Lars Lentz Audio™

Setting your music up for success on streaming platforms can be overwhelming, but here are some things I’ve learned, and maybe they can help you, too.

Updated: 11-June-2026

tl;dr

(“Too long; didn’t read” or “Here’s a summary in the table below so you don’t need to read more.”)

Also, if you make ambient electronic drone and noise types of music, I added a section down below that will help explain why a lot of these rules in the table below do not apply, so skip down to that if this is you.

Integrated Loudness (I)	-16 to -12 LUFS. -14+0.2 is the recommended loudest track of an album by the AES D1008v3 document, but I think you should be in the -14 to -12 LUFS range to be heard well and be competitive. Some say quieter tracks on an album, down to -16 LUFS can be acceptable, but you risk not being able to hear them well on all playback systems (the quiet parts may be too quiet for the system – think classical music being played back in a lumber truck). But if you’re making classical music or that kind of stuff, then o.k., go to -16 at the lowest.
PSR (peak-to-short-term-loudness ratio) levels throughout a track	A “spiky” waveform with minimums not lower than 8 or 10 is great. This is a highly visual (using a meter) and subjective judgment, so a good meter capable of displaying the PSR level vs. time is needed. Inherently non-dynamic tracks (drones, for example) may not achieve this, but this is okay because that is in their nature.
PSRmin (minimum peak-to-short-term-loudness ratio)	It is good to know the minimum, but this single value often has no significance. Instead, watch that there are not too many minimums for dynamic music. Drone music or music that is inherently dense, on the other hand, can have a lot of minimums. I like to have my minimum be, repetitively, not less than 8, but some tracks are just going to be less by the nature of their sound profile, and this includes drones, ambient, or otherwise densely packed musical tracks.
PLR (peak-to-integrated loudness-ratio) for a track	>12 should be maintained to have a perceived high dynamic quality of any track. (Bob Katz says above 10.) Too high of PLR can fatigue the ear and can also be turned down by streaming platforms. Too low is bland and not very dynamic, and you risk losing your audience’s interest.
DR or DRi (Dynamic Range, Dynamic Range Integrated)	Developed by Friedemann Tischmeyer and referenced by Bob Katz in his book, this measure is similar to the measurement LRA, but takes into account all of the range, and not just for broadcast. It is used in the Maat metering systems and is said to be a better measure of dynamics than LRA or PLR.
Short-term loudness level for any track (Smax)	I believe that =<-10 LUFS should be the loudest level throughout any track. I found this by personal trials. I think that most of an album should be in the -10 to -12 range, but if you go over into the -9.x area, don’t sweat it.
True peak levels for any track	=<0 dBTP, listen for codec distortion and reduce as needed. I set the master limiter to -0.1 dBTP to be safe; with some limiters, this will not produce distortion (Sonnox, in particular, is one of these). Some tracks will not hit this level, and if the PLR is okay (see above), then this is okay too. Many say it should be at =< -1 dBTP, but this is not necessary for all music and is contrary to the current AES document.
LRA (loudness range)	If you are in the 1-3 LU range, you may be overcompressed. However, it is highly dependent on the music. Music with higher LRA values (6-12 LU) tends to be more impactful. Below 4 LU, and a track is considered somewhat static in nature (not much variation in loudness). But this low LU can be o.k., so don’t get too wound up on this number for most music tracks. If you’re going for high dynamic range, though, look at this number more closely and keep it above 6 if possible.
K scale metering	Not used any more – the creator of this scale, Bob Katz, does not recommend it anymore.
VU metering	Good for gain staging. Great on most instruments. It can be tricky on drums, but then use rms/peak if needed.
Crest value or Crest Factor of a track	Not used anymore – the creator of this scale, Bob Katz, does not recommend it anymore.
File Formats for Uploading to Streaming	Wave files (*.wav) of 24-bit and 128 sinc resampling.

