One of the primary reasons for me choosing z9 over other monsters is its dac section which implements the best in any avr. I had to wait long for z9 but finally when I had it and after listening I understood it is a special avr with a very special pre-amp and dac section which is why it sounds so good.
High quality digital audio reproduction isnt cheap and AV amplifiers like DSP-Z9 have eleven separate digital channels (7.1 + 2 front presence + 1 additional subwoofer) to deal with. I believe there are much more to Yamaha DSP-Z9 than just excellent DACs in the digital audio board. The sound quality which you are referring is primarily due to exemplary circuit designing of its digital audio boards by some of the most renowned experts in this field. Along with the combination of top quality electronic parts, advanced digital signal processors (DSP), top of the line digital to analog converters (DAC), analog to digital converters (ADC), operational amplifiers (op-amps) and volume controllers made it truly shine above the others.
The first generation of Uber AV amplifiers kicked off with Denons 90th anniversary celebration in 2000. Denon AVC-A1SE was born. Pioneer quickly followed next year with Pioneer VSA-AX10. Mighty as they were at US $4,000 price tag, on release knocked off many well known two times more expensive separates in terms of performance, sonic quality, industry leading video capabilities and groundbreaking new technologies like auto acoustic calibration (for Pioneer VSA-AX10 only). They literally got vanished the moment they arrived on dealers shelves. They sported some of the most advanced DSPs of those times using primarily the first generation of SHARC DSPs from Analog Devices. These 32-bit DSPs where powerful enough to perform DVD quality audio decoding/ processing of 24 bit depth and up to 48 kHz sampling frequency from THX, Dolby & DTS. Post-processing like auto acoustic calibration, acoustically simulated environments and digital audio management like bass management were also supported. But before we talk about DSPs and their processing capabilities lets have a walk through the meaning of audio sampling frequency and bit depth.
------------------------------------------------------
Sampling Frequency: As per theory you can accurately measure an audio wave up to half the sampling frequency. So if we consider DVD quality 48,000 samples per second (48 kHz), it is good enough for a 24 kHz audio wave covering the full audible frequency range. A DSP capable of handling 48 kHz sampling frequency is good enough for all we can ever hear. On the other hand if we consider higher quality DVD-A/ SACD/ Dolby TrueHD/ DTS-HD MA capable more is better DSPs then it must handle 96kHz sampling rates, thus processing 96,000 samples per second (96 kHz) and being good enough for a 48 kHz tone going well above the audible frequency range. I will not even talk about anything higher than 96 kHz DSPs, because to me it is pure overkill.
Bit depth: It denotes word length, i.e. how many ones and zeros in the digital "word" that records each measurement. 24 bit has 24 ones and zeros and can record 2 to the power 24 or about 16 million different volume levels between dead silent to the loudest note. 24 bit have been the de-facto standard in the audio industry for a long time.
------------------------------------------------------
A high quality BD player will decode Dolby TrueHD/ DTS-HD MA sound from BD disc into eight individual PCM digital audio channels, all at 96 kHz/24-bit. For analog transmission of this 7.1 channel signal it will then use 96 kHz/24-bit DACs to convert the 8 channel PCM digital signal to analog audio waveform over its 7.1 channel analog outs. An AV amplifier accepting this audio transmission over its 7.1 channel analog ins should ideally use 96 kHz/24-bit ADCs to regain the same nearest to original 8-channel PCM digital audio signal (as generated in the BD player) and then continue through its digital chain to DSPs which should also be capable to process 96 kHz/24-bit signals. In the end of the chain the DACs should also be capable of 96 kHz/24-bit. So the same bit depth and sampling audio frequency is maintained throughout, from BD player to the internal analog power amplifiers of the AV amplifier. Same goes for DVD-A.
