[HeyManslowdown97]

I am the last person to comment on those since I removed them

Exactly.

Vocals are usually attenuated around 2-7k the zishan stock is lowered to avoid sibilance! If fact the slope is slightly elevated from 1-2k and should be lowered its very noticeable on some tracks running high in the tremble which lots of people have found similar results in the DSD stock tuning.

 

[Ivan TT]

Vocals are usually attenuated around 2-7k the zishan stock is lowered to avoid sibilance!

Any DAC or DAP is designed to deliver as close to straight line FR as possible, this is axiomatic.

My FR response chart you misinterpreted shows FR under a real life load (as opposed to FR into the input stage of a soundcard used to take measurements, which usually provides better looking straight line FR measurement, but does not reflect the FR under load, especially such o complex one as a 6 BA IEM, where complex interplay between crossovers and drivers inductance makes it a challenge to maintain even FR as evident from the measurement) and the unevenness across 2-7kHz is due to coloration caused by output DC blocking caps and BA drivers due to reasons outlined above. LF attenuation is caused by HPF which is totally unnecessary but is accidental and formed by DC blocking caps and a combination of output resistors and source resistance (and if the latter is low, the cut off frequency creeps in).

As to the LPF let me remind you that: 1. DAC has a built in LPF 2. Stock LPF cutoff frequency is around 182kHz (so will not show on any measurements unless they are done at just under 400kHz sampling rate) 3. Subtractor has 1nf caps in FB loop forming first order LPF at just under 200kHz 4. LPF filters in AK series datasheets are there to satisfy scarlet book requirements, which are clearly based on legacy x8 oversampling (as opposed to x256 in recent AKM DACs).

 

[HeyManslowdown97]

The frequency cutoffs are for the noise outside the audio frequency on the AK4497EQ the filters on the DAC and datasheet are for the pre-amps. It’s not to satisfy the scarlets book which is moot point since you MUST downsampling to Audio signal sampling rate and bit resolutions in the output!

LLPF filters in AK series datasheets are there to satisfy scarlet book requirements, which are clearly based on legacy x8 oversampling (as opposed to x256 in recent AKM DACs).“

DSD64, DSD128, DSD256 natively. I have no idea why you brought this up compared to your chart! Misdirect?? Again that’s what the Digital filter cutoffs are for: 39, 76, 176, 302, and 708 Hz on the chip!

Outside the audible signal 39 and 76 are digital filters which are the “cutoffs”and so are the 6 roll offs that are connected directly to the chip not that’s tuned as a preamp for the output audio signal. (Please!)

( 39 cutoff digital filter is almost doubled what the 20 that the human ear can hear)

These are PCM digital filters. Just so we are clear the Ak4497eq decodes the PCM from DSD specification it’s not a R2R Multibit Dac. I hope you know this!

Also, You do know what the DSD Pro is, what DSD is and and DXD are players are? And what the Ak4497eq chip made for?

There are two filtering systems one for DSD and PCM!

It’s not a “nod” to old DAC specification. Please take the time and READ what the Ak4497eq is and actually marketed for!

Opinion and speculation and only discusses PCM digital filtrations and built in filters operate using the firmware:

“Reading frequency response graphs in AK4497EQ datasheet show about 100dB attenuation by built in filter, page 15 for example:
https://www.akm.com/akm/en/file/datasheet/AK4497EQ.pdf
So extra LPF makes even less sense.

I believe it is there as a nod to standards that were developed for old DAC, where filters running at 1/2 of sampling frequency were critically important, but not so much for delta-sigma convertors with built in filters.“

https://www.head-fi.org/threads/the-zishan-dsds-corner.826185/page-129#post-14802116

In fact a lot of the frequencies in the Delta Sigma are inaudible in fact most are! The AK4497EQ is NOT bit perfect chip in fact which is run on STM32F Delta Sigma that’s why the cutoffs are on the firmware! It’s the only development board that I can think of that can run it!

It’s a Digital-to-analog converter you still needs those filters to shape the frequencies and tune them.

Laymen’s you need convert binary forms into audio frequencies that are created by filters!

I explained the slopes and how they work. The elevated flatlined is the headphone amp sans the filters. That’s how its read as the output and how you bragged that it was tuned that way. You elevated all the audio filters not the cutoff inaudible ones. It’s the analog signal filters were are talking not the noise-shaping ones you can’t hear!

That’s what is being charted.

You mean when your playback your highly compressed lossy MP3 files that may cause the coloration? My Zishan DSD doesn’t have that coloration. Maybe it’s your files? Or your mods? It’s gotta be one of them.

