Mar 2, 2025 at 6:13 PM Post #146,116 of 151,730
There are no limitations as it can be scaled down with smaller transistors, inductors, capacitors, and power supply and with its innate higher efficiency, size and battery isn't an issue. The only limitation is knowledge and expertise in class D designs.
Outside of Chinese amp manufacturers, most class D amps use off-the-shelf modules that are pretty hefty size-wise. Few seasoned designers seem to be able to manage class D, which is why they nearly all use the off-the-shelf modules. I'm assuming it all comes down to noise filtering. I haven't tried any of the ultra-cheap Chinese class D speaker amps myself, but I have to imagine that if it were possible and cheap enough to do, one of the Chinese brands would have already made a push for class D headphone amps and portables.

Just curious what the big trade-off is. Must be a reason more than just "no one's done it" or "audiophiles don't like class D." May be as simple as the fact that class AB is already cheap enough to do in small form factors with low power output (for headphones and IEMs). Perhaps there's simply no need for the efficiency of class D when you don't need that much power to begin with?
 
Last edited:
Mar 2, 2025 at 7:45 PM Post #146,117 of 151,730
Outside of Chinese amp manufacturers, most class D amps use off-the-shelf modules that are pretty hefty size-wise. Few seasoned designers seem to be able to manage class D, which is why they nearly all use the off-the-shelf modules. I'm assuming it all comes down to noise filtering. I haven't tried any of the ultra-cheap Chinese class D speaker amps myself, but I have to imagine that if it were possible and cheap enough to do, one of the Chinese brands would have already made a push for class D headphone amps and portables.

Just curious what the big trade-off is. Must be a reason more than just "no one's done it" or "audiophiles don't like class D." May be as simple as the fact that class AB is already cheap enough to do in small form factors with low power output (for headphones and IEMs). Perhaps there's simply no need for the efficiency of class D when you don't need that much power to begin with?
I agree but I would appreciate a cm18p that doesn't run as hot and many portables are now pushing the power envelope so they should benefit from considering class d and from what ibasso has said, the amp17 should have ganfets which are only usable in class d designs.
 
Mar 2, 2025 at 8:45 PM Post #146,118 of 151,730
Been enjoying the Clarinet out of USB c, while my DX170 charges. It's brighter than Coax, but as you know, I can handle brightness with ease.

But it might be time to switch over to TOOBS, before my work day begins?

I used to hear people saying Coax is "warmer and more analogue" whilst USB is "more detailed and digital". Surprised to see it's actually not all hot air, as per your comment.

I checked around the internet a little bit and landed in certain "science" forum to hunt for some specs of Coax. Immediately repulsed by the attitude :dt880smile:


Outside of Chinese amp manufacturers, most class D amps use off-the-shelf modules that are pretty hefty size-wise. Few seasoned designers seem to be able to manage class D, which is why they nearly all use the off-the-shelf modules. I'm assuming it all comes down to noise filtering. I haven't tried any of the ultra-cheap Chinese class D speaker amps myself, but I have to imagine that if it were possible and cheap enough to do, one of the Chinese brands would have already made a push for class D headphone amps and portables.

Just curious what the big trade-off is. Must be a reason more than just "no one's done it" or "audiophiles don't like class D." May be as simple as the fact that class AB is already cheap enough to do in small form factors with low power output (for headphones and IEMs). Perhaps there's simply no need for the efficiency of class D when you don't need that much power to begin with?
I agree but I would appreciate a cm18p that doesn't run as hot and many portables are now pushing the power envelope so they should benefit from considering class d and from what ibasso has said, the amp17 should have ganfets which are only usable in class d designs.

You gentlemen should get together and open an online workshop "DAC and amp 101: what's inside them". I will watch.
 
Mar 2, 2025 at 9:04 PM Post #146,119 of 151,730
I used to hear people saying Coax is "warmer and more analogue" whilst USB is "more detailed and digital". Surprised to see it's actually not all hot air, as per your comment.