Ubers: 1st Gen
Even though the first generation Ubers came laden with 96 kHz/24-bit ADCs and 192 kHz/24-bit DACs they were all bottlenecked by their DSPs, which were capable to process 48 kHz/24-bit max. The ADCs used in AV amplifier are required to feed the incoming analog audio signal to their digital domain. As their DSPs maxed out at 48 kHz so the ADCs were pre-configured to work at constant rate of 48 kHz, as any higher sampling frequency from ADC were not supported. At the end of the digital chain the DSPs fed the DACs at 48 kHz max. If only digital audio data were fed through optical/ coaxial inputs then these DSPs would process any sampling frequency up to 48 kHz untouched, but as 64 kHz, 88.2 kHz and 96 kHz signals were received it will under sample by half to 32 kHz, 44.1 kHz and 48 kHz respectively to perform any kind of processing. Any signal higher than 96 kHz, these DSPs will simply shut off as under sampling a 128 kHz, 176.4 kHz or 192 kHz by half would yield 64 kHz, 88.2 kHz or 96 kHz and are unsupported by 1st Gen SHARC DSPs. It is only when you use Direct-PCM/ DSD modes on these AV amplifiers did they by-pass the DSPs and you be able to retain up to 192 kHz signals from coaxial inputs, but as the DSPs are shut down you lose valuable setting like acoustic calibration that you performed by auto acoustic calibration (YPAO for Yamaha), bass management and DSP post-processing capabilities like Movie Theater Sc-Fi Enhanced effects.
Ubers: 2nd Gen
Enter Uber second generation AV amplifiers. Denon and Pioneer had a great base with their 1st Gen, so they just upgraded their entire digital audio boards to perform full blown 96 kHz/24-bit DSP processing. It was mainly possible due to release of second generation SHARC DSPs by Analog Devices which where more than three times powerful than 1st Gen SHARCs used in 1st Gen Ubers. The low cost variant was SHARC Hammerhead ADSP-21161 while the full blown version of the same with higher memory capacity, better core optimizations was SHARC Melody ULTRA ADSST-21161. Both where 32- bit floating point quantization microprocessors with cores running at 100 MHz having computing power up to 600 MFLOPS. A year later Texas Instruments introduced another DSP known as Aureus TMS320DA610 which took the crown of being most powerful, this time with core running at 225 MHz having computing power up to 1200 MFLOPS, it had two times more computing power than SHARC. These DSPs were mainly targeted for DVD-A/ SACD quality 96 kHz/24-bit DSP processing and the choice from AV amplifier manufactures where either to use two SHARCs in tandem or an Aureus, both yielding identical performance.
Denon AVC-A1SRA
Pioneer VSA-AX10Ai-S
Onkyo TX-NR1000
Yamaha DSP-Z9
Retailing at US $4,500 minimum in 2004, these 2nd Gen Ubers bought new support for i.Link opening the door for multi-channel digital audio transport from SACD/ DVD-A players. They upgraded the DACs & ADCs in their digital audio boards to support full 192 kHz/24-bit with targeted minimum dynamic range of 110 dB, as it was deemed necessary to scale the enormous dynamic range these new formats had in offer.
So now in 2004 we had Denon AVC-A1SRA & Pioneer VSA-AX10Ai-S on offer. Yamaha finally entered the scene with DSP-Z9 while Onkyo following the year after with TX-NR1000. The benefit from all these upgrades was they performed analog to digital conversion at a constant rate of 96 kHz through their pre-configured 192 kHz/24-bit ADCs (192 kHz spec ADCs were used just for marketing reasons) from 7.1 channel analog inputs. Thus nearest to identical 8-channel PCM digital audio signal (as generated in the BD player) was maintained and continued digital audio processing through the 2nd Gen SHARC or Aureus DSPs capable to handle 96 kHz/24-bit signals. In the end of the chain DACs capable of 192 kHz/24-bit were applied to perform digital to analog conversion at 96 kHz sampling frequency. So the same bit depth and sampling audio frequency were maintained from the BD player to the internal analog power amplifiers of these AV amplifiers which are what it should be.
