Faders

[-mjk-]

@David Porter your not hearing the distortion comment is very revealing. Now I really wish I was back home in my studio! I'm in south Taiwan until Monday.

The actual level of the mixdown does not matter. It needs to be as clean as possible and if there are heavy peaks, that low level might be what it takes to keep those from clipping. I'm going to check out the mix in RX8 and then later, I'll see what I can do with it using Ozone 9 and Finalizer.

 

[BazzBass]

+I hear no distortion on the second, mastered mp3.

When I was recording in Reaper and dropping the file into dropbox to share with my bandmates, it always was too soft. Reaper showed it at normal volumes,and it played fine on my pc,but as soon as I shared it, the volume dropped.
I blame windoze and it's why i bought an old 2488 mk I lol

 

[Arjan P]

OK so I analyzed your files in Wavelab, and the original mix peaks at -34 dB! Like I said, that's unnecessary, and though it is a 24 bit file, I don't see why you'd throw away 30 dB of the (theoretical) 144 dB that these 24 bits represent? In practice you may have 90 dB before you get into noise (that's with very good hardware!) so these levels are simply too low for a stereo mix.

The mp3 peaks at -0.1 dB True Peak, has an integrated loudness of -12.5 LUFS (which is quite nice IMO) and has a loudness range of 6.8 LU. These last two values don't look strange at all.

The -0.1 dB peak however is troublesome, especially since this is a reading after decoding the mp3 (which is what Wavelab always does). I think this is what your distortion comes from. Common practice in mastering is to keep digital peaks of the stereo master at -1.0 dB at least, so this is what most mastering engineers set their limiters for (or an even bigger margin). The point is, that when encoding an mp3, information gets lost, and also the peaks of the recording change. If you have a 1 dB margin in the master before conversion, the encoding process may leave you with a file that has a digital peak of -0.5 - and that's a safe value for most players.

 

[Di Ba]

Thanks Arjan....... so in a nutshell it's recorded too low which is where the distortion is coming from, if I understand correctly?

Will I be able to boost the file using digital trim and what's left on the faders to get a stronger signal to mix or will I need to start again do you think?

 

[Freebird]

Yeah... thanx Arjan... VERY interesting information indeed! Will copy/paste this info to my personal growing "database" of useful knowledge!!!

 

[-mjk-]

Wow @Arjan P that is well below what it should be (when I said the level "does not matter" I was assuming a basic amount above the noise floor!). Good work analyzing the track. I'm still down south and away from the studio.

 

[Arjan P]

Thanks Arjan....... so in a nutshell it's recorded too low which is where the distortion is coming from, if I understand correctly?

Will I be able to boost the file using digital trim and what's left on the faders to get a stronger signal to mix or will I need to start again do you think?

@Di Ba No, I would say you have two different issues:

1. The mix is way too soft - but this just goes against the common practice of getting good levels while still staying clear far enough from 0 dB full scale or true peak. Let's say it's better to aim for -6 dB instead of -34.

2. Your mix was mastered too hot, getting too close to 0 dB. I don't know what the value was of your mastered .wav file (if you got that file in the first place), but it must have been close to 0 dB - too close to it anyway. I can only tell from the decoded mp3, which has a true peak of -0.1 dB. Since true peak is an estimation, my guess is that in different players this too hot master causes distortion. So try to see if you can tell your mastering service to stay further away from 0 dB with the .wav, so it will also be safe after converting to mp3.

BTW, there is this great site from Ian Shepherd, where you can drop a mastered file to see whether it will be turned down by most streaming services, and by how much. BTW this is not really related to your issue, since this is about loudness. Check it out: https://www.loudnesspenalty.com/

 

[skier]

One of the things about which we need to be careful is to not normalize too close to 0 db. While normalizing under, or even at 0 db will keep the digital signal from going over, that signal needs to be converted to analog and smoothed at the end of the chain. When it undergoes that process, the integration and smoothing functions can, and often do, raise the amplitude of some of the analog output peaks several db over 0 and that is caused by inter-sample distortion. For that reason, I like to limit my peaks to -6 - 9 db. That provides pretty much maximum dynamic range like Arjan described, but under the 0 db limit even after ADC.

 

[Di Ba]

Ok guys that's informative. Thanks ever so much for taking the time to look at the problem.
I'm new to the dp 24 so I reckon that I've not been paying attention to the numbers marked on the db meter .... I was trying to keep levels well below max and still have what I considered a" decent sound" in my headphones and monitors.

I was under the assumption that a lower level signal could be boosted without problems because there was a lot of headroom but never considered the negative impact of too low a signal.... oops!

I'm also guilty of assuming the Dp24 would react like the Dp01 that I've been using for years when it came to metering . If the peaks came anywhere near half way up the meter on the Dp01 ( there's no db marked on that, only a max /ol level) I would get distortion so levels would be kept deliberately low..... I've had great results with the DP01 .... and the last album I made had a lot of airplay on blues stations in europe, uk and the Us..... if anybody wants to hear some of the tracks I'll post a link, but I'm not on this forum for that.

So I think I've decided to start again and redo all the numbers I've recorded since getting the 24sd.
But please let me know if you think anything can be done to salvage the tracks I've already done, as that would save a lot of work

Thanks again for all your help .. it's great to be on a forum that doesn't go off on obscure tangents .... Thank You !

 

[Arjan P]

@Di Ba Maybe you got the wrong impression from what I wrote after my analysis in Wavelab, but there is nothing wrong with your stereo mix - other than that it is not using the full benefit of the dynamic range you have with 24 bits. Otherwise the mix is fine!

The real issue with the distortion is the too hot mastering. So if anything, only the mastering would need a new go. Don't go redoing something that was already good, but only not using the full potential of the medium.

 

[Di Ba]

@Arjan P ...... ah ok! I understood you to mean that maybe there wasn't enough information in the mix because of the low recording levels to give a good master AND the mastering was too hot.

The problem with these online mastering sevices is that they don't tell you to what level they master the tracks...... on the one I use, they give you a choice of low medium and high intensity mastering, same with comp eq etc ............. I'll give it another go with lower intensity all round.
Thanks again Arjan.

 

[-mjk-]

@Di Ba This video explains the issue better than anything I've yet seen:

I just got back from south Taiwan (later than I thought). I'm going to use Ozone 9 on your tune and see what I can do with it. I have all day tomorrow open.

I'll give it another go with lower intensity all round.

Don't spend any more money on that online Mastering service just yet.

 

[BazzBass]

thanks for sharing that link,a lot of good advice there