Saturday 18 March 2023

REVIEW/MEASUREMENTS: AOSHIDA A7 Class D TI TPA3255 amplifier. Good sound. Good power. Economical.

Above, you see the small AOSHIDA A7 Class D amplifier which we'll go in depth with in this post. It's inexpensive and can be purchased for less than US$150 these days. This design is also sold as the Douk Audio H7 with a different color scheme, presumably the innards must be very similar if not identical. I preferred the look of the black AOSHIDA version with orange-ringed VU meter. I purchased this amplifier through the usual retail channels.

Over the years, it has been good to see the evolution of Class D designs and the inexpensive "chip" amplifiers. Internally, this is based on the TI TPA3255 "PurePath" chip (released in 2016) rated to provide up to 315Wpc into 4Ω stereo but at 10% THD when provided with adequate power (data sheet max is 53.5V). Notice the italics to emphasize that these numbers must be put into context. For audiophiles who want lower distortion in their amps (like <1% THD), we would be looking at much more conservative power figures than that 300+W being promoted for these amps.

(The chip also has a mono configuration that's spec'ed at up to 600W into 2Ω which would not apply for this review.)

From the front, other than the prominent centered VU meter, we can see independent volume controls for each stereo channel. The knobs feel quite good and it's finely articulated (feel the "clicks" as you turn) rather than continuous. While we can visually approximate turning the knobs to the same level, without a direct Vrms reading, there's a little bit of guesswork and listening involved for the end user to achieve exact output level matching.

With two separate volume controls, logically, we can easily adjust R/L balance. However, with no remote control or unified volume setting, I would recommend using this in combination with a preamp or using the volume control from your playback device such as the digital volume control of your DAC, leaving the AOSHIDA's knobs alone once you've set them to an appropriate level. Exactly how much is enough will depend on the output level of your XLR/RCA device as well as sensitivity of your speakers. As a point of reference, in my system downstairs with the Paradigm Signature S8 v.3 main front speakers, about 3 clicks below 75% on the volume knob is loud enough for all my listening. (As usual, I encourage you to measure your own power needs.)

The smaller knobs are VU-Sensitivity setting to the left of the meter (turn clockwise and the VU meter responds more to lower levels), and to the right is the combined OFF / XLR / RCA input switch.


The amp came in a generic-looking white cardboard box with adequate foam padding. In the box is the amplifier itself which has a rounded side-tapered shape. The amp is approximately 17.0cm wide x 12.5cm deep x 4.2cm tall which makes it quite compact. There's also the stock switching power included. Here's a look:


It came with a 32V, 5.0A DC adaptor which can potentially provide up to 160VA total if it lives up to the specs. Here's the back of the unit:

Note the plug for the power supply indicating that it can accept up to 48V. Upgrading to a 48V power supply can increase output wattage significantly.

Here's a chart from Aiyima for their A07 model which uses the same TI TPA3255 chip amplifier showing power supply specs and potential power out which approximately applies to this A7 also:

The stock 32V/5A unit should get us around 80Wpc into 4Ω and 60Wpc into 8Ω. If you look around, there are inexpensive 36V/10A power supplies that could improve this to >150Wpc, and going all the way to 48V/10A will potentially achieve >200Wpc if you really need that kind of wattage. The trick is to find high quality supplies that actually perform to the rated specifications. For today's testing, I'll just use the stock power supply and in the future, might try out some upgrades if I find high quality, good value recommendations.

The speaker output binding posts are reasonably sturdy and I had no issues with any of my banana plugs. Both XLR balanced and RCA single-ended inputs available. There are handy line-level RCA outputs as well if you want to pass along the signal to your subwoofer(s) for a 2.2 configuration if you have two subs.

The enclosure is made of an aluminum alloy shell and feels sturdy, weighing in at just over 800g (excluding the power supply).

