Yanyin Discussion thread
Nov 5, 2022 at 1:05 PM Post #2,056 of 2,418
PEQ should do the trick. Go to hbb.squig.link and pull up both the moonlight and the mahina and you can auto EQ the moonlight to mahina. It will give you the PEQ settings
I don't understand how the app works. Auto eq only eqs any iem to the Badguy Target, not from IEM A to B.

There is a button for "upload target", but I'm not sure what format it needs to be in, or how I can get it

----

EDIT: I think I have figured it out, you need to delete the reference Badguy curve, then with only 2 iem, whichever one is not selected is the target of the one that is selected.

This seem like it is a really powerful tool. Assuming I find a suitable PEQ program for my computer, I could sample every other tribrid out there just from having one of them...
 
Last edited:
Nov 5, 2022 at 2:57 PM Post #2,057 of 2,418
I don't understand how the app works. Auto eq only eqs any iem to the Badguy Target, not from IEM A to B.

There is a button for "upload target", but I'm not sure what format it needs to be in, or how I can get it

----

EDIT: I think I have figured it out, you need to delete the reference Badguy curve, then with only 2 iem, whichever one is not selected is the target of the one that is selected.

This seem like it is a really powerful tool. Assuming I find a suitable PEQ program for my computer, I could sample every other tribrid out there just from having one of them...
yep! It's a pretty amazing feature for squig. I have a bunch of profiles for EQing the Moonlight to various other tribids, EJ07M KL, EJ07M, Mahina, Variations, Oracle, etc. I haven't listened to any of those so idk how accurate it is but @MMag1983 had both the Moonlight and Mahina and said when he EQ'ed the Moonlight to the Mahina he couldn't tell a difference.
 
Nov 5, 2022 at 4:15 PM Post #2,058 of 2,418
I don't understand how the app works. Auto eq only eqs any iem to the Badguy Target, not from IEM A to B.

There is a button for "upload target", but I'm not sure what format it needs to be in, or how I can get it

----

EDIT: I think I have figured it out, you need to delete the reference Badguy curve, then with only 2 iem, whichever one is not selected is the target of the one that is selected.

This seem like it is a really powerful tool. Assuming I find a suitable PEQ program for my computer, I could sample every other tribrid out there just from having one of them...
That's it!! In other IEMs the curve is not everything, you can try to imitate the tonality, but never the technical aspects. In the case of Mahina/Moonlight they are the same drivers and box. I imagine that the HBB tuning has been done by changing some filters, or in the ventilation hole, but in this case the EQ should leave the moonlight almost equal to the mahina.

On PC I use Roon's PEQ. I have also used the one from JRiver. In Android I use the Neutron Player or UAPP
 
Nov 5, 2022 at 10:23 PM Post #2,059 of 2,418
Just ordered the Canon on AliExpress 11.11 sale via Penon Audio Store for just under $270 USD using "US20" code for $20 off. Probably take 3 weeks to get to USA, but that's OK with the good sale price.
 
Nov 5, 2022 at 10:39 PM Post #2,060 of 2,418
Just ordered the Canon on AliExpress 11.11 sale via Penon Audio Store for just under $270 USD using "US20" code for $20 off. Probably take 3 weeks to get to USA, but that's OK with the good sale price.
Congrats outstanding,most days I reach for the Canon over the EJ's You will enjoy. I have settled on 010 but the beautiful thing with the Canons is that you don't have to change it whenever you feel like it . I bought a few of theese for just that purpose!

C$ 0.81 36%OFF | 1PC Funny Anti-Lost Card Pin For IPhone X XS XR Xiaomi Samsung Universal Sim Card Remover Tray To Open The Sim Card Eject Tool
https://a.aliexpress.com/_mK1nsEo
Screenshot_20221105-194024_AliExpress.jpg
 
Nov 6, 2022 at 1:32 AM Post #2,062 of 2,418
Congrats outstanding,most days I reach for the Canon over the EJ's You will enjoy. I have settled on 010 but the beautiful thing with the Canons is that you don't have to change it whenever you feel like it . I bought a few of theese for just that purpose!

