• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Topping G5 Review (Portable DAC & HP Amp)

Rate this DAC & HP Amp

  • 1. Poor (headless panther)

    Votes: 11 3.0%
  • 2. Not terrible (postman panther)

    Votes: 20 5.5%
  • 3. Fine (happy panther)

    Votes: 75 20.7%
  • 4. Great (golfing panther)

    Votes: 256 70.7%

  • Total voters
    362

KeithPhantom

Addicted to Fun and Learning
Forum Donor
Joined
May 8, 2020
Messages
641
Likes
658
Thanks for your explanation. Since the corner frequencies in the 44.1k and 48k sampling graphs appear to be different, I'm wondering whether this is the same filter exhibiting different responses under different sampling rates, or the firmware specifically selecting different filters when the sampling rate changes.
That's what I am not sure yet, but without having access to the documentation or some kind of other source, I can't say for sure. I think it doesn't use the exact same filter, but it may be moving the Fc as you change the sample rate. Documentation rarely has transfer functions for filters, but if I can find something I will post it.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
I’m sure everybody noticed that bit, to me it is the most unsettling part of the whole review. As far as I can see (but I certainly have not read all, or even the majority of, Amir’s reviews) this is unprecedented. An objective measurement is being overruled by a subjective and uncontrolled listening test. I had to scroll up to check for a byline like “Contributing editor: Darko” ;)
The only thing unsettling is to read comments like this. This is a world-class portable product. It sets new bar by far when it comes to distortion and noise. It has plenty of power for a portable product to drive difficult headphones. You expect me to throw all this away and rate it lower just like other products with much less performance?

Darko would tell you that the less performant Chord Hugo 2 sounds more musical. Don't insult me with nonsense like that.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Oh, the relative difference of the chosen corner frequency [Fc] (usually 20,000 Hz but it can be any other if within Fs/2) and the stop frequency will increase (stop frequency will increase from 22,050 to 24,000 Hz if keeping a constant Fc of 20,000 Hz), giving you a filter with different results (different transition band slope, thus different transition band). This is possible, but I can't confirm this is the case. Most likely, they are using a filter depending on the input sample rate by using the different registers in the package, but I can't find an extensive ESS 9068AS datasheet to confirm.
I already showed this to be the case on G5: https://www.audiosciencereview.com/...portable-dac-hp-amp.36542/page-7#post-1283962

index.php
 

KeithPhantom

Addicted to Fun and Learning
Forum Donor
Joined
May 8, 2020
Messages
641
Likes
658
The fact that the objective measurement being cast aside here is one where the device does not exhibit SOTA performance makes the whole affair seem like grading on a curve. Otherwise, if “sounds great” was the final arbiter in ASR reviews, there should be a hell of a lot more golfing panthers.
The majority of the measurements indicate that the device achieves State-Of-The-Art performance for its category. It is the reference to compare other devices in their category and use case against. And this is using the data not produced by ears, but by a machine that has sensitivity down to -160 dBFS (and can be further calibrated). So I think it is fair to consider it as SOTA. State-Of-The-Art does not mean perfection, it just means that it is the current best benchmark to compare to similar products; what was SOTA 10 years ago is pretty much average (or even below depending on your cutoff metrics).
 
Last edited:

Labjr

Major Contributor
Joined
Dec 14, 2018
Messages
1,051
Likes
949
Nice review but not really a product for me. I'm not willing to schelp another device around for listening on my phone. I don't even like carrying a phone. Also, this DAC/Amp has no line level outputs. So not usefull for much else. My guess is they'll drip-feed feauures like line outs and then balanced connections in future product cycles.
 

jerm_ph

Member
Joined
Feb 16, 2021
Messages
14
Likes
34
Location
Philippines
Nice review but not really a product for me. I'm not willing to schelp another device around for listening on my phone. I don't even like carrying a phone. Also, this DAC/Amp has no line level outputs. So not usefull for much else. My guess is they'll drip-feed feauures like line outs and then balanced connections in future product cycles.
E1750841-F281-4CFA-A985-641BCA8B7F83.jpeg
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
You mean many non-Topping devices! ;)
At any rate, the filter wouldn't bother "me".
But this device indicates, that making a portable DAC amp, with the right size, weight and battery life, and having exemplary measured performance, is not that easy.
If one really feels the need to compare it to original Mojo, it is a little bigger (by 70%) and heavier (guessing, how heavy is it?), same sort of battery life (9h as a dac amp), does have balanced output (not Mojo), more max output, but higher output impedance, and an analogue input (what on earth, for?).
G5 is 70% bigger than Mojo, Hugo2 is 70% larger than G5.
Measured performance, over all, similar! Though Mojo has no specific area in the measurements, one can put a finger on.
The adverts keep repeating 2 X 1200mW output, does it genuinely have two output amplifiers?
Or is it the same output, just wired in single ended mode to both jacks, akin Chord DACs?
2x1200mW means both channel driven(left + right). There were people/manufacturers using power measurements with only single(left or right) channel driven.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
I don't think this is specific to the unit you tested Amir, as in the published product manual (very small) I believe the same issue is shown;