When reading this, please keep in mind that rules can be broken, and I am not an authority on any of this. This is what I have found out and done for my music, and I present it here only so you can potentially build from it and adapt it for your own purposes. Lots of people have studied this crap to death, so you can find almost any opinion you want out there. I would say Ian Shepherd is one of the better, if not the best, on this subject.

AES Recommendations

There is a guideline for what you should be doing with your music, and it comes from the AES Technical Council. Careful reading of the document offers insightful guidance that can be tested and backed by real-world examples.

Technical Document

The technical document from the AES outlines the recommended loudnesses for various media and streaming.

The AES site also has a section on loudness that is helpful: https://aes2.org/audio-topics/loudness-2/

Integrated Loudness (I)

Integrated loudness is what everyone I know wants to focus on. But it is actually one of the least important. It’s not unimportant.

The AES guideline says streaming platforms should keep integrated loudness around -14 LUFS for the loudest tracks of an album and -16 LUFS for single tracks or the quietest tracks of an album. But this is for the streaming platforms to adjust and not the artist.

Except for Wide Dynamic Range Content, it is recommended that distributors (Spotify, etc.) normalize the loudest music track of an album to -14^+0.2 LUFS (integrated, I), per Table 1. This will typically result in an integrated album loudness of -16^+0.2 LUFS (integrated, I) since the loudest tracks are typically 2 LU louder than the quietest track of an album.

I’ll concentrate on album normalization as I read this AES guideline for streaming services. Album normalization says that the loudest track should be around -14, and the quieter tracks should be around -16 integrated loudness (which is also the track normalization target). And, of course, anywhere in between is okay too.

I’ve surveyed lots of music, and the weakest music out there as far as playability and perception is the music that stays steadfast around -14 or -16 LUFS Integrated with low PSRs and low PLRs (more on that below). For this reason, I would suggest being in the -14 to -12 LUFS range.

Something greater than -14 LUFS is necessary to be competitive and heard on a variety of playback devices (car, earbuds, lo-fi, hi-fi, noisy environments, clubs, etc.). Going above -12 LUFS integrated loudness will degrade the PLR to less than 12, so I recommend that integrated loudness be in the -14 LUFS to -12 LUFS range. If you want to stay with the AES standard, then you’ll want to be in the -16 to -14 range. But in my mind, it is okay to be above -14, and up to -12 seems about right as an upper maximum.

-14 LUFS to -12 Integrated LUFS will keep music competitive with what else is out there and will not degrade the PLR, which also depends on the integrated loudness in the calculation of PLR. PSR is an equally important measure that needs to be sufficiently high most of the time during the playback of a track (more on that below, also).

If you look back at integrated loudness after you have adjusted the PLR and PSR to be in good ranges, you will succeed in mastering good dynamic music. When PLR and PSR are properly adjusted, it is a simple act to bring the integrated loudness of a track within a good range (-12 to -14 LUFS) by adjusting level on the limiter and possibly also adjusting compression on the master bus (you can also use the term “buss” here instead of “bus”).

I’ve also found that the soft knee of a limiter, if adjustable, when brought to a level where limiting is gradually occurring, will sacrifice some loudness for dynamics, so this is a good option if you’re consistently coming in high on the short-term loudness (more later on that) or on the integrated loudness.

Loud Mastering

Yes, there’s lots of music out there that is way more than -12 LUFS and is in the -6 LUFS to -10 LUFS range. For certain types of music, this is probably acceptable. It becomes an artistic decision. EDM music, for example, can be in this higher LUFS range. Dynamics decrease as the integrated LUFS values increase, so if dynamic music is not what you are after, then by all means master louder. Just know that the streaming services will turn down music that is louder. Using the plugin, Loudness Penalty by Meterplugs, will be instructive and useful in judging where you want to be.

PSR

The Peak-to-Short-Term Ratio (PSR) measures the dynamic range of an audio track, specifically quantifying the ratio between the short-term loudness and the maximum true peak value. This ratio is crucial for understanding a track’s dynamics, which are essential for its overall sound quality and perception.