C$ 0.81 36%OFF | 1PC Funny Anti-Lost Card Pin For IPhone X XS XR Xiaomi Samsung Universal Sim Card Remover Tray To Open The Sim Card Eject Tool
https://a.aliexpress.com/_mK1nsEoScreenshot_20221105-194024_AliExpress.jpg
Gotcha, I have a couple of those laying around, the Canon switches do look a bit small!
 
Nov 6, 2022 at 4:11 AM Post #2,063 of 2,418
yep! It's a pretty amazing feature for squig. I have a bunch of profiles for EQing the Moonlight to various other tribids, EJ07M KL, EJ07M, Mahina, Variations, Oracle, etc. I haven't listened to any of those so idk how accurate it is but @MMag1983 had both the Moonlight and Mahina and said when he EQ'ed the Moonlight to the Mahina he couldn't tell a difference.

That's it!! In other IEMs the curve is not everything, you can try to imitate the tonality, but never the technical aspects. In the case of Mahina/Moonlight they are the same drivers and box. I imagine that the HBB tuning has been done by changing some filters, or in the ventilation hole, but in this case the EQ should leave the moonlight almost equal to the mahina.

On PC I use Roon's PEQ. I have also used the one from JRiver. In Android I use the Neutron Player or UAPP

**EDIT** I HAVE SINCE TOOK MY OWN ADVICE ABOUT CORRECTLY MATCHING BASS LEVELS BECAUSE MY SUBWOOFER WAS RUNNING 3DB TOO HOT AT 63HZ CROSSOVER. THE BELOW IMPRESSIONS NEED TO BE ADJUSTED FOR LESS EXPECTATIONS OF BASS.

Right, I have spent Saturday night at home playing around with EQ-ing the Moonlight. It's not as sad as you think since it's Bonfire night in the UK and I generally avoid it due to air pollution and breathing difficulties.

Basically, I have went onto HBB Squig and got the PEQs needed to approximate the Moonlight to various other IEM's tonalities. The ones I have picked were <Mahina, Monarch II, Oracle II, Xenns Up, Variations, Shuoer x HBB EJ07M (is that Kinda Lava??) and Senn IE600/900. These were chosen because of either curiosity, approximate price, or fellow tribrids. If anyone has any other suggestions after reading this through, please chip in.

Essentially, my main complaint with the Moonlight is that it lacks in bass. How do I determine that? Well my testing track for bass levels, funnily enough isn't EDM or anything, it's particular track by Celine Dion titled <Because you loved me>. The reason for this is there is a strong backing bass in the track along with high vocal peaks from Celine's voice. Using Adobe Audition's frequency analyzer function to analyze the FLAC file of the song with regards to what frequency is at what -db level in the mix, I have used the snapshot tool to take various snapshots in time of the exact frequency spectrum and the max peaks. So overall, you see for a large part of the sound track, the bass level is supposed to be a lot higher than midrange / treble, it's mixed that way and so it should sound deep / warm. And roughly speaking, the max peak of bass at approx 35hz is the same -db level mixed as Celine's vocals peaks at 1kz.

Given that information, I play the track back on my speaker + subwoofer setup, which is a pair of professional studio monitors plus a 155lb subwoofer with a 8.33 cubic ft ported box enclosure capable of flat extension in-room of 16hz. I then measure the max peaks of playing back Celine Dion <Because you loved me>using a spectrum analyser app on my mobile to measure the max peaks detected for every frequency. essentially, I am looking for the measured 35hz and 1khz peak to be roughly equal in height, and how *I perceive* that level of bass determines for me, what the correct level of (low) bass *should* sound like (for me).

Given that methodology, I installed Equaliser APO for Windows, and exported .txt files of all PEQs needed for Moonlight approximations with the pre-amp gain reduction disabled. Auto EQ on HBB Squig was set from 20 to 20000hz, not 15000 by default, though the algorithm didn't always go above 15k, but if you set the limit higher it can do for a better fit. I also added more filters until the algorithm determined no more could be optimised (for one of them, I manually added PEQs myself to further improve the laziness of the algorithm and used the outputted line graph as a guide for fit). I added about 5 filter rows from the web app's default.