I was hoping to open the box up to have a look. Unfortunately, those XLR tabs made it tough to open up without forcing anything so I elected to not risk any damage. As a "chip amp" with an external power supply, I expect that the circuit board would be similar to other TPA3255 devices out there. Thankfully, someone else has done a teardown of a similar device (Suca Audio U7) - check it out here.

I noticed that AOSHIDA advertises the use of Sagami-brand inductors in the output path. These days with stuff like cheap, fake parts, I wonder if different brands might use different components even if the circuit is the same. Charging a reasonable bit more on a trusted brand ensuring quality components could be a way to differentiate otherwise very similar products.

Objective:

For my testing procedure, unless specified, I'll use a combination of the Topping D10 Balanced to feed the amplifier for these measurements (the Topping D10s with RCA output was also tested later). I used the Intona USB isolator to keep the signal as clean as possible.

I'll use the Linear Audio Autoranger for the measurements and my RME ADI-2 Pro FS; no need for the extra precision of the E1DA Cosmos products for measuring amps. General testing system described previously.

A peek at the AOSHIDA A7 on the test bench warming up with the Topping D10 Balanced as source. 4Ω power resistor, RME ADI-2 Pro FS ADC and Linear Audio Autoranger behind right.


Amplifier Basics: Oscilloscope, Damping Factor, Frequency Response

To start, I checked the volume controls to get a sense of the amount of gain available. Using a low voltage XLR signal, with the volume turned up to 100%, I'm seeing +34dB gain (equivalent both channels). This should be more than enough sensitivity even with low-output analogue signal sources.

Let's have a look at what the oscilloscope shows for this amplifier when sent some 1kHz signals.

With the output levels for both channels equally set, we see that channel balance can be excellent. Notice that this is indeed a switching Class D amplifier and ultrasonic switching noise can be seen in the tracings.

Let's have a peek using a wide bandwidth oscilloscope FFT at the ultrasonic frequencies up to 600kHz:

Using 5kHz & 93kHz tones as markers, we can see that there are a few prominent ultrasonic frequencies starting at 300kHz with a peak at ~460kHz. While such high frequencies are inaudible and would be attenuated by inductance in the audio system (ie. your speakers and cables) I think it's worth being aware of them. Over the years, I have not noticed any significant interference from switching devices affecting other hi-fi gear. [There was once when I noticed the switching power supply of a device causing issue with DAC/ADC performance but that was 8 years ago.]

A useful characteristic to be mindful of with amplifiers is the damping factor which gives us an idea of how much "control" the amp can provide over the transducer. Here's the A7's damping factor graph measured across audible frequencies using a 4Ω load:

Notice the drop in damping factor above 2kHz. It's still pretty good at >30x at 5kHz but dips below 10 above 10kHz and is a weak 3.5x by 20kHz. Average damping factor across the 7 frequencies is 49x. This same data can be recalculated as the output impedance:

This will have repercussions when it comes to handling speakers with wide impedance variations in the treble frequencies particular above 5kHz. While for a solid-state amp this damping ratio is alright but not great, the increase in high-frequency output impedance is rather "tube-like" I suppose and 3.5x (almost 1.2Ω) at 20kHz is still higher than some "audiophile" amps like say the Pass Labs SIT-2 utilizing minimal/zero feedback.

The increasing output impedance of the TPA3255 amp at high frequencies was discussed awhile back on the QuantAsylum page using the TI TPA3255 Evaluation Module.

Let's have a look at the frequency response using 4Ω and 8Ω resistive loads, as well as the little Sony SS-H1600 bookshelf speaker I've been using in my other amplifier reviews over the years.


Interesting. It looks like the output was optimized for flat response into a 4Ω resistive load (only -0.25dB at 20kHz). High frequencies lift modestly into 8Ω and we can see the effect of the reduction in damping factor with the Sony speaker in the higher frequencies as well. While there are many other variables to consider, compared to other amps, we might notice a slightly brighter sound signature with this amp. Perhaps optimizing the frequency response into 6Ω would be a nice compromise?