It seems correct before you got rid of both filters. Now it reads like a “stock” eq you would find in any Dap or music playback program leveled all above 0db I might add!

Other people who have charted their T1 the slopes are almost aligned but are underpowered. That’s voltage/current that the output stage is showing but they nearly match. Maybe you can ticker with the resistance levels or output cap BUT you still need those filters that you removed!

 

[Ivan TT]

These are PCM digital filters. Just so we are clear the Ak4497eq decodes the PCM from DSD specification it’s not a R2R Multibit Dac. I hope you know this!

AK4497 is delta-sigma DAC. So there is nothing "related" to PCM, including filters as PCM data stream is converted to DSD and then converted to analog signal (including application of digital filters that reduce out band noise introduced by delta-sigma modulator and its noise-shaping). Please review your post above taking into account this important piece of information (and toning it down would help too).

I see you don't understand what measurements show, but your interpretation is absolutely incorrect and based on wrong assumptions.
To clarify:
1. The measurement is of DSD's output under load (using sound card), not IEM's output (using measurement microphone)
2. Difference in level is due to removal of output resistors, and the level itself is rather arbitrary (as it depends on output volume and soundcard's gain) I set output volume at my comfortable listening level, way below what Zishan’s can handle.
3. Stock measurement is performed using STOCK DSD, no mods.
4. Measurement is of my 2nd DSD, that has LPF filters intact
5. Measurements are performed using dedicated software, as you should be aware, so lowly jab regarding my "highly compressed lossy MP3 files" comes from ignorance and is misplaced.

Finally, analogue LPF cutoff frequency is just under 190kHz. This is not only inaudible, but effect of it could only measured using audio interface operating at twice the sampling rate. What frequencies does it "shape" and "tune"?

 

[HeyManslowdown97]

The filters are as follows the PCM ones are actually away from the audio signal and must be selected by the firmware.

Both DAC and Ak4497eq is a high resolution player that requires files of the same quality and will also expose imperfections in the recordings (if they occurring making them sound worse!)

More issues ie coloration, DC offset, sibilance, etc in playback is generally coming from the files themselves. Mp3 for instance is a highly compressed file that has low resolution and less data to sample resulting in more quantitative errors in the playback!

(A perfect illustration of this is playing SD or inferior video playback on a HD Tv or if the original video wasn’t shot in HD)

There’s a plethora of material about why DACs have lpf and hpf filters and it’s very common if not mandatory to do. It also in the Ak4497eq datasheet evaluation board and isn’t placed there as some type of homage nostalgic “old dacs” like it was suggested! It’s a necessary step in the playback of the DAC with the ak4497eq!

Dithering, which acts like a filter, is introduced into the less used frequencies and has been scientifically proven to improve the audio when the the resolution is below 16 bit.


There are DACs that DO NOT upsample and read bit perfect which are called R2R and do not need to go through any processing which cost thousands of dollars!



https://www.allaboutcircuits.com/te...s-filter-a-pwm-signal-into-an-analog-voltage/

 

[HeyManslowdown97]

...

 

[fabien32]

which cost thousands of dollars!

Have you heard of TENTO Porta DAC 1866?

https://www.headfonia.com/tento-porta-dac-1866-all-your-bases/

 

[HeyManslowdown97]

Have you heard of TENTO Porta DAC 1866?

https://www.headfonia.com/tento-porta-dac-1866-all-your-bases/

Can’t say that I have but it looks interesting.

The R2R DAC is ALL about resistance (Resistance to Resistance) which makes them a challenge to make based on what I read.

The really great ones I’ve heard sound fantastic and lifelike! But it’s all about how well implemented they are same with Sigma Delta although the microprocessors have come a long way to sounding great.

My Zishan DSD sounds amazing and I can’t wait to listen to my T1 when it arrives!

Fyt: there’s also a lpf on this dac! Removing them seems odd to me!

 

[HeyManslowdown97]

First Impressions:

Soundwise: The T1 Ak4497eq has a different soundstage compared to DSD Pro I wanna say it’s “V Shaped” with the vocals in foreground and at center stage like they would be in “Mono” center track in stereo. It’s also isn’t as harsh in the tremble like the DSD Pro sound at times or on some tracks with lots of tremble elevated.

Without any modding the sound is VERY impressive with newer recordings it sounds great but can sound veiled at times especially in the top end. This can be remedied with some minor modification upgrades (op amp, caps, LDOs, clocks, etc). It’s actually impressive how much it packs a punch in such a small unit!

The touchscreen isn’t as bad once you know the “tricks” but it’s definitely not as touch sensitive as your modern smartphone or tablet (it’s more like circa 2010 touchscreen sensitivity) and the prompts aren’t seamless and fluid but as an “experienced” Zishan User you expect this and overlook this flaw because of their sound in the 100 dollar budget players that’s plays music like it’s triple or quadruple the budget.