I checked around the internet a little bit and landed in certain "science" forum to hunt for some specs of Coax. Immediately repulsed by the attitude :dt880smile:





You gentlemen should get together and open an online workshop "DAC and amp 101: what's inside them". I will watch.
Yep. That's exactly how it is. How I would describe it is the Coax comes off as if you were listening to a well implemented r2r dac, and the USB c would be a saber DAC chip solid state. This is as close as I can come to describing the differences in sound. Also the soundstage is fuller in coax mode. More intimate in USB mode, but not tiny, just more intimate than coax.
 
Mar 2, 2025 at 9:05 PM Post #146,120 of 151,730
I used to hear people saying Coax is "warmer and more analogue" whilst USB is "more detailed and digital". Surprised to see it's actually not all hot air, as per your comment.

I suspect that in this case the Dethonray Clarinet is intentionally tuned to sound different on the coaxial and optical input.

I have several DAC and DAC/amp that sound the same to me via USB, coaxial or optical, and that would seem to be a design ideal, but the Clarinet does indeed sound different.

Dethonray offer three different firmware that also sound different, they are intended to be that way.

I suspect that the different sound signatures would show up on frequency response graphs just as subtle EQ would, I suspect unlike some of the sound differences reported between other devices.

Have you ever graphed one IEM with a number of sources and seen indications of changes in measurable sound ?
 
Last edited:
Mar 2, 2025 at 9:29 PM Post #146,121 of 151,730
I suspect that in this case the Dethonray Clarinet is intentionally tuned to sound different on the coaxial and optical input.

I have several DAC and DAC/amp that sound the same to me via USB, coaxial or optical, and that would seem to be a design ideal, but the Clarinet does indeed sound different.

Dethonray offer three different firmware that also sound different, they are intended to be that way.

I suspect that the different sound signatures would show up on frequency response graphs just as subtle EQ would, I suspect unlike some of the sound differences reported between other devices.

Have you ever graphed one IEM with a number of sources and seen indications of changes in measurable sound ?

I know that the tube amp of iBasso has a noticeable bass boost. Other than that, even with E5000 on apple dongle (extremely bad pairing) vs desktop amp on high gain, I didn’t detect any FR change. (I exclude andromeda here because it’s such an easy-to-change IEM due to its ugly impedance curve.)

Despite the same FR, I would never consider listening to E5000 on apple dongle. It’s noticeably bad vs a decent dongle. And decent dongle still sounds noticeably bad vs a portable amp at least. So FR is not the answer I’m looking for to explain this phenomenon.

Maybe “dynamic” amp does something similar to Joe Bloggs dynamic extender plugin 🤔 maybe IEM lose bass or treble precisely at the transient if the amp is not strong enough? I don’t get it. On paper, we have enough output power for even 110dB peak.



I remember there was a bluetooth dethoray that was absolutely punching bag a few years back due to very bad measurements. Good to see they fix that (or not)
 
Mar 2, 2025 at 9:33 PM Post #146,122 of 151,730
I suspect that in this case the Dethonray Clarinet is intentionally tuned to sound different on the coaxial and optical input.

I have several DAC and DAC/amp that sound the same to me via USB, coaxial or optical, and that would seem to be a design ideal, but the Clarinet does indeed sound different.

Dethonray offer three different firmware that also sound different, they are intended to be that way.

I suspect that the different sound signatures would show up on frequency response graphs just as subtle EQ would, I suspect unlike some of the sound differences reported between other devices.

Have you ever graphed one IEM with a number of sources and seen indications of changes in measurable sound ?
It's not just the sound (tonality), it's also the dynamics and stage. If it wasn't for that difference in dynamic stage/spatial cues in coax mode, I would be using the USB c mode and skipping coax, because technically, USB c + stock fw are my preferred signature, but you can't unhear the difference in dynamics, even if the coax is warmer and less treble forward than I typically prefer (but this still hasn't stopped me from using USB c, since I can take my thicc'r and more "safe" tuned IEMs, and throw them in USB, and they sound fantastic, aka Letshuoer Cadenza 4, ISN EBC80, Pula Anvil 114, and Tanchjim Bunny).
 
