Are audio companies all involved in a huge conspiracy? - Page 39

No, what Ethan means by frequency balance does not affect pitch. Frequency affects pitch. Frequency balance means that all the frequencies are being reproduced at the same relative levels. Since timbre is determined by the relative weight of harmonics, frequency balance, or flat frequency response, properly maintains that weight, and therefore accurately reproduces timbre.

Arny, I don't remember its designation from JBL, but the driver in that sub isn't related to the 2235H at all. It could be considered related to the 2226H, in that it has the 2226H's three "Vented Gap Cooling" tunnels that replaced the 2235H-style single large pole vent.



The driver in the HB5000 is actually an evolution of the driver originally found in the Revel Ultima Sub15.

PE closed out Harman's stock of the Revel sub's drivers as the " Jbl "Sub1500," so we have specs on those.

I would guess that the HB5000 driver has a little more throw than the Sub1500's 18mm geometric xmax, because the basket (shared with their W15GTi car sub) has greater clearance than the Sub1500's old-style basket.

While there are current 15" drivers that have more xmax...personally, I'd take a JBL driver with "only" ~18mm rated xmax but a 4" voicecoil, well-implemented shorting rings, and obvious attention to heat management over pretty much any of them.

Freq,Hz Max SPL, DB

10 89
20 101
30 108
40 113
50 117
60 120
70 123
80 125
90 127
100 129
130 133


I hate speculation, especially when reliable facts are at hand. With little or no speculation larger displacement woofers are readily available:

http://www.jlaudio.com/13w7ae-d1-5-car-audio-w7ae-subwoofer-drivers-92116

Diameter = 13 inches, Xmax = 32 mm

Freq,Hz Max SPL, DB

10 91
20 103
30 110
40 115
50 119
60 122
70 125
80 127
90 129
100 131
130 136

Pretty close by advantage: JL Audio (slight)

or:

http://store.ficaraudio.com/q15/

Diameter = 15 inches, Xmax = 28 mm

Freq,Hz Max SPL, DB

10 93
20 105
30 112
40 117
50 121
60 124
70 126
80 129
90 131
100 133
130 137

Even a little better, and check the pricing!

Come on, Arny! What is this statement:

"I've seen pictures of the 15" driver contained within, and it reminds me of the JBL 2235."

The only things that driver has in common with the old 2235 are the logo and the voicecoil diameter! The basket allows considerably more throw, the cooling system is different, the cone is different, the surround is much wider.

JBL 2235H


JBL W1500H



Point well taken, that the very, very expensive (and flawed for most uses, as I'll explain below) Jello 13W7 is the equal of even top-tier 15's in terms of volume displacement. And also, we agree that a few cheap, crudely engineered-and-built drivers can do get loud down low, too.

But assuming we're talking about drivers for use above the first-mode region (i.e. above 40-50Hz), there's more to it than just displacement.

For example, I've used the 13W7. What your theoretical displacement-limited SPL numbers miss is that while the 13W7 has an extraordinarily long strong stroke and damn near perfect BL linearity over that stroke (per Klippel measurements), it also has a giant inductor (voicecoil) in a motor with no Faraday shielding whatsoever to linearize inductance over that prodigious stroke. (IMO, no shorting rings is a a shocking omission at the price Jello asks for those.) So you have a massive inductive hump right in the middle of the driver's passband in most systems, with a rolloff on the top side that severely limits its usable bandwidth.

The Fi driver is cheaper, but Fi doesn't do what I'd consider acceptable build quality. The Ascendant Avalanche 18 I bought from Fi didn't even have the surround glued on properly; I had to puncture it just to access the basket's screw holes! A their motor design is as crude as it gets. Just big. Here's0 the kind of frequency response one gets from such a crude motor, due to the very high inductance from the crude motor.

By contrast, the JBL driver has low inductance, and Faraday shielding for low inductance linearity over its stroke. I've measured older version (Sub1500*) as pretty flat out to 300Hz, which mean it's easily usable up to 150Hz or so if one's conservative. (And higher if one is more adventurous with filtering in that perceptually important upper bass region. JBL claims a 400Hz top end for the HB5000, to which I say, "maybe.") Not to mention the thermal management (perimeter venting, large coil, etc.) to let it play full tilt for longer periods.

