Before we dive into measurements, please keep in mind that everybody hears frequency range differently. Measurements are just simulations that are trying to represent equal (or flat) frequency response best. But in reality, there is no way of solving this mathematically as there is always a mathematical unknown - your ears' anatomy (you can find more on the science behind it at the end of this blog post). So the only way to use this in a way that makes sense is to join the mathematics with experience and experiment. 


Why does it make sense to understand measurements of frequency response (and its deviations from "flat") of your headphones or any playback system you use?

You can compensate for it! The starting point is always identification - you have to know where your playback system lies to you. You can achieve that by measuring your room acoustics or by getting the measurements of the frequency response of your headphones. Some headphones and monitoring manufacturers offer the curves online, or you can always make your own measurements using readily available tools. 

The OLLO S4 and S4R frequency response

Let's say you find out, your room has a bump at 70 Hz. You can compensate for it in two ways. Just acknowledging it and taking it into consideration when mixing will take you a long way. As when you hear a boost in your mix on 70 Hz, you will know it's all fine since it's from the room response not actually in your mix. If you want more tangible results, we suggest you use EQ digital signal processing on your master channels to flatten the curve based on your room frequency response measurements.

In the case of headphones, room acoustics have no role. You can find measurements of almost any headphones online. In our case, you can find your individual measurements in the log in area on our webpage. If you find that our S4 or S4R model does not sound flat to your ears than you can compensate to a standard used by Mini DSP by downloading the measurements and inverting them in EQ on your master channel. We recommend only making half measure changes. The reason is how acoustics in our S4 and S4R headphones work. When you’re removing low frequencies that leaves more room for higher frequencies to develop, hence the response changes the moment you touch the EQ or input signal. In our research, we found out that a nice round compensation curves are about halfway in amplitude. For example, if your chart shows a 4dB boost at 200 Hz then you cut it down by 2dB, not more. Feel free to experiment with this and you’re welcome to report back to us for further research. Write to with your findings.


Each and every pair of OLLO headphones comes with its specific frequency response measurements. You will find a printed version in your package when receiving the headphones (from 15th April 2019 on), but to make more profound use of the measurements, you need a raw version of it.

Note: All measurements from 15th April 2019 were done using Mini DSP HEARS system with their compensation curves and standards. All measurements from 15th February 2020 are made using G.R.A.S 45CC using IEC 60318-1 simulation ear and Dewesoft AD conversion Sirius system. A massive upgrade to gear used in top research labs like Nasa, CERN and others. The comparison of measurements between the two systems is not possible in a simple way. You can’t open a measurement file from one system and compare it directly to another. Keep that in mind!

Here is how you get them (click here for video):

  1. Create an account here on the website:
    PLEASE NOTICE! With our newest website migration, all the accounts made prior December 2019 were deleted according to GDPR laws and you'll have to register again. We apologize for the inconvenience caused. 
  2. When in your account, click on the "get measurements" link that will take you to our Google Drive
  3. Find your serial number (it's written in your manual and on the printed version of the graph) then download the file.
  4. Return to our website and click on "get REW software" to install the free software that will open the measurements.


  1. Adjusting the scale

Once you have your measurements opened in REW software, the first thing is to adjust the scale of the graph to a resolution that can actually provide you with some insightful data. In a too broad scale, your response will be just a flat line, that doesn't tell you anything. Zooming in too much will reveal all the little deviations from 0dB, but you can’t really use that as those differences are so small your ears won't even perceive them, and there is no need to complicate your life by compensating for them.

Every headphone has some deviations from 0, you need to be able to see that deviations in parameters that help you do something with them.

10 dB scale is something that, in our opinion, represents how headphones sound the best and offers the most useful view to benefit from when trying to improve the overall response.

2. Identify the room for improvement.

Find deeps and boosts in the curve that is greater than 3dB. Find the center of the plot. Doing that with your eyes is accurate enough, no need to go crazy with a ruler. :)

4. Go to your EQ and mimic the situation, just in reverse.

So if your measurements have a boost with the highest point on 200 Hz, you should set the EQ to that frequency and cut it down a half step. If it’s a 4 dB boost you cut it down 2 dB. Also try using low Q factors, below 1. High Q factors will cause more phase issues than anything else. We are intentionally using half measures because of the way speakers acoustics responds to the input signals. If we take the low frequency with a lot of energy out, that would also affect how much harmonics we have left, how much room it's going to be in the earcup for other frequencies to develop. Do this with all the significant bumps and boosts. Probably 2-3 areas on most headphones.

 The whole process is described in video below:

 As said, measurements are just simulations that are trying to represent equal (or flat) frequency response best. But in reality, there is no way of solving this mathematically as there is always a mathematical unknown - your ears' anatomy. Here is a little bit on the science behind it.


The Journal of Neuroscience research on this topic was published on 28th March 2018. The hypothesis was that our spatial hearing depends on the shape of our ears and once we learn how to interpret changes in frequency and amplitude of sounds we would have problems adjusting our spatial positioning of sounds if your ears changed. In short, the research used test subjects and control subjects that were asked to locate a sound source around them. It was recorded that spatial localization from left to right (horizontal) was not affected much with rubber ear implants that changed the anatomy of the ear. On the flip side, vertical localization in the auditory cortex was very much changed. After a week of wearing implants, all test subjects learned the change in anatomy and correlated sound changes. Their brain takes over and learns to interpret the changed sound. Their spatial positioning was back to 100%.

 This suggests that how we hear is different for everyone but at the same time our brain acts as a DAC with DSP if you will. This finding correlates to well-known advice from experienced audio engineers that you can use almost any monitoring system and learn to hear it flat. On the flip side, it also opens a question on how much is an acceptable tolerance or deviation of such system before simple physics of sound or mechanical vibrations start masking frequencies in harmonic ranges. Hence our brain could not do what they do in the absence of all frequency.

We tried finding proper research on this topic but was not able to find one just yet. For now, we can only put out an assumption that a double of SPL will start the masking effect. Our hypothesis is that a 3dB deviation in frequency response between harmonics is the limit. In short, if a system has a response of +3dB at 80Hz it must not read 0dB or + 6dB at 160Hz. We'll continue to research this topic.

The basics were explained by many researchers on a topic of human hearing sensitivity in the free field starting with well-known Fletcher and Munson curves from the early 20th century. It was standardized with ISO organization in 226 standard that tried achieving a useful standard for what is considered flat in everyday terminology by calculating the median of many researchers from different time, universities, private labs and test subjects from different global regions. What we know for a fact is that the attempt was good and we do have a standard but it still has massive tolerances in input data. For example, in the 200Hz area at 80phon, the tolerance is easily 10dB or more. This means the masking effect is in full swing.


The ISOv226 standard and how the OLLO S4 curve fits its allowed tolerance.

To spice this up just a touch, research on how we hear sound and frequency ranges in headphones is in very early days. To our knowledge, the most advanced researcher in this area is Dr. Sean Olive from the Harman Corporation. You can find many of his research in the AES members area.

What does it all mean?

It means that a flat line in the graph won't necessarily bring the best results. Some of our customers experienced this first hand when they tried to flatten the curve with the software for response flattening. They reported that the response curve looked more flat on the screen, but in reality, the results when mixing with such altered responses were not as good as with unaltered response. This is exactly the confusion that using standards create. Flat line on a screen is always based on some standard and compensation curves for headphones. It's based on some research that was not necessarily done in an empiric way. In simple terms, no standard is an absolute truth.

There is always room for improvement and we are eager to get there with all our endorsers and customers sending us feedback on what they experience using our products. Join in and send whatever thoughts, experiments to

More OLLO Life episodes