmers that made the JS meters did so incorrectly!

It's in getting these details right where Reaper could really distinguish itself, because every other DAW is not bothering. How about this, instead of half a dozen arbitrary peak meter preferences, Reaper had two: 1) Peak, 2) VU with adjustable reference (-20 dBFS to -12 dBFS). Imagine a DAW with VU meters! Like the API's and SSL's of yesteryear! Having a VU meter could give people a better understanding of loudness and lead to better recordings. At the very least it would be very pro.

https://forum.cockos.com/showthread.php?t=101210
Old 31st July 2017
  #3
Lives for gear
 
stinkyfingers's Avatar
 

Quote:
Originally Posted by trashman View Post
What's going on here with the gain staging?

It's supposed to sound like a mangled horrid mess but it doesn't.
1)You should reference peak level, not RMS.
2)Sometimes clipping sounds good.
Old 31st July 2017
  #4
Lives for gear
 
Lumbergh's Avatar
 

Quote:
Originally Posted by trashman View Post
... a sine wave should read the same value for peak and RMS.
Old 31st July 2017
  #5
Lives for gear
 
stinkyfingers's Avatar
 

Quote:
Originally Posted by Lumbergh View Post
that is the "Bob Katz" RMS scale...it adds +3 dB to the normal RMS value...
pretty stupid.
Old 31st July 2017
  #6
Lives for gear
 
Lumbergh's Avatar
 

Great, thats not confusing much!
Old 31st July 2017
  #7
Lives for gear
 
stinkyfingers's Avatar
 

Quote:
Originally Posted by Lumbergh View Post
Great, thats not confusing much!
The irony here is that there was confusion when speaking about VU calibration and whether people are using peak vs RMS...so bob decided to just make them (peak/RMS) the same value to avoid confusion.

Old 31st July 2017
  #8
Deleted User
Guest
I've been going over this crap all day with LUFS and RMS and Peaks.

I ran tests, too. More this evening.

There is some relationship between the driver, headroom on your DAC and the peak meter in Reaper.

My mixdown is peaking ar -6db. Fader at 0.
I flip the RMS only meter on the master.
I crank the fader.
I watch the meter on my board for clipping and watch RMS.
When I push the RMS to -1 my board starts to clip and so does monitoring.
Pull back a touch.
Render.
Loud.
Old 5th August 2017
  #9
Gear Addict
 
Pollo's Avatar
 

I am trying to understand what you are saying but your use of language is not very precise which makes it hard. I just tested the Reaper meters on the master track and I find nothing wrong with them. I have set the Master VU settings to Peak + RMS, windows size is 300 ms and display offset is 18 dB. I am feeding it with a sine tone of -18 dBFS RMS. It shows as 0 dB RMS on the Reaper meter, exactly as expected. The peaks show as -15.0 dB which is also correct, given that the peak scale goes from 0 dB down. I don't really understand why they would use different scales but it is not wrong.

If the JS plugins are not correct then my advice would be to not use them. I've never used them.

Now adding 3 dB (3.01 dB to be precise) to the RMS level to make it the same as the peak level for a sine signal, that is what I call wrong. I don't care if it is AES17 standard, it doesn't make sense to me.
Old 5th August 2017
  #10
Gear Addict
 
Pollo's Avatar
 

Quote:
Originally Posted by stinkyfingers View Post
The irony here is that there was confusion when speaking about VU calibration and whether people are using peak vs RMS...so bob decided to just make them (peak/RMS) the same value to avoid confusion.

It will only be the same for a sine wave. On any other signal peak and RMS values will not be the same. So this is indeed a pretty dumb solution. Probably only enhancing the confusion. But we are on the same page here, I can tell.
Old 5th August 2017
  #11
Deleted User
Guest
Quote:
Originally Posted by Pollo View Post
I am trying to understand what you are saying but your use of language is not very precise which makes it hard. I just tested the Reaper meters on the master track and I find nothing wrong with them. I have set the Master VU settings to Peak + RMS, windows size is 300 ms and display offset is 18 dB. I am feeding it with a sine tone of -18 dBFS RMS. It shows as 0 dB RMS on the Reaper meter, exactly as expected. The peaks show as -15.0 dB which is also correct, given that the peak scale goes from 0 dB down. I don't really understand why they would use different scales but it is not wrong.