These SHARC/ Aureus DSPs were able to offer decoding & post-processing capabilities all up to 96 kHz with full auto acoustic calibration on a much denser level and bass management. If digital audio was fed through optical/ coaxial/ i.Link inputs then they would process any sampling frequency up to 96 kHz untouched, but as 128 kHz, 176.4 kHz or 192 kHz are received it will under sample by half to 64 kHz, 88.2 kHz or 96 kHz respectively and then perform processing. So these DSPs were able to handle any sampling frequencies you can throw at them without getting shut down. While using Direct-PCM/ DSD modes on these AV amplifiers will you be able to pass a 128 kHz, 176.4 kHz or 192 kHz digital audio signal from coaxial/ i.Link inputs directly to feed the DACs, untouched. For analog inputs, one way to feed the internal power amplifiers with untouched 2-channel/ multi-channel audio is to use Pure-Direct mode which bypasses the digital board.
The digital audio boards of these AV amplifiers also implemented the analog I/V converter for each channel that deals with raw current being generated by DACs at their very end of audio signal chain. The I/V (current to voltage) conversion stage is just one of several critical analog stages that can impact the performance of a DAC. Typical I/V-convertors are separated into two stages: one for current to voltage conversion, and another for the low pass filtering. Each of these analog stages uses high quality audio operational amplifiers (op-amps) that just contribute slight noise and distortion to the analog audio signal. With careful design, these artifacts are minimized to levels that are well below audibility having distortion less than 0.001%. Ubers used very well designed and robust analog stages for I/V conversion.
I also thought to mention and this often goes un-noticed, is the digitally regulated analog volume controllers in AV amplifiers. No matter how good or pure the analog audio signals are emanated after I/V conversion from these highly complex and expensive digital audio boards, they have to finally pass through these volume controllers to be send to power amplifier domain. As such the importance of high quality volume controllers become a necessity so to incur minimum total harmonic distortion (THD) on the outgoing analog signals to internal power amplifiers. Ubers used high quality stereo volume controllers which had THD of 0.001% while some went to extreme levels to use them in differential mode to obtain full channel separation and THD levels as low as 0.0003% on all channels. Today the 2nd Gen Ubers are most compatible with current generation BD players supporting 7.1 channel analog outs if we keep HDMI out of the equation.
Ubers: 3rd Gen
These HDMI capable Uber AV amplifiers like Yamaha DSP-Z11 or Denon AVC-A1HDA performed identical 96 kHz/24-bit DSP signal processing from the incoming HDMI signal as there was no need to go any higher with regards to Dolby TrueHD/ DTS-HD MA.
-----------------------------------------------------
Some light to the internal digital audio circuit boards as implemented by various 2nd Gen Ubers:
Denon AVC-A1SRA
Digital Audio Chain
DSPs used are a pair of 32-bit SHARC Hammerhead ADSP-21161 from Analog Devices. These are low cost version of 2nd Gen SHARC DSPs, and sincerely Denon never needed any more computational power as they have the simplest implementation of this quad Uber AV amplifier shootout. One DSP is dedicated to Dolby/ DTS decoding sending the 8-channel decoded digital data to other SHARC which performs additional post processing like THX, AL24 Plus, bass management and some surround simulation effects. No auto acoustic calibrations are implemented even though DSPs have support for them. When operational through 7.1 channel analog inputs, four 192 kHz/24-bit Burr-Brown PCM1804 stereo ADCs in single-ended configuration are used to feed the SHARCs.
In a single-ended configuration a single discrete stereo ADC/ DAC/ Volume Controller is used to handle two audio channels, which tends to be the simplest and cheapest implementation. These ADCs capable of delivering 112 dB dynamic range are pre-configured to work at a constant rate of 96 kHz to feed these DSPs.
Digital Audio Board
The 8-channel processed digital audio signal from DSPs is then sent to four 192 kHz/24-bit Burr-Brown PCM1738 stereo DACs in single-ended configuration delivering117 db dynamic range. For I/V conversion stages from these DACs, following op-amps are used: Analog Devices OP275 followed by JRC NJM5532 followed by JRC NJM2068. In the end the 8-channel analog audio signal from these I/V conversion stages are digitally regulated by eight individual Toshiba TC94A17 stereo volume controllers in differential configuration on all channels.
In a differential configuration a single discrete stereo ADC/ DAC/ Volume Controller is configured to handle a single audio channel rather than two, which tends to produce better performance at the cost of design complexity and price.