The peak-to-short-term ratio is important but must be considered overall. It is usually graphed either horizontally or vertically as a bipolar signal with a value of 0 in the center and guide lines (in my case, at the value of 10) on either side.

PSR as a single value does not mean much to me, but a spiky-looking graph of the values where most stay above 10 is really good. This indicates a track that has inherent dynamics and is not overly compressed.

But, depending on the material, a non-spiky graph plotted below the 10 value is okay if the track has no inherent dynamics. You cannot pull dynamics out of something that does not have dynamics. This is usually exemplified by drone or distortion music and rarely with drums or percussion involved. Or, if there are only some spots that are low in a track, it is because a lot of sounds fill that space in time, and if there are not many of these, it is absolutely okay and can give a track great dynamism.

The Klangfreund Multimeter plot of the track shown below is about ideal for me. It has PSR values that go low in places but maintains a spiky appearance above 10 LUFS most of the time. Look at this in the green and briefly yellow horizontal plot. See the lines that extend from the 10’s. Those are the 10 LUF guardrails of PSR. The PSR meter is momentary so does not mean much right now in this image, but at the top of it is the PSR minimum value which is 7.3 for this particular track. This 7.3 probably occurs in those yellow areas of the horizontal PSR plot. Now also look at the short-term loudness value of -10.5. That is below the -10 maximum so is colored in green. Also at the top is the true peak maximum that is below 0 so is colored green also. Going to the bottom of the meter, you can see PLR is 13.3 that is a very healthy value for dynamic music. Just to the right of it is the integrated LUFS of -13.4 that is in that sweet spot between -12 and -14 for what I call “competitiveness” with other music out there. The rest of the values are greyed out on this plot because they are momentary values that are nice to see while the music is playing in your DAW but have little meaning afterward. The exception is LRA, which is cumulative over time and can indicate the impactfulness of a track but can be hard to interpret correctly as well.

Overcompression can reduce the running PSR values, so if your PSR values are consistently low, the first trick to try with any track is to shut off all the compressors and see where the PSR values go. Then, I selectively turn on compression at lower ratios (1.5-2 at the maximum) and see if I can compress effectively without losing dynamics and reducing my PSR values. Alternatively, I consider gating (downward expansion) or upward expansion to counter a low-running PSR.

I think that the impact of PSR on music quality can be significant:

• Clarity and Perception: A recording with a high PSR is often perceived as clearer and more focused, as the contrast between loud and soft sections allows for a better definition of individual elements within the mix.
• Dynamic Range: PSR helps to maintain a consistent dynamic range throughout the track, avoiding excessive compression that can lead to a ‘flat’ sound or ear fatigue due to high short-term loudness levels.
• Musical Character: Different genres have varying aesthetic ideals regarding dynamics, and PSR reflects these ideals. For instance, classical music or acoustic tracks typically aim for increased dynamics and little distortion. In contrast, metal or hard EDM tracks are characterized by distortion and a dense, powerful sound pattern.
• Technical Limits: PSR influences the technical aspects of mastering, such as the peak-to-long-term-loudness (LTLR) ratio, which affects the perceived sound quality and compatibility with various playback systems.

Lower PSR values (regularly occurring below 10 on dynamic music) result in louder passages in tracks and can cause ear fatigue. They are also monitored by streaming services and can get your music turned down in some cases, causing a perceived drop in quality when played back.

PLR

PLR may be the most important factor for me in both listening and producing music.

PLR has replaced Crest Factor for many people, so it is worth explaining what PLR is and how it is different from Crest Factor. The Crest Factor is the difference in the average sound level in dB of a whole track and the track’s peak level in dB. PLR is almost the same, but it is calculated by the difference between the integrated loudness of the whole track (in LUFS) and the true peak level (in dB).

Peaks affect perceived sound quality. A recording with a high peak-to-long-term-loudness (integrated) ratio (PLR) is often perceived as clearer and less fatiguing than an excessively peak-limited one. But, it must be kept in mind that, simplified, PLR is simply the subtraction of the peak level from the integrated level, so it is not something that you want to control explicitly as it is a product of the other factors.