If the JS plugins are not correct then my advice would be to not use them. I've never used them.

Now adding 3 dB (3.01 dB to be precise) to the RMS level to make it the same as the peak level for a sine signal, that is what I call wrong. I don't care if it is AES17 standard, it doesn't make sense to me.
I pulled that info from the Reaper forum.
I've found that to be incorrect also.

Still, I am pushing my master fader well into clipping territory and get much louder and punchier mixes.

I have the BX Master on the master bus. All set to 0. Im driving it with the master fader. It's pushing the plug so it grabs a few peaks. The master fader is beyond clipping. But the mix sounds good. When I reverse this process it sounds like ass. I also have much lower LUFS.

It's this way with each limiter I use:
FG-X
Ozone
BX Master
Elysia
Precision

It's exactly the opposite from what I thought.

I have Izotope Insight on the end, so my metering is now very clear and precise.

What I've learned:

Drive the master fader into your master FX bus. Set it to RMS and push it as far as your ears and DAC meters tell you.

So now I'm wondering: if my DAC has higher headroom can I push it harder?

The Dangerous Audio The Source has very high headroom. I'm wondering if I can get even louder mixes.

In the end none of this will matter with new LUFS standards.
Old 6th August 2017
  #12
Lives for gear
 

DAcs that use passive i/v stage with op amp gain/buffer stages have generally 3 db headroom above odbfs. This area was originally used for the slew rate effect of the i/v+buffer to restore transients. how ever in the beginning of the end of cds that had dynamic range to them, They learned how to soft clip enough to cause the typical DAC to restore part of the waveform plus its transients, the result measured was a higher peak-to-peak voltage. Some mastering guys call this "Transparent Limiting". If you go overboard it results in a lot of undesirable harmonics that results in a very nasty edgy signal that is very over blasting, distorted sound a lower levels.

You see, you don't really have to race to 0dbfs or go over either, because its the distance between the loudest and softest sound in the track will dictate its perceived loudness.

but I also want to tell you too about these mixers in the newer daws. After 2005, Daws were built to mixer channels that have a higher bit rate. on the masters and mix groups the least significant bits are chopped or reduced when the signal goes over 0dbfs. the more bits over 0dbfs, the more lower bits are chopped or reduced. This also results in a transparent limiting signal. Before this, if you went over, bad crunchy, spitting, nasty noises would happen if you barely touched the red in DAWS.
Old 6th August 2017
  #13
Deleted User
Guest
Quote:
Originally Posted by Audiospacific2 View Post
DAcs that use passive i/v stage with op amp gain/buffer stages have generally 3 db headroom above odbfs. This area was originally used for the slew rate effect of the i/v+buffer to restore transients. how ever in the beginning of the end of cds that had dynamic range to them, They learned how to soft clip enough to cause the typical DAC to restore part of the waveform plus its transients, the result measured was a higher peak-to-peak voltage. Some mastering guys call this "Transparent Limiting". If you go overboard it results in a lot of undesirable harmonics that results in a very nasty edgy signal that is very over blasting, distorted sound a lower levels.

You see, you don't really have to race to 0dbfs or go over either, because its the distance between the loudest and softest sound in the track will dictate its perceived loudness.

but I also want to tell you too about these mixers in the newer daws. After 2005, Daws were built to mixer channels that have a higher bit rate. on the masters and mix groups the least significant bits are chopped or reduced when the signal goes over 0dbfs. the more bits over 0dbfs, the more lower bits are chopped or reduced. This also results in a transparent limiting signal. Before this, if you went over, bad crunchy, spitting, nasty noises would happen if you barely touched the red in DAWS.
Very very interesting.

It's amazing how much beefier and louder my basic mixdown can be by reversing the order of peak monitoring with the FX bus and using Insight.

I had LUFS of -16 with the normal master fader keeping peaks below the 0dbfs. I was pushing the limiter on the FX bus with processing behind it and some processing on the track.

When I pushed the track itself it was clipping crap. But not on the master bus.

It must be this buffer you spoke of.