View attachment 224537

In the same manual reference is made to;

View attachment 224539

This unit uses the ESS ES9068AS which actually has three preset filters available;

View attachment 224538

So unsure if the filter is one of the presets or a custom filter @JohnYang1997;

... is capable of rendering MQA too.

Still, a good product and well priced.

Thanks for testing of course, your time is appreciated.


JSmith
The published fr is tested with 96khz. So the result is a much flatter response than with 44.1khz. The consideration was that people were caring about the cutoff 22.05khz so f3 was chosen but if we want flat 20khz more then yeah we can put out a new firmware for that.
 

audiofun

Active Member
Joined
Jul 4, 2021
Messages
242
Likes
243
The published fr is tested with 96khz. So the result is a much flatter response than with 44.1khz. The consideration was that people were caring about the cutoff 22.05khz so f3 was chosen but if we want flat 20khz more then yeah we can put out a new firmware for that.
All Topping's ESS products have high noise issue when playing under DSD128 or DSD 256. Is this something you can fix via firmware?

 
Last edited:

Labjr

Major Contributor
Joined
Dec 14, 2018
Messages
1,051
Likes
949
All Topping's ESS products have high noise issue when playing under DSD128 or DSD 256. Is this something you can fix via firmware?


Archimago measured good DSD results with the SMSL DO100.

http://archimago.blogspot.com/2022/06/review-smsl-do100-dac-part-iii-dsd.html
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,682
Likes
4,220
Location
Liège, Belgium
.Afaik there's no equivalent to HQPlayer for Android or macOS, especially not one that works with streaming services like Apple music and Spotify, which are likely to be used with the G5.
On Android, USB Audio Player Pro can do that, I think.

I use it with Qobuz in bitperfect through USB, but you may as well force resampling to highest frequency rate the DAC will accept (or the closest integer multiple of the source frequency).
I didn't try it though.
It also works with Tidal.

This app has some bugs, but, overall, it works fine.
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,682
Likes
4,220
Location
Liège, Belgium
There are players (such as Neutron) that can play hires from local files, DLNA, SMB shares, etc (and also do PEQ and upsampling), but don’t work with streaming services. Apple Music, Qobuz and the like have hires content that will work with USB DACs, but the majority of their content is still 44.1kHz.
See just above
 

ririt

Senior Member
Joined
May 5, 2020
Messages
363
Likes
342
Location
France
I had a quick look on the websites which are offering the G5 (audiophonics, shenzenaudio…). None of them seem to offer the lightning version of the G5 which includes the suitable cable. Does anyone know how to get it?
 

Tom C

Major Contributor
Joined
Jun 16, 2019
Messages
1,501
Likes
1,370
Location
Wisconsin, USA
I’m sure everybody noticed that bit, to me it is the most unsettling part of the whole review. As far as I can see (but I certainly have not read all, or even the majority of, Amir’s reviews) this is unprecedented. An objective measurement is being overruled by a subjective and uncontrolled listening test. I had to scroll up to check for a byline like “Contributing editor: Darko” ;)

The fact that the objective measurement being cast aside here is one where the device does not exhibit SOTA performance makes the whole affair seem like grading on a curve. Otherwise, if “sounds great” was the final arbiter in ASR reviews, there should be a hell of a lot more golfing panthers.
The whole point of this entire exercise IS grading on a curve. Competition breeds excellence.
 