“I would say as a guide, If the peak to loudness ratio (PLR) of your mix is above 10 dB (LU), more typically above 12 or even 13-14 dB, then you likely have a candidate for a good mix.” Bob Katz

Music with a high Peak-to-Long-Term-Loudness Ratio (PLR) is often perceived as clearer and less fatiguing than music with an excessively peak-limited one. This perception is influenced by the role of peaks in affecting perceived sound quality. When a recording has a high PLR, it can maintain a consistent level of detail and clarity throughout the piece, contributing to the overall listening experience and satisfaction.

Here are specific reasons why I think that music with a high PLR might sound better than other music:

• Dynamic Contrast: High PLR values indicate significant differences between a piece of music’s softest and loudest parts. This dynamic contrast is crucial for creating a sense of depth and realism in the sound.
• Perceived Loudness: The PLR value is a long-term measurement that gives an overall value for a song or album, which can influence the perceived loudness of the music. A higher PLR can mean the music maintains its intensity and volume level without becoming overwhelmed or distorted.
• Dynamics Enhancement: Research has shown that listeners prefer music with a PLR above 10 dB (LU), and this preference increases as the PLR value goes up. This suggests that a higher PLR can enhance the dynamics of the music, making it more enjoyable to listen to.
• Genre and Production Quality: The PLR value is affected by genre, and classical or ambient piano music, for instance, can vary widely in PLR values. Higher PLR values tend to correlate with higher production and mixing qualities, which can significantly impact the overall sound quality of the music.
• Mastering and Delivery: During mastering, the goal is often to balance maximizing the perceived loudness and maintaining the integrity of the music’s dynamics. A higher PLR can allow for a more aggressive mastering approach, producing music that sounds better to many listeners.
• Audience Expectations: Over time, audiences have come to expect certain standards in music production, including the use of PLR values that align with the perceived quality of the music. This expectation can influence the reception of music with a high PLR, as it may be seen as meeting or exceeding these expectations.

Music with a high PLR sounds better to me than other music because it benefits from enhanced dynamics and perceived loudness. The higher PLR seems to imply a higher production quality. These elements combine to create a more immersive and satisfying listening experience, so high PLR values remain a key target for audiophiles and producers alike.

PLR, or peak-to-loudness ratio, plays a significant role in shaping the overall listening experience of a music track. It affects the dynamic range and contrast of the music, which can impact the listener’s emotional response and engagement.

• Emotional Connection: A higher PLR indicates a more dynamic track, with a greater contrast between loud and soft moments. This can create a more engaging and emotional listening experience, as the listener is more likely to be surprised and drawn in by the sudden changes in volume and intensity. On the other hand, a lower PLR can result in a more consistent but less dynamic listening experience, which may lead to a sense of monotony and disengagement.
• Loudness and Fatigue: PLR also affects the perceived loudness of the music. A higher PLR can result in a louder overall volume, which can be overwhelming or fatiguing for the listener. Conversely, a lower PLR can result in a softer overall volume, which may be more comfortable for the listener.
• Mood and Atmosphere: The dynamic range and contrast created by PLR can also influence the mood and atmosphere of the music. A more dynamic track with a higher PLR can create a sense of tension and release, while a more consistent track with a lower PLR can create a sense of calm and relaxation.

Too high of a PLR can be a problem. I have studied classical recordings where the integrated loudness is around -20 LUFS, and the PLR is correspondingly in the 20’s, and these, while very dynamic, lack enough volume in the recording to have good playability in most noisy situations (in the car, home stereo, etc.). The lower level passages of tracks are barely heard in noisy environments. They would be great for listening in headphones, but still there is a limit and I find that too high of PLR in headphones can even be annoying dut to the low level of parts of tracks. So there is a balance to be maintained between high PLR and suitable overall loudness (integrated).