So the track is super hot and loud on my DAC anolog output and the master fader is cranked yet the limiter is just grabbing those peaks. More tests soon.

Going to see about squeezing more on the track and using a clipper plug to see if I can get more "LUFS".

Thank you for your insight and learning.
Old 6th August 2017
  #14
Lives for gear
 

yes, but you don't have to crank it to get a loud track. Just shorten the distance between loud and soft in each track, then mix them together at around the same proportion.
Old 6th August 2017
  #15
Deleted User
Guest
Quote:
Originally Posted by Audiospacific2 View Post
yes, but you don't have to crank it to get a loud track. Just shorten the distance between loud and soft in each track, then mix them together at around the same proportion.
You mean compress?

Anyway, there's something else going on here with the gain staging that I never understood before.

The meters on your DAC are what's important for the final mix to be loud.

This is my conclusion. This is my actual results.

I'll post some audio clips later as a reference.
Old 6th August 2017
  #16
Lives for gear
 

Here is a snippet of one of my mixes. I use compression and limiting in various stages of tracking and mixing. Notice I leave 3db of room on purpose for mastering.


I usually end somewhere -6 peak on my mixes. but when I want a little bit more dynamics, I do a -3 peak.

Of course there isn't really a key method. I can do this at -10 or even -20 peak.

the trick is dynamic range control and its ratios to the other tracks in the mix.
Old 6th August 2017
  #17
Deleted User
Guest
Quote:
Originally Posted by Audiospacific2 View Post
Here is a snippet of one of my mixes. I use compression and limiting in various stages of tracking and mixing. Notice I leave 3db of room on purpose for mastering.


I usually end somewhere -6 peak on my mixes. but when I want a little bit more dynamics, I do a -3 peak.

Of course there isn't really a key method. I can do this at -10 or even -20 peak.

the trick is dynamic range control and its ratios to the other tracks in the mix.
What's the RMS on this? LUFS?
Old 6th August 2017
  #18
Lives for gear
 

Quote:
Originally Posted by trashman View Post
What's the RMS on this? LUFS?
I don't look at that stuff, probably 17 lufs with rms [email protected]
Old 6th August 2017
  #19
Deleted User
Guest
Quote:
Originally Posted by Audiospacific2 View Post
I don't look at that stuff, probably 17 lufs with rms [email protected]
Oh ok.

My last mixdown was -5RMS -8LUFS

But this was pretty loud for sure.

However the same mix reversed order is significantly lower.

Last edited by Deleted User; 6th August 2017 at 08:56 PM..
Old 6th August 2017
  #20
Lives for gear
 

I think my lufs are positive instead of negative, but I never understand lufs, my ears say loud and the distance between I use, not meters.
Old 6th August 2017
  #21
Lives for gear
 

Quote:
Originally Posted by Audiospacific2 View Post
I think my lufs are positive instead of negative, but I never understand lufs,
Don't think you can even have positive LUFS. It's a loudness measurement of a digital signal. Its reference is 0dBFS in a sense, and since you can't go over that you can't have positive LUFS, at least not something like +17LUFS.
Old 6th August 2017
  #22
Lives for gear
 

Quote:
Originally Posted by mattiasnyc View Post
Don't think you can even have positive LUFS. It's a loudness measurement of a digital signal. Its reference is 0dBFS in a sense, and since you can't go over that you can't have positive LUFS, at least not something like +17LUFS.
ok, I'm still clueless about lufs. That came around several years after I started doing this stuff.

To be honest, I don't really look at my meters, but I believe things hover around the -6 area when I mixed that one above.
Old 6th August 2017
  #23
Lives for gear
 

Quote:
Originally Posted by Audiospacific2 View Post
ok, I'm still clueless about lufs. That came around several years after I started doing this stuff.

To be honest, I don't really look at my meters, but I believe things hover around the -6 area when I mixed that one above.
I hear you. The measurement using either LUFS or LKFS came about to manage broadcast content on TV. So it's a development of a more "accurately" weighted measurement that corresponds better with how we perceive loudness. Sort of like an improved RMS if you will. The same algorithm is used in the US as in Europe as long as people use those specific meters that generate readings called "LUFS" or "LKFS".

