Topping DX5 II

[tccalvin]

Ask mix or mastering engineers if ringing and phase distortion are not that big of a deal. It’s a massive deal. Now I’m not sure the average listen can even spot either, especially without having as mental reference the appropriate sound, also I don’t think (at least I hope) neither are too bad in the case of the topping filters. Most likely they’re all decent and with some trade-offs among them. Also I invite you to spot a +- 0.3db difference at 20kHz, heck even a 1db difference. It’s not practically relevant. Now the same way add or remove a phase rotation that starts at 1kHz in something like a rock or metal song, and I’d argue even the average bear would hear a difference. I’d trade-off a bit of linearity in the frequency domain for phase linearity and clean transient any day.
Unfortunately you can’t have everything in life, there’s always a bit to trade off, even if new FIR and IIR filters are improving in design.

There is a big difference between equalizing a signal in the passband vs using a low pass reconstruction filter. Ringing and phase distortion artifacts caused by a DAC's LPF are tiny and concentrated around the band edge, where you likely don't hear anything anyway.

Group delay and phase delay - Wikipedia

en.wikipedia.org

 

[czAudiophile]

So in other words use, whatever I like correct?

 

[tccalvin]

So in other words use, whatever I like correct?

Of course. But if you aim for accuracy and want to minimize artifacts, use fast roll-off filters (linear phase, or minimum phase for low latency).

Shaving off ultrasonic content also helps keeping your other audio equipment alive.

 

[Dandelion]

There is a big difference between equalizing a signal in the passband vs using a low pass reconstruction filter. Ringing and phase distortion artifacts caused by a DAC's LPF are tiny and concentrated around the band edge, where you likely don't hear anything anyway.

Group delay and phase delay - Wikipedia

en.wikipedia.org

We’ve been talking about the decimator filter in first place, no need to bring in equalizers into the discussion. You’re wrong, more then once I’ve encountered anti-aliasing LPFs in minimum phase with phase rotations starting much much lower in the audible spectrum (a couple of times even as low as 1kHz), that’s not “tiny” nor “concentrated around the band edge”. It doesn’t mean is ubiquitous but you shouldn’t take filters’ quality as guaranteed.

 

[tccalvin]

We’ve been talking about the decimator filter in first place, no need to bring in equalizers into the discussion. You’re wrong, more then once I’ve encountered anti-aliasing LPFs in minimum phase with phase rotations starting much much lower in the audible spectrum (a couple of times even as low as 1kHz), that’s not “tiny” nor “concentrated around the band edge”. It doesn’t mean is ubiquitous but you shouldn’t take filters’ quality as guaranteed.

That's interesting. If you manage to gather any data feel free to post it. Would be much appreciated.

 

[Dandelion]

Of course. But if you aim for accuracy and want to minimize artifacts, use fast roll-off filters (linear phase, or minimum phase for low latency).

Shaving off ultrasonic content also helps keeping your other audio equipment alive.

See… Steeper filters come with trade-offs, ringing in case of linear phase and much more pronounced phase shift in case of minimum phase. You really just can’t say: it doesn’t metter, just use a steep one. You’re just assuming that filters are just about frequency response and ignoring all the rest. If that was the case people wouldn’t have hated the quality of the first period of digital audio (beginning of CD era) or wouldn’t care about DSD (which is all about the decimator filter acting way way up in frequency without disturbing the audible range the slightest).
Also how does aliasing effect the life of “other audio equipment”? It’s just audio folding down, in most of these filter what’s left is at volumes lower than 24-bit (with your hearing capped at 20-bit in the best case scenario).

 

[Dandelion]

That's interesting. If you manage to gather any data feel free to post it. Would be much appreciated.

Yeah it will take a while before I get my unit but it will be the first thing I’ll do. I would appreciate if someone can measure phase response of the different filters before then.

 

[tccalvin]

Yeah it will take a while before I get my unit but it will be the first thing I’ll do. I would appreciate if someone can measure phase response of the different filters before then.