I’m finding that 12-15 is a good range for PLR. Music sounds sufficiently dynamic but also listenable. Too high of dynamics can also fatigue the ear due to the high and low sound levels making the internal workings of the ear work harder (going back and forth).

Also, the streaming platforms have a desired PLR by default due to their normalization protocols. For example, Spotify may have -14 LUFS as its integrated target, with -1 dB as its maximum peak, so that would be 13 for PLR, and if you exceed it, you will get turned down a bit.

Always know that PLR values can be low if the dynamic content of the material is inherently low and/or the overall true peak value is low. This is because PLR depends on the peak value and the integrated level. If, for example, the integrated loudness is -13 LUFS, but the true peak is -3 dBTP, then the PLR will be 10 (-13 – -3 = 10).

LRA

LRA or Loudness Range is a measure, in my interpretation, of the impactfulness of a track. It can also indicate overcompression, but I find that PSR is a better indicator of that. However, if a track has an overall LRA of 1-3 LU then it could be overcompresssed and should be looked after.

Tracks with LRA below 4 LU are considered static in loudness, while those between 6 LU and 12 LU show considerable loudness variation. Aiming for an LRA above 5 LU can create tracks with more tension and dynamic contrast.

I find that tracks with a constant drum beat throughout them tend to be around 4 LU for LRA values and that is regardless of compression – it is inherent in the music itself. On the same track when I turn off the drums completely, the synths, pianos, guitar, etc., are the only things playing, and the LRA values soar up to 8 or 9 LU. Try this for yourself on a track with drums or constant percussion – turn them off and see how your LRA values change.

I’ve also found that for a track with lower LRA values, trying expansion techniques did not work for me, as they changed the music too drastically. Yes, I had higher LRA values (marginally), but the music sounded choppy, with high and low points that were not the original intent of the track.

I have also tried automation of various instruments in a track to achieve higher LRA values and have had limited success. Some tweaking like this can change the LRA upwards by 1 or 2 LUFS, in my experience, and still maintain the original character of the track. So I would say this is worth trying.

So the inevitable question is, “Does LRA matter?” It is a subjective answer at best and vague at worst, so I would conclude that for impact, a track can have a higher LRA and be perceived as being “better” somehow. Still, lower LRA on a track does not necessarily make it “worse,” but to a point – less than 4 LU and maybe you should take a look at the track volume differences, but maybe not if it is techno or drone music that tends to have constant levels throughout the tracks.

Bob Katz describes LRA as “more descriptive rather than prescriptive,” meaning that it is not something that you should aim for, and I would agree with that with the caveats that I noted above.

“Esthetically speaking, the LRA value should only be considered to be a very general guide.” – Bob Katz

It is also worthwhile to know how LRA is measured, and there is a technical paper that explains all of it (see linked document below).

S_max (Short-Term Loudness Maximum)

S_max, or short-term loudness, is measured over three seconds. This is enough time for your ears to hear it well, so it is a good judge of a track’s perceived loudness. It is also a good measure of ear fatigue. I have found that too many excursions of the audio into a short-term loudness of above -10 LUFS cause my ears to fatigue. I lose interest in the track quickly because the dynamics are no longer perceived by my ears, even if they’re there.

Not every track on an album needs to hit that -10 LUFS of short-term loudness (S), but in my opinion, at least some can. If I listen to a whole album of material, the highest S_max value will be the judgment point for my ears, so I want at least one of my tracks to hit that maximum. Sometimes, I have most of my tracks at this level. It all depends on the genre, track material, and my taste.

Another thing to think of is whether the track is very short. Short tracks of less than 30 seconds could have higher Short-Term Loudness and possibly higher Integrated Loudness. Remember, there are no hard-and-fast rules.

True Peak Levels

Will Exceeding -1 dBTP Cause Distortion?

The answer is “probably not.” “Can you hear the distortion?” would be the better question, and it is usually a solid “no.” I use the Sonnox Oxford Limiter, which does not produce distortion unless driven to the extreme (and then negligibly), so maybe that is why I never hear distortion, but listening for distortion with your ears subjectively and measuring objectively with the codec (see below) or some other means is about all you can do.