For some uses, there may be better tools than the JBL woofers that cost less. (Other competitive drivers, such the Aurasound ones, are more expensive.) But that JBL woofer is very, very good. And very well-rounded, too. Unlike a Jello W7 or Fi piece, one needn't design the system around the woofer's weaknesses.
Edited by DS-21 - 12/23/12 at 3:58pm

If you could illustrate that by means other than hand waving, rumor, questionable claims, and speculation, I would be happy to agree.

Actual performance measurements would suffice. ;-)

Seeing nothing reliable, my opinion remains unchanged.

A skewed frequency response in an audio device does not change the pitch (frequencies) of audio passing through it. Only the frequency balance changes, such that you get too much or too little bass, midrange, or treble. If distortion is great enough to be audible then sure, that will affect the tonal balance too by adding harmonics. This is why low distortion is an important goal of audio equipment.

--Ethan

Arny, your response puzzles me, especially since I provided data in the form of pictures showing conclusively that the only thing about the JBL W1500h in the sub discussed above that would legitimately "remind" anyone of their old 2235 is the logo on the back! The basket, soft parts, and motor are beyond-any-reasonable-doubt very different.

Also, I provided a link to measured data of a system with a driver that has the same basic configuration as the Fi unit. with a very long overhung coil and no shorting rings in the motor. (The eD driver has a stamped basket, and less throw. The former, I think you'll agree, is a distinction without a difference in this context, and the latter, if anything, only casts further doubt on the Fi driver's real world performance, given that its longer coil is be an even bigger inductor.)

Perhaps I should've offered a contrast with data from subwoofer system based on a woofer designed for high-fidelity reproduction from the same source. Here's a correction to that oversight. Here are Klippel measurements for a variant of the driver in that SVS sub:


Yes, that particular sub is a 12 with a lot less displacement. I went to it because I wanted to show measurements from something reasonably affordable. (As opposed to, say, the LMS Ultra driver on that page.) Also, it's worth noting that a decade-ish ago Keith Yates measured four of an earlier variant of that Peerless woofer in a large vented box with 250W each, and found that Genelec system sufficient to generate "cinema reference" bass levels levels down to 20Hz. So greater volume displacement won't necessarily provide any performance differences for people who listen at or below "cinema reference" levels.

Also, here's some "data" about Fi's build quality and attention to detail, from my old Ava18, that I posted to AVS a few years ago:


But more generally, which of these statements do you find objectionable:

1) A low-efficiency woofer with a basket that provides lots of mechanical clearance, a wide surround, and multiple gap venting (JBL W1500h) is very different from a high-efficiency woofer with a basket that provides a lot less mechanical clearance, a narrow surround, and a central pole vent (JBL 2235H).

2) Considering the observed similarities in the soft parts of the W1500h and the old JBL/Revel Sub1500,, but that the W1500h has more motor (double-stack of magnets) and a basket with visibly more clearance (shared with, inter alia, Harman's W15GTi car sub), it's not really "speculation" but more "inference" that it's probably going to have a little more throw than the earlier driver (but not much more, because the soft parts seem the same).

3) For a given stroke, a woofer with low normalized inductance (Le/Re) will have a flatter native frequency response in the upper bass than one with high normalized inductance.

4) A driver with a long voicecoil and no shorting rings has both high normalized inductance and large inductance variation over its stroke or Le(x) in Klippel terms. The higher inductance and Le(x) creates audible distortion (for lack of a less-stupid word, a "sluggish" or "plodding" sound, like the fundamental shakes the room just slightly after one has already heard its overtones) compared to a driver with well-controlled Le(x) in the top half of a typical subwoofer's passband.

I expect #4 is the only one you might reasonably find controversial. Fair enough. It does depend on how one uses the driver, and the cheap/crude ones are great in terms of price-performance for use as dedicated air pumps where human hearing is relatively insensitive. But when asked to play the whole bandwidth expected of a typical subwoofer (LF cutoff up to 80-120Hz), they just don't do the job well in my experience.