This said, no gapless playback Zishan makers (what’s up with that!)...c’mon on already! Also, there’s this a second static noise that plays between transitioning songs on DSD files that should be fixed!

Attention BA iems Users!!!

I may have found an op amp that may solve your underpowered issues with the T1.

The AD8397 has a 310ma output current and comes in SOIC package which can be soldered/hot air soldered in lieu of the OP275 in output stage of the SE or in lieu of replacing the LM4562 2x in the lpf with an adapter in balanced if you like the sound over the SE??

The spec are as follows.

Features and Benefits

Dual operational amplifier

Voltage feedback

Wide supply range from 3 V to 24 V
Rail-to-rail output

Output swing to within 0.5 V of supply rails

High linear output current
310 mA peak into 32 Ω on ±12 V supplies while maintaining −80 dBc SFDR

Low noise

4.5 nV/√Hz voltage noise density at 100 kHz

1.5 pA/√Hz current noise density at 100 kHz

High speed

69 MHz bandwidth (G = 1, −3 dB)

53 V/μs slew rate (RLOAD = 25 Ω)

I have not tried this op amp so I don’t know how it sounds but it definitely satisfies the high current requirements needed to power BA iems!

I plan in the near future to experiment some op amp rolling combos and the AD8397 is my target to see which ones I like on my T1.

https://www.analog.com/en/products/ad8397.html

Battery life/“Hot” Zishan after hours of playback

I only play DSD files and so far on the T1 I have not experienced the very hot T1 after long playback hour(s) as many other T1 users have experienced. I believe the unit gets über hot for some users because converting PCM requires lots of cpu processing to decoding PCM especially in lower resolution. I haven’t tried my DXD files yet so I don’t know if heats up the unit like others’ have experienced with MP3 playback??

However, The battery life isn’t the greatest only getting about 2-2 1/2 hours per sesh with my DSD files but the unit’s small portability and fast charging usb-c makes a external battery charger not as inconvenient nor as much a nuisance to carry on-the-go!

 

[IcedFrosty]

First Impressions:

Soundwise: The T1 Ak4497eq has a different soundstage compared to DSD Pro I wanna say it’s “V Shaped” with the vocals in foreground and at center stage like they would be in “Mono” center track in stereo. It’s also isn’t as harsh in the tremble like the DSD Pro sound at times or on some tracks with lots of tremble elevated.

Without any modding the sound is VERY impressive with newer recordings it sounds great but can sound veiled at times especially in the top end. This can be remedied with some minor modification upgrades (op amp, caps, LDOs, clocks, etc). It’s actually impressive how much it packs a punch in such a small unit!

The touchscreen isn’t as bad once you know the “tricks” but it’s definitely not as touch sensitive as your modern smartphone or tablet (it’s more like circa 2010 touchscreen sensitivity) and the prompts aren’t seamless and fluid but as an “experienced” Zishan User you expect this and overlook this flaw because of their sound in the 100 dollar budget players that’s plays music like it’s triple or quadruple the budget.

This said, no gapless playback Zishan makers (what’s up with that!)...c’mon on already! Also, there’s this a second static noise that plays between transitioning songs on DSD files that should fixed!

Attention BA iems Users!!!

I may have found an op amp that may solve your underpowered issues with the T1.

The AD8397 has a 310ma output current and comes in SOIC package which can be soldered/hot air soldered in lieu of the OP275 in output stage of the SE or in lieu of replacing the LM4562 2x in the lpf with an adapter in balanced if you like the sound over the SE??

The spec are as follows.

Features and Benefits

Dual operational amplifier

Voltage feedback

Wide supply range from 3 V to 24 V
Rail-to-rail output

Output swing to within 0.5 V of supply rails

High linear output current
310 mA peak into 32 Ω on ±12 V supplies while maintaining −80 dBc SFDR

Low noise

4.5 nV/√Hz voltage noise density at 100 kHz

1.5 pA/√Hz current noise density at 100 kHz

High speed

69 MHz bandwidth (G = 1, −3 dB)

53 V/μs slew rate (RLOAD = 25 Ω)

I have not tried this op amp so I don’t know how it sounds but it definitely satisfies the high current requirements needed to power BA iems!

I plan in the near future to experiment some op amp rolling combos and the AD8397 is my target to see which ones I like on my T1.

https://www.analog.com/en/products/ad8397.html

Hello, SlowMan,
I'm glad to know that you have received your T1. And thanks for the quick impression. Once you feel like you're ready, please write a review on the review page of the T1.