---------------------------------------------------------------------------------------
Pioneer VSA-AX10Ai-S
Digital Audio Chain
DSPs used are a pair of 32-bit SHARC Melody ULTRA ADSST-21161 from Analog Devices along with a 44-bit Motorola DSP56367. These are high-end version of 2nd Gen SHARC DSPs, Pioneer needing them as their digital audio boards implemented much advanced technologies like auto acoustic calibration over Denon. One SHARC is dedicated to Dolby/ DTS decoding sending 8-channel decoded digital signal to the Motorola DSP to perform additional post-processing activities like THX and some surround simulation effects. Finally the 8-channel post processed digital signal from Motorola is fed to the other SHARC to perform Pioneers own auto acoustic calibration activity better known as Advanced MCACC. VSA-AX10Ai-S was able to perform much denser acoustic calibration over VSA-AX10, courtesy to its SHARCs being 3x powerful. When operational through 7.1 channel analog inputs, four 192 kHz/24-bit AKM AK5385 stereo ADCs in single-ended configuration are used to feed the SHARCs. These ADCs capable of delivering 114 dB dynamic range is pre-configured to work at a constant rate of 96 kHz to feed the DSPs.
The 8-channel digital audio signal from DSPs is then passed to four discrete 192 kHz/24-bit Burr-Brown DF1706 stereo digital filters to perform oversampling and then be sent to eight 24-bit Burr-Brown PCM1704 mono DACs. These are renowned Burr-Brown DACs among audiophiles (other than the ominous PCM1792) which is capable of 112 dB dynamic range while supporting 8x oversampling at 96 kHz i.e. sampling frequencies up to 768 kHz. Generally DACs have digital filters built in like PCM1738 or PCM1792, but PCM1794 required discrete digital filters. Burr-Brown made two of them: 96 kHz/24-bit DF1704 and 192 kHz/24-bit DF1706 to complement a pair of mono PCM1704. Pioneer clearly stated VSA-AX10Ai-S had 192 kHz/24-bit DACs, it means DF1706 capable to accept up to 192 kHz sampling rates from any digital audio source (DSPs/ i.Link/ coaxial) is most probably set at 1x oversampling, literally sending digital data to the DACs at sampling frequencies identical to what is received with no oversampling. I presume they were just used as PCM1704 required discrete filters to operate. This is good in one way as with no oversampling there is no need to reduce the word length, either by truncating, rounding, or dithering. For I/V conversion stages from these DACs, following op-amps are used: JRC NJM5534 followed by JRC NJM2068. In the end the 8-channel analog audio signal from these I/V conversion stages are digitally regulated by four individual Toshiba TC94A07 stereo volume controllers in single-ended configuration on all channels. Without doubt Pioneers implementation for digital audio part in VSA-AX10Ai-S is superior as well as expensive compared to Denon AVC-A1SRA.
---------------------------------------------------------------------------------------
Yamaha DSP-AZ1
Digital Audio Chain
Yamaha DSP-AZ1 do not fall under US $4,500+ Uber category we are discussing here, but I thought to include this US $2,800 TOTL as it has some uncanny internal structural design pattern similarities with 1st Gen Uber Pioneer VSA-AX10. Mind you both were released in 2001, and if you inspect their internals very closely in terms of placement of main components, VSA-AX10 seems to be a much refined version of DSP-AZ1 negating all known compromises of the Yamaha solution. Yamaha had to live with those compromises in DSP-AZ1 as it used a decade old internal structural design pattern from DSP-A1000. Even the digital audio boards of Pioneer VSA-AX10 and its upgraded versions VSA-AX10Ai and VSA-AX10Ai-S uses the same Digital ToP-ART DAC configuration that Yamaha introduced with DSP-AX1 in 1999, later to be followed by DSP-AZ1. Though it may be pure speculation from my side, it seems Pioneer was inspired by Yamahas proven base structural design principle and Digital ToP-ART design philosophy in their digital audio boards and thought to elevate this principle to higher levels with VSA-AX10. Fortunately this similarity ends at a much higher internal structural level. When you start comparing at circuit designing level they are vastly different, VSA-AX10 being superior is every which way you compare which is what it should be for an Uber. So full marks to Pioneer as I find nothing wrong to be inspired from some proven technical philosophy followed by their competitors like Yamaha.