While I can't offer hard data - I don't have any example of what I'd consider an inferior bass unit on hand, and haven't bought such a thing in years - here is my most recent experience. Many years ago so I bought the above-pictured Ava18 (made by Fi). It did not have shorting rings in the motor, and the claimed normalized Le was under 1mh/Ω. I had previous experience that high-inductance drivers were bad news. Specifically, the old Stryke Audio HE15, which was a TC Sounds driver based on an earlier variant of the motor used in this driver. It sounded horrible compared to a much smaller Peerless XLS-based sub that had similar capabilities down to 25Hz (passive radiator 12 compared to closed box 15). But at the same time I bought the Ava18 I had also purchased bass EQ capabilities I didn't have during that comparison, in the form of a Velodyne SMS-1 and two Crown XTi2000 amps. So I convinced myself that the Ava18, considering its stroke (dual-gap motor with a more-or-less evenhung coil, not a very long overhung coil), would make it a basically different experience from that old Stryke-marketed TC Sounds woofer with smart use of parametric EQ.

Oops. While the Ava18 provided more pressurization on movie special effects such as the helicopter drone in Black Hawk Down than anything I had previously experienced, when playing music with lots of upper-bass content it just sounded bad. There was that same "plodding" (for lack of a better word) quality to the bass like the old HE15, regardless of how much EQ I threw at the top end of its passband. (I later measured my driver, and found that the published value of Le was less than half of what I measured on my unit.) That was after EQ, not before it. I ended up getting good bass at the primary listening position (but nowhere else) by adding a big and ugly pro audio sub behind the listening position, and only playing the Ava18 very low. The big pro box behind my couch was ugly, but it worked, and a twentysomething grad student can get away with "ugly." That was good in a way, because it led me down the path of reading about how to get good-sounding bass over a wider listening area in a real room, and ultimately to Dr. Geddes' setup methods for multiple subwoofers. (Ironically, my two "broadband" subs in Geddes-speak at the time used the aforementioned JBL 2235. It was fine for that use, though a little light on throw compared to modern drivers.)

Later, I was able to eliminate the giant ugly sub behind the seat (while keeping the other "Broadband" subs) when I replaced the Ava18 with a driver of better design, Exodus Audio's Maelstrom-X. The Mael-X was very similar to the Ava18 in most ways: both were long-throw 18's with foam suspensions and dual-gap motors, and they were drop-in replacements for each other. But low inductance and Le(x) variation were design goals for the Mael-X.

FWIW, here's how the models compare for both drivers, in 120L cabinets with 2kW of power:


As a postscript, I subsequently replaced the Mael-X with an Aurasound NS18-992-4a. It did so for the sole reason that the other subs in my room used Aurasound drivers, and I wanted all my bass drivers to look the same. I also hoped for slightly better performance, as the Aurasound NS drivers basically define the current state-of-the-art (underhung motors with wide but relatively short voicecoils, copper-sleeved poles for very low inductance and very linear Le(x), neo magnets, etc.). The NS18 has much lower inductance than anything of its displacement. After actually doing the swap...I doubt I could distinguish the Mael-X from the NS18 in a blind test. So there is a threshold level of where "better" is "no different." I've found that for most uses, very-long-coil overhung drivers with no shorting rings do not meet that threshold, though.

Arny, your response puzzles me, especially since I provided data in the form of pictures showing conclusively that the only thing about the JBL W1500h in the sub discussed above that would legitimately "remind" anyone of their old 2235 is the logo on the back! The basket, soft parts, and motor are beyond-any-reasonable-doubt very different.

Also, I provided a link to measured data of a system with a driver that has the same basic configuration as the Fi unit. with a very long overhung coil and no shorting rings in the motor. (The eD driver has a stamped basket, and less throw. The former, I think you'll agree, is a distinction without a difference in this context, and the latter, if anything, only casts further doubt on the Fi driver's real world performance, given that its longer coil makes an even bigger inductor, with larger inductance variation over its stroke.)

