Hegel H95 Review (Streaming Amplifier)

[garbulky]

Oh man...another regarded, good subjective reviews audiophile brand with a feet of clay product...The whole thing is rather bad in so many metrics there is just no saving, even if I like the aesthetics...sad.

I listened to it subjectively and was very underwhelmed. The thing sounded completely uninvolving.

 

[amirm]

I'd like to ask if it has been mentioned anywhere why amirm measures the SINAD (at 5W output) of integrated amplifiers at low input signal levels. For both analog and digital input. In this case ~0.157V for analog input and -24dBFS for digital input.
I mean I use sources with an output of around 2V for an integrated amplifier.

The reasoning is a solution to a tough problem. Namely, how to compare integrated amplifiers with volume control to power amplifiers without. Anything else won't be fair. The gain on power amplifiers is not variable of course. But it is generally around the 29 dB that I test for. Where did that number come from? It is the THX certification criteria for amplifiers. So what I do is adjust the volume on an integrated amplifier until it hits 29 dB. I then adjust the input until I get 5 watts. That results in the type of input values you are seeing.

Yes, in a modern world I like to see us move to lower gain amplification and take advantage of the higher output of modern DACs. But the industry with some rare exceptions, is not going there.

Fortunately the difference is not large in many tests I have performed. Most amplifiers are distortion limited, not noise so gain won't matter. When it does change, it is only a few dB.

 

[ahofer]

@Hegel Music System clearly thinks the amp sounds better this way, I’d genuinely be interested in the listening tests they used to reach this conclusion - who listened, how, and how was it controlled? I’m always open to the possibility of a genuine preference difference overwhelming differences that, while measurable by @amirm , may not be audible.

 

[Rottmannash]

We have some data for vol = 98.

What if the It would reach SINAD of ~90 at vol = 50?
Can you show from this data it would not?

So a new owner can only listen to his/her new $3000 integrated amp at half volume? I'll pass.

 

[Lambda]

So a new owner can only listen to his/her new $3000 integrated amp at half volume? I'll pass.

What is halve volume? is why would you care abut the number in the display?

i normally adjust my volume so by ear not number on the display.
but the manufacture explained customers don’t like turning it all the way up... so they increased the gain that customers don't need to turn it all the way up to get the same power
Because louder sells better...
and if the amp is louder at "lower volumes" customers think its more powerful and has more headroom.

 

[PeteL]

Then you would have measured the DAC itself and got quite decent numbers. That is assuming the volume control stage isn't active and adds noise. On top of that the 0.5V is 12dB below 2V so there will be a 12dB lower S/N ratio opposite the input level that is normally seen (2V for S.E.).
The analog inputs can also handle 2V input voltage. This is no problem either because the volume control (attenuator only and the reason for the distortion ?) can attenuate.

As it is measured the last 11dB headroom is distorted (a bit). Yes it is at -60dB or so but higher harmonics. To get rid of it in the way it is measured by Amir one would have to lower the digital input by 11dB.

I understand, but I assume that the majority of users will feed the digital in at Full scale, and adjust the volume to taste, and "to taste" will be lower than volume at 98. We do not know that there will be a 12 dB lower S/N ratio. Maybe, but attenuating, an analog signal at this point, also lower the noise level at this amp stage, unless as you said it creates noise of it's own but we don't know. Yes it's a design where the power amp section is quite a bit more than usual, but as far as the actual preformance in term of Sinad, the measurments are inconclusive since it's been measured beyound it's linear behaviour zone. Again, not saying it's a good design The Sinad we can see at -13 dBFS or so is not a best case sinad. The Best case Sinad would be to feed 0dBFS and reduce the volume to half a volt. For what it<s worth to know that.

 

[Rottmannash]

How many hifi amps do you know of have that have an accurately labeled volume control scale (accurate to the reference level published in the specs or labeled next near the input)?

And how many amps do you know of that allow you to plug-in a CD player or streamer and turn the volume control up to 100% without causing clipping? You know they sell RCA attenuation adaptors to solve this issue a lot of people experience?

Just saying, gain structure in HiFi is messy and you can be assured it's not only Hegel's problem. It maybe is an additional thing to test for in the reviews, so the whole industry can improve onto it.

My Onkyo RZ810. You do mean "ref level" right?

 

[Rottmannash]

What is halve volume? is why would you care abut the number in the display?

i normally adjust my volume so by ear not number on the display.
but the manufacture explained customers don’t like turning it all the way up... so they increased the gain that customers don't need to turn it all the way up to get the same power
Because louder sells better...
and if the amp is louder at "lower volumes" customers think its more powerful and has more headroom.