Filters are the ES9039Q2M's. I don't think there's anything special about the Topping implementation.

SMSL filters mislabeled on ES9039 DACs

Several SMSL DACs using versions of the ESS ES9039 chip have mislabeled filters. First, I will use the case of the DL200 using one ES9039Q2M chip to show this is the case and correctly identify the filters. Then, I will show that other SMSL DACs using ES9039 chips namely the D-6S, DO100 Pro and...

audiosciencereview.com

Also how does aliasing effect the life of “other audio equipment”? It’s just audio folding down, in most of these filter what’s left is at volumes lower than 24-bit (with your hearing capped at 20-bit in the best case scenario).

Not the folds, but the true ultrasonic images that D/A conversion inevitably generates. To us, these are inaudible, but the unneeded ultrasonic signal is still in the current and causes your speakers to move accordingly (to name one example). I'm not an expert on the subject but I've read it's good practice to filter that current out to avoid IMD or component damage. At best, it's useless current.

 

[Dandelion]

Filters are the ES9039Q2M's. I don't think there's anything special about the Topping implementation.

SMSL filters mislabeled on ES9039 DACs

Several SMSL DACs using versions of the ESS ES9039 chip have mislabeled filters. First, I will use the case of the DL200 using one ES9039Q2M chip to show this is the case and correctly identify the filters. Then, I will show that other SMSL DACs using ES9039 chips namely the D-6S, DO100 Pro and...

audiosciencereview.com

Not the folds, but the true ultrasonic images that D/A conversion inevitably generates. To us, these are inaudible, but the unneeded ultrasonic signal is still in the current and causes your speakers to move accordingly (to name one example). I'm not an expert on the subject but I've read it's good practice to filter that current out to avoid IMD or component damage. At best, it's useless current.

Speakers reproduce whatever they can reproduce, they have natural roll-offs at their frequency bounds. I understand the concerns but electronics are built to handle stuff by themselves, they aren't built of fragile crystal. It reminds me of this thing in the macbook community where some people put all sorts of protections for keyboard, display, ports, etc. to protect the machines and those people end up ruining the devices the most obtaining the opposite because the laptops are built to be closed shut without nothing in the middle or without stuff obstructing the ports area, etc.

Thanks for the link, is partly data from the ESS spec sheets but is a decent deep down, only thing missing which is what (for my personal use case) I care the most is the phase distortion of the minimum phase filters, I'll try to add to that once I get my unit.

 

[wilhelm2015]

I haven't been able to hear an ounce of difference switching between the different PCM filters.

its normal, most people cant really hear above 16khz

 

[Tuner]

Is there any news on using Topping Tune with MAC OS? Or an APP to make PEQ adjustments? Supposedly Topping has been working on this for a while.

 

[ziggurcat]

Is there any news on using Topping Tune with MAC OS? Or an APP to make PEQ adjustments? Supposedly Topping has been working on this for a while.

I don't think anyone has any idea when Topping plans to release the Mac OS version. They are going to release a phone app for Android/iOS, but there's also no known ETA.

 

[Erata]

Has anyone tried driving a HE6se V2 or similar low sensitivity headphones using the DX5 II? At max volume and gain, the HE6se V2 on DX5 II XLR output is only slightly louder than on the Schiit Magni Unity I have. And the iFi micro iDSD Signature Finale can make the HE6se V2 way louder compared to the DX5 II.

According to the reported specs, the Magni Unity outputs 1.5W at 50 ohms and the iFi micro iDSD Signature Finale outputs 1.56W at 64 ohms. The DX5 II should have way more power. Is it possible I have a defective unit?

 

[zonk]

Has anyone tried driving a HE6se V2 or similar low sensitivity headphones using the DX5 II? At max volume and gain, the HE6se V2 on DX5 II XLR output is only slightly louder than on the Schiit Magni Unity I have. And the iFi micro iDSD Signature Finale can make the HE6se V2 way louder compared to the DX5 II.

