Topping DX5 II

[Dandelion]

Can't see a sine on your chart at -4.5dBFS.
Can't be the 20Hz one as no H2 and H3 are at 40Hz and 60Hz.

Also there's no normal visible noise floor.
Is it some kind of cross-correlation method used?

The whole pattern looks weird, are you sure is not some kind of triangle or square wave here?

Plugindoctor

Discover DDMF’s range of high-quality audio software for mixing, mastering & sound design. Pro sound, intuitive design, and great value.

ddmf.eu

 

[Sokel]

Plugindoctor

Discover DDMF’s range of high-quality audio software for mixing, mastering & sound design. Pro sound, intuitive design, and great value.

ddmf.eu

I got the plug-in, but is just as good as the quality of the signal you feed it.

Look at this noise floor:

Looks familiar?

It's actually an accident, Direct monitoring accidentally activated (on purpose for this test though) fighting with software monitoring.
Two signals fighting each other essentially, and the result is of course clipped too.

Whenever you see such weird noise floor something is amiss.

For clarity, what's what it looks like if Direct is deactivated:

 

[Dandelion]

I got the plug-in, but is just as good as the quality of the signal you feed it.

Look at this noise floor:

View attachment 465559

Looks familiar?

It's actually an accident, Direct monitoring accidentally activated (on purpose for this test though) fighting with software monitoring.
Two signals fighting each other essentially, and the result is of course clipped too.

Whenever you see such weird noise floor something is amiss.

For clarity, what's what it looks like if Direct is deactivated:

View attachment 465560

Let me get this straight: you've put a testing plugin inside a testing plugin? Yes it plots the noise floor of the plugin (or machine) being tested inside it, it feeds a pure signal inside what you want to test, you don't have to feed tones in it yourself. Hence if for example you throw in it a poorly coded plugin it will show a higher noise floor due to quantization noise.

 

[Sokel]

Let me get this straight: you've put a testing plugin inside a testing plugin? Yes it plots the noise floor of the plugin (or machine) being tested inside it, it feeds a pure signal inside what you want to test, you don't have to feed tones in it yourself. Hence if for example you throw in it a poorly coded plugin it will show a higher noise floor due to quantization noise.

No, I just replicated the noise floor I see at the chart you posted.
No machine can do that (that noise floor) on its own unless is broken.

The why this plug-in shows this result needs to be searched upwards, at the signal and the settings.
To me it just shows conflict. Not a valid way to test gear.

In short: the plug-in we see there seems broken by design unless is some kind of deliberate effect.

 

[Dandelion]

No, I just replicated the noise floor I see at the chart you posted.
No machine can do that (that noise floor) on its own unless is broken.

The why this plug-in shows this result needs to be searched upwards, at the signal and the settings.
To me it just shows conflict. Not a valid way to test gear.

Good lord sir, you're turning something easy into something difficult going through the useless. Save yourself the time. PluginDoctor serves to test plugins, the harmonic blend I posted is not equipment and you didn't replicate any of it, do this: open PluginDoctor and put inside a plugin that does nothing to the audio signal, something like Hofa 4U+ ProjectTime or BlindTest and you'll see the noise floor at 64-bit precision. (-175db or something) and the pure test signal as line (you can setup the volume and the frequency of such test sine tone as you can see). What I posted is the harmonic analysis of a plugin doing saturation blends, which on its settings had a bunch of parameters, gain staging, wet/dry etc. don't obsess over it, it was just complementary to my post. I understand without context is harder to interpret, but the point is: you have 2 signals (a guitar, a full mix, a drum, whatever) same perceived loudness, same frequency response one sounds fuller and more mid forward than the other (because of harmonic distortion). Yes, a headphones preamp is not as colored as distortion box, obviously, but color and tone in preamps is a thing.

I would really enjoy users having both machines doing blind tests, and no, 90% is unnecessary and irrealistic, I could put a literal eq bell boost in one of the 2 signals and I guarantee most people wouldn't spot it 70% of the times, let alone 90%

 

[RHO]

you have 2 signals (a guitar, a full mix, a drum, whatever) same perceived loudness, same frequency response

Before or after it goes trough the RME?
If it's at the output, then certainly, no. Audible amounts of THD are certainly measurable and will result in a different frequency response.

one sounds fuller and more mid forward than the other (because of harmonic distortion).

Because of inaudible levels of harmonic distortion the RME adds?

I would really enjoy users having both machines doing blind tests, and no, 90% is unnecessary and irrealistic, I could put a literal eq bell boost in one of the 2 signals and I guarantee most people wouldn't spot it 70% of the times, let alone 90%

70% is too close to guessing.
I was actually being lenient here, because in general a 95% (or 5% if you accidentally switch the samples) score is expected.

 

[JanHantschke]

That doesn't make any sense to me - why would they remove that text display? And user experience shows that that was added in 1.49 not removed.

If you check Chinese changelog, it says "修复：耳放直流量检测到异常后，主界面不显示"HP DC abnormal"" which when translated to English with Google (as I don't understand Chinese) is "Fixed: After the headphone amplifier DC current is detected to be abnormal, the main interface does not display "HP DC abnormal"".

