Stereo file too low

[-mjk-]

@Arjan P truthfully, I like your production values and clearly you know what you are doing. It comes down to style and how we were trained. I like to have more going on with the limiter so I like to have tracks that are around -20/-18. Streaky shared his mastering template for Pro Tools and he starts out with a track where he imports the mix. The first thing he does it turn it down to give his limiter more room to maneuver. Then he sends the output of that track to another track where he adds his chain. I have a similar approach with the limiting.

We should do a new thread on mastering techniques in the Recording 101 forum.

 

[Arjan P]

@-mjk- It was more your remark saying 'the lower the better' that triggered my response. I try to stimulate mixers to use the headroom they have, so in my view a final mix @ -8 dB is better than a final mix @ -18 dB. Which doesn't mean that I never lower a track's initial level. My mastering approach has the limiter as a last plugin, and anything before it may already (well, I should say, will) increase the level - this would basically be EQ and compression, either in combination (multiband compressor or dynamic EQ) or separate. The best final mixes hardly need anything at all though and I will only bring them at a desirable loudness level and have the true peaks at - 1 dB.

Good idea to start a mastering topic!

 

[Arjan P]

But I'm starting to think that maybe normalizing my file I could just get what I need. I was dubious of this option as I thought it altered the signal in other ways that just raising the volume. But maybe if I normalize using Peak normalization I can preserve the natural sound of the instruments and the inter-relation between them, only that everything sounding at a more professional level, right?

@Bambi It very much depends on your type of music. If there's acoustic drums or percussion involved, normalizing will probably get you nowhere. Normalization does nothing else than raising the volume up to the limit of the signal range. So if a song has one snare hit at -1.5 dB and you normalize to -1 dB, the whole file will have its level increased with 0.5 dB. Not much use there.. If I master a track by someone else I always ask them not to normalize, sine I'm going to need the headroom anyway in mastering.

 

[Bambi]

Hello MJK and Arjan. MJK, thank you very much for encouraging me to try different things. I've tried EQ sometimes and I may try compression just for curiosity, and compare with the same recording uncompressed. What I'll sure need is normalization to get a better final volume. I'm looking for software or online options with enough quality so not to worsen the signal to noise ratio or other variables. And I'm wondering if there is any kind of external device to normalize without the computer. It would obviously be a much more expensive option, but I don't even know if it exists or what's the name of the device.

Arjan, thank you for the info. I do have a very soft percussion. I also have a main keyboard line and a keyboard bass accompaniment below. I've measured my files levels with an online sound meter software and it gave me back different results depending on the file, but all of them with peaks lower in volume than -10 LUFS; for example one song had the highest peak at -11 LUFS and an integrated value of -17 LUFS, and another reached a maximum peak of -20 LUFS with integrated value of -25 LUFS. I guess that "integrated" means something like "average". The results are not completely reliable as, in addition to being different songs, I'm experimenting with my mixing techniques from one to another. The little difference between the peaks and the average level I guess it's due to the distorted sound I use for the keyboard and bass lines; the keyboard line has a very narrow dynamic range (around 8 dB), and the keyboard bass line barely 2 or 3 dB. My soft percussion has a relatively narrow range too, compared to other noisier percussions.

The reason for me explaining this to you is to show you the inevitable need of rising my files final volume. I may be still doing some things wrong, but I don't know if I'll be able to get a louder mix from where to start. However, there may be no need of it if I can just normalize my final mix and get a more professional volume level, right?

 

[-mjk-]

@Bambi your levels seem pretty normal to a bit low.

One of the easiest and more transparent mastering solutions is TC-Electronic Finalizer. Of course one can over-do it, but after using it for quite some time, I find that I use it for more acoustic or folk type music. The best thing about it that you can just drag the slider on the limiter for whatever LUFS-I level you need (and you are right: the integrated level is calculated across the length of the file).

Why don't you post a link to one of your mixes and let some of the engineers on the forum take a crack at doing the mastering? It won't cost you anything and you can see what happens.

 

[Arjan P]

@Bambi I agree with MJK but want to add that since we're now talking about loudness: normalization will get you nowhere - unless we're talking about loudness normalization. That's very different from 'plain' normalization.

FYI; most music streaming services are using a loudness normalization level of -14 LUFS integrated, so with -17 you're not too far behind. A -25 LUFS song would need some work IMO.