Last edited:
Mar 2, 2025 at 9:33 PM Post #146,123 of 151,730
Anyone know anything about a new Beoplay TWS. The elevens were an upgrade, but not a strong one. But there is rumor floating around on Reddit that new ones are supposed to be coming in March. And everywhere but the B&O website is sold out - which is always fishy. Any thoughts on this
 
Mar 2, 2025 at 9:50 PM Post #146,124 of 151,730
I know that the tube amp of iBasso has a noticeable bass boost. Other than that, even with E5000 on apple dongle (extremely bad pairing) vs desktop amp on high gain, I didn’t detect any FR change. (I exclude andromeda here because it’s such an easy-to-change IEM due to its ugly impedance curve.)

Despite the same FR, I would never consider listening to E5000 on apple dongle. It’s noticeably bad vs a decent dongle. And decent dongle still sounds noticeably bad vs a portable amp at least. So FR is not the answer I’m looking for to explain this phenomenon.

Maybe “dynamic” amp does something similar to Joe Bloggs dynamic extender plugin 🤔 maybe IEM lose bass or treble precisely at the transient if the amp is not strong enough? I don’t get it. On paper, we have enough output power for even 110dB peak.



I remember there was a bluetooth dethoray that was absolutely punching bag a few years back due to very bad measurements. Good to see they fix that (or not)

I would be interested to understand what technically can create a difference in sound between devices that show the same measured FR from a given IEM and the power required is tiny.

I know personally I have thought I heard differences but volume matching and blind listening make the differences disappear.

You have younger ears and almost certainly listen for minutiae that I don't but I have had numerous situations that have all but convinced me that, beyond tubes and outliers like the Clarinet that is clearly tuned differently, differences between most solid state devices working within their design range are psychological.

I know you won't agree but, as I said, various experiences have led me to that belief despite that sighted listening with an awareness of listening for differences creates differences for me, differences just like others report.

Dethonray measurements - I wouldn't mind betting the Clarinet measures poorly, just a hunch, never bothered to check.



@FreeWheelinAudioLuv2 - I get that with the Clarinet because it does something different with the coaxial input and different firmware, I believe essentially not a lot different to built in EQ and I would put money on it showing up on a FR graph.

For me the coaxial certainly has more weight down low and I believe that different tonality is what creates stage etc. Some Campfire IEM are famous for having good soundstage but at the expense of wonky tuning that messes up natural tonality to achieve the stage perception. Boosted bass will close the stage in making it seem more intimate.

On other devices I don't get that difference and any perceived difference I think I hear can be made to disappear by blind volume matched listening. The fact that the tonality is likely not measurably different as Genesis pointed out above makes me feel what I am hearing is supported by measurements.

Of course mileages vary and we don't want to go too far down this road :relaxed:



Edited to add - at the end of the day we are comparing personal listening experiences and those are very compelling regardless of the reality of how these nuances work. I respect your gents opinions and thoughts but I am going toi take some shifting from my current position on this stuff :relaxed:

I would be very grateful if somebody with a Clarinet could measure the FR of an IEM with the different input options and firmware.
 
Last edited:
Mar 2, 2025 at 10:05 PM Post #146,125 of 151,730
I know that the tube amp of iBasso has a noticeable bass boost. Other than that, even with E5000 on apple dongle (extremely bad pairing) vs desktop amp on high gain, I didn’t detect any FR change. (I exclude andromeda here because it’s such an easy-to-change IEM due to its ugly impedance curve.)

Despite the same FR, I would never consider listening to E5000 on apple dongle. It’s noticeably bad vs a decent dongle. And decent dongle still sounds noticeably bad vs a portable amp at least. So FR is not the answer I’m looking for to explain this phenomenon.