An ideal Class D comparison would be the Hypex nCore NC252 where we see consistently high damping factor >200 across the audible spectrum with tight control over the Sony speaker in the frequency response. The ICEpower is pretty good as well although it also has a damping factor dip into the high frequencies but still >20 with the 50ASX2 Class D module.

For completeness, here's the frequency response and phase into the 4Ω load from 10-48kHz:

-0.9dB dip at 20Hz; bass quantity is still strong. Phase is nice and flat - no filter or tone control in the path. Notice that it's up at 180° which means it's inverted polarity. Feel free to flip speaker cable polarity if this bothers you. ;-)

As a stereo amp, let's have a quick peek at the crosstalk. Here's a signal with 0dBFS 4kHz in the left channel, and 0dBFS 300Hz in the right channel and look for seepage into the other channel:

With almost 100dB channel separation at 2Vrms output, that's really good. In fact, that's a better result than I got with the ICEpower and even the Hypex nCore NC252MP. Of course, there's only so much channel separation needed - you'll be fortunate to get 30dB at 1kHz out of a phono cartridge.

Distortion & Power:

Using a standard 1kHz Sine Sweep into 4Ω and 8Ω loads, let's see the THD(+N) vs. Voltage graphs to see the limits of this little amp using the stock power supply:


I often think in dB and V for these kinds of graphs (like I do with DAC measurements) rather than % and W, hence the use of those units although I have put some equivalents in red on the axes as you can see. Even though the human ear/mind is not very sensitive to harmonic distortion, as an audiophile who values high fidelity, I typically put my threshold at 0.1%/-60dB THD(+N) for reporting power measurements.

This little amplifier with stock power supply performed surprisingly well! Even though it's just a 32V/5A spec'ed switching supply, with both channels driven into , it was able to achieve up to 83Wpc at 0.1% THD+N. Into , it was able to provide almost 50Wpc at 0.1% THD+N. This suggests that the power supply is doing a good job providing the necessary current. I trust that this should be enough power for the vast majority of home systems. Pushing the output to >21.8V (clipped) into 4Ω both channels driven resulted in the A7 going into protection shutdown.

Measurements above were done with the left channel. Let's overlay both the 4/8Ω and left/right THD+N:

In the graph above, left and right are denoted by the colors blue and red, 4Ω solid lines, 8Ω dotted lines. As usual, I'll put the -60dB/0.1% cursor up. Notice that between 3V to about 16V (starting to clipping), the right channel has a bit more distortion than the left. Total distortion + noise still very low at 0.02%/-74dB or better so I wouldn't worry about this being audible. I don't know if this right/left difference is consistent among other A7 amps or is just idiosyncratic of my unit here.

Extending the data down to -120dB from clipping at 0dB (~22Vrms into 4Ω), we can examine the amplifier's output level linearity:

Note the clipping at ~22Vrms on the right. Linearity is maintained with <0.5dB error down to a minuscule 83μV signal which is around -108dB from the clipped peak level. Excellent.

Here's the THD vs. Frequency sine sweep across a few power levels to check for any frequency-related issues:

Looks pretty good. THD tends toward lowest amount between 100-200Hz. THD then increases into the higher frequencies on account of rising odd-order harmonics (3rd and 5th) as output level increases, otherwise benign 2nd order harmonics predominate. Nothing scary here looking at 250mW to 23W into the 4Ω load.

Alright then, here are some 1kHz THD+N FFT into 4Ω to see how the distortions look with the 1kHz signal at various output levels:

Low order 2nd and 3rd harmonics predominate. I am reminded of an article about amplifiers and harmonics I recently read on JWT Audio; check out the section on "Amplification and Distortion" as an interesting discussion about relative harmonic amounts and sound quality.

In the same way we can have a quick peek at the 1kHz THD+N FFTs into 8Ω:

With higher output level, we can see the improved signal-to-noise, but also higher harmonics.

Multi-Tone Distortion:

As usual, we can now look at some multi-tone intermodulation distortion results. Realistically, most of the time we're not listening at massive wattage levels. 2V into 4Ω (1W) is usually quite loud already for most audiophiles with small to medium rooms and reasonably sensitive speakers.

2V IMD results are in the -80dB region - good. How about the old Transient Inter-Modulation distortion (TIM) test?

Nothing to worry about here either. Very clean reproduction of transients with a 96kHz bandwidth signal (192kHz test file samplerate) at 2Vrms out. Intermodulation still below -100dB in the higher power 6.3Vrms (~10W) condition.

Finally, let's have a look at the Triple Tone TD+N, 2V into 4Ω using a higher resolution 256k-point FFT:


While up to now I've been using the XLR input to the amplifier, notice that I've also grabbed a couple of measurements from the RCA inputs (using Topping D10s DAC). As you can see, there's basically no difference between the XLR and RCA inputs; at most maybe a slight 1dB TD+N benefit to XLR. We can see a tiny 60Hz hum that is no different between the XLR and RCA inputs; maybe this would be more important with longer cables than 6'.

A -80dB Triple-Tone TD+N at 1W into 4Ω is again really quite good.

Subjective:

In the picture above, we see the AOSHIDA A7 amplifier connected to my sound system being fed audio from the Sabaj A20d 2022 DAC running either through my Yamaha AVR (not shown, using RCA pre-outs) or directly through XLR with its own volume control. Music played through MeLE Quieter2Q fanless PC as Roon endpoint (running Linux with or without HQPlayer upsampling to PCM352.8/384 and DSD128 to taste as discussed here). Notice the relatively diminutive size of the AOSHIDA compared to the DIY Hypex nCore NC252MP assembly that it's sitting on top of.

I listened to the amplifier for about a week before putting it on the objective testbench allowing me to form some initial subjective impressions and I suppose provides a little bit of "break-in" (not that I have any evidence for this).

These days, I am enjoying more multichannel music in my system so one of the first things I did was to integrate this amp to drive the front channels using the pre-outs from the Yamaha RX-V781 receiver. As far as I could detect, the tonality of these amps were flat with my Paradigm speakers and sounded good with the Yamaha driving the other channels once I dialed in relative output levels and speaker distance in the multichannel set-up menu.

The AOSHIDA A7 performed very well playing multichannel material. The spaciousness of Lisa Jacobs & The String Soloists' Haydn: Violin Concertos (2017, DR14, DSD128 5.0 - converted to 5.0 16/48) was easily heard captured in the Cunerakerk church in Rhenen, Netherlands. As audiophiles, we might bemoan multichannel effect used excessively as if the listener is unnaturally situated in the middle of the band or use of rear channels with instruments placed for dramatic effect. The use of multichannels to "expand" the sound of the room whether subtly like with this recording or in a live album with the sound of attendees certainly adds to the sense of "reality" that the recording can convey.

While multichannel sound is a combination of all the speakers playing, the AOSHIDA A7 did a fine job performing its duty with the front speakers rendering the lead violin with excellent clarity and subtlety, likewise I love the finesse on the reproduction of the harpsichord as it adds highlights, almost floating in space mid-right-rear in the sound stage through "Violin Concerto in C major". Emotionally, I'm entranced by the melodic beauty of the "Adagio moderato" movement in the "Concerto in A major".

This amp had no problem with handling those "microdynamics" and quick transients. Noise level is low with no hiss or hum to distract from the music.

Eva Cassady's Live at Blues Alley has been released as a 25th Anniversary (2021, DR8, note color front cover), with the same 13 tracks from the original 1997 album. Notice the drop in DR from the original recording's DR11 down to DR8 with this remaster. As expected, this will have repercussions resulting in a noticeably "louder" sound.

While the vocal quality is nice and the live nature of the album provides a different "feel" to the music from unblemished studio-produced content, the fact that this is a live performance makes me prefer the original 1997 release. What they did with the reduced dynamic range was accentuated Cassidy's vocals and diminished the incidental sounds of the audience and the room ambiance. In fact, they even took out the intro at the start of "Cheek to Cheek" with the hand claps and cheers!

While I guess some might prefer the sound of the new, more-compressed, remaster when listening in a car or walking around with wireless headphones, for a high-fidelity system, we've unfortunately lost much of the energy of the live venue. In effect, it took a step towards sounding more like the studio esthetic which is a shame; this is supposed to be Live at Blues Alley.

I assume the performance wasn't captured with more microphones to be able to produce a true multichannel version. Now that would have been a worthy anniversary edition and upgrade to the live sound!

The A7 had no difficulty highlighting the differences in dynamic range or change in the feel of the performance when flipping between the versions of this recording. I found the midrange on the A7 to be very satisfying.

Next let's talk about the album above. It has been a long time since I've listened to new country music and I see that this Morgan Wallen (controversy in 2021 - notice what happened to music sales) album is currently #1 on the charts - One Thing At A Time (2023, DR6). 

Well, this ain't the sound of Garth Brooks ;-). And it looks like Auto-Tune is alive and well in albums like this, with the strong vocal twang of course. A song like "Last Night" could have easily been a Maroon 5 track and "One Thing at a Time" could have naturally been included in one of Taylor Swift's older country-oriented albums (remember when her songs were like this?).

I find the digital manipulation and obvious Auto-Tune in the music a bit odd. Even with slower songs like "Devil Don't Know", there's a synthetic glare that's a bit jarring and makes the song less down-to-earth when talking about sadness, heartbreak with spiritual overtones that are supposed to be "human". I heard that the producer Joey Moi likes this kind of coloration in his productions. If you're not typically into current popular country, have a listen to "Thought You Should Know" which currently is riding high on the country charts along with "Last Night" for a taste.

I must say that I'm blown away that this is a double-album with a total of 36 new songs of which I've so far only really concentrated on the first 18 on CD1 and CD2 ended up being background music one evening. Despite the sound quality of this album, there's certainly a sweet accessibility to the lyrics and joy to be had in country music (as well as a lot of songs about alcohol...). Whether you embrace lyrics like "But what do you expect from a redneck? / Hell, I was born with a beer in my hand" or not, remember that country music encompasses around 13% of album sales in the U.S. That demands respect and overshadows genres like the 1% classical music sales these days.

The AOSHIDA A7 did not have any trouble reproducing this kind of production.

Let's talk about an album that doesn't sound great fidelity-wise but I love the music. Icehouse is an Australian band that I got into back in the '80s in my high school years. I enjoy the remix album Meltdown (2002, DR7). The "Electric Blue" remix is a bit bright and harsh already, I felt this was accentuated through this little Class D amp, sounding even thinner and at times unpleasant. A little EQ to roll off the top from 10kHz helped significantly here. The remix of "Hey Little Girl" is enjoyable and it's great to hear the little AOSHIDA pump some strong bass to the speakers without the sub.

I'm not one of those audiophiles who think the hardware should have any special powers to make all recordings sound great; simply not possible regardless of how much money you put into an amp or what some reviewers hype. Whether it's a DAC or amplifier or speakers, achieving high-fidelity, transparent sound IMO is the ultimate, simple goal even if collecting the parts of the system to achieve this might require some researching. Rendering the album faithfully will show limitations, "warts and all", as they say, of the production quality. Romantic notions of a little bit of tweaking this or adding that but not substantially changing the device or room (as beautifully exemplified in this thread) is great for stimulating the placebo effect, nothing more. This goes for silly miniscule "tuning" devices like these that are powered by faith rather than physics.

I've mentioned this album over the years. The other day I came across Boris Blank's Electrified (2014, DR10) album again in Roon's Discover panel and listened to the track "Inner Mountain"; very cool demo-quality track. Lots of bass, extremely wide soundstage that envelops the listener, multilayered depth, great test of clarity for the amplifier and your speakers. Just one of many excellent, accessible pieces of electronica produced with good dynamic range, including track 1 "The Time Tunnel" I've brought to audio shows and dealerships for a listen. "A Touch of Magic" is the instrumental version of the song "Celestial Echo" with Malia on Convergence and great as well.

I really like the subjectively clean "speed" I hear with my Paradigm's beryllium tweeters handling this kind of music and it sounded very good with this amp!

Okay, enough of my own subjective musings... This amp sounds very good to me delivering more than enough amplitude using the stock power supply in my room. Good bass reproduction, the treble is well extended and soundstage excellent, able to convey depth and width to 2-channel music.

Notice in this picture I turned "Vol-R" down to see if this would give me lower distortion or more power through the left channel only. Alas, no significant difference.

Summary:

In a world where prices seem to be soaring all over the place - including the "High End" audio world of course - I think it's important for audiophiles to just take a deep breath and think about the technology goals we're actually after. I'm sure there will be some hobbyists and audio reviewers who would take a glance at this little AOSHIDA A7 TI TPA3255 amplifier, then the low MSRP, and just dismiss it as being unworthy for anything beyond garage duties.

Based on what I heard and what I measured, this would not be fair if we're truly honest about sound quality being of paramount importance. For example, I can easily say that I prefer the sound from this amp than the Pass ACA or Pass SIT-2 even though I can understand that some might like those and the heritage of the design/designer. Heck, I'll even dare to say that this little amp is superior to the Vitus Audio SS-010 in sound quality with more power and lower noise than that pretentious 25W into 8Ω Class A "$500/watt" monstrosity (which my friend already sold off for a good price). Likewise, I can appreciate the quality some audiophiles point to as why they prefer tube amps, but honestly, it's highly unlikely I'd ever feel the desire to use those in my system other than for temporary testing for a number of technical and practical reasons.

Here's my summary AMOAR Score graphic which I use with my amplifier reviews, capturing the key characteristics I look for:

Audiophiles, take note of the units on the Y-axis - the range is 4dB only. Below 20kHz, the 8Ω and Sony speaker deviations are less than +1dB, so don't get freaked out. Frequency response variation is typically subtle...

What we're seeing is a competent amplifier with good average damping factor through most of the audio spectrum, low distortion resulting in a great -80dB Triple-Tone TD+N score (2V into 4Ω, similar to the Emotiva XPA-1L Class AB monoblock), and capable of around 80Wpc into 4Ω ≤0.1% THD+N (and clean 48Wpc 8Ω ≤0.1%) with the stock power supply which I was pleased by. Plus one has the option to upgrade the external power supply to a 48V/10A unit for potentially >250Wpc into 4Ω if you really need that! [I know some audiophiles spend all kinds of money upgrading power supplies of DACs and stuff like that which likely makes little difference. This is one situation where an upgrade to the power supply actually can make a significant change.]

However, although the average damping factor is good, it dips very dramatically into 20kHz which is why we see that loss of "control" of the frequency response into the Sony bookshelf speaker, plus it has a tendency to accentuate high frequencies above a 4Ω load. Be mindful of this if you're using speakers that have good high-frequency extension and may sound excessively "bright". On the other hand, if you have "dull" speakers, you might find that this amp adds some sparkle in a pleasant way - I think the fancy word is "synergy". :-) Feel free to use DSP/EQ to taste.

As expected, a Class D amp like this is highly efficient. Turned on, at rest, it draws 6W from the mains outlet and I've only seen it go up transiently to 10W with highly dynamic movie soundtracks at loud "reference" volume levels. While we can easily measure amplifiers producing hundreds of watts per channel, let's be reasonable and appreciate that for most music listening using speakers of normal sensitivity (like 85-90dB/2.8V/m anechoic), in small-to-medium rooms, we're typically only using a handful of watts if that.