I think you may have missed my use of sarcasm.

 

[nagster]

@Hegel Music System
I pay tribute to your reply.


Everyone, barrage is not good.
amirm just made the usual measurements and added trial and error.
Their explanation was useful, but certainly their excuses are non-scientific and the measurements are non-HiFi. Some consider it a defect.
However, there is hope just for appearing.

Change to the current standard design instead of 40 to 50 years ago for compatibility. And if they don't lie, the problem goes away.
Only those who believe in Non-NFB will buy it.

If they are honest and reborn as manufacturers that sell high-performance, easy-to-use products, we have more choices. It is even better if the price is reasonable.
Good luck Hegel.

 

[Everett T]

Dont agree.
Offering 96 kHz is maybe good enough if the jitter is low but not up to date 2021.

The TE7022 was probably chosen only because of low cost. This tested amplifier is expensive. The Chinese brands Topping and S.M.S.L offers xu208 and xu216 chips with super low jitter, with sometimes good volume implementation in the digital domain thats superior to TE7022 and does 192 kHz or 384 kHz.

Why the need when cost to improve is so low? If you're charging SOTA prices, at least specs could reflect the attempt. I'm not saying it's an end all, but it's a poor reflection on the manufacturer.

We know that other products of theirs measure well with amplifier tests, so when you add the low performance of this side of a product it drags them down.

Just my $.02

 

[Geert]

My Onkyo RZ810. You do mean "ref level" right?

Do you have measurements that support that? Just out of interest. It's possible because it's a THX certified amp, and THX has strict requirements when it comes to levels and headroom.

 

[PeteL]

My Onkyo RZ810. You do mean "ref level" right?

This Receiver has an input sensitivity of 200 mV on the line inputs. If you would feed it a signal at two volts, it has to be signifcantly attenuated to not clip the amp. 200 mV at full volume will reach the amp clipping point. On this amp I understand that we are griping about a digital input, but for that matter, once converted it's the same thing in the end, it have to be attenuated like most, almost all 2V input have to in any integrated amp, some more than others.

 

[Everett T]

Do you have measurements that support that? Just out of interest. It's possible because it's a THX certified amp, and THX has strict requirements when it comes to levels and headroom.

It would be nice if THX released the requirements, but what you mentioned would follow what seems to be out there, gain structure, crossover slopes, and output voltage . Their speaker requirements have always seemed more demanding to fulfill.

 

[djkrimsom]

It is so easy to believe in the hype for Hegel. Thanks for an objective review.

 

[restorer-john]

I was being silly, but yeah I remember real Player. Never cared for it. I was all about some Winamp back then.

I remember the Real Player- it was a nasty piece of malware, almost a virus. Embedded itself into my installation in a bad way. No proper uninstaller IIRC.

 

[AudioSceptic]

There was digital audio in the 60s?

Distortion and S/N measurements are not a new thing, y'know. ;-)

 

[phoenixdogfan]

Oh, don't worry, I've been here a long time.

I'm pretty comfortable being accused of stupid spending.

@Sal1950 has been scoffing at my vinyl for half a decade.

Although I draw the line at cable voodoo and power conditioners.

Well, I do understand your reason. You feel the performance difference between the Devialet and anything else is insignificant or inaudible AND you really like the look and are willing to spend whatever to get it, even though it's almost certainly not a Veblen good to most everyone else.

Me, I'd rather buy a generic looking component with top notch numbers (whether audible or not) especially if it's not materially more expensive than adequately performing, commodity priced components, and place everything out of sight in something like a BDI cabinet. To each his own.

Enjoy your purchase, sir.

 

[MaxBuck]

Maybe Devialet.

Devialets don't measure that well. And the Expert 140 costs $6,500 US. So, pretty much what I said.

 

[MaxBuck]

Why do you think Lyngdorf TDAI-3400 is a good integrated amp? It appears on Amir's SINAD chart here and at 68dB SINAD it seems to have a much poorer standing than Hegel H95, or am I missing something?
View attachment 168176

Well, I guess I was wrong about the Lyngdorf.

 

[watchnerd]

Devialets don't measure that well. And the Expert 140 costs $6,500 US. So, pretty much what I said.

That's just the cost of the baby one.

I got mine used so didn't pay list, but the dual mono version I have lists for about triple that.