Maybe “dynamic” amp does something similar to Joe Bloggs dynamic extender plugin 🤔 maybe IEM lose bass or treble precisely at the transient if the amp is not strong enough? I don’t get it. On paper, we have enough output power for even 110dB peak.
If there were a way to measure or see why different amps sound different... I was going to say, we'd all have nothing left to talk about. But then I realized that audio hobbyists will always find something to argue about.

But, yeah, there's something to Joe Blogg's DRX plugin. I don't like its default setting since it swings way too hard for me. But if I tone it down a bit, it really does make my R4 DAP sound a lot closer to my desktop gear. Doesn't close the gap obviously, but it does move in the right direction. No idea how it works, but it's some serious technical magic that goes way beyond what I can understand.
 
Mar 2, 2025 at 10:40 PM Post #146,126 of 151,730
Finally, it's getting normal for studios to release material at 24 bit depth. 24 bits of accuracy slightly exceeds the tolerance of capacitors, which all of your gear requires. This steps us up from the range of relative volume variation your TT or CD players had by 50%, and probably even exceeds the capabilities of reel to reel wide and fast tape recorders we could never afford to pay for copies of albums as, after they beat recording straight to vinyl.
If you want to hear why digital audio spells money VS a TT setup, play a newer 24 bit track, (it was probably recorded straight to digital, and can meet the full 24 bits compared to tape), and listen for the ability to play the supporting sounds that aren't maximum volume at lower and therefore more richly accurate volumes than 16 bits allows to be represented.
Studios are getting away with this simple gear toggle to finally give us recordings that can exceed our playback gear's potential, even though 24 bits of variation only slightly exceeds our gear's tolerance. It works really well as an argument that beats TT's and even reel to reel, except for the part where only digital plays only samples for an only-ever too long of a time to bring out the variation of details, which requires the truly ever-changing time flow of non-digital audio. That's the part where studios are screwing their artists by not making the final wide-releases recordings that max out our newest gear, currently 384khz, or DSD 512 for our transfers from tape, (which will require no editing, just like DSD demands to prove it's superiority).
I've always thought that Pink Floyd's "The Wall" was clever garbage, and wasted some good musicianship from Pink Floyd, as they usually did. However, it came from late in tape's era, and Floyd is also good about not playing everything at max volume, so if you want to hear a transfer from tape to your new 24 bits of volume accuracy capabilities, that's a good recording to test it out on, currently only at (but at least) 96khz of sample rates. Plus, if you're like me, and were 10 years old when that came out, you've probably only ever heard it on crap, compared to your systems now, so you could be happy. Especially if you just received your box to put in front of your DAC that will rebuild any digital output you feed it into a really well built custom dedicated I2S output for your DAC. No more need for the uber-streamer instead of my jack-of-all-trades PC! It's true, once you've spent more than originally planned for a DAC that exceeds your expectations, you will get another 10% performance by buying a transport with a HQ digital output port on it for less $ than your DAC, instead of having to pay the insane 2x as much for the next DAC up, which your 1/2 price unit can still do 90% of. It's just that that same transport will give that DAC an extra 10% performance boost, too, so...
Now that everyone's digital playback has left physical CD's, there's no reason not to max out everyone's gear capabilities. Everyone can do 192khz of detail variation, by now, even your cheap phone's line out. But 24 bit recordings are ending the era of stunted sounding volume accuracy. Remember, 1 bit recordings would only be able to play all sounds that happen at 100%. 16 bit recordings just "wing" their non-max volume's accuracy, and round everything up, making it sound cheap, if you only knew the 24 bit versions first. DSD figured out ho to get past that, using millions of samples per second, instead. Except NO EDITING of DSD files! (The right one for old tape transfers, though). We need the mid-late studio Beatles re-releases in DSD 512, particularly for Abbey Road studios' era-busting recordings. Do you hear Paul's hard-soled shoe tapping the hard rubber coating on the floor of that studio? You didn't hear too much in the way of unnatural guitar reverb, thanks to that coating getting applied. If you ever get into a recording studio, you'd better not spill your drink on the rubber floor due to actually going with a girl for a change in there, either!
Welcome, 24 bit era, for maxing out our gear's volume sensitivity. Now, please focus on improving digital's detail sample rate. Details are always the harder part to max out our playback gear with, anyhow, although it's always possible for a favorite artist to produce a track that is a good simple sounding one, too.
Mmm, my EQ-plugin perfected HD800's for reference home monitoring. I've never heard anything else near this good. Really, the $3k Audeze planars are still only naturally EQ friendly, but with punchy bass instead of being analytical like these? I can add 2dB or more of sub bass, if I want more but unnatural amounts of bass. You'll never take my analytical ability after EQ away from me. I also have gigantic but lightweight cups that take up almost half of my head each, and my apparently my soundstage is king of headphones-sized. The word from users of the king of soundstages is that, to the dismay of the overly-bright best selling Grado fans, open backed headphones, when pushed to infinite volume levels in your own indoor listening room, can eventually make your cups the size of your walls, if you want reflections, but my soundstage with this current gear is unaltered 4 inches outside of my drivers, at this volume level, and it sounds really great, having a continually running deep bass line going on during this whole track. It's probably not always 4 inches out, I'm just cranked up listening to my new box-in-fronts output. A better digital signal into your DAC than a PC's USB port provides will make your soundstage maximum and open, and I got more bass slam after that, too. But if you don't want to spend $800, a 7 USB 3.1 port expansion card for $35 already kept my audio from sounding like it came from the cheapest output port possible, and made me wonder if it's still true that eventually I'll want that $300-500 audiophile USB port addon card. Unfortunately, those guys still have common sense on their side, even after this rebuilder.
 
