[Quaddy]

great post i am spending some time with my deq2496 and have made some nice inroads for eqing my hd555 at this time, will read through and see what i an glean!

 

[Liver]

Quote:

Originally Posted by Quaddy /img/forum/go_quote.gif great post i am spending some time with my deq2496 and have made some nice inroads for eqing my hd555 at this time, will read through and see what i an glean!

I am quite curious about an external one as well. Seemingly if you correct the headphone / head part, then you could take it to a new source and not be terrriblely far off.

 

[PiccoloNamek]

Yes. This will, of course, also work with hardware parametric equalizers. The same technique can be applied, only you will be playing with real knobs and sliders instead of software knobs and sliders. In fact, this may even be preferable because then the response would be applied to all outgoing audio data, not just audio played through the media player.

 

[Max F]

I know that everyone's ears are different, but can you share some of your settings you use for your headphones. Maybe that would be a good starting point for other folks.

 

[PiccoloNamek]

Certainly. I will add them to the OP as well.

The settings I use for my STAX Lambdas are:

Mode: Linear Phase
Peak type: S-plane type II

(I always use the same mode and peak type, BTW)

-8.5dB at 7400Hz with a BW of 0.750
-9.7db at 13100Hz with a BW of 0.950

Note that this response will probably work with any Lambda series headphone, from the original Lambda all the way to the SR-404, because they all have the same shape and hence, the same transducer-to-eardrum distance.

The settings I use for my ER4S are: (These settings are not final, I am still working on them.)

-2.5dB at 2500Hz with a BW of 1.500 (In this case, the 2500 peak is actually built in to the headphones and is not due to resonance. Its actual amplitude is 5dB. I find that not taking the peak completely out sounds better than removing all of it.)

-5dB at 6000Hz with a BW of 0.850 (Notice that the resonance is lower with the ER4S than with my full size headphones. I thought this was a little odd, but I have verified with sine waves and pink noise that it is indeed 6000Hz. I think the notch may need to go a little deeper, though.)

-6dB at 12550Hz with a BW of 0.950 (The ER4S has just a little too much high-frequency boost. This helps to make it sounds more normal.)

Sometimes I will add a small bass boost on certain songs. I use an S-plane type II peak at 100hz at 2dB with a BW of 6.00. Both of the above responses, while greatly different from each other, give nearly the same perceived response with both sine waves and pink noise.

 

[Max F]

Awesome, thanks!

 

[Acix]

LOL...You just need a good flat FR headphones.

 

[PiccoloNamek]

But that's just the problem. Even if the headphone had absolutely equal output at every frequency, there would still be resonance peaks for anyone who used it. That is why we need proper equalization.

 

[Planar_head]

Interesting...

I've spent about an hour tuning it, and I must say, it's interesting.
It sure helps, if your music has music content around your resonance area (I think I'm saying that right). Fortunately, it seems most of my music doesn't have too much music around that area, so the improvements were minimal. However, on music with sound around 6.8-7.1khz there is significant reduction in a sound that was a bit peaky, unnatural, and harsh to say the least.

Good stuff. I don't need this all the time, but when its needed it's good. Probably could live without it, because of my music habits.

It wasn't too hard to get rid of with my Grados either, and the program is easy enough to tune to avoid any high-frequency rolloff.

 

[Lad27]

@OP

Very nice. I have a good hardware parametric equalizer and in the past I did spend some considerable time tweaking my ED9's.

Back to your original tutorial, I'd like to clarify a few things, that may benefit people using software tweaks:

Your screengrab of the blue screen with some white curve on it is not referenced in your text. From what I can see it seems like it is "Cool Edit Pro" software. I also assume that the following screengrab with EQ box is from the same sofware. Very nice, user friendly and surprisingly powerful package. I use it to prepare files for FM radio broadcast. Maybe you should provide links to Cool Edit Pro as well.

When running the whole setup via computer, do you assume that your soundcard response is flat? I realize that this is relative measurement taking the whole audio path including phone drivers. What I'm trying to get across, that if someone "normalize" their headphones using computer and software, and then these phones are plugged to let's say rockboxed iRiver with a parametric EQ built in and set to exactly the same settings as done via software - it won't have the same flat response.

I would like to attest that this kind of EQ has its place in audio world and I would strongly recommend to everyone inclined to do so to try it. The results are quite noticeable and for me worthwhile of my investment in spent time and hardware.

 

[PiccoloNamek]

Yes, the blue screen with the white curve is from Cool Edit Pro 2.0, one of the best audio editors ever released. The picture is the log sweep sine wave in viewed in spectral mode.

I really wanted to include a link to the Cool Edit software. I figured it would be OK since it is no longer on the market, but I'm really not sure what Head-Fi's rules on that are, so I left it out.

As for the soundcard issue, no, I did not take that into account. To be honest, it did not occur to me that one soundcard could have such a drastically different frequency response from another that it would make a perceptible difference.

 

[Lad27]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif I really wanted to include a link to the Cool Edit software. I figured it would be OK since it is no longer on the market, but I'm really not sure what Head-Fi's rules on that are, so I left it out.

Understood. Better be safe than sorry.

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif As for the soundcard issue, no, I did not take that into account. To be honest, it did not occur to me that one soundcard could have such a drastically different frequency response from another that it would make a perceptible difference.

The more I think about it...You are probably right. The souncards nowadays would have some opamps in audio path. These opamps (even the most basic ones) are pretty flat across the audio bandwidth. Unless very poorly designed audio circuit, one probably doesn't need to worry too much about frequency spikes or notches of the soundcard.

 

[PiccoloNamek]

I just noticed something you said that I overlooked before:

Quote:

I use it [Cool Edit Pro] to prepare files for FM radio broadcast.

Funny you should say that, because that is exactly what I used it for! My job involves recording sermons for churches and editing their content (doing things like removing mistakes, pops and clicks, noise, etc) and mastering them for eventual broadcast on radio stations. For the longest time, I only used Cool Edit Pro 2.0. I only upgraded to Adobe Audition 3 when I read about the Healing Brush tool, which I now view as utterly indispensable.

 

[oqvist]

Interesting. However how can you get equalizers to work and still run bit perfect if you have an external DAC? I have an Elite PRO which have equalizers built in. Seems like you can´t use equalizers with foobar with asio for example. I guess getting an external EQ costs a lot or is there a good bang for the buck solution here

 

[PiccoloNamek]

I just found a freeware program that could prove exceptionally useful for equalizing your headphones. It is called "SineGen" and it can produce a sine wave that is smoothly sweepable throughout the entire frequency range. Using this program, it should be exceptionally easy to detect where there are response peaks in your headphones! I will have to update my tutorial to reflect this discovery.

http://www.softplatz.com/Soft/Audio-Multimedia/Other/SineGen.html

It is too bad the master control of the program isn't properly interfaced yet. If it was, you could set up a macro that would play a selected group of frequencies automatically, with each frequency having its own amplitude setting. This would make it exceptionally easy to quickly discover your personal headphone response, completely bypassing the first part pink noise process. (Which, while valid, is a little tedious.) The pink noise will still be needed when actually equalizing, because you will need to hear all frequencies at once so that you can have something to reference the problem frequency to. In the meantime, the program will still be exceptionally useful and can replace the current sine wave technique.