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
what is happening here? everyone that complains about there being problem with the engineering should really take a look at amir's replies.
some people here are trying to spread false information of topping trying to cheat by using this filter for better measurement, which is not the case at all, confirmed by amir with proof. so why are there still argument regarding this?
as for the filter itself, you may argue this product have this one flaw, but not due to its engineering.
in fact, they engineered it to the best possible, in the condition that this specific dac chip being utilized.
the only argument you could make is whether or not they should choose this specific dac chip, thus inevitably resulting in a product with flaw, one way or another.
if you want to argue there is engineering at fault, it's ESS's, not Topping's.
once the dac chip choice is set, the rest is a design choice!! as clarified by topping & amir.
what you don't seem to understand is, there is only 3 filters provided by ess for this dac chip, & the one chosen by topping is the only one that doesn't merely have -10dB'ish attenuation at 22.05kHz, which is not ideal, just like not being flat to 20kHz, too, not ideal.
if there is a filter that is flat to 20kHz & not just -10dB attenuation at 22.05kHz, topping would use that instead as seen in the default filter of many other topping's products not utilizing this specific dac chip, but there isn't.
the funny thing is, e50 (same exact dac chip) got this problem too with no filter getting both the flatness & attenuation right, let alone a sharp accurate one. but because it defaults the one prioritizing flatness, everybody's happy, no one got a problem with it. so maybe the thing that topping should do with G5 is, including a dedicated button & indicator just for users to switch filters, confusing the hell out of a lot of customers in the process, but then topping don't really have to choose & the choice is left to the customers, just like the e50, so everyone's happy.
as for why they chose this filter, i have no idea, maybe g5 being a portable device they find it particularly prone to ultrasonic noise, or other reasons, i must say i completely do not know.
all i know is, it's 2 situation both being not ideal & in this particular case, THERE IS NO RIGHT CHOICE. it comes down to preference. it's practically INAUDIBLE, FOR GOD'S SAKE. i for one can hear up to 20kHz, but everyone who can do that & isn't extremely young knows that, even if you can hear those frequencies in those frequency test, you're still not sensitive to it, you have to ramp it up to barely hear it, & if you can't hear it, then no matter what volume, you just can't. either way why would anybody care about 'music content' in there? not to mention they rarely exist if at all.
i personally also would prefer they use the filter that is flat to 20kHz, in a sense that i will be happier if they do so, but if you can't get the argument straight & give credit where it's due, instead you just talk without any willingness to understand what you're actually talking about, then you really should shut up, not saying you must, since you have the right to talk, to express opinion, but you SHOULD shut up, for the right is all that you have left, as in this case your opinion is so non-constructive that the world is better off without it, therefore you expressing them would be morally wrong, you just have the right to do it, nothing more.
so now, to those in question: would you please kindly shut up?

tl;dr: 1) topping didn't cheat for better measurement by applying the filter in use. 2) there is no problem whatsoever on topping's side of engineering.

for reference:
Wow, what an over-the-top self-important diatribe! Keep it in perspective - there are people here that think flat from 20Hz-20kHz is an important thing to get right (I'm one of them), especially if we're so hard on DAC & amp reviews in terms of focussing on beyond audible SINAD improvements, then such design & engineering rigour should 100% extend to the expectation of a flat frequency response from 20Hz-20kHz.....and even more so given the fact some humans can hear up that high, so it's arguably more important than rating even higher up the inaudible SINAD scale. I think there is zero problem for people pointing out this weakness of this DAC.....I think they should choose the other filter that's flat to 20kHz.
Filter #1 in the E50 (red line):
index.php
 
Last edited:

VariousArtists

Active Member
Joined
Jan 24, 2021
Messages
117
Likes
95
The published fr is tested with 96khz. So the result is a much flatter response than with 44.1khz. The consideration was that people were caring about the cutoff 22.05khz so f3 was chosen but if we want flat 20khz more then yeah we can put out a new firmware for that.

I probably won’t hear a difference between f1 and f3 but now that I understand the rationale for choosing f3 for the G5, I am happy with this choice as a soon-to-be user.

If you do release a new firmware for the f1, please offer it only as an alternative and not one with more features and bug fixes so both f1 and f3 versions get the same level of support.
 

TRDK

Member
Joined
Aug 17, 2022
Messages
86
Likes
41
Location
Hamburg - Germany
I had a quick look on the websites which are offering the G5 (audiophonics, shenzenaudio…). None of them seem to offer the lightning version of the G5 which includes the suitable cable. Does anyone know how to get it?
Have the same problem. Did order the G5 at Shenzenaudio but did forget the lightning cable :) So I am looking :cool:
 
Top Bottom