That makes more sense: i.e. there was a problem that the text wasn't being displayed and that problem has now been fixed. Seems to me that whoever wrote English changelog doesn't know English that well and made an error there, i.e. "no longer displays" instead of "does not display".

Thanks for checking. I also wanted to point out that this is likely an translation error but forgot it in my original post.

 

[Dandelion]

Before or after it goes trough the RME?
If it's at the output, then certainly, no. Audible amounts of THD are certainly measurable and will result in a different frequency response.

Because of inaudible levels of harmonic distortion the RME adds?

70% is too close to guessing.
I was actually being lenient here, because in general a 95% (or 5% if you accidentally switch the samples) score is expected.

Obviously at the output which gets compensated in level, and I'm talking in general, not about a specific device.Technically yes, practically can often be too small to trigger eq matching tools like fabfilter or ozone (state of the art) or others to trigger any adjustment, nor you can spot any difference by eyes when freezing spectrum analyzers.

Then you would have large % of people fail to recognize devices that cost $100 from ones that cost $3000 in a blind test, guaranteed. I'm telling you, even with actual small eq changes people would still not spot them, let alone harmonic saturation.

 

[JIW]

Just installed it. Not bricked. Menu colours have been fixed. Relay clicking on HPA SE witrh headphone unconnected has been fixed.

Preamp/DAC mode volume memory has not been fixed. However, I found out it needs about 3 seconds to memorise the volume. Still, in preamp mode, it turns on where it was last set in DAC mode if volume is not changed in preamp mode. In DAC mode, it works fine. So, overall, this is not that big of an issue.

I may test more at a later time.

Topping replied to my message about it.

The device currently takes about 15 seconds to save any changes, including volume adjustments, so please test again to see if all modifications can be successfully saved.

Waiting 15 seconds between switches from preamp to DAC mode with volume adjustment immediately after switching and power cycling does not change anything.

Volume is memorised after about 3 seconds as it is sometimes maintained after a power cycle about 3 seconds after the volume change. At 4 seconds, it is always maintained and at 2 seconds it is never maintained.

 

[RHO]

Then you would have large % of people fail to recognize devices that cost $100 from ones that cost $3000 in a blind test, guaranteed. I'm telling you, even with actual small eq changes people would still not spot them, let alone harmonic saturation.

Exactly!
But the result would be statistically valid.

 

[Dandelion]

Exactly!
But the result would be statistically valid.

This opens a whole can of worms about what is actually statistically valid, nonlinear systems, normal distribution, mandelbrot, black swans, p-hacking and what not which covers everything from finance to medicine to artificial inteligence and most things in life which is way beyond the scope of this forum.

I would tend to say that just like any other human physical features hearing follows normal distribution hence 95% CI is enough to establish something as valid and something above 1 sigma deviation (68%) as relevant.
But there's indeed people with a more trained ear. Recently a mix/mastering education platform and plugin manufacturer released an eq/ear training tool with levels of difficulty that go from 1 to 100, I fiddled around a couple of afternoons and reached 91, 5th by difficulty in global leaderboard and 13th in absolute position (people who reach a score point earlier are on top of others which share the same score). If that means something I would say I maximized what mother nature gave me with decades of experience, and I'm perhaps near the peak before age starts to show its toll on hearing.

 

[ziggurcat]

Just had dispatch notice from Shenzenaudio for my White DX5 II, so the new stock has finally turned up of at least the white models.

My silver DX5 II just shipped this morning as well.

 

[The Noob Guy]

Just had dispatch notice from Shenzenaudio for my White DX5 II as well

 

[ngs428]

Just had dispatch notice from Shenzenaudio for my White DX5 II as well

Congrats! My black one was just pushed out to late August.

 

[The Noob Guy]

Congrats! My black one was just pushed out to late August.

Awe that sucks! Sorry to hear

Edit: If anyone wanted to know when I ordered it was June 10th

 

[Nezil]

OK, so I'm pretty convinced that the PEQ memory handling still isn't 100%. I cannot get the PEQ setting to track the output, it w

Giving full details of my setup here in case someone else can replicate.

I have my unit set up for two output options, LO ALL and HPA BAL, so these are what I'm toggling between.
I have 3 PEQ memories configured, 1. is for my headphones, and 2 and 3 are for my speakers, with 3 being what I usually use.

When I switch outputs, which I have set to shortcut on the volume push, the PEQ config remains what it was set to for the other input.

What I think is weird, is that I can tell the UI and the EQ itself match, in that they show the EQ that really is being applied, and it is the EQ setting from the other output, but... when I try to change the EQ to the correct one, the first click selects the next memory from where it should have been. I'll try to explain this by going step by step:

1. Configure HPA BAL to PEQ 1.
2. Switch to LO ALL and configure to PEQ 3.
3. Switch to HPA BAL and note that the menu, and the EQ that is actually being applied is PEQ 3.
4. Press the button to configure the PEQ and it steps to PEQ 2, not PEQ 1 which would be the logical next step from the menu which displays PEQ 3.