Since it's measuring digital signals the ceiling should be 0LUFS. I've never ever seen a signal go over that. This measurement is on a scale where the reference can't really be moved, so whereas people move their dBVU around in their DAWs using plugins one can't do that with this method.

I didn't want to disturb your discussion btw, just wanted to point that out so there's no confusion.
Old 6th August 2017
  #24
Lives for gear
 

So if lufs is a distance measurement from 0dbfs, then I guess it want to quantify the distance between loud and soft. If that is the case, then you want that number to be greater for dynamic range. but if that is so, then a lower number meaning would deliver a lower dynamic range and less impact because the acceleration potential isn't there which would be loud but no definition, and could be distorted sounding at lower listening levels.

dynamic range is good. but I do believe Alan Parsons once said: "There is a Balance between too much and too little dynamic range"
Old 6th August 2017
  #25
Lives for gear
 

Quote:
Originally Posted by Audiospacific2 View Post
So if lufs is a distance measurement from 0dbfs, then I guess it want to quantify the distance between loud and soft. If that is the case, then you want that number to be greater for dynamic range. but if that is so, then a lower number meaning would deliver a lower dynamic range and less impact because the acceleration potential isn't there which would be loud but no definition, and could be distorted sounding at lower listening levels.
I think you're misunderstanding it actually.

It is a measurement of loudness relative to 0dBFS. But the term "dynamic range" is a bit tricky here because of the measurement 'period' used. In post-production for TV we typically measure the entire program from start to end, and it needs to be -24LUFS/LKFS (US) +/-2LU (LU is "Loudness Units", which basically means +/-2dB). But the algorithm ignores silence. So if there's quiet in the program, either because no sound is heard within the show for some reason, or if there's a commercial break, that doesn't get included in the measurement.

So basically it's an average loudness over whatever timeperiod you measure minus any complete silence.

This means that you can have a wide dynamic range and have a reading of -24LUFS. But if your absolute maximum peak is -10dBFS, then you could hypothetically increase the level of everything with 10dB and not go over 0dBFS. You follow? So you've just increased the gain of everything by 10dB, but the relationship between loud and soft within that signal hasn't changed, so your new loudness measurement will be -14LUFS with maximum (absolute) peak of 0dBFS. But your dynamic range of your content is the same.

You see the difference?
Old 6th August 2017
  #26
Lives for gear
 

ok, so why bother looking at this? because everything mixed at the same relative level is not always going to make musical sense.
Old 6th August 2017
  #27
Lives for gear
 

maybe I just let the mastering gods worry about it.
Old 6th August 2017
  #28
Lives for gear
 

but if anyone wants to know, this was mixed ITB using reaper.
Old 6th August 2017
  #29
Lives for gear
 

Quote:
Originally Posted by Audiospacific2 View Post
ok, so why bother looking at this? because everything mixed at the same relative level is not always going to make musical sense.
What do you mean by "relative level"? Relative to maximum level?

For music mixing/mastering I suppose it's up to you to figure out if it's useful. If you're mixing/mastering a song for example and you start suspecting that it's too soft but aren't sure you could measure your entire song and compare it to a song that is your reference. The value (LUFS) would tell you what the perceived loudness would be for each (over the measured period). It also measures "loudness range" which is similar to dynamic range, so that's another possibly useful parameter.

If you do that to several pieces of music then perhaps you have a guide for when you're mixing in general, meaning that for genre X you want to reach a value of Y LUFS. So when you're working you can push towards that value without having to A/B with other tracks/songs. You know that if you're close using that measurement you're close to the other. But if you find this a bit confusing or unnecessary I don't entirely disagree.

For broadcast it's different; the point of it is standardizing loudness so that the viewer at home doesn't have to reach for the volume control between content. It's all written down and there are production guidelines from the networks etc.
Old 6th August 2017
  #30
Lives for gear
 

well relative to the loudest track that was mixed.

I can see why is there a school of thought around this: LUFS. But applying it as a paint by numbers method for one genre might make those mixes not have the individuality they deserved.

Don't give me wrong, but couldn't this be done using a lookahead AGC limiter? Granted there could be a guideline defining what is required, but does that guideline work with every sound, no matter if it is dialogue or music ?
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump