Hegel H95 Review (Streaming Amplifier)

[amirm]

Of course, sure, but think about it Amir, for exemple, look at this curve:

And look at the delta between 0 dBFS and -24 dBFS.

The SINAD in that DAC at -24 dBFS is around 90 dB. It would be transparent to an 81 dB amplifier. Not knock it down to 72 dB. The DAC in H95 does that because unlike the above graph, it has this kind of noise floor:

Now SINAD is around 70 dB and hence the reason it sucked the performance of the amp down. Which precisely makes my point. The internal DAC is way too noisy. Both DAC tests and integrated pipeline of the dac and amp show this.

 

[amirm]

You know better than draw conclusions based on comparison that can't compare.

I was going to say the same to you but doesn't seem know you know better. See the above DAC graph? The last 11 dB of its dynamic range is useless. So -24 dB is not what it seems. It is actually just -13 dB below its max level -- very reasonable and representative of what could happen in music.

 

[Fleuch]

The SoundEngine2 is the basic building block used in all current Hegel amplifiers, giving a high output damping factor of over 2,000 for the H95 and H120 amplifiers and over 4,000 for the remaining three integrated amplifiers (H190, H390 and H590). Retailers use these figures to suggest that a Hegel amplifier has no problem driving any loudspeaker, and "buy" the company line that the circuit design leads to ultra low distortion. So no difficulty, then, driving a nominal 4-ohm load or less?

The H95 is an "entry level" product into the Hegel range. Looking back over recent reviews, the H95 in the Hegel range is the equivalent of the IO in the Rega range - neither are a credit to the respective manufacturers on the basis of test bench results. In an ideal world the up-market kit from both manufacturers would be available for testing to determine if similar problems are found to be intrinsic to each marque.

Was this a "rogue" sample? or is there an engineering explanation? A digital circuit is really an analogue circuit operating at just two defined signal levels, with timing in those circuits now measured in femto-seconds. When feedback, or feed-forward, is applied around an analogue circuit there must be a time delay between the amplified signal and the feedback (feed-forward) signal arriving at the summing point. If (very) small time delays contribute to distorted performance in a digital circuit, does the same principle apply to analogue circuits? Perhaps this is more critical in the SoundEngine2 circuitry used by Hegel.

Reviews in the consumer media are bland, vanilla bland, with the reviewer seldom calling a spade "a spade", making each review little use as a "critical friend". Only individual specifications are provided and very little comparison (if any) between products. In contrast Amir offers test bench results and a "green, amber, red" conclusion based on those results alone. It is a truly independent source of information (with other members such as Wolf contributing similar work in a different part of the globe). The concept that gear is sent for testing is core to that independence, rather than a reviewer being offered an expenses paid trip to a manufacturer's lair. YouTube reviews should carry a health warning!

 

[PeteL]

I was going to say the same to you but doesn't seem know you know better. See the above DAC graph? The last 11 dB of its dynamic range is useless. So -24 dB is not what it seems. It is actually just -13 dB below its max level -- very reasonable and representative of what could happen in music.

I am still not sure I understand your point. The digital noise floor as 29 dB of gain applied to it, how do you calculate that it should be transparent I am not sure, what we know is that it's higher than the noise floor of the 154 mV analog signal, for the rest I am ready to agree but please explain to me your maths. I think we can say it's not state of the art, and yes higher than the SOTA DAC that I used as an exemple. But how high? Let me know your calculation. The last 11 dB is useless because it's clipped. I's not useless at a lower volume!

Edit: well I guess from the last curve we can extrapolate and approximate a SINAD from this DAC around 90 to 95 dB, so sure definitely not great. But not 70 dB, not close, and this is combination of The DAC SINAD And the "preamp that is not really a preamp" SINAD.

 

[valerianf]

It seems that after Rega it is Hegel that is falling from the "audiophile" pedestal.
Once again it proves that ranking electronic audio equipment is impossible without serious real life measurements.
Thanks Amir.

 

[zeppzeppzepp]

Conclusions
Measuring Hegel H95 as intended, i.e. with analog or digital input producing just 5 watts, produces far more noise with digital input than analog. This is a bad thing in even mass market products. It is really, really bad in the case of a premium product.

I hope Hegel looks at sources of this performance loss with digital input and remedies it. Had it not shown this problem, it would have gotten an acceptable review from me.

It seems like a good business strategy of showing we can't do any better for a small integrated device, please go buying our separates and connecting them with thousand dollar cables to get better sound.

 

[zeppzeppzepp]

It seems that after Rega it is Hegel that is falling from the "audiophile" pedestal.
Once again it proves that ranking electronic audio equipment is impossible without serious real life measurements.
Thanks Amir.

Not so severe for them.
Rega's vinyl players and Hegel's functional restricted amplifiers are still OK with competition.
They just can't keep up to make competitive products beyond their profession.

 

[amirm]

It seems like a good business strategy of showing we can't do any better for a small integrated device, please go buying our separates and connecting them with thousand dollar cables to get better sound.

Only if they perform better. Until then, we don't now. Maybe the company will share their measurements with us and not let us find out later!

 

[ShiZo]

Aside from frequency response, ew...

 

[JiiPee]

The H95 is an integrated amplifier with a DAC and a streamer. I think it is safe to assume that most consumers who buy it will use it at that role i.e. they have it connected to digital and analog sources plus passive speakers. Consequently, I think the main method to measure it should be to feed it with typical digital and analog source signals and measure the output from speaker terminals. If this means that the measurements are not directly comparable with the measurements for pure power amplifiers, then so be it. They belong to different categories.

As the H95 also has something marked as "variable line output", it is good to measure that too, and it is obvious that Hegel should have provided more information about the intended use case and specifications concerning this output to avoid misunderstandings.

Also, I think there are two different ways to look at the results. One is to judge them from the "audio science merits" point of view, where all key metrics should be as good as possible, never mind will they improve the sound in a way that is detectable to humans, or not. From this point of view, the H95 appears to be a rather mediocre product and the touted "SoundEngine2" fails to provide measurable improvements.

The other way of looking at it is from the consumer point of view, where the measured results should be good enough taking into account the restrictions of human hearing capability. I don't claim to know where the line for "good enough" is, but I suspect that the H95 is not quite the abhorrent disaster some forum members paint it to be.

 

[Eetu]

a high output damping factor of over 2,000 for the H95 and H120 amplifiers and over 4,000 for the remaining three integrated amplifiers (H190, H390 and H590).

Soundstage measured an average damping factor of ~400 for the H95 and max ~700 for the H590.

 

[Bruce Morgen]

I suspect that the H95 is not quite the abhorrent disaster some forum members paint it to be.

It's not any such thing -- the thing is only disastrous in terms of price/performance ratio. IMO a consumer spending well into four figures in USD deserves more for their money than near-meaningless audiophile platitudes and rather unremarkable performance that can be had for a good deal less moolah elsewhere, YMMV.

 

[eyes-on-you]

The local distributor had sent my home H590 and KEF Ref 1 for us to listen.

Me and my two friends were amazed at the difference when I connected my Benchmark AHB2 and DAC2 duo after the H590.

The Hegel H590 was squashed like an insect and, despite the AHB2, it played awful. In fact, you wouldn't have bought the H590 for $100 after hearing Benchmark.

That day, my audiophile friends thought that there was a production problem specific to our model in the H590, but it seems that Hegel does not deserve the price.

One of them was considering purchasing the H590 in the future. He had given up on it that night.

Got it cheap.

 

[valerianf]

JiiPee said:


I suspect that the H95 is not quite the abhorrent disaster some forum members paint it to be.

A few months ago I discarded the choice of the H95 because of its price and its form factor.
I finally bought a Yamaha WXA 50.
After reading both reviews from Amir I understand that there is a clear winner technically.
I am very satisfied with the WXA 50 for only $550.
I use it as a power amplifier using the analog input, all digital treatments disabled.
What the H95 would have bring as improvement?

 

[ah-ra]

It's fun each morning to check the comments on this review. This is one of the forums where I actually can learn something (not saying I understand everything though..).
Looking at the test results I would really like to see how some of the actual amps of the major players perform. Thinking about Marantz (especially given the discussion here about their HDAM modules in AVRs), Denon, Yamaha.....