The general look and feel of the unit I think is good given the MSRP. Sometimes it's enjoyable to just sit and watch the VU meter dance with the signal strength. Other than using it to know that a signal is being offered to the amplifier, it's mainly cosmetic but audiophiles love their meters (as per McIntosh and D'Agostino). One annoyance at times is that the brightness can be a bit too much especially when I put it in my home theater - distracting when watching a movie with lights off. I would have appreciated a dimming option.

As I mentioned in the beginning of this article, while the independent front channel volume controls are a direct way to adjust channel balance, I think this is too cumbersome for use as an "integrated amp" (much better to have a single volume knob, and balance control). I just think of the volume knobs as balance + gain/sensitivity adjustment up to +34dB for the analogue input you intend to use. From there adjust the VU meter sensitivity to taste and use either a preamp, your DAC's volume setting, or computer audio mixer control to change volume in day-to-day use.

The job of technology is to solve problems first and foremost. For DACs, flat frequency response, very low noise, even jitter IMO have been well resolved. In the amplifier world, modern Class D devices have also solved problems like size, cost, power efficiency to the point where we have devices like this one. I think as audiophiles, the appropriate response to the growing market of inexpensive Class D amplifiers is one of joy given the sound quality we can now so easily obtain. Value has taken a very significantly leap forward in the last few years with these chip amplifiers such that we can focus on other considerations like esthetics (VU meter!) and features (Bluetooth, USB in, DSP, etc...). Or just free to enjoy music and not worry about the diminishing returns on fidelity going up the price ladder!

This TPA3255 amplifier has handily surpassed the Yeeco TPA3116-based amp in sound quality and clean power from a few years back without exacerbating bank runs. I suspect not long from now we'll be seeing more GaN FET-based Class D amps at very reasonable prices as well (and switching rates probably up around 750kHz - 1MHz).

Other than the usual availability on Amazon (Canadian site), Douk H7 alternative (Canadian site), this amp can also be purchased directly from AOSHIDA; a worthy consideration among the TPA3255 amplifiers out there.

[For an even lower price, so long as you don't need the VU meter or XLR input, the older Aiyima A07 might be worth trying.]

--------------------


To end, for those of us reviewing amplifiers "more objectively" and incorporating subjective listening as well, I don't think it's at all surprising when folks run somewhat controlled trials of amplifiers and don't find much of a difference - like this. Audiophiles should know from all the decades of writings out there that within the linear range of a hi-fi (low noise, low distortion) amplifier, sonic differences are going to be subtle given the limitations of human ears, speakers, and rooms. I bet if we were to measure the Aiyima T9 vs. Yamaha AS801 as in that video, we would easily find substantial differences even if not well heard! I'm certainly not saying that all amps sound the same (in fact I've shown that they don't in my system), I just don't think there are any surprises that price and sound quality don't necessarily correlate and it's not hard to predict potentially audible differences using technical means.

There's no reason these days that an inexpensive $150 amplifier like the A7 should sound any worse than a $1000 amp or even a "high-end" luxury $10,000+ amp like the Vitus Audio. To a certain point, higher prices should get us better quality components and workmanship for reliability and longevity, but beyond that, price will be primarily governed by "non-utilitarian", non-sound-quality factors - whether certain reviewers admit it or think they hear massive differences. :-)

Have a wonderful springtime (in the Northern Hemisphere). I've got quite a bit of real work to do before heading off to some "spring break" R&R. Also checking out U2's just-released Songs of Surrender (2023, DR8-9 average) re-recording of some of their classics spanning 4 CDs and 40 tracks.

Hope you're enjoying some wonderful music, audiophiles!

And don't forget to give me your 16-bit vs. 24-bit Part Deux Blind Listening impressions and Systems/Services Survey before May!