Mar 2, 2025 at 10:45 PM Post #146,127 of 151,730
Finally, it's getting normal for studios to release material at 24 bit depth. 24 bits of accuracy slightly exceeds the tolerance of capacitors, which all of your gear requires. This steps us up from the range of relative volume variation your TT or CD players had by 50%, and probably even exceeds the capabilities of reel to reel wide and fast tape recorders we could never afford to pay for copies of albums as, after they beat recording straight to vinyl.
If you want to hear why digital audio spells money VS a TT setup, play a newer 24 bit track, (it was probably recorded straight to digital, and can meet the full 24 bits compared to tape), and listen for the ability to play the supporting sounds that aren't maximum volume at lower and therefore more richly accurate volumes than 16 bits allows to be represented.
Studios are getting away with this simple gear toggle to finally give us recordings that can exceed our playback gear's potential, even though 24 bits of variation only slightly exceeds our gear's tolerance. It works really well as an argument that beats TT's and even reel to reel, except for the part where only digital plays only samples for an only-ever too long of a time to bring out the variation of details, which requires the truly ever-changing time flow of non-digital audio. That's the part where studios are screwing their artists by not making the final wide-releases recordings that max out our newest gear, currently 384khz, or DSD 512 for our transfers from tape, (which will require no editing, just like DSD demands to prove it's superiority).
I've always thought that Pink Floyd's "The Wall" was clever garbage, and wasted some good musicianship from Pink Floyd, as they usually did. However, it came from late in tape's era, and Floyd is also good about not playing everything at max volume, so if you want to hear a transfer from tape to your new 24 bits of volume accuracy capabilities, that's a good recording to test it out on, currently only at (but at least) 96khz of sample rates. Plus, if you're like me, and were 10 years old when that came out, you've probably only ever heard it on crap, compared to your systems now, so you could be happy. Especially if you just received your box to put in front of your DAC that will rebuild any digital output you feed it into a really well built custom dedicated I2S output for your DAC. No more need for the uber-streamer instead of my jack-of-all-trades PC! It's true, once you've spent more than originally planned for a DAC that exceeds your expectations, you will get another 10% performance by buying a transport with a HQ digital output port on it for less $ than your DAC, instead of having to pay the insane 2x as much for the next DAC up, which your 1/2 price unit can still do 90% of. It's just that that same transport will give that DAC an extra 10% performance boost, too, so...
Now that everyone's digital playback has left physical CD's, there's no reason not to max out everyone's gear capabilities. Everyone can do 192khz of detail variation, by now, even your cheap phone's line out. But 24 bit recordings are ending the era of stunted sounding volume accuracy. Remember, 1 bit recordings would only be able to play all sounds that happen at 100%. 16 bit recordings just "wing" their non-max volume's accuracy, and round everything up, making it sound cheap, if you only knew the 24 bit versions first. DSD figured out ho to get past that, using millions of samples per second, instead. Except NO EDITING of DSD files! (The right one for old tape transfers, though). We need the mid-late studio Beatles re-releases in DSD 512, particularly for Abbey Road studios' era-busting recordings. Do you hear Paul's hard-soled shoe tapping the hard rubber coating on the floor of that studio? You didn't hear too much in the way of unnatural guitar reverb, thanks to that coating getting applied. If you ever get into a recording studio, you'd better not spill your drink on the rubber floor due to actually going with a girl for a change in there, either!
Welcome, 24 bit era, for maxing out our gear's volume sensitivity. Now, please focus on improving digital's detail sample rate. Details are always the harder part to max out our playback gear with, anyhow, although it's always possible for a favorite artist to produce a track that is a good simple sounding one, too.
Mmm, my EQ-plugin perfected HD800's for reference home monitoring. I've never heard anything else near this good. Really, the $3k Audeze planars are still only naturally EQ friendly, but with punchy bass instead of being analytical like these? I can add 2dB or more of sub bass, if I want more but unnatural amounts of bass. You'll never take my analytical ability after EQ away from me. I also have gigantic but lightweight cups that take up almost half of my head each, and my apparently my soundstage is king of headphones-sized. The word from users of the king of soundstages is that, to the dismay of the overly-bright best selling Grado fans, open backed headphones, when pushed to infinite volume levels in your own indoor listening room, can eventually make your cups the size of your walls, if you want reflections, but my soundstage with this current gear is unaltered 4 inches outside of my drivers, at this volume level, and it sounds really great, having a continually running deep bass line going on during this whole track. It's probably not always 4 inches out, I'm just cranked up listening to my new box-in-fronts output. A better digital signal into your DAC than a PC's USB port provides will make your soundstage maximum and open, and I got more bass slam after that, too. But if you don't want to spend $800, a 7 USB 3.1 port expansion card for $35 already kept my audio from sounding like it came from the cheapest output port possible, and made me wonder if it's still true that eventually I'll want that $300-500 audiophile USB port addon card. Unfortunately, those guys still have common sense on their side, even after this rebuilder.
Didn't you already post this lovely ramble already?
 
Mar 2, 2025 at 11:11 PM Post #146,128 of 151,730
My hearing time for the past several months has been occupied by two IEMs. Lime Ears Incognita and the Maris. Both these IEMS are so damn good, I have a new found respect for Emil at Lime Ears. Guy clearly knows what he is doing. Both these IEMs will be more than competitive in the higher end segment and the Maris caps off what is his masterpiece from the Power Trio line up.
1740974799997.png

Now I get to listen to more down to earth stuff well I am getting the new ISN H20 maybe next week. Will report about those when I get them in my ears.
In the mean while. My take on the Maris
https://www.head-fi.org/showcase/lime-ears-maris-of-the-sea.27988/review/37239/ These are mind-blowingly great as they should be for the price. My poor ears are now ruined. Lol.
 
Mar 2, 2025 at 11:11 PM Post #146,129 of 151,730
Didn't you already post this lovely ramble already?
Oh shoot, did I? I didn't know where to spout about the 24 bit improvement, all 24 bit threads were pre-2010 dead, and people just said stick your gear to 24 bits, it can only be better than 16.
Sorry if you're finding it a boring subject. It's a recording upgrade, so it should be normal for people to be talking fondly of it's merits, from time to time on an audio forum, and will be until nobody listens to 16 bit or vinyl recordings anymore, since 24 bits will have taken over as the new normal.
 
Last edited:

Users who are viewing this thread

Back
Top