 

[amirm]

The H95 is an integrated amplifier with a DAC and a streamer. I think it is safe to assume that most consumers who buy it will use it at that role i.e. they have it connected to digital and analog sources plus passive speakers. Consequently, I think the main method to measure it should be to feed it with typical digital and analog source signals and measure the output from speaker terminals. If this means that the measurements are not directly comparable with the measurements for pure power amplifiers, then so be it. They belong to different categories.

That is how I measured it. See the amplifier section in the review.

 

[Mnyb]

They have dig themself a hole they can’t come out off imho.

They market thier sound engine and low global feedback approach as the way to good performance that’s their IP and part of what they are selling.

But much if is just audiophile folklore and does not hold up .

I Dow wonder if it this exactly why there is no buffer on the variable output ? Because another audiophile classic “ op amps are bad “ some one at Hegel actually believing to much of this audiophile nonsense.
I would not blame Hegel as a “snake oil company” that’s oversimplification. Sadly I think the audiophile faith systems also creates real believers .

There are many ways to Rome , yet can get decent amp performance in many possible configurations , when you reach low distortion low noise and flat frequency response they are all very very similar until driven to hard.

Wonder if this is the price the pay for the THD wars of the late 70’s and 80’s when everyone fixated on driving an 8 ohm resistor to perfection and nothing else . amps could have tons of other issues .

Hence bad implementation in the past have made some audiophiles shun the methodology completely ?

Just like inti vaxxers brings up bad implementation off vaccine science , but “forgets” how darn good it actually works when done rigth

Can we send Hegel a Christmas gift of Doug Selfs publications about blameless amplifiers

 

[DanielT]

Can we send Hegel a Christmas gift of Doug Selfs publications about blameless amplifiers I

I wish Douglas Self a Merry Christmas so please do not.

Hegel can contact him and ask a lot of basic questions, so "do not wake that bear". Douglas Self probably wants to take it easy during Christmas.He he.

By the way, I think your questions and thoughts are interesting. Hope someone responds to them.There is material for a number of new threads. Feel free to start someone, some. Exciting.

 

[KSTR]

IMHO it's very clear and natural choice -- used throughout the industry -- to measure a DAC vs. its input levels, not its output level like @amirm does. So, you dial in 0dBFS or the more typical -3dBFS to see its behavior at high levels where it usually is worst wrt distortion.

If the DAC has a variable output based on analog stage after the DAC chip (which we have in this case), you set that to a reasonable level you have to inspect that seperately if and how much this affects the measurements. It clear a close to zero volume will degrade signal to noise ratio, whereas too high a level can add additional distortion.

Since normally ASR measurement procedures aim to extract the best performance the device is capable of in full isolation on the bench, a DAC output should measured in this way:
- Set up a typical high send level you want to test at, for example -3dBFS, say, for a 1kHz spectrum,
- Scan through DAC output level settings, if any, to find the region where the performance is stable and constant and then leave it there (during that scan one would have quickly seen that the DAC is not the problem at higher output levels, the level/buffer stage is. The same would have come to light if the transfer line-in --> line-out were inspected).

Amir's referencing to output levels (as of late) has produced misleading data at times. What makes it even more problematic is that Amir seldom states digital input levels when he adjusted the input level to get his "nominal" output levels of 2Vrms for unbalanced and 4Vrms for unbalanced... these choices don't make much sense either because it automatically assumes that a) a balanced output is signal-balanced (which is NOT the definition of a balanced output, impedance balance is) and b) the balanced output is derived from the unbalanced output with an added inverter. You shall not make such assumptions, output level is output level and when the device doubles the output level for balanced then measure at this level, if not then don't (repeat: the reference is input level).

This are just a few points why I'm saying Amir's measurement lack some thoroughness. Not specifying most details of his measurement setups doesn't help either to have confidence. Why not create and attach a professional test report?

A real point that speaks for Amir, though, is that he strives for some sort of consistency so comparisons and ranking (for those who feel the need for such) are somewhat easier to make, but we must never forget that his data doesn't always represent "the truth". I personally wished Amir produced less tests per month but more complete and thorough ones.

I still think his data is very useful, despite all the criticism, just don't jump to conclusions, and fellow posters please don't be silly claqueurs.

 

[sarumbear]

It can filter out the ups and downs that my measurement was showing.

I understand that but why would it matter?