Perhaps I should've offered a contrast with data from subwoofer system based on a woofer designed for high-fidelity reproduction measured by the same source. Here's a correction to that oversight. Here are Klippel measurements for a variant of that driver:


Yes, that particular sub is a 12 with a lot less displacement. I went to it because I wanted to show measurements from something reasonably affordable. (As opposed to, say, the LMS Ultra driver on that page.) Also, it's worth noting that a decade-ish ago Keith Yates measured four of an earlier variant of that Peerless woofer in a large vented box with 250W each, and found that Genelec system sufficient to generate "cinema reference" bass levels levels down to 20Hz. So greater volume displacement won't necessarily provide any performance differences for people who listen at or below "cinema reference" levels.

Also, here's some "data" about Fi's build quality and attention to detail, from my old Ava18, that I posted to AVS a few years ago:


But more generally, which of these statements do you find objectionable:

1) A low-efficiency woofer with a basket that provides lots of mechanical clearance, a wide surround, and multiple gap venting (JBL W1500h) is very different from a high-efficiency woofer with a basket that provides a lot less mechanical clearance, a narrow surround, and a central pole vent (JBL 2235H).

2) Considering the observed similarities in the soft parts of the W1500h and the old JBL/Revel Sub1500,, but that the W1500h has more motor (double-stack of magnets) and a basket with visibly more clearance (shared with, inter alia, Harman's W15GTi car sub), it's not really "speculation" but more "inference" that it's probably going to have a little more throw than the earlier driver (but not much more, because the soft parts seem the same).

3) For a given stroke, a woofer with low normalized inductance (Le/Re) will have a flatter native frequency response in the upper bass than one with high normalized inductance.

4) A driver with a long voicecoil and no shorting rings has both high normalized inductance and large inductance variation over its stroke or Le(x) in Klippel terms. The higher inductance and Le(x) creates audible distortion (for lack of a less-stupid word, a "sluggish" or "plodding" sound, like the fundamental shakes the room just slightly after one has already heard its overtones) compared to a driver with well-controlled Le(x) in the top half of a typical subwoofer's passband.

I expect #4 is the only one you might reasonably find controversial. Fair enough. It does depend on how one uses the driver, and the cheap/crude ones are great in terms of price-performance for use as dedicated air pumps where human hearing is relatively insensitive. But when asked to play the whole bandwidth expected of a typical subwoofer (LF cutoff up to 80-120Hz), they just don't do the job well in my experience.

While I can't offer hard data - I don't have any example of what I'd consider an inferior bass unit on hand, and haven't bought such a thing in years - here is my most recent experience. Many years ago so I bought the above-pictured Ava18 (made by Fi). It did not have shorting rings in the motor, and the claimed normalized Le was under 1mh/Ω. I had previous experience that high-inductance drivers were bad news. Specifically, the old Stryke Audio HE15, which was a TC Sounds driver based on an earlier variant of the motor used in this driver. It sounded horrible compared to a much smaller Peerless XLS-based sub that had similar capabilities down to 25Hz (passive radiator 12 compared to closed box 15). But at the same time I bought the Ava18 I had also purchased bass EQ capabilities I didn't have during that comparison, in the form of a Velodyne SMS-1 and two Crown XTi2000 amps. So I convinced myself that the Ava18, considering its stroke (dual-gap motor with a more-or-less evenhung coil, not a very long overhung coil), would make it a basically different experience from that old Stryke-marketed TC Sounds woofer with smart use of parametric EQ.

Oops. While the Ava18 provided more pressurization on movie special effects such as the helicopter drone in Black Hawk Down than anything I had previously experienced, when playing music with lots of upper-bass content it just sounded bad. There was that same "plodding" (for lack of a better word) quality to the bass like the old HE15, regardless of how much EQ I threw at the top end of its passband. (I later measured my driver, and found that the published value of Le was less than half of what I measured on my unit.) That was after EQ, not before it. I ended up getting good bass at the primary listening position (but nowhere else) by adding a big and ugly pro audio sub behind the listening position, and only playing the Ava18 very low. The big pro box behind my couch was ugly, but it worked, and a twentysomething grad student can get away with "ugly." That was good in a way, because it led me down the path of reading about how to get good-sounding bass over a wider listening area in a real room, and ultimately to Dr. Geddes' setup methods for multiple subwoofers. (Ironically, my two "broadband" subs in Geddes-speak at the time used the aforementioned JBL 2235. It was fine for that use, though a little light on throw compared to modern drivers.)