Aside from that one Celine Dion track, I have also used Lene Marlin <A place nearby> to test for bass level, as there is a constant pulsing of bass throughout. To test for tonality of female voices (I don't care for the men, sorry guys), I used Celine Dion, Katie Melua and Karen Carpenter.

To anyone reading this, the obvious disclaimer is that all the impressions are based from a baseline of Moonlight.

First of all, both the Senns IE600/900 were kicked off because I didn't like the treble, with IE900 being a much worse offender than IE600. It was a U shape, and the treble was too "spicy", not much of an improvement from my Senn IE300 which I only keep around because I got it half price.

Both the Thieaudio Monarch II and Oracle II were kicked off again because I didnt like the treble, in particular in the way it added a brittle, crystalline edge to female voices. Eg the first 20-40 seconds of <Call off the Search> by Katie Melua. The Oracle II was a far worse offender for this.

I believe the biggest contributor as to why those sets were kicked off were due to various gains in the treble at 5k and above. Which you may wonder, why I have left the Moondrop Variations on the graph, it's to prove a point essentially, which I'll come to later.

You will see from the screenshotted graph that all of them have higher bass than the Moonlight which I definite feel is lacking.

So now, does the Mahina sound very different from the Moonlight? No is the simple answer. There is however, a perceptible audio difference, more noticeable if you A/B sections of a track. Is the difference worth it for me to take a hit on restocking / resell? I'm not sure yet. But I *am* sure that I should have bought the Mahina over the Moonlight to start with. And that HBB preference tuning is definitely closer to what I would consider to be the "good and correct" sound profile.

However, if I just want more bass at a level but keeping the vocal profile as close to the Moonlight / Mahina as possible, it is actually the Xenns Up which fits the best. It has a level of bass output / kick that I feel is good for the Celine Dion track and the Lene Marlin one, it is how I would expect my subwoofer to sound in the audible frequencies without physically shaking the room and its contents. However, I'm unsure whether the higher level of bass when carried over to every song is exactly how my speakers + subwoofers sound. With the Xenns up sound profile on, the bassline in every track is clear and audible. I could live with it, but I will have to listen to my speaker subwoofer after I get up from bed as I couldn't do it in the night to see whether the level of bass in the Xenns up is "correct" or not. It's certainly pleasant though.

I will now explain why I have rejected Moondrop Variations, but not EJ07M x HBB (Kinda Lava?) even though they both add to the Moonlight's vocal / treble range above 1k. Essentially, the Variation has higher level between 5k to 9k, which I find interferes with female vocals and make them sound brittle / crystal / harsh.

Both the Variations and EJ07M x HBB (Kinda Lava?) also has a boost relative to the Moonlight peaking at 2.8khz which again affects female vocals. I find that for the Kinda Lava? the gain at 2.8khz but not between 5-9k adds extra "bite" to female voices without brittleness, and actually sounds very nice to me. It's like the Moonlight / Mahina vocals with extra seasoning. And this model's bass boost is a happy medium between the Moonlight and Xenns UP, bass levels tested to me also sound to be good and possibly more "correct" than the Xenns UP.

The only thing left is a bit of confusion as on the HBB Squig site, there are EJ07, EJ07M, EJ07M x HBB and EJ07M 2 x HBB. I don't understand the difference and I intend to ask this in a Shuoer thread, but the curve I used was EJ07M x HBB, which had the large bass gain out of those 4 relative to the Moonlight.

I hope my impressions here helps anyone else as opposed to me just pissing in the wind. lol.
 

Attachments

  • 1667722699782.jpg
    1667722699782.jpg
    1.1 MB · Views: 0
  • graph.png
    graph.png
    405.7 KB · Views: 0
Last edited:
Nov 6, 2022 at 4:38 AM Post #2,064 of 2,418
Right, I have spent Saturday night at home playing around with EQ-ing the Moonlight. It's not as sad as you think since it's Bonfire night in the UK and I generally avoid it due to air pollution and breathing difficulties.

Basically, I have went onto HBB Squig and got the PEQs needed to approximate the Moonlight to various other IEM's tonalities. The ones I have picked were <Mahina, Monarch II, Oracle II, Xenns Up, Variations, Shuoer x HBB EJ07M (is that Kinda Lava??) and Senn IE600/900. These were chosen because of either curiosity, or fellow approximate price, or tribrids. If anyone has any other suggestions after reading this through, please chip in.