 

[Bambi]

Hello MJK and Arjan, I'll tell you my latest experiments. I've downloaded Audacity just to try and see how normalization works. It has several options; one of theme is plain peak normalization, but then it has 2 types of loudness normalization: RMS and another called "perceived volume". In both you can set the exact levels of dB's or LUFS to which you want to normalize. I'm trying all the different options to understand how they work and see what results I get. By the moment, as Arjan suggested, my favourite option is using Loudness normalization, especifically the "perceived volume" type. I've targeted at -14 LUFS and it sounds great. This is starting to sound much closer to what I had in mind.

Regarding the idea of sharing a link to get it mastered by some engineer here at the forum, it's sounds nice. It sure would serve me as a clue of how well my music can end up sounding. It will take me sometime until I make a complete recording and mixing of my 3 instruments, but I'll try to upload a link when I have it. Thank you for your suggestion

 

[-mjk-]

Sounds good @Bambi. You'll get this by doing exactly what you're doing. Experiment. It's just electrons

 

[Bambi]

I've watched closer to my normalization results, and there are some things that I don't like and others that I don't know if I'm doing right. I've been able to check in Audacity the levels of one of my mixes before normalizing. It has several percussion peaks around -3 dBFS, as Arjan had foreseen, so if I normalize by peak I don't get any improvement, and when I do loudness normalization I get a much higher sounding result but my dynamic range has been greatly reduced. That means that limitation has been applied in the process of loudness normalization. My original mix had the keyboard introduction peaking at -21 dBFS and then, when the percussion comes in, the peaks reach -3 dB. If I loudness normalize to -14 LUFS, my keyboard intro peaks at -8 dBFS and, when percussion comes in, there is 1 single hit past 0 dBFS, while the rest of the peaks only reach o dBFS.

I've compared my original mix with the one loudness normalized, trying to match the volumes approximately by ear with my headphones and, while the loudness normalized sounds very good, I kind of like my original mix better for the dynamic range, but I don't know if I may be unconsciously biased. I also find the keyboard bass more annoying on the loudness normalized track, perhaps because the loudest peaks affected by the limiting correspond, in my case, to a high pitched percussion drum; so part of the treble frequencies of the song have been lost in favor of the lower ones.

Anyway, as this is difficult to explain with words, I'll try to record something in the following weeks and share it with you so you can tell me what you think

 

[Arjan P]

@Bambi Yes, you found the downside of loudness normalization - you lose dynamic range, and the EQ balance has changed because of the way the energy in the original signal was present (the bass sounds of the keyboard). There is therefor more to mastering than just going through a limiter.

The signal going through 0 dBFS should not happen in a mastering plugin, but probably the loudness normalization in Audacity only does digital peaks and not true peaks (which is more or less a prediction value that anticipates where a 0 dBFS digital signal will result in a higher value after DA conversion). Looking in my Audacity copy, I see that there's not much to set, only the perceived loudness value. In a more sophisticated plugin, you should also be able to set the digital (or even true) peak level, which I normally set at - 1 dB. This for future conversion to lossy formats, which can increase peak levels.

 

[-mjk-]

@Bambi that is the issue right there. You nailed it. Mastering engineers are always trying to balance loudness and dynamic range. The truth is, Mastering can achieve some amazing results that ruin the listenability of the mix. That's not supposed to happen. Sometimes when mixing, the engineer has to consider what is likely to go on in Mastering down the road. For acoustic music I use the TC-Helicon Finalizer because it's very transparent and has some great "soft" and neutral presets to get you started.

What are you experiencing is because you don't have such a wide range of adjustment. It's not all or nothing. You can dial in the amount of compression and the gain of the limiter to reach the target you want. Having control over the EQ is very important in Mastering as the limiting itself changes the spectrum. Streaky has said that if you take care of the bottom end properly, "the tops pretty much take care of themselves." Most of the energy is at the low end, so compression and limiting have more effect in that part of the spectrum. With the limiter in the chain, by compressing the audio, that itself will generally have the effect of making the whole track brighter. You have to work with your Mastering engineer as well as your Mixing engineer to get what you want. Often they have good suggestions for making your stuff sound "like a record" and get you into the realm of a commercial mix.

FWIW, I thought you articulated that very well. As @Arjan P has confirmed, Audacity is probably not the best thing for Mastering. Finalizer has a 30 day free trial and it's worth looking at.

 

[Bambi]

Hello Arjan and MJK, thank you very much for your comments. I hadn't realized that limiting affected EQ, not only dynamics. It's a very interesting finding. I hadn't thought of Audicity as a mastering tool but rather as a free and easy way to try normalization. But, on one hand, it's normalization options are limited, as Arjan says, because being able to set the true peaks it's important. And, on the other hand, i imagine that it's mastering options must be very limited too. I've taken a look at the Finalizer but I don't have any DAW to use it with, as I don't like to work with the computer. Maybe only for the normalization part. I like the idea of anticipating the normalization results to set the balance, EQ, compression, etc, according to that. I'm going to take a look at normalization dedicated software, too. And, for curiosity, is there anything as a external hardware normalizer? I don't know if hardware limiters have options like precise loudness setting to an specific LUFS value, limiting algorithms similar to RMS or ear perceived loudness, and things like that

 

[-mjk-]

Finalizer is a standalone application, @Bambi.