Later, I was able to eliminate the giant ugly sub behind the seat (while keeping the other "Broadband" subs) when I replaced the Ava18 with a driver of better design, Exodus Audio's Maelstrom-X. The Mael-X was very similar to the Ava18 in most ways: both were long-throw 18's with foam suspensions and dual-gap motors, and they were drop-in replacements for each other. But low inductance and Le(x) variation were design goals for the Mael-X.

FWIW, here's how the models compare for both drivers, in 120L cabinets with 2kW of power:


As a postscript, I subsequently replaced the Mael-X with an Aurasound NS18-992-4a. It did so for the sole reason that the other subs in my room used Aurasound drivers, and I wanted all my bass drivers to look the same. I also hoped for slightly better performance, as the Aurasound NS drivers basically define the current state-of-the-art (underhung motors with wide but relatively short voicecoils, copper-sleeved poles for very low inductance and very linear Le(x), neo magnets, etc.). The NS18 has much lower inductance than anything of its displacement. After actually doing the swap...I doubt I could distinguish the Mael-X from the NS18 in a blind test, even though the NS18 technically has a fair bit less throw. Still I sold the Mael-X fairly recently, simply because I like knowing that all of my subwoofer drivers have the same design. Prior to buying the NS18, I briefly owned a TC Sounds LMS Ultra, which is the immediate predecessor to this driver. Despite the longer stroke, it did not appear to be an upgrade for my uses, so I took advantage of one of TC Sounds' periodic bouts of insolvency to unload it at a profit. So there is a threshold level of where "better" is "no different." I've found that for most uses, very-long-coil overhung drivers with no shorting rings do not meet that threshold, though.

You seem to be very confused about what constitute performance measurements for a subwoofer. You seem to be lost in the woods and confused by all the trees.

An example of "Performance measurements" is things like SPL versus frequency versus nonlinear distortion.

So far I've seen a ton of bandwidth wasted with all kinds of details, and no data about how things actually work when the rubber hits or the road or in this case the SPL hits the air (at say 20 Hz).

Three subwoofers were mentioned:

JBL W1500H

JL Audio 13W7

Fi Car Q15

It's really pretty simple. Just provide data for something like THD versus SPL @ 20 Hz as SPL varies from 80 dB to 130 dB. All three drivers. Or even 3 other comparable drivers.

You see you can brag about the polish on the pole pieces or shorting rings or whatever, but if it doesn't give more clean SPL at 20 Hz, why should I or anybody else care?

Because one doesn't just use a sub at 20Hz. If we're talking just the super-low stuff, then I agree with you that all one needs is a big air pump. In which case, a big crude driver with a very long voice coil (like the Fi) is probably the best choice, because it'll pump the most air at the lowest cost.

But someone who does not low-pass a sub at 40-50Hz should consider how a given device performs in the top half of its passband, too. And beyond, given that a lowpass is not a brick wall.

Of those three drivers (Fi Q, Jello W7, JBL W1500h), the only one useful as anything but an extreme ULF air pump is the JBL device.

The Jello has the volume displacement and BL(x) linearity, and is beautifully built, but has too much inductance and too much inductance variance over its stroke. Same goes for the Fi, except for the parts about build quality and BL(x) linearity. The JBL may be beaten by cheaper drivers (such as the Fi) in terms of SPL at 20Hz, and maybe equalled or bettered for most home audio uses overall by cheaper drivers, too. But of those three, it's the only one with low enough inductance to be a good performer over all of the typical bandwidth of a subwoofer in a domestic audio system.
Edited by DS-21 - 12/26/12 at 7:29am

For that comment to be relevant, you'd have to show some kind of fairly common systematic problem where subs have good performance at 20 Hz, but fall apart at 80 or 140 Hz.

I see the hands waving and not leaving the ends of arms.

I don't see a cogent discussion or compelling evidence.

