Questions and Answers Concerning Neurofeedback Protocols


With NFB training we train and correct the amplitude size of specific frequencies and not the number of the specific waves. Is that correct?

That would mean that dysfunctions are not related to the number of the specific frequency bands when we talk about „too much or too little theta“, but that the voltage is too high or too low – the amplitude size? There the neurons “fire” too much or too little – and training would reduce/reinforce in these areas and thereby distribute the energy more equally (through chemical and electrical processes)?


In classic neurofeedback amplitudes are trained in specific frequency bands. The amplitude of the EEG signal in one frequency band is measured and feedback is given proportional to this amplitude. Of course this amplitude differs from person to person. Therefore, thresholds are implemented that can be adjusted for each person individually. When the amplitude is above the threshold the feedback is different from when it is below the threshold. Most neurofeedback approaches work according to this method. It is possible to measure and analyse several EEGs at the same time and then to control the feedback with the overall result. A method that is very common since a long time is the beta-SMR training. Here, two EEGs are measured, one on the left side, one on the right side. The left side gets rewards for frequencies in the range of 15-18 Hz, the right side for 12-15 Hz. Furthermore, in this classic protocol, thresholds are also set for the remaining frequency bands, but with the opposite effect, i.e. they are inhibited. 

Interestingly, the amplitude in a frequency band does not necessarily change towards the direction in which we want to train it. That means, for example: We train 12-15 Hz “up”, but the amplitude can either stay the same or even drop within a session or throughout the whole training. The reason seems to be that our brain isn´t really interested in absolute amplitudes. We rather give the brain a very quick and direct feedback through the comparison with the threshold – we analyse the activity within one frequency band and display it to the brain. This is something more like a workout, we want to achieve an increased ability of the brain to selfregulate. The EEG of a well regulated brain shows small amplitudes, therefore, the amplitudes often decrease even though we are training „up“ in a specific frequency band.

Over the years other protocols were developed from the mentioned classic beta-SMR protocol, e.g. the bipolar ones. These protocols do not train the absolute amplitude on one or more locations, but the difference between two EEGs of different locations. The brain seems to be interested in the phase relationships between the electrode positions that are hidden in the resulting signal.

Relatively new are training protocols with a bipolar montage at ultra low frequencies – here signals are evaluated that have time constants of several seconds. The SCP training (Slow Cortical Potentials) also belongs to this „frequency range“ and was studied extensively in Germany.

Personally, I do not think much of  'too much alpha' or not enough of this or too much of that – not because this approach is basically wrong, but because it is often more a guess than anything else if there is not enough experience or the equipment is not sophisticated enough. Each brain is different – people with a diagnosed mental disorder can have a “perfect” EEG, “normal” people can have an abnormal one. For comprehensive information a QEEG analysis (quantitative EEG, which means at least 19 electrodes and a standardized measurement process), evoked potentials and a comparison to a normative database is absolutely necessary. But also simpler methods can achieve good outcomes – A solid diagnosis and closely watching the patient during the training and from session to session can be sufficient for optimizing the training parameters and getting excellent results (symptom-based approach, e.g. the Othmer method). 

We have to be a bit careful – there are lots of statements and discussions going on in this field that are full of technical terms and neurophysiological half knowledge, but nonetheless are presented as generally accepted reality. The brain is a highly complex organ, which we are “listening into” with extremely primitive methods. But empirical results show that these primitive signals work really fine – we probably just shouldn´t interpret too much into this data unless we use more sophisticated equipment.

EEG Info Neurofeedback certification


Visit us from 13.-15. September at the SGPP Congress in Bern/Switzerland,... [more]

New Cygnet® version released

The new Cygnet® version 2.0.5 now has been released [more]

About EEG Info
We use cookies on our website to offer several services and for analyzing purposes.
By using our website, you are agreeing to our use of cookies. More information about our
Privacy Terms