Essentially, my main complaint with the Moonlight is that it lacks in bass. How do I determine that? Well my testing track for bass levels, funnily enough isn't EDM or anything, it's particular track by Celine Dion titled <Because you loved me>. The reason for this is there is a strong backing bass in the track along with high vocal peaks from Celine's voice. Using Adobe Audition's frequency analyzer function to analyze the FLAC file of the song with regards to what frequency is at what -db level in the mix, I have used the snapshot tool to take various snapshots in time of the exact frequency spectrum and the max peaks. So overall, you see for a large part of the sound track, the bass level is supposed to be a lot higher than midrange / treble, it's mixed that way and so it should sound low / warm. And roughly speaking, the max peak of bass at approx 35hz is the same -db level mixed as Celine's vocals peaks at 1kz.

Given that information, I play the track back on my speaker + subwoofer setup, which is a pair of professional studio monitors plus a 155lb subwoofer with a 8.33 cubic ft ported box enclosure capable of flat extension in-room of 16hz. I then measure the max peaks of playing back Celine Dion <Because you loved me>using a spectrum analyser app on my mobile to measure the max peaks detected for every frequency. essentially, I am looking for the measured 35hz and 1khz peak to be roughly equal in height, and how *I perceive* that level of bass determines for me, what the correct level of (low) bass *should* sound like (for me).

Given that methodology, I installed Equaliser APO for Windows, and exported .txt files of all PEQs needed for Moonlight approximations with the pre-amp gain reduction disabled. Auto EQ on HBB Squig was set from 20 to 20000hz, not 15000 by default, though the algorithm didn't always go above 15k, but if you set the limit higher it can do for a better fit. I also added more filters until the algorithm determined no more could be optimised (for one of them, I manually added PEQs myself to further improve the laziness of the algorithm and used the outputted line graph as a guide for fit). I added about 5 filter rows from the web app's default.

Aside from that one Celine Dion track, I have also used Lene Marlin <A place nearby> to test for bass level, as there is a constant pulsing of bass throughout. To test for tonality of female voices (I don't care for the men, sorry guys), I used Celine Dion, Katie Melua and Karen Carpenter.

To anyone reading this, the obvious disclaimer is that all the impressions are based from a baseline of Moonlight.

First of all, both the Senns IE600/900 were kicked off because I didn't like the treble, with IE900 being a much worse offender than IE600. It was a U shape, and the treble was too "spicy", not much of an improvement from my Senn IE300 which I only keep around because I got it half price.

Both the Thieaudio Monarch II and Oracle II were kicked off again because I didnt like the treble, in particular in the way it added a brittle, crystalline edge to female voices. Eg the first 20-40 seconds of <Call off the Search> by Katie Melua.

I believe the biggest contributor as to why those sets were kicked off were due to various gains in the treble at 5k and above. Which you may wonder, why I have left the Moondrop Variations on the graph, it's to prove a point essentially, which I'll come to later.

You will see from the screenshotted graph that all of them have higher bass than the Moonlight which I definite feel is lacking.

So now, does the Mahina sound very different from the Moonlight? No is the simple answer. There is however, a perceptible audio difference, more noticeable if you A/B sections of a track. Is the difference worth it for me to take a hit on restocking / resell? I'm not sure yet. But I *am* sure that I should have bought the Mahina over the Moonlight to start with. And that HBB preference tuning is definitely closer to what I would consider to be the "good and correct" sound profile.

However, if I just want more bass at a level but keeping the vocal profile as close to the Moonlight / Mahina as possible, it is actually the Xenns Up which fits the best. It has a level of bass output / kick that I feel is good for the Celine Dion track and the Lene Marlin one, it is how I would expect my subwoofer to sound in the audible frequencies without physically shaking the room and its contents. However, I'm unsure whether the higher level of bass when carried over to every song is exactly how my speakers + subwoofers sound. With the Xenns up sound profile on, the bassline in every track is clear and audible. I could live with it, but I will have to listen to my speaker subwoofer after I get up from bed as I couldn't do it in the night to see whether the level of bass in the Xenns up is "correct" or not. It's certainly pleasant though.