Yamaha DSP-AZ1
Pioneer VSA-AX10Ai-S
1 -> Power Supply Chamber
2 -> Video Chamber
3 -> Analog Audio Chamber
4 -> Power Amplifier Chamber
5 -> Digital Audio Chamber
For DSP-AZ1, the 2, 3 & 4 chambers are merged, rest are same as VSA-AX10AiS
----------------------------------------------------
Yamaha used their own DSPs for decoding until DSP-AZ1, after which their decoding LSis lost steam in comparison to latest and greatest on offer from Analog Devices and Texas Instruments. For the future models they still retained their proprietary DSPs for their own renowned post-processing technology better known as Digital Sound Field Processing. In DSP-AZ1 a 32-bit Yamaha YSS-938 is the mother decoder dedicated to Dolby decoding and if it detects DTS signals a pass is made to child DSP named Fujitsu MB87J0470 which returned the data to YSS-938 after DTS decoding. Finally the 6.1 channel decoded digital audio signal [Dolby Digital EX, DTS ES] is fed to two 44-bit Yamaha YSS-910 Digital Sound Field Processing DSPs which combined have equal computing power of a 1st Gen SHARC. They performed many additional post-processing activities like actual sound-field generation, creation of additional font presence channels with their Digital Sound Field Processing. Many surround simulation effects are added like HiFi DSP, Cinema DSP, Silent Cinema & Virtual Cinema technologies. The DSP audio processing is able to handle a maximum sampling rate of 24-bit, 48 kHz (CD & DVD quality), that of Gen 1 Ubers.
Digital Audio Board
DSP-AZ1 uses a similar DAC implementation like Pioneer VSA-AX10Ai-S in digital audio board where five discrete DF1704 stereo digital filters are used in tandem with ten PCM1704 mono DACs for its 10 digital audio channels (6.1 channel + 2 front presence channels + 1 additional subwoofer) from YSS-910. But as these digital filters are capable of accepting up to 96 kHz sampling rates from any digital audio source, it is preset to 2x oversampling. Thus incoming sampling frequencies like 32 kHz, 44.1 kHz, 48 kHz, 64 kHz, 88.2 kHz and 96 kHz are raised to 64 kHz, 88.2 kHz, 96 kHz, 128 kHz, 176.4 kHz & 192 kHz respectively by the digital filters and then fed to DACs. Even though DSP-AZ1 never really accepted anything higher than 96 kHz, it oversampled and fed the DACs at up to 192 kHz thus making the marketing team very happy to publicize the presence of 192 kHz/24-bit DACs. Personally I tend to stay away from oversampling as there may be a need to reduce the word length, either by truncating, rounding, or dithering. For I/V conversion stages from these DACs, following op-amps are used: NEC uPC45700 followed by NEC uPC45700. In the end the 10-channel analog audio signal from these I/V conversion stages are digitally regulated by five Cirrus Logic CS3310 stereo volume controllers in single-ended configuration on all channels. DSP-AZ1 has no support for digital processing from 5.1 channel analog inputs. DSP-AZ1 is an excellent AV amplifier for its price, being nearly there with VSA-AX10, though the Pioneer shined in every aspect due to a better engineered product.
---------------------------------------------------------------------------------------
Onkyo TX-NR1000
When Onkyo released their only Uber TX-NR1000 in 2005, the catch phrase was
It would take a meteor to wipe this out. Such was the built quality which extends to all other Ubers. The TX-NR1000 is also known as TX-NA1000, TX-NR5000E and Integra DTR-10.5. To me it is also an example of Mother of flexibility as the entire AV processor part is a card-based system. You want HDMI with 4K capability? Just take out the HDMI card from the back and insert a new one with 4K capability available from Onkyo. Voila, now you have 4K!!! This bold move from Onkyo was a great concept, but backfired. More of it can be found in audioholics.com:
Onkyo/Integra Create Upgradable Receiver... Haha, Just Kidding!
