Topping G5 Review (Portable DAC & HP Amp)

[GXAlan]

The engineering is superb. They have simply decided on one dac filter vs another.

I clearly noted the response issue in the review with additional testing. Everything is there for you to decide otherwise.

I know you’re super busy but it would be really educational to show a multitone done at 44.1 vs 96 kHz.

What I am hearing here is that their digital filter decision may sound better (why they ESS or Topping chose it) but it may also give better SINAD at 1 kHz by having this roll off.

I wonder with a 32 tone test or even a SINAD at 10 kHz instead of 1 kHz at 44.1 vs 96 would show the difference. I imagine you can take your current 32 tone and downsample with and without dither.

It’s purely academic and for learning, but it would be interesting as a YouTube video to understand how some of these tests look different if you change the test conditions.

 

[elvisizer]

I’m not interested in the root causes

that's fine, but let's keep the info in the thread accurate- there's no ban on eq from apple on iOS as was speculated earlier.

it’s just a plain fact (unless someone shows otherwise) that on iOS eq can’t be applied to qobuz

i'm kind of being a punk here but it can be done even though they don't support it in their own apps- use roon with Qobuz or tidal and roon can do the DSP

 

[Joe Smith]

I gave it a 'great', although the form factor of this is a little off for a good portable solution for most folks, at least coupled with a phone/tablet. Nice portable solution for a laptop, though. Amazing how things have continued to improve.

Now that I'm only in the office one day a week, and streaming Spotify, I am content with my lower-oomph DAC/Headphone combos (Schiit/Fiio), but if I were buying today, I'd probably get one of these.

 

[levimax]

It's the principle thought, ESS (and Topping for choosing it) otherwise consistent with performance include such a filter is somewhat strange.

The problem I see is that this filter was created and chosen for one reason only which is to "game" SINAD. The product is so good it didn't need to make that choice and it kind of ruins it in my eyes (I'm sure my ears would be more than satisfied). I down graded it 2 levels for an "intentional gaming" violation.

 

[SuicideSquid]

I am willing to bet you won't hear below -96dB SINAD either. Still no reason for them to cheat on FR knowing they will get the prize for totally inaudible dB gained in return. It's a bad sign gaming the system, trading some qualities for the one that ezoteric numbers. Cheating should not make one a champion. For that alone this product I find to be less than 'fine'.

And you should care about it exactly as much.

We have a product that tests almost perfectly across the board except that it rolls off ever so slightly too early in a manner that will be inaudible in all circumstances except maybe to someone with perfect ears listening to certain synthetic tests. If it had perfectly flat response to 20kHz and a jitter spike at -110dB, or if it had perfectly flat response and no jitter and a bit of noise at -96dB, for a $300 portable DAC and headphone amp I'd still say "wow" and I'd expect the device to get a thumbs up from Amir.

I've got no issue with someone saying "despite being inaudible that's a bit disappointing, hopefully they can fix that minor flaw in an update" but throwing a hissy fit like restorer_john over a slight rolloff at the very top end of the audible spectrum is just ridiculous histrionics.

 

[Sokel]

Also where's this 7khz number coming from that people are throwing around? Looks to me like rolloff starts around 18kHz.

Never said has 12db rolloff @7Khz,I said that starts there (VERY gradually until 18Khz) and then comes the cliff as you can see.

Also Amir says "The strong filter makes great showing in THD+N vs frequency:" above the THD+N vs frequency graph.
So can conclude if ESS (and whoever uses it,I don't think it's only Topping) takes benefit by it.

Edit,the graph doesn't even say -12db @20KHZ,it says -14.377db on the Y if I can see it right.

 

[NDRQ]

RIP Chord Mojo.

 

[Berwhale]

Without eq the iPhone is just not a compelling solution for phone users who care about headphone tonality.

Fixed it for you

 

[Phorize]

Fixed it for you

You got me there I’m actually looking at getting an android phone for this reason-at which point the G5 becomes the best thing since sliced bread.