Where Should I Limit

Peak limit the true peaks to be below 0 dB. This aligns with the AES recommendations. Many have said limit to -1.0 dBTP, but I find music with this lower level of peak limiting lacks a certain crispness (probably from reduced PLR or maybe distortion) and is perceived as inferior to those limited to 0 dB. Also the AES recommendations are not for -1 dBTP as many think. Careful reading is required – see below.

True Peak Level and Perceived Quality

I have found that, through listening to a lot of music and then measuring the true peak levels of that music, I prefer music with true peak levels nearly reaching the 0 mark (and sometimes exceeding). I do not know why this is, but I often find that music with peaks nearly reaching the 0 mark or somewhere between -0.1 and -1 dBTP is some of the best music for its clarity and strength (for lack of a better word).

A lot of people will tell you to set your limiter at -1 dBTP, but from the AES document, here is why I do not set my limiter to -1 dBTP (I use -0.1 dBTP on my limiter):

“A maximum of 1 dB of limiting prior to final encoding is recommended as a starting point. Do this only if frequently-occurring peaks cause audible distortion or artifacts at the decoder. Use the ear as the final judge, because limiting more than about 1 dB may produce more audible artifacts than simply letting the audio material clip on occasional transients. However, when judging, beware of Players’ Handling of True-Peak Overloads, discussed below. Peak limiting, not clipping, may be occurring in the player. Different player designs can sound different when handling peak overshoots, so it is wise to audition several types of players.”
AESTD 1008

What this document is saying is not to set the limiter to -1 dBTP, but to use 1 dB as the starting point for limiting. This is far different than setting the peak. Think of this 1 dB limiting as the foldback amount in your limiter and not as the peak set point.

In the image below, this limiting of 1 dB is circled in red and you can see the yellow meter showing a limiting of about 1 dB of gain reduction or foldback. Circled in green is the true peak limit which in this case is set at -0.1 dBTP.

Spotify says to set the peak limiting at -1 dBTP and -2 dBTP for “loud” music, but this is clearly not what the AES document (see excerpt above) says.

Sonnox Fraunhofer Codec Toolbox

I check for audible distortion while mastering, and I use the Sonnox Codec Toolbox to do that. It is fairly low-cost when on sale (think Black Friday, for example), and I only use the low-cost version and not the professional one because I don’t need a high-level product. I really only need my ears, but the codec toolbox helps me tune into my track better and can hear more than I do.

I look for “overs” while I play back the track through this codec toolbox, and when I see them on the meter, I listen to see if I can hear them. If I cannot hear them, I will do nothing. If I can, I reduce my levels. It is really that simple.

The NMR meter shows the loss of fidelity in various frequency ranges due to the lossy encoding (in this case, MP3 at 128 kbps). This is unavoidable, but it is interesting to know how my tracks will sound in the various encodings.

ADPTER Streamliner Codex

I’m also using the handy Adpter Streamliner to check different codexes fast.

I can see all of my levels on one screen and check the codex, too – without leaving the same application. The only drawback is that it doesn’t save the values after a render, so I cannot see the values for what I have rendered. Instead, I have to listen to it while watching the meters, and this is o.k., but it doesn’t serve me when I render tracks. To get the values after rendering, I use Klangfreund’s Multimeter.

File Formats

The only true non-lossy file formats that should be uploaded to streaming are wave files (*.wav). I’ve tried others like FLAC that are supposed to be lossless, but they really are not and they always give me less than pristine recordings. My advice is to always stick with wave files for uploading to streaming platforms.

Wave files with 24-bit and 128 sinc resampling are noticeably better than 16-bit and 32 sinc resampling.

Limiting

Settings On My Sonnox Oxford Limiter

I’ll set my limiter to -0.10 dB and do true peak limiting. Unlike other limiters, the Sonnox Oxford Limiter I use does not clip if the threshold is set below 0 dB. This is nice, as you will not get the clip distortion that other limiters produce. It does this by adjusting the level internally so that peaks that would produce clipping distortion in other limiters will not be heard in this one. It also has a soft knee setting, so the onset of limiting can be more gradual. This is one of the closest limiters to an analog version I have come across, and that’s one of the reasons I use it.