In 2005 Onkyo/Integra released their new TX-NR1000 and DTR-10.5 AV Receivers. It was a card-based system that took the industry by storm. It touted HDMI functionality, Net-tune support for Internet radio and, most importantly, to be "The Future-Proof Anchor to Your Entertainment Needs" - a platform that could be updated simply by purchasing new cards from Onkyo/Integra. It seemed like a dream come true, and many folks flocked to the company to pick up this AV receiver with a base price of $3500.
All the latest encoding formatsTHX Surround EX, DTS 96/24, Dolby Pro Logic IIx, and moreare here along with the knowledge that youre ready for the future with upgradable flash memory
Of course, there was a problem, er... more than one, actually. You see, first off, the marketing person and product manager who thought that building a single product that was upgradable was a great business decision must have been on some heavy drugs that day... as were any persons responsible for giving them the green light. You see, while Onkyo was touting the genius of a receiver that never required replacement, some accountant up at headquarters was scratching his head saying "What?!?!" That's right, while Yamaha and Denon were on 9 month new product release cycles, Onkyo was about to trade in its product cycle for add-on or replacement "card" purchases. Not exactly what we'd call a smooth business decision - nor one that could be sustained for any length of time due to new formats and processing requirements (room EQ, anyone?).
At the same time, technology went through an incredible and fast expansion from HDMI 1.0 to HDMI 1.3 in just a short period of time. Silicon for HDMI 1.3 was particularly problematic - and Sony buying up every available piece for its overpriced, undersoftwared (is that a word?) PS3 didn't help. AV manufacturers were stuck in limbo and the new emerging technologies such as automatic room EQ systems, Dolby Digital Plus, Dolby TrueHD and DTS-HD made mincemeat of existing platforms and their audio processing chipsets.
As for firmware upgrades? Well, the Onkyo TX-NR1000 and Integra DTR-10.5 could ONLY be upgraded at a service center. Users could not download firmware updates online from the Onkyo website. And within a year, iMerge, the Internet Radio service available through Net-Tune capable Onkyo brand products, announced that it discontinued the iMerge service and would no longer be supporting Internet radio streaming or the Internet Radio database. Internet Radio stations available through Net-Tune would decrease and eventually cease.
As the expected occurred, Onkyo apparently realized their mistake and scrambled to their next product... all the while product managers were left making empty promises that would later get pulled out from under their feet. Take the HDMI 1.2a card that was promised (though not officially) "in late summer" of 2006. Around mid-September it was finally leaked that there would indeed be no card, but that the company would release an HDMI 1.3 card sometime "next year".
After waiting until mid-2007, users were greeted with this emailed response:
Sorry, at this time we have no info on when the card will be out suggest contacting us at a later date for more info.
Soon after that the TX-NR1000 and DTR-10.5 were removed form the Onkyo and Integra websites... eliminating all hope for an HDMI upgrade of any kind. As it turns out, Onkyo was saved from being completely decimated with false advertising claims thanks to releasing a single new card after the product debuted... an HD-Radio/XM card.
To this day no other manufacturer has even hinted at the prospect of a card-based system.
Link:
The Twelve Biggest Industry Mistakes of the Digital Age
In the end how much we thrash the TX-NR1000, it is still a great AV amplifier the best Onkyo ever built. Its flexibility also extended with two separate sets of 7.1 channel speaker outs for two separate zones. A 32-bit Aureus TMS320DA610 from Texas Instruments is used for Zone A. These are two times as powerful as SHARC Melody ULTRA, hence only one is used to perform both Dolby/ DTS decoding as well as post-processing like THX, surround simulation effects and bass management on Zone A. Consider a situation where out of seven powered channels of TX-NR1000, five are used in a 5.1 channel Zone A setup while the rest two channels are used up in Zone B stereo setup. Now if someone is watching a DTS encoded movie in Zone A, the above Aureus DSP gets busy at work. So what will happen if someone simultaneously wants to hear a DTS encoded CD on Zone B? Another DSP will be needed right? Thus another 32-bit Aureus TMS320DA610 handled the task of decoding the DTS encoded CD for Zone B simultaneously with Zone A. So TX-NR1000 came with true simultaneous multi-zone support, though it is a pity with the power of Aureus no auto acoustic calibrations are implemented even though DSPs have full support for them.