6 comments:

  1. Eva Cassidy. Qobuz has two versions of the 25th Anniversary Nightbird in 24/44.1 "high res". One is called "Nightbird (Audience Muted)" - apparently the one you have - and the other is simply "Nightbird". When they became available I sampled both and definitely preferred the sound of the unmuted version. It just seemed more musical. Original mix? I notice that Qobuz also has a version showing the black & white cover art in 16/44.1. Original mix? I will try a comparison.
    Also, did I read that Qobuz is coming to Canada?

    ReplyDelete
    Replies
    1. Thanks Mark,
      Ahhh, that clears things up quite a bit. I see that Nightbird is the live compilation and there are these "audience muted", lower DR, versions they're using in the 25th anniversary.

      Yeah, I see articles like this about Qobuz in Canada:
      https://www.hiresaudio.online/qobuz-to-launch-in-canada/

      They say it's "Coming in May 2023". Hopefully competitive in price and offers a decent amount of multichannel content that can be played over Roon!

      Delete
    2. My bad! Replace "Nightbird" in my comment with "Live At Blues Alley". I blame being on my first cup of coffee this morning when I wrote my reply. (I do have Nightbird in my collection.) Everything I said applies to Live At Blues Alley. The Qubuz albums include the two 24 bit versions of the 25th Anniversary release, with and without audience sounds, and the B&W cover art version in 16 bit.

      Anyway, this afternoon I did do some comparisons using mostly the first track, Cheek to Cheek.
      But first let me say - What a great album! Boy can that Eva ever sing!
      I started with the 16 bit version and it sounded like I remembered. To stay in line with your comparison I then put on the audience removed version of the 25th. Initial impression was it sounded as good, until about one minute in, when Eva really starts going. There the compression (or whatever) robbed the music of its impact. Back and forth comparisons throughout the track confirmed this impression. I then went to the non-edited 25th version. It sounded pretty much the same as the 16 bit version. Maybe it sounded "cleaned up" a touch, but if so it was ever so slightly. I could happily listen to either.

      Which I did. I couldn't stop, so I poured myself a beverage and listened to the whole album and fully enjoyed it. So thanks for pointing it out. This is the second time in a couple of weeks that you have inspired my listening choices. Once I got into assessing the Daft Punk samples in your blind test I couldn't help but listen to that whole album too. I had not put it on in a long while. Very enjoyable, but not quite on par with Eva.

      Keep up the good work.

      Delete
    3. Glad you had a great time with the music Mark!

      That's what it's really about ultimately anyways. ;-)

      As for these remasters and audience-removed versions, I guess they're an option but I'm quite happy to stick with the original. Appears to be the fashion of the day to provide 24-bit versions even though the 25th Anniversary has a lower DR than the original and therefore likely doesn't even benefit from having the extra bits.

      Delete
  2. Very interesting review Arch, especially your subjective results. Can we envision a day when Class D becomes asymptotically "close enough" to audible perfection at reasonable price points? This seems to have largely happened with DACs.

    ReplyDelete
    Replies
    1. Hey there wtnh,
      Yup, I think it's happening. Just a matter of time...

      These days, I think it's hard to get excited over DAC sound quality given the performance of devices over the last couple of years whether it's an inexpensive Topping D10s/B or higher up in the DAC ladder with ES9038-type devices or the newer AKMs. More important to consider whether the DAC fits the needs and has the features we want.

      So too amps. I think the day is coming if not here in many situations. With such a wide range of speakers to drive, likely will need to have a good range of amps as well. With reasonably efficient speakers, I think the job is not difficult. Most of the time, I think it would be unwise for audiophiles to lust over those monster 500W amps but it's good that they're available for those who might need them!

      As consumers, I think it's good to know what we need so as to get what we want without being unduly influenced by advertising hype (especially of the very questionable variety!).

      Delete