[sparkyhx]

Hi,
I've recently been looking for new headphones (highish end) and during my research I was reading ' xxxx headpnone' sounds better with a bit of EQ at 'so and so' frequency range(s) etc etc..

So I started investigating EQing out of curiosity, not cos I thought I needed it with my current sets of cans. I ended up downloadingd EquiliserAPO with Peace GUI, and entered the world of EQ.

This got me thinking, how the F@#* does it work?

My setup is Computer and ifi Zen Dac V2, currently connected by USB cable. I'm using the headphones single ended but looking at balanced cable for my new headphones.
- How can it EQ the signal, but the Zen Dac still do DAC duties?
- Are the DAC duties now being done by the computer and the audio signal effectively bypasses the DAC when you EQ?
- does the EQ decode, and re-encode and the Zen DAC decode again?

Really confused? Cos the 'hifi' guiding rule was always mess around with the sound as little as possible. My Class A amp I had for 20 years had on/off, source and volume...........no 'tone or balance' controls, observing that principle.

So how does EQ work with external DAC.............can someone explain in simple terms?

 

[Roseval]

All EQ is done in the digital domain, just like you can have digital volume control.
So file > EQ or volume control or any other type of DSP> modified digital file > USB (or SPDIF) > DAC
Looks like you are mixing up the graphic equalisers of the previous century (fully analog hence adding distortion) with digital EQ.

 

[sparkyhx]

All EQ is done in the digital domain, just like you can have digital volume control.
So file > EQ or volume control or any other type of DSP> modified digital file > USB (or SPDIF) > DAC
Looks like you are mixing up the graphic equalisers of the previous century (fully analog hence adding distortion) with digital EQ.

You went all technical on me there!
The file is a Flac, this is digital and needs to be converted to a sound signal by the DAC
The flac 'describes' the music in digital terms. In a simplistic way its saying "play x frequency, at thisamplitude for this length of time" just lots of them all at the same time. or is this wrong?

How does EQ change the sound if not by manipulating the Flac and by doing so does that not involve some level of decoding and re encoding? or is it as simple as change a few bits here and there a 1 to a 0 or vice versa. If so, how does it know what to change?

 

[VNandor]

Ask a technical question, get a technical answer.

To understand anything about how a digital eq works, you first have to understand at least a little bit about how "the flac describes music in digital terms". Your idea right now is far from how that really works. The flac is decoded into a pcm stream at some point during the playback, and the digital eq operates on that pcm stream.
Here's what i mean by pcm. The first paragraph should already clear up your misconception. To (over) simplify it, it is a series of samples representing amplitudes over time. It's not amplitude over frequencies.
On a philosophical note, it does not describe music either (because amplitude is just amplitude, it is not music), music sheet is what's used to describe music.

You input what frequencies you want to cut/boost into your eq and it calculates how the samples should change according to that. The EQ receives the samples from the playback software and outputs the changed samples (towards) to the DAC. The flac file gets left alone, its only read, not written.

EqualizerApo is programmed to use biquad filters. The flowcharts describe what the filters do in general. x(n) is the input sample point y(n) is the output sample. The equation could be written down by looking at the flowchart.
The flowchart shows that the current output sample y(n) is calculated by taking the current input sample x(n) multiplied by b0, then you add the previous sample x(n-1) multiplied by the b1 constant to that and so on...

When you input your eq settings, the program calculates the coefficients specific to that setting once, and then it starts solving the equation for every sample (44100 times a second) using those coefficients.

At the end, the filter does elementary school level math (delay, add, multiply) but providing a coherent explanation on how to get there is not easy, so sorry about the not simple terms.

 

[sparkyhx]

Thanks for that, I’m not going to pretend I understood everything, but I understood enough, greatly appreciated