This sort of weirdness has been going on for a week. How long does this go on before it gets interesting or even useful? ;-)

Maybe They are involved in a Conspiracy. I abx'ed both lossless Dts Hd Master Audio & Dolby True Hd Tracks & lossy DD & DTS(Including 96/24 DTS) by converting them to wav, Flac, Ogg Vorbis, & (Yikes!) even Mp3! I then discovered that there was no major differense between the Lossy & Lossless sourced audio once converted to another format. the only exception being that the

I think they want you to buy & keep replacing processors, amps tuners, etc

I will post a link of 2 Lossy DD tracks that were converted to mp3 that sound as good as any lossless track i've heard. The first is Goldeneye(Downmix From 5.1 to 2.0 Pro Logic Stereo) www.packupload.com/3YW2T2IQW6E

The Second is Star Wars Episode 1: The Phantom Menace(2.0 Pro-Logic Mix From The 2001 dvd) www.packupload.com/6BDLEMC79Q1 Play them on your home theater systems for the best results.

I also converted hi-res 96khz 192khz Flac/Wav Music files to 44khz, & 48khz Ogg Vorbis files & I heard no difference in quality.

I think the general thought is that DTS and DD encoding as typically implemented on DVDs (not highest bitrate) is nearly but not quite fully transparent to the original source. But DD and DTS at higher bitrates (though still lossy) included on most BDs is very nearly entirely indistinguishable from the original source. FWIW, when I was a new BD convert and motivated to do some testing, it seemed to me that (after level matching) the rain that opens (the second?) Pirates of the Carribean movie sounded noticeably different on my old DVD than in lossless (PCM, IIRC) on BD. IIRC, there was more "detail" to the overall sound and little muances like rain striking a china cup stood out better in the lossless format. Rain, being close to a pink/white noise source, might make a pretty good "test signal," as most real world content is significantly easier for lossy systems to handle since there's less "going on" at least from a frequency response perspective, and more things are potentially masked and thus "losable."

I thought the "win" with the hi-rez codecs was in the transfer process and added channel capability, not so much the lossy/lossless argument. (Specifically over HDMI.)
I don't know how it works, but I thought there was a benefit to having the unpacking and clocking done at the DSP in the preamp rather than the DSP in the player & then being sent..

I can't use movies for reference, that just seems too out there for me, but I would bet dollars to donuts that I noticed an improvement in live concerts I have on Blu Ray in DTS HD MA or True HD over DVD in DD or DTS.
But, there's no way of knowing if it's just the natural progression of the actual recording processes, the higher resolution, or if it's what I wanted to hear....

There is good data that shows that AC-3 (Dolby Digital) is not transparent at the data rates used for DVD. Here is an example double blind test at 384 kbps:



As the top line shows, performance varies based on content and falls short of transparency. Many people are not sensitive to such differences though so what the other poster said may very well be true for his ears.

Usually, when there are fairly and generally audible differences, the most likely cause is mastering and mixdown, both of which are fairly likely to change when comparing a DVD to a BD.

The more reasonable comparison would be between the lossless and lossy tracks with the same basic surround configuration on the same disc.

BTW I'm not saying that DD tracks are sonically transparent because it has long been known that DD at the bitrates on common DVDs can be picked out in DBTs comparing them to their lossless direct equivalents.

One reason why there are newer coders such as AC3 is real, scientifically proven concern over the transparency of DD.

However, a lot of this lack of transparency is not easy to hear and is only really audible upon direct comparison with the lossless tracks that were encoded to create the DD tracks.

One of the commercial goals of re-releasing media that was remastered with new encoders is to coax people into buying artistic works that they already own. This was a bonanza for the audio CD, and everybody would like to see that lightening strike again. Trouble is, very few mainstream commercial formats are as obviously audibly flawed as the LP or consumer analog tape particularly cassette.

AC-3 is the older name for DD.

I'm late to this party, but find your posts fascinating. I have the opportunity for an experiment I'd be interested in your prediction of the outcome.

I have a system and amps which I think would meet the criteria of reasonably designed. I use an OPPO 95 for CD/DVD/BluRay playback. For use with my wife's TV, I'm getting a $50 Samsung BluRay player. I'll have an opportunity to test it against my OPPO. Unfortunately it will be a sighted comparison. What would you expect a DBT to yield?

This above was for Arnyk and related to his comments on amps and digital sound differences from a few months back. Sorry for not being able to run this forum better.

If you compare the digital outputs of both an OPPO 95 and a good low end blu ray player the sound should generally be the same. The bitstreams should be the same unless one of the players does not support all of the audio formats involved. The video may vary because of differences in upscaling, etc.