It's like the menu and the PEQ memory itself is not getting applied correctly, but some part of the software thinks it is.

Can anyone else reproduce this?

In addition to this concern, I'm also not certain that the PEQ is being applied to the SE LO output. I have a sub connected to the SE RCAs and my main speakers connected to the Balanced output. I've set the same PEQ on the DX5 II as I had been running on my Mac, and I am getting a lot more bass from the sub than I was previously.

Measurements would confirm this, but again, putting it out there in case anyone else had noticed this.

 

[jodrim]

OK, so I'm pretty convinced that the PEQ memory handling still isn't 100%. I cannot get the PEQ setting to track the output, it w

Giving full details of my setup here in case someone else can replicate.

I have my unit set up for two output options, LO ALL and HPA BAL, so these are what I'm toggling between.
I have 3 PEQ memories configured, 1. is for my headphones, and 2 and 3 are for my speakers, with 3 being what I usually use.

When I switch outputs, which I have set to shortcut on the volume push, the PEQ config remains what it was set to for the other input.

What I think is weird, is that I can tell the UI and the EQ itself match, in that they show the EQ that really is being applied, and it is the EQ setting from the other output, but... when I try to change the EQ to the correct one, the first click selects the next memory from where it should have been. I'll try to explain this by going step by step:

1. Configure HPA BAL to PEQ 1.
2. Switch to LO ALL and configure to PEQ 3.
3. Switch to HPA BAL and note that the menu, and the EQ that is actually being applied is PEQ 3.
4. Press the button to configure the PEQ and it steps to PEQ 2, not PEQ 1 which would be the logical next step from the menu which displays PEQ 3.

It's like the menu and the PEQ memory itself is not getting applied correctly, but some part of the software thinks it is.

Can anyone else reproduce this?

In addition to this concern, I'm also not certain that the PEQ is being applied to the SE LO output. I have a sub connected to the SE RCAs and my main speakers connected to the Balanced output. I've set the same PEQ on the DX5 II as I had been running on my Mac, and I am getting a lot more bass from the sub than I was previously.

Measurements would confirm this, but again, putting it out there in case anyone else had noticed this.

Is it the same issue as this?

Yeah I can hear it changing. Visually when listening to Arya through balanced it says EQ1 top left HPA BAL, when I change to HD620S with HPA SE it says EQ2 and listening to music through either it sounds wonky because of these ridiculous EQ presets that I set to test - The EQ profiles were too similar to hear the difference

I hear the same issue. With these profiles EQ1 has a distinct weird bright sound and EQ2 sounds like music playing in another room. With EQ1 set for HPA BAL and EQ2 for HPA SE, if I switch from HPA BAL with EQ1 to HPA SE I hear EQ1 even though it says EQ2. I have to manually switch through the profiles to EQ2 to hear the "other room" EQ sound. So this is definitely a bug that needs to be fixed.

A workaround for now would be to store the EQ1 and HPA BAL on C1 for example and then EQ2 for HPA SE on C2. Then switch between them via C1 and C2 buttons on the remote as the EQ switch works when switching between the configs.

View attachment 465099View attachment 465100

These are two good EQ configs to hear if it's working or not. It either sounds like it's playing in a pringles tube or from a Nokia phone

 

[Nezil]

Is it the same issue as this?



These are two good EQ configs to hear if it's working or not. It either sounds like it's playing in a pringles tube or from a Nokia phone

I think so, yes.

At least my first issue; I'll have to do some more debugging to confirm if the PEQ isn't getting applied to RCA.

I'm confused though, because it seems like you're having this issue @jodrim, but others are saying that PEQ memory and moving with the output is working fine.

 

[jodrim]

I think so, yes.

At least my first issue; I'll have to do some more debugging to confirm if the PEQ isn't getting applied to RCA.

I'm confused though, because it seems like you're having this issue @jodrim, but others are saying that PEQ memory and moving with the output is working fine.

Yeah I did a full factory reset and chose to remove EQ profiles and it's still happening. Maybe people think it's working because the EQ configs are fairly similar just like I thought at first? Hopefully just a bug in the firmware because it seems fairly trivial to fix.

Edit: The same thing also happens with the input memory option. So with "PEQ memory" set to "Input" instead of "Output" if I have high-shelf on USB and low-shelf on BT switching from one to the other the previous EQ profile plays but the UI says the saved one. I've emailed Topping about the issue so I'll see what they say.

Edit 2: When changing the outputs and the UI on the DX5 updates to the intended saved EQ profile Topping Tune stays on the old one also so I'm pretty sure this is just a bug of it not changing the PEQ config in use when changing inputs or outputs.

 

[ziggurcat]

In addition to this concern, I'm also not certain that the PEQ is being applied to the SE LO output. I have a sub connected to the SE RCAs and my main speakers connected to the Balanced output.

To my knowledge, you can't have two PEQ filters active at the same time, if that's what you're trying to accomplish.