When operational through 7.1 channel analog inputs, two 96 kHz/24-bit AKM AK5384 quadraphonic ADCs in single-ended configuration are used to feed any one of the Aureus with pre-selected Zone. These ADCs are of very basic types having dynamic range of just 102 dB (DSP-Z9 did 114 dB) and are pre-configured to work at a constant rate of 96 kHz. Onkyo realizing these multi-channel ADCs being not good enough implemented a separate pair of higher quality 192 kHz/24-bit AKM AK5385 stereo ADCs with 114 dB dynamic range for stereo analog inputs and Zone B analog stereo inputs. The 8-channel digital audio signal from either of these DSPs (Zone A or Zone B) is then sent to four 192 kHz/24-bit Wolfson WM8719 stereo DACs in single-ended configuration delivering 109 dB dynamic range (DSP-Z9 did 129 dB). So the entire 7.1 channel chain fails to meet the minimum target of 110 dB dynamic range as set by other Ubers. For I/V conversion stages from these DACs, following op-amps are used: mPC4570 followed by mPC4570. These are unknown op-amps. In the end the 8-channel analog audio signal from I/V conversion stages are digitally regulated by eight individual Wolfson WM8816 stereo volume controllers in differential configuration on all channels.
To me TX-NR1000 is a surplus of everything, multiple ADCs for the same audio channels with varying performance, three additional DACs along with the above four which should have been used in dual-differential configuration for fronts and center channels thus finally breaking 110 dB barrier, unknown op-amps and multiple DSPs for decoding with nothing exceptional other than real multi-zone support. The ADCs, DACs and op-amps are nothing to write about while the powerful DSPs are kept wanting with no acoustic calibration implemented. In the end it seems the tight coupling and optimized maximization of all its resources (electronic components) to gain maximum performance is missing as compared to other Ubers, and surplus of parts are used to do the same job. Sound quality is a different matter though, it should never be judged by hardware implementation.
---------------------------------------------------------------------------------------
Yamaha DSP-Z9
Digital Audio Board
My DSP-Z9: Digital Audio Board
Digital Audio Chain
The digital audio board of DSP-Z9 boasts the most powerful DSP section in comparison to above three Uber implementation. All DSP IC chips and related circuitry are located on this 4-layered digital audio board which makes dimensions smaller leading to lower digital interference and impedance. It is also implements the most exotic ADCs, DACs and other electronic components in this circuit board. But first you have to understand what Yamahas Digital Sound Field Processing actually means. AV amplifiers from other makes base their sound on Dolby/ DTS decoding, using matrix and steering technologies to create surround sound effects and then perform post processing to recreate simulated environments that are not obtained from real venues with real world data. Yamaha Digital Sound Field Processing on its first phase also does Dolby/ DTS decoding but then in post-processing phase draws a huge amount of digital data representing real-world data obtained by visiting, measuring and collecting from actual venues (say Rock Concert HiFI DSP mode was drawn from real rock concert venue). This valued propriety data is stored in internal banks of memory (ROM chips) adjacent to specially developed DSP microcomputers by Yamaha which process this data. DSP-Z9 then process this data and is able to recreate the actual acoustical characteristics of live performance environments by digitally reproducing the depth, imaging and spatial of those environments using recorded reverberation and echo patters. In order to make the sound-fields more realistic they not only post-processes the 7.1 channels with their processed data but also adds front presence speakers to create additional depth. Yamaha has a lot data that they collected from different venues, so it is a question of how powerful the Yamaha DSPs are, how many of them are used in tandem and how much they can bite on this data to perform a denser processing so that the effects are more realistic. With DSP-Z9 being their first Uber, the first thing they did was to throw away DSP-AZ1s refined digital audio board. With DSP-Z9s focus on pure audio fidelity they built from scratch a new Digital ToP-ART design worthy of HiFi pedigree. They also wanted to implement their own auto acoustic calibration technology (YPAO) for the first time with DSP-Z9.