The Program I used to convert the Audio tracks to 5.1 WAV/Flac/Vorbis is called DVD Audio Extractor.

I wonder if he was thinking of AAC. That would make sense.

AAC is 5 years newer than DD (1997 vs 1992, respectively), but an odd choice of codec to cite it in the context of discussing audio on DVD and BD.

Thanks. I've tried it. Haven't listened carefully yet, but the Samsung sounded pretty good.

BTW: Cruising around the AVS forum (my Onyko 886 is so complicated I need verification of my switching and hookup) I found a derogatory comment per the 886's analog inputs penned by one Kal Rubinson, of "Music in the Round" from Stereophile. I PM'd him to ask whether his comment applied to both the mulitchannel inputs and the separate stereo inputs. He said it did. I further asked whether his comment was based on general principle or had he some specific knowledge. So far no response to that.

That got me motivated to look around for tests I found this review in Hi Fi World: http://www.hi-fiworld.co.uk/index.php/av-receivers/39-av-reviews/199-onkyo-pr-sc886.html?start=1

A key point: Distortion was very low at 0.0002% and noise also low at -90dB at full gain, Direct or Pure Direct selected, lowering to 85dB through the input A/D.

So I queried how this could be considered mediocre performance by any standard. Still no response.

I suspect it is either his own personal preference, the unit under review had a problem or more than likely it didn't cost enough $$$$ so he thought there was no way it could sound anything but bad.

Frankly -90 dB dynamic range is pretty mediocre technical performance from a modern DAC. It is unlikely to be an audible problem, but given that $3 chips beat that by like 15 dB...

http://www.ti.com/product/pcm1681

http://avnetexpress.avnet.com/store/em/EMController?action=products&catalogId=500201&storeId=500201&N=0&langId=-1&slnk=e&term=PCM1681PWP&mfr=TIS&hrf=http%3A%2F%2Fwww.ti.com%2Fgeneral%2Fdocs%2Fusercart.tsp%3Faction%3Dadd%26elementId%3DPCM1681PWP%26elementTypeId%3DPR%26quantityType%3Dreg%26localeId%3DEN%26apUserId%3DC_EN_133341874%26ref%3Dpcm1681

It's not a measurement of the DAC, it's a measurement of the receiver through the DAC. Lots of other places for noise to creep in. I'm guessing there's an extra gain stage or ihigher gain from some intermediate amplification device that's adding hisssss. I'd agree with Arny that the noise is very very unlikely to be related to the DAC chip itself . . .

Edit: Bein' a curious fellow, and since a receiver includes a power amp section, I went lookin for power amp performance. Here's a quote from J Atkinson's testing of a Classe CT-M600 amp "The Classé CT-M600 is among the quietest amplifiers I have measured. Its wideband, unweighted signal/noise ratio (with the input shorted and ref. 1W into 8 ohms) was a superb 79.6dB. This increased to 93.1dB when the measurement was restricted to the audioband, and 95.8dB when an A-weighting filter was switched into circuit. For reference, the respective S/N ratios for the Bryston 7B SST2 were 75.6dB, 92dB, and 95.7dBA."

For whatever that's worth.
Edited by JHAz - 1/8/13 at 9:43am

After measuring the performance of a ton of equipment that contained DACs, the performance of the internal DAC chip is often represented at the output terminals within a dB or two at most. Even power amps often have > 110 dB dynamic range so they often don't degrade the performance of the DACs that drive them that much.

The review states the noise is measure with the output volume set to max, which is 15 dB gain. The volume when listening normally would be 20-40 dB lower than that. The noise floor will undoubtedly drop at reduced volumes. So we cannot know if the Onkyo's noise will dominate at the speaker terminals from the data at hand.

The amp's noise is around -90 dB below 1W (1V in 8 ohms) and has 29 dB gain. If the Onkyo's output noise were 90 dB less than 36 mV (the input voltage needed for a 1V output), then it would be similar to the amp's noise floor. That will put it 124 dB below a nominal 2V line level signal. All this to say that we cannot compare these figures between a preamp and an amp and come to any conclusion other than the tougher job lies with the preamp.