Earbud target curve tests
Jan 21, 2021 at 8:52 AM Post #256 of 318
And thank you as well to @furyossa for pushing you to do a compensation file that led to that "Cinderella transformation" of Datura Pro, you 2 are an invaluable asset to the earbuds community here on Head-Fi! :thumbsup: :relaxed:
Thank you for your kind words. :beerchug:
But don't forget that you, like other members, participated in it in your own way. Any support is welcome, it's what moves and inspires people,
yet we are all here because we love this hobby (music, gears, buds and other cool stuff) :L3000:
 
Jan 21, 2021 at 10:42 PM Post #257 of 318
Hi friends, here is the product link of KB EAR Stellar! You can find both no mic and with mic versions! Don't hesitate, and just have a try! :ksc75smile:

KB EAR Stellar: https://www.aliexpress.com/item/1005001737624099.html?spm=2114.12010615.8148356.6.7b55468flqZQRS


7(1).jpg



5.jpg
 
Jan 24, 2021 at 5:14 AM Post #258 of 318
I play with some new buds that I bought last week and today I bought a spare pair. I just test to see if there some changes after the >40h break-in period between these two.
I will post more info in a few days about this interesting "retro" model with great tonal balance.
burnin-test.jpg
Which earbud graph is this?
 
Jan 24, 2021 at 8:17 PM Post #259 of 318
Last edited:
Jan 26, 2021 at 7:32 AM Post #260 of 318
Toneking TP16 32 ohm
v1.5 bass and neutral editions


These compensation files are limited to 20hz to 10khz as I'm testing how I'll be processing community submitted measurements (more on that in the future.) One of the advantages of utilizing a 10k ceiling is lower eq strength or multiple presses of the "compression" button in Peace/APO is not strictly required. You might still need to adjust eq strength to personal preference.

Here's what these files are doing to the frequency response:

toneking tp16 stock vs 1.5 bass vs 1.5 neutral.jpg

The FixedBandEQ file has 10 band GEQ settings.
 

Attachments

  • toneking tp16 FixedBandEQ v1.5 bass 10k.txt
    470 bytes · Views: 0
  • toneking tp16 PEQ bass 10k.txt
    325 bytes · Views: 0
  • toneking tp16 wavelet bass 10k.txt
    1.3 KB · Views: 0
  • toneking tp16 FixedBandEQ v1.5 neutral 10k.txt
    471 bytes · Views: 0
  • toneking tp16 PEQ v1.5 neutral 10k.txt
    325 bytes · Views: 0
  • toneking tp16 wavelet v1.5 neutral 10k.txt
    1.3 KB · Views: 0
Last edited:
Jan 28, 2021 at 4:32 AM Post #261 of 318
If you have premium wavelet and the Rambo 2, please try this out and let me know what you think.

Run the file at 100% eq strength and the bass tuner feature with these settings:

Screenshot_20210128-012920_Wavelet.jpg

I'm testing for a basshead sound profile. @Nool do you have wavelet premium?

**edit**
You can also try to lower the cutoff to 60hz. I'm not sure where to land but go ahead and leave feedback in this thread and we'll figure it out.

***edit***
The qudelix 5k handles 5dB bass boost with no distortion. The btr3 distorts, so I have to lower the boost to 3dB.

****edit****
adding updated GEQ and PEQ files to this post. @rprodrigues the "fixedband" file has 10 band GEQ settings.
 

Attachments

  • isn rambo 2 wavelet v1.5 bass 10k.txt
    1.3 KB · Views: 0
  • isn rambo 2 FixedBandEQ v1.5 bass 10k.txt
    470 bytes · Views: 0
  • isn rambo 2 PEQ v1.5 bass 10k.txt
    231 bytes · Views: 0
Last edited:
Jan 28, 2021 at 4:40 AM Post #262 of 318
I expected you to start with the DaturaPro model with a new update for EQ. :wink:
I already asked on the previous thread, which is the reason why you have to correct all compensation files?
 
Jan 28, 2021 at 5:00 AM Post #263 of 318
I expected you to start with the DaturaPro model with a new update for EQ. :wink:
I already asked on the previous thread, which is the reason why you have to correct all compensation files?
ahh.... started with the rambo 2 because the bass is, in general, more resonant compared to the datura pro. And for my late-night experimenting (2am!) I was curious about capturing real sub-bass with an earbud.

Throughout this journey of iterating on a target curve for earbuds I made two embarrassingly silly mistakes:

1. When I first started (v1.0) I modeled the test curve with IEM Harmon target and kept modifying the curve with IEM characteristics in mind. I should have just started with Headphone curves first.

2. My recent mistake was to generate my compensation files all the way up to 20khz. I feel like an idiot because I've known all along that post 10k accuracy is really hard to maintain. There's also a 12k resonance point that forces AutoEQ to correct (incorrectly). This is the main reason why "eq strength" in wavelet and the "compress" button had to be used for good results. As you already know, the Datura Pro and EM5 are the only 2 buds I've made files for that can handle 100% eq strength without destroying timbre post 5k or introducing upper frequency distortion.

After I get caught up on re-compiling compensation files, the more common use case will be to keep eq strength near 100% and most likely never require going below 85%.
 
Last edited:
Jan 28, 2021 at 5:17 AM Post #264 of 318
1. When I first started (v1.0) I modeled the test curve with IEM Harmon target and kept modifying the curve with IEM characteristics in mind. I should have just started with Headphone curves first.
You are right, it's really more logical to use a target curve that is closer to the headphones, because the sound that comes out the shells partly enters the ear canal and the rest is reflected from the ear shell so the percentage of sound entering the channel is closer to headphones than IEMs, if you understand what I mean.
2. My recent mistake was to generate my compensation files all the way up to 20khz. This is the main reason why "eq strength" in wavelet and the "compress" button had to be used for good results. As you already know, the Datura Pro and EM5 are the only 2 buds I've made files for that can handle 100% eq strength without destroying timbre post 5k or introducing upper frequency distortion.
I hope we both agree that the goal here is to use 100% EQ strength
After I get caught up on re-compiling compensation files, the more common use case will be to keep eq strength near 100% and most likely never require going below 85%.
This would also make it easier for users to just load the file without any additional setup.
 
Last edited:
Jan 28, 2021 at 2:43 PM Post #265 of 318
@Sam L
Is Wavelet the single way to apply your compensations files on Android?
 
Jan 28, 2021 at 2:57 PM Post #266 of 318
@Sam L
Is Wavelet the single way to apply your compensations files on Android?
no, on android I've also used my GEQ files in the powereq app in the playstore. There are a number of other GEQ apps on the google play store that will work. Wavelet has a better implementation because it's not limited to 10 bands (wavelet behaves more like a 250 band EQ). Wavelet also has a strangely effective bass tuner feature. I can't figure out how it's doing what it's doing, because it's not just a simple boost. (it also has a filter as well as whatever other magic @jaakkopasanen programmed into AutoEQ).

@The3DCie uses my files with GEQ.
 
Last edited:
Jan 28, 2021 at 3:00 PM Post #267 of 318
@Sam L

Thank you.

Unfortunately, Wavelet doesn't work well in my phone. I will try your GEQ files instead.
 
Jan 28, 2021 at 3:02 PM Post #268 of 318
Jan 28, 2021 at 3:03 PM Post #269 of 318
there are also a couple quirks. After you import the file, you have to still search for it and select it before the file works.
 

Users who are viewing this thread

Back
Top