According to the reported specs, the Magni Unity outputs 1.5W at 50 ohms and the iFi micro iDSD Signature Finale outputs 1.56W at 64 ohms. The DX5 II should have way more power. Is it possible I have a defective unit?

According to the DX5 II specs the output on 4.4mm/XLR connector is
- 4300mW x 2 @64Ω THD+N<1%

on the 6.35mm connector:
- 1160mW x 2 @64Ω THD+N<1%

Did you use High Gain?

 

[Erata]

According to the DX5 II specs the output on 4.4mm/XLR connector is
- 4300mW x 2 @64Ω THD+N<1%

on the 6.35mm connector:
- 1160mW x 2 @64Ω THD+N<1%

Did you use High Gain?

Yep, I used high gain. I also have PEQ disabled.

 

[JIW]

Has anyone tried driving a HE6se V2 or similar low sensitivity headphones using the DX5 II? At max volume and gain, the HE6se V2 on DX5 II XLR output is only slightly louder than on the Schiit Magni Unity I have. And the iFi micro iDSD Signature Finale can make the HE6se V2 way louder compared to the DX5 II.

According to the reported specs, the Magni Unity outputs 1.5W at 50 ohms and the iFi micro iDSD Signature Finale outputs 1.56W at 64 ohms. The DX5 II should have way more power. Is it possible I have a defective unit?

Relative to its unbalanced line-outs, the DX5 II balanced HPA high gain is 16 dB while the Magni Unity high gain is 15 dB. The iFi is specified to have a maximum output voltage in excess of 10 V RMS - less than the DX5 II - but the gain is not specified.

For a 50 Ohm load, the Magni Unity in high gain is clipping at around 10*log10(1.5*50) -(8 + 15)= -4.24 dBFS when at full volume and the DX5 II at full volume as well. For a 32 Ohm load, the DX5 II balanced HPA in high gain clips at around 10*log10(6.4*32) - 24.7 = -1.59 dBFS at full volume. For a 64 Ohm load, it clips around 10*log10(4.3*64) - 24.7 = -0.30 dBFS. With demanding enough material, you should hear distortion from the Magni at lower levels. Also, note that odd-order harmonic distortion typically arising from clipping solid state amplifiers tends to increase loudness.

 

[czAudiophile]

Recently changed from Ananda Stealth to Arya stealth.
On my DX5 2, and I can hear a clear difference in sound.
Looks like planars are the same, but cup rotation and aligment is different. Thus bass sounds way more plesant.
Now they are close to 600$ which is a steal from my point of view.
Bass finally has a punch that I adore, and finally some weight there.

 

[czAudiophile]

Not long ago, I criticized Hifiman for its lack of innovation. In particular for the Unveiled model series. I then bought the Arya Stealth (for the third time) and tried out Crinacles preset.
From this I was able to build a preset for myself, which made me like the Arya Stealth better than the He6SE. So the HE6SE had to go.

Today I had the opportunity to compare the Arya Stealth, HE1000SE, and Arya Unveiled. Using the EQ profiles from Crin and Resolve, I tried to match all three headphones to the same FR.

View attachment 461200

For me, the Arya Unveiled is the clear winner.
It sounds by far the most authentic. The timbre is sensational.

As of today, I can understand the hype surrounding the Arya Unveiled.
I also agree with what some reviewers have said: the Arya Unveiled sounds incredibly authentic. There is no harshness or edgyness to be heard.
This has become clear in direct comparison with the other two headphones.
No matter what I did in the range between one and 15 kHz, I couldn't correct the basic character in the form of sharpness and edgyness that both have.
The Arya Unveiled simply doesn't have any of that.
It just sounds incredibly pleasant.
I didn't miss anything. Not punch, not resolution.
OK. I digress...

Just saw this. Can you please provide your EQ settings. Cause I've tried ones from autoeq, and do not like them at all.

 

[Bow_Wazoo]

For stealth or Unveiled?

Btw. the best headphone I've ever heard:

 

[czAudiophile]

for stealth