I will now explain why I have rejected Moondrop Variations, but not EJ07M x HBB (Kinda Lava?) even though they both add to the Moonlight's vocal / treble range above 1k. Essentially, the Variation has higher level between 5k to 9k, which I find interferes with female vocals and make them sound brittle / crystal / harsh.

Both the Variations and EJ07M x HBB (Kinda Lava?) also has a boost relative to the Moonlight peaking at 2.8khz which again affects female vocals. I find that for the Kinda Lava? the gain at 2.8khz but not between 5-9k adds extra "bite" to female voices without brittleness, and actually sounds very nice to me. It's like the Moonlight / Mahina vocals with extra seasoning. And this model's bass boost is a happy medium between the Moonlight and Xenns UP, bass levels tested to me also sound to be good and possibly more "correct" than the Xenns UP.

The only thing left is a bit of confusion as on the HBB Squig site, there are EJ07, EJ07M, EJ07M x HBB and EJ07M 2 x HBB. I don't understand the difference and I intend to ask this in a Shuoer thread, but the curve I used was EJ07M x HBB, which had the large bass gain out of those 4 relative to the Moonlight.

I hope my impressions here helps anyone else as opposed to me just pissing in the wind. lol.
Excellent research!
As I said before, not everything is curves. The technical aspects of each IEM cannot be imitated, but you have managed to have a clear idea of what you like or don't, and make decisions about it. My congratulations.
 
Nov 6, 2022 at 5:00 AM Post #2,065 of 2,418
Excellent research!
As I said before, not everything is curves. The technical aspects of each IEM cannot be imitated, but you have managed to have a clear idea of what you like or don't, and make decisions about it. My congratulations.
I am actually now more confused about the Kinda Lava than ever. First of all, it seems to have been a custom edition that has sold out. Then on the Squib Link, it has 2 different curves, I like the curve with the extra bass. However, on Crinacle's database, the measured bass was actually less than the Mahina and the same as Moonlight.

But thanks for your comment. I'm still undecided what to do atm. But will think about it and play around some more.

There is also an addition factor to consider, in the sense that I am using the Moonlights "unamped" straight out of a phone (admitted an LG V20 with a decent "Quad Dac" circuitry and my PC, which to be fair was very high end when custom built so headphone circuitry will be better than a trash office HP box. If there is a portable source with PEQ options, then I could just simply EQ the Moonlight to my liking whenever. My LG V20 is too old and even a basic equaliser will not work due to Android OS being outdated. My current Xiaomi phone is too budget, and headphone circuitry is likely to be poor.
 
Nov 6, 2022 at 5:19 AM Post #2,066 of 2,418
I am actually now more confused about the Kinda Lava than ever. First of all, it seems to have been a custom edition that has sold out. Then on the Squib Link, it has 2 different curves, I like the curve with the extra bass. However, on Crinacle's database, the measured bass was actually less than the Mahina and the same as Moonlight.

But thanks for your comment. I'm still undecided what to do atm. But will think about it and play around some more.

There is also an addition factor to consider, in the sense that I am using the Moonlights "unamped" straight out of a phone (admitted an LG V20 with a decent "Quad Dac" circuitry and my PC, which to be fair was very high end when custom built so headphone circuitry will be better than a trash office HP box. If there is a portable source with PEQ options, then I could just simply EQ the Moonlight to my liking whenever. My LG V20 is too old and even a basic equaliser will not work due to Android OS being outdated. My current Xiaomi phone is too budget, and headphone circuitry is likely to be poor.
Yeah, on HBB’s squib link, The first HBB kinda lava graph is pre-production. That’s the signature it was supposed to be. More mid bass without losing the sub bass. When it came to actual production Letshouer reduced the sub bass without boosting the midbass. Similar low end curve, just less of it. Some folks were pissed about that, but some liked it. It is/was a limited release, but there’s still some around.