Decoding DSP -> SHARC Melody ULTRA ADSST-21161 + "Sound-field" Processing -> YSS-930
Eight Yamaha YSS-930 in tandem (Post Processing + YPAO)
For decoding Dolby/ DTS/ THX the high-end version 32-bit SHARC Melody ULTRA ADSST-21161 from Analog Devices is used and sends 8-channel decoded data for post-processing, where it gets more interesting. Yamaha generally used a pair of latest available Yamaha LSis for their sound field processing in their TOTLs. But DSP-Z9 being their first Uber ended up with four of their latest 32-bit Yamaha YSS-930 capable of handling 96 kHz sampling frequencies. This YSS-930s in quad configuration have three times more computational power over dual YSS-910 implementation in DSP-AZ1. Also YSS-910 is able to handle only up to 48 kHz sampling frequencies. Thus with DSP-Z9, Yamaha is able to obtain three times more computational power at twice the sampling frequency which resulted in six times denser Digital Sound Field Processing as compared to DSP-AZ1. This was a very big jump for DSP-Z9 over previous DSP-AZ1. So the first owners of DSP-Z9 immediately reported how realistic and more natural the sound field effects are in comparison to previous DSP-AZ1. It was all due to six times denser sound-field processing of DSP-Z9 allowing it to gulp gobs of data stored in memory banks and process in real time.
Six times denser DSP Processing
After post-processing this 11 channel digital audio signal (7.1 + 2 front presence + 1 additional subwoofer) arrive to their newest invention YPAO. For their proprietary auto acoustic calibration they included another four numbers of YSS-930 LSis in tandem after post-processing phase in the digital chain with complete bass management included. In doing so they ended up with the highest number of DSP processors ever used in any AV amplifier, nine in total, a SHARC followed by quad YSS-930 LSis for sound-field processing followed by quad YSS-930 LSis for their own auto acoustic calibration. When operational through 7.1 channel analog inputs, four high quality 192 kHz/24-bit Cirrus Logic CS5361 stereo ADCs in single-ended configuration are used to feed these DSPs. These ADCs capable of delivering 114 dB dynamic range is pre-configured to work at a constant rate of 96 kHz.
ADC -> Cirrus Logic CS5361
DAC -> Burr-Brown PCM1792
Op-amps: Analog Devices OP275 and JRC NJM2068
Finally the 11-channel digital audio signal from DSPs is then sent to six ultra-exotic 192 kHz/24-bit Burr-Brown PCM1792 stereo DACs. These are the highest performing DACs from Burr-Brown while being the most expensive implementation in any AV amplifier till date. 4x as expensive as Denon or Onkyo implementation and 2x as expensive as Pioneer implementation. Five of these PCM1792 are used for fronts, surrounds, surround backs, stereo subwoofers and front presence channels in single-ended configuration to obtain sky high 129 dB dynamic range while the center channel is in exotic dual-differential configuration obtaining 132 dB dynamic range. For I/V conversion stages from these DACs, following op-amps are used: Analog Devices OP275 followed by JRC NJM2068. To retain this level of analog signal purity from these DACs I/V conversion stages, Yamaha developed their own stereo volume controller Yamaha YAC-520. Nine in total, seven of those were used in dual-differential configuration for the main seven channels (fronts, center, surrounds, surround backs) to offer higher performance, the rest two being used in single ended configuration for two subwoofers and two front presence channels. It is also to note the very best and expensive Nichicon Gold Tune electrolytic capacitors were used all around along with military grade solid capacitors in this board.
---------------------------------------------------------------------------------------
Summary
With all these objective views of different digital audio board implementation on various Ubers the natural conclusion drawn will be DSP-Z9 is the best. DSP-Z9 is using the most expensive hardware while an AVC-A1SRA is using pretty modest ones which lead to clear conclusion, at least objectively: Yamaha is way better than Denon. In real world though, this is not the case. Sound quality I believe is a subjective element, and as such very simplest of audio implementations can sound very good in my ears. In reality when I was exposed to both these Ubers the Denon was just there with Yamaha DSP-Z9, the DSP-Z9 having a slight edge when it came to audio fidelity. So my point is, it is good to understand the technology that went behind making of these Uber AV amplifiers, but never to draw a conclusion as of which is best in terms of audio fidelity until you have heard and compared them with the same or various setups. This is what I have learnt.