Thanks,
Me

 

[Ivan TT]

Attention BA iems Users!!!

I may have found an op amp that may solve your underpowered issues with the T1.

There’s no under powering issues with T1, as it has a discrete transistor based class AB buffered output stage (under two larger capacitors). So not only it can provide sufficient current for low impedance earphones, but it also makes using high output current op amp irrelevant.

The problem with low impedance BA’s is in output protection resistors, which increase the output impedance and decrease the damping factor, making some BAs sounding thin and brittle.
Unfortunately this also applies to the balanced output, where 10ohm protection resistors are used.

Interestingly enough few modded versions of DSD that were available for purchase (eBay and taobao) used AD8620 in LPF. While I used it in LPF for its sound, it is also quite a decent HP driver: negligible output impedance, ability to drive capacitive loads, acceptable output current (for balanced configuration), built in short protection.

 

[HeyManslowdown97]

There’s no under powering issues with T1, as it has a discrete transistor based class AB buffered output stage (under two larger capacitors). So not only it can provide sufficient current for low impedance earphones, but it also makes using high output current op amp irrelevant.

He doesn’t have T1. I’m gonna try it and don’t care about this opinion.

 

[endia]

Interestingly enough few modded versions of DSD that were available for purchase (eBay and taobao) used AD8620 in LPF. While I used it in LPF for its sound, it is also quite a decent HP driver: negligible output impedance, ability to drive capacitive loads, acceptable output current (for balanced configuration), built in short protection.

Your post made me realize that 4497 is not using buf634s which drives headphones in "dsd 4495seq". maybe that is the reason for why 4495 was pricier than 4497..

 

[HeyManslowdown97]

Don’t think the T1 is a “most high-end audiophile headphone sources are well under 2 ohms” at budget price of 117 US dollars. Here’s what the famed O2 Headphone maker says about the Ohms ranging from 0-20 Ohms and what the T1 is within the consumer/prosumer market. But if you have homemade iems that are poorly tuned made for 40 bucks I’m sure they performed better than 500-1,000+ ones made by professionals with endless R & D dollars vs the diyer who reads stuff online as a knowable hobbyist who gives untested arbitrary figures with speculative subjective results playing MP3s on a high res player.

There is no standard for powering low impedance headphones that can find!

The 1/8 rule is by the O2 headphone maker.

PSUEDO STANDARDS: A lot of professional gear has a 20 – 50 ohm headphone output impedance. I’m not aware of any that follows the 120 ohm IEC standard. Consumer gear tends to be in the range of 0 – 20 ohms and, with the exception of tube and certain other esoteric designs, most high-end audiophile headphone sources are well under 2 ohms.”

“HEADPHONE DESIGN: So what output impedance do headphone manufactures design for? As of 2009 well over 220 million iPods had been sold. The iPod, and similar portable players, are the 800 pound gorillas in the headphone market. So, not surprisingly, most manufactures started designing many or all of their headphones to work well with the iPod. That means they’re designed to work with an output impedance under 10 ohms. And higher-end full size cans are most often designed for sources that follow the 1/8th Rule or have a near zero output impedance. I’m not aware of any current audiophile headphones intended for home use designed to the ancient 120 ohm standard.”

http://nwavguy.blogspot.com/2011/02/headphone-amp-impedance.html?m=1

 

[DBaldock9]

There’s no under powering issues with T1, as it has a discrete transistor based class AB buffered output stage (under two larger capacitors). So not only it can provide sufficient current for low impedance earphones, but it also makes using high output current op amp irrelevant.

The problem with low impedance BA’s is in output protection resistors, which increase the output impedance and decrease the damping factor, making some BAs sounding thin and brittle.
Unfortunately this also applies to the balanced output, where 10ohm protection resistors are used.

Interestingly enough few modded versions of DSD that were available for purchase (eBay and taobao) used AD8620 in LPF. While I used it for its sound, it is also quite a decent HP driver: negligible output impedance, ability to drive capacitive loads, acceptable output current (for balanced configuration), built in short protection. A

I've purchased 2 pairs of AD8397ARZ (310mA high output current) op-amps, and installed one pair on SOIC to DIP-8 adapters.
That pair has been used in several of my devices - and sounds good, with a solid low Bass drive capability.
NOTE: The maximum supply voltage for the AD8397ARZ is ±12.6VDC, so it shouldn't be used in devices with ±15VDC rails.

The other pair is destined to be used in the LPF of one of my DSDs (or possibly in a T1).
Is the Balanced circuit of the T1 like the DSD, in that there are two dual op-amps in the LPF, which drive the 2.5mm TRRS jack?