Great write-up! I have the Mahina, sometimes the bass is too much, almost. Haven’t messed around with PEQ with it, sounds like a good idea though
 
Last edited:
Nov 6, 2022 at 5:34 AM Post #2,067 of 2,418
Yeah, on HBB’s squib link, The first HBB kinda lava graph is pre-production. That’s the signature it was supposed to be. More mid bass without losing the sub bass. When it came to actual production Letshouer reduced the sub bass without boosting the midbass. Similar low end curve, just less of it. Some folks were pissed about that, but some liked it. It is/was a limited release, but there’s still some around.

Great write-up! I have the Mahina, sometimes the bass is almost too much, almost.
Oh I see. Well that's a bummer. If so, the the (2) actually produced model has less bass than the Mahina then between extra bass and extra 2.8khz bite to vocals, I'd actually take the extra bass than vocals as sounding "deficient" is worse then not sounding "extra". So Mahina > actual production Kinda Lava. The pre-production curve I used on my EQ though definitely would have been better. Shouer f*ed up lol

(then that explains why on Crinacle database, the Kinda Lava graph is the #2 from the HBB database)

Then that leaves the Xenns Up with the extra bass and also slight increase in bite of female vocals or the Mahina, building on the Moonlight vocals which I already know, and love? haha since I've had only them for just over a day.

Not sure if many other iems would fulfil my taste of extra bass without much extra treble from 5k+
 
Last edited:
Nov 6, 2022 at 6:17 AM Post #2,068 of 2,418
I am actually now more confused about the Kinda Lava than ever. First of all, it seems to have been a custom edition that has sold out. Then on the Squib Link, it has 2 different curves, I like the curve with the extra bass. However, on Crinacle's database, the measured bass was actually less than the Mahina and the same as Moonlight.

But thanks for your comment. I'm still undecided what to do atm. But will think about it and play around some more.

There is also an addition factor to consider, in the sense that I am using the Moonlights "unamped" straight out of a phone (admitted an LG V20 with a decent "Quad Dac" circuitry and my PC, which to be fair was very high end when custom built so headphone circuitry will be better than a trash office HP box. If there is a portable source with PEQ options, then I could just simply EQ the Moonlight to my liking whenever. My LG V20 is too old and even a basic equaliser will not work due to Android OS being outdated. My current Xiaomi phone is too budget, and headphone circuitry is likely to be poor.
The moonlights are easy to drive, but for that, what you have to watch is the output impedance of the source, since they have a low impedance (8ohm) as I said in a previous comment. You don't need a big amp to drive them.

Both Neutron Player and UAPP should work for you on older androids as they work on my DAP running Android 9, and they have excellent PEQs.
 
Nov 6, 2022 at 8:45 AM Post #2,069 of 2,418
The moonlights are easy to drive, but for that, what you have to watch is the output impedance of the source, since they have a low impedance (8ohm) as I said in a previous comment. You don't need a big amp to drive them.

Both Neutron Player and UAPP should work for you on older androids as they work on my DAP running Android 9, and they have excellent PEQs.
So from looking at previous posts, it is said you want the IEM impedance to be at least 8x higher than the output impedance of the DAC / Amp?

Therefore looking at 1ohm impedance or less from source.

Well, how would I know what is the output impedance if it is not measured as a public specification? There doesn't seem to be any data on the output impedance of my phone's 3.5mm jack and zero information about the headphone jack on my computer.

Can it be measured? But my phone "changes gain" depending on the impedance it measures from the headset plugged into it, does that mean its own output impedance changes also?
 
Nov 6, 2022 at 8:47 AM Post #2,070 of 2,418
yep! It's a pretty amazing feature for squig. I have a bunch of profiles for EQing the Moonlight to various other tribids, EJ07M KL, EJ07M, Mahina, Variations, Oracle, etc. I haven't listened to any of those so idk how accurate it is but @MMag1983 had both the Moonlight and Mahina and said when he EQ'ed the Moonlight to the Mahina he couldn't tell a difference.
@The-One I did have both side by side for a few days. I honestly couldn't tell a difference when I EQd Mahina to Moonlight with squigs.occasionally Unfortunately I didn't think to try the other way around. Don't see why it wouldn't work though if it did one way.
 

Users who are viewing this thread

Back
Top