 

[Ken Tajalli]

Looks like it has single ended and balanced out.
4000mAh battery.
Like Mojo, can play and charge (5h).
Volume control is a potentiometer, naturally since it has analogue input.
I wonder , if an artificial ground loop noise can be created, to see how well it behaves.
Personaly , I wouldn't mind the 20K roll off.

 

[Liya]

I would rather buy new Sennheiser Momentum 4 than carry another box + wires

 

[Berwhale]

I would rather buy new Sennheiser Momentum 4 than carry another box + wires

Seen any measurements for these yet?

 

[Ken Tajalli]

If Amir's measurement is right and the device not broken (aren't the reviewers samples cherry-picked?) they should correct the above ASAP or else they won't be much different than the rest mambo-jambo claimers.

Is it supposed to play music or pass lab tests?
If I was in the market, that wouldn't bother me at all.
I am not a bat.

 

[Liya]

Seen any measurements for these yet?

Yeah.

7 x 7.67 x 1.82 inches

 

[Sokel]

Is it supposed to play music or pass lab tests?
If I was in the market, that wouldn't bother me at all.
I am not a bat.

I'm way more tolerant than it seems,doesn't bother me to listen to music even from a broken radio if I like the performance and the play.

I admit I'm maybe infected by the measurements virus a little and a little ASR biased towards the published specs vs Amir's findings but nothing I can't suppress when it comes to buying stuff,durability and good looks comes first to me then,everything else is a bonus.

 

[staticV3]

Seen any measurements for these yet?

 

[Ken Tajalli]

RIP Chord Mojo.

Aah , that is so sweet.
That you consider Chords seven year old, discontinued device as a reference to beat.
Chord, is humbled by your sentiment.

 

[Tks]

Because achieving this is a measure of engineering excellence, not really a matter of relative audibility. In terms of converters, we are limited at the reconstruction filter order, slope characteristics, and ripple level. As you increase the order and steepness of the reconstruction filter, you start losing bandwidth to achieve the desired stopband (22,050 Hz for 44,100 Fs/s). This is an over-simplification of the process going on, but this is the trend that appears when designing a filter. Digital filters help a lot, but still, we can't get an infinite slope transition band (the theoretical brickwall filter), thus we are limited in that aspect.

But just to be perfectly clear, there is zero audibility concern on this front pragmatically speaking? Am I correct on that front? Like there's not going to be a case if I fired up HQ Player, and stacked a few filters, or simply increased the slope factor to where my filter will be forced to start only be able to stop band at like 10kHz and no higher. Correct?

I know we can't get true theoretical brickwall practically speaking of course. You did mention as you increase steepness you start losing bandwidth. Sorry, but where can this be observed to be a practical issue? Software filters (and even Chord products) can throw enough processing power to where this becomes a non-issue in the practical sense, and since processing power for such filters is miniscule with today's processing hardware, I'm not seeing where this "starting to lose bandwidth" can ever be a factor outside of academic benchmarking or something. Like, to me personally, a Chord DAC's brickwall (which can also be achieved with software) is as performant as I'd ever practically ever care brickwalls be if I had to point out a hard-limit to where my concern goes out the window.

And I mean this also from an engineering excellence perspective as well. I don't take a DAC that achieves -110dB of attenuation at 22.1 to be compelling enough compared to a DAC that can only do -105dB of attenuation at that frequency if a purchase decision is being made. I will of course admit it is more performant. But there will never be a case in which those two differences could mean more than virtually any other aspect of the devices being compared. And this is coming from someone who's constantly roasting devices somewhat regularly if the same company makes regressions with filter offerings or things like that btw.

 

[amirm]

What I am hearing here is that their digital filter decision may sound better (why they ESS or Topping chose it) but it may also give better SINAD at 1 kHz by having this roll off.

There is no such impact on SINAD. That is a tiny bit of attenuation. Here is the SINAD at 48 kHz sampling which has wider bandwidth:

And filter response:

Performance is the same. So the filter selection has no impact on SINAD.

 

[amirm]

Topping got back to me and said the filter choice is by design. They are willing to change it if we/I ask for it. Do we want to do it?

The trade off here is getting rid of out of band noise with current filter vs flatter frequency response but potentially more bleeding due to ultrasonic leakage.