Measurement Examples

Ambient Drone Noise and Electronic Music Exceptions

For ambient electronic and drone music where there are not beats but long passages of noise or drone, the rules are not going to strictly apply because

• Drones and noise are not dynamic so there will not be high and low excursions of the signal. This means that the true peak (5 in red circle in the image below) will not be reaching -0.1 dB or anywhere close. In the example below it is at -2.27 dB.
• Because the true peak will not be very high, that means your PLR (2 in image below) will not be high either. Here it is 10.6.
• Instead of watching these metrics closely, instead focus on making sure the short-term loudness maximum, Smax, (3 in image) is not exceeding -10 LUFS because you do not want to fatigue the listener’s ear. Here it is at -10.1 which is almost exactly right.
• Likewise, keep your integrated loudness, I, (4 in the image below) is a listenable range centered on about -13 LUFS because, like Smax, this integrated loudness is what your listener is going to be “hearing” as their ears are not going to be receiving dynamic sounds like drums. In the image below, it is at -12.9 LUFS which is about right for a good long-term listening experience.
• Another factor comes into play with drone type of music and that is the loudness range, LRA, which will be a very low value for the static nature of long-term drones. Try to bring this range upward by applying some volume automation throughout your track. LRA will affect the listener more with drone-like music because with long drones there is a risk of losing the interest of the listener. Plus, it sounds weird, but LRA increases the listener’s engagement with the track at a subconscious level. We are naturally wired to tune out things that are static (think of the hum of a refrigerator that you hardly notice because you’ve been living with it for so long), and adding a little complexity or volume ups and downs to bring up LRA will hold your listener’s interest way more.

Artists and Recordings I Studied

Here are some artists and recordings that are exemplary of the kind of dynamics and sound I am looking for.

1. The Black Dog — I have found their music to have a punchy yet deep sound. It seems loud enough to be heard well in headphones or a car. I studied their albums and levels and found that their PLR is usually quite high. Notice how they have tracks that are in the -9 to -10 LUFS short-term but only a few on any album. I highlighted these in red (but not because they are wrong). Note the high PLR values.
2. Biosphere — His music is often with very high PLR values and a very low integrated loudness, whilst still hitting the high peaks (hence the high PLR). The same thing applies here for the short-term loudness – there are few that are in the -9 to -10 LUFS range. Note the high PLR values.

My Process

Here is a block diagram of my basic process and the two major loops I get stuck in – The Mix Loop and The Master Loop:

What About K-Metering?

Bob Katz developed the K scale for metering, and I used it for a long time – until I read his book, “Mastering Audio, the Art and the Science, Third Edition” where he said on page 261 that he no longer endorses it (see excerpt image below):

Bob Katz is a master at mastering, and his advice is priceless. I’d highly recommend you check this article: https://tapeop.com/interviews/116/bob-katz-bonus/

VU Metering

VU metering is where it’s at for me right now. I like to have 0 VU (-18 calibration) on my instruments before going into my mix bus. Drums can be a bit tricky to meter this way, but I like to have them ideally hitting 0 but without going over -6 dB peaks. For this, I switch to an RMS/peak meter sometimes or just rely on the -6 dB indicator on the meter.

After having adjusted the inputs to be hitting 0 on the VU meters, this only means that everything is coming in at the same level. It does not mean that I want them all to be at the same level for any given track. For example, drums may be too loud and may be overpowering my synths. If that is the case, I pull the fader down for the drums. I do not adjust the metering and my VU meters read the same. This is when I use the faders for each instrument. I don’t raise anything using the faders – I only lower the faders where needed.

By following this VU strategy, I’ve been able to not only come out with a more balanced mix but also it gain stages everything nicely.

Other Informative Sources

---