The original Finalizer was a hardware unit. Mastering hardware is frighteningly expensive.

For free, you can upload stuff to the Finalizer website and have it analyze it for you. Also, loudnesspenalty.com is another site where you can see where your levels are.

 

[Mark Richards]

At about $200US, TC-Electronic Finalizer is a gift.

As mj says, it's a free-standing computer program. All you need do is export your Model 12 stereo mix, move it to your computer, and import it to Finalizer. Finalizer works its magic, and you then export the result as mastered recording. No D.A.W. required.

 

[Bambi]

Hello guys, thank you very much for the info. In case I opt for a software mastering tool, I'll consider the Finalizer. I've also tried the Loudness Penalty web and it's interesting. Regarding normalization options, I'll tell you what would be my ideal in case that you know any solution. Doing the whole mastering process would be too complicated for me, as I don't know anything and feel reticent to learn about it. So I would need something similar to what I'll describe next:

1) At risk of resulting very stubborn, what I really am, I'd like an external device in case it exists

2) As MJK says, complete mastering hardware are luxury devices. However, what I would really need would just be a dedicated normalizing tool, be it a normalizer, if it even exists, or a limiter with the possibility of setting the exact LUFS and true peak value, and any other important value there might be

3) That it had some kind of preset or automatic function to execute the normalization

Thinking about it, I've remembered that the 24SD Portastudio meets all that needs. In fact I seriously considered the option of buying it instead of the Model 12. But the Model 12 had so many important features I needed, that I thought I could always complete the production process by adding any necessary gear later in time.

So these are my intentions. I don't know if there even exists something like that, but I think it may be very possible it does

 

[-mjk-]

@Bambi at least you admit you're stubborn, lol. Many of us are. Often, being stubborn comes with a price. That is also the case here. https://vintageking.com/tc-electronic-finalizer-96k Good luck finding one. Besides, you have to learn how to use it. The software does the same thing and costs $200 USD. If you can learn how to use the hardware unit, it's even easier to learn how to use the software. There is a reason why even the most famous mastering engineers use computers these days.

There are online mastering services where you just upload your file and press a button. That is a completely automatic function. I suggest that you review the many reviews on YouTube that have tested those services. If I were going to do online mastering, I would consider these guys: https://www.abbeyroad.com/online-mastering
As you can see, have them do 2 songs and you could have just about bought the software.

 

[Bambi]

Hello MJK, thank you very much for the links. Mastering through online platforms is a very easy option too. However, I agree that soon it would become a much more expensive solution than buying the software and doing it myself. I have to find some option that convinces me, although it will take some time. One of the possibilities I thought of when buying the Model 12 was creating the 2 track stereo master on it, and then importing it somehow to apply compression to the mix. Doing that process several times I could get something similar to a limiting effect on my mix. I don't know if it will work. I know it is far from doing a professional mastering, and it won't give me the exact level of LUFS or true peaks, but it's the best option I can think of to come near what I have in mind (external hardware, easy to use, etc). I kind of remember that there was a post on the forum regarding this way of proceeding, but I'm not sure. Do you know if there is?

 

[-mjk-]

@Bambi there are quite a few posts (particularly on the DP forums) about Mastering and issues with mixdown files being too low. In my sig, there is a link on how to search the forums.

An alternative is to upload your mixdown file to a cloud drive and let one of us help you with it.

 

[Bambi]

Ok, MJK. I'll take a look at those DP threads about mastering. And I'll probably try your suggestion of uploading my mix so somebody in the Forum can help me. Thank you for your advice

 

[Bambi]

I've heard Mark Richards talk a couple of times about your own MJK's production system based on submixing, applying repeated compression to the submix, and then doing a final mix with all the compressed subgroups. It sounds very interesting. It reminds me to paralell compression when applied on drums, except that in your method you apply it to instrument subgroups and then to the final mix, instead of to an individual drum and then to the whole drumset.

I'm going to think of a possible way of doing something similar on the Model 12, although it will probably be a very laborius one, specially if without using a computer. The most tedious part will be the exporting and importing sequence I'll need to follow to apply compression twice to individual tracks or subgroups, and then once to the final mix. I don't even know if it will work, but I may think about it. As you said, being stubborn comes with a price