Trinnov Altitude 32 Review and Measurements
View attachment 373841
This is a review and some measurements of a Trinnov Altitude 32.
This is an older unit with components that are several generations old. What makes the Altitude 32 different from the Altitude 16 is the ability to use AES/EBU digital outputs for all channels. The Altitude "32" can be spec'd with as little as 8+4 channels or all the way up to 32+4 channels in a single box. The
x+4 means that you can choose "
x analog" and "
x <=16 digital" outputs or "
x analog and 4 digital outputs." Whereas the Altitude 16 only does 24/96, the Altitude 32 handles 24/192 kHz including full room and speaker EQ,
We've often seen super-expensive luxury audiophile products fall flat on their face when measured. Let's see how the Altitude 32 does...
Test Configuration
Trinnov Altitude 32
24+4 channel version
DAC board Version 1
Version 1: Burr-Brown PCM4104 based board (max output at 4V; unit tested)
Version 2: Tested by
@amirm here in the
Altitude 16. Reaches 6V, 100 dB SINAD; I believe also a Burr-Brown PCM4104
Version 3: ESS Sabre DACs
HDMI input board Version 2
Version 1: Momentum Data Systems HDMI 1.4
Version 2: Momentum Data Systems HDMI 2.0 (unit tested)
Version 3: Cypress HDMI 2.0
Version 4: Cypress HDMI 2.1
It is possible to upgrade the DACs or HDMI input boards as a paid upgrade.
Test configuration
I used the E1DA Cosmos ADC for analog testing and a Hifime S2 Digi SPDIF recorder to measure the SPDIF out from the TOSLINK connector. The generic optical cable that came with the Hifime only seemed to work at 96 kHz and I got dropouts at 192 kHz. TOSLINK is officially spec'd to 96 kHz only but I didn't try other optical cables since the Hifime uses a mini TOSLINK input. A HP Envy 15 laptop was used as the signal generator using a Monoprice HDMI cable to the Altitude 32.
@pkane and
@JohnPM 's free software Multitone and REW were used to generate the graphs.
Drumroll...
s'il vous plaît
149.1 dB
SINAD (digital AES, 96 kHz, no room correction)
View attachment 373821
Superbe!
Digital in, Digital Out. Of course the SINAD should be good, right? I've bypassed all processing. A common concern is that with many AVR and AVPs is that there needs to be some headroom for DSP processing. This is because there are different "circuit blocks" often from different vendors to handle anything from volume control to dialog normalization to room EQ. Preserving headroom is critical. Trinnov is different. They have a full software stack that runs on an x86 or ARM based general purpose CPU.
Let's first run some REW sweeps to see how the unit performs with room correction turned on.
View attachment 373822
The Trinnov Optimizer(TM) can be configured with all sorts of parameters. The resolution of the energy response can be corrected with just 1/3rd octave or all the way to 1/24 octave. You can adjust FIR lengths and change a ton of different settings to your choosing. Some sample screenshots of the configuration are shown here:
View attachment 373823View attachment 373825View attachment 373827View attachment 373828
So going back to the REW sweep, you can see that I ran a bypass which is a straight line, an "Optimizer Bypass" which removes room EQ but maintains the crossover for my main channel speaker, and then I have shown two sweeps with the Trinnov room/speaker correction turned on with the LFE off. The two different levels of correction are just to show you the graphs; you don't automatically want to boost the correction level to the max since it can introduce pre-ringing and exacerbate problems that from measurement artifacts rather than a need to correct speaker. With good speakers, correcting full range but limiting yourself to the broad peaks and dips generally works well on the Trinnov system.
You can basically see that there's no "excess" distortion even with room correction turned on to excessive levels of correction. Again, to be clear, this is an electronic measurement where I have used HDMI out from a laptop sending a 20Hz to 20 kHz linear sweep, let Trinnov apply its room EQ to correct my speaker and then measured what is being output via SPDIF out.
View attachment 373829
And if you like looking at percentages for distortion...
View attachment 373846
So how does the 1 kHz SINAD look when we have room correction turned on?
144.9 dB
SINAD (digital AES, 96 kHz, FULL Trinnov Room Correction including 3D remapping)
When you have room correction turned on, running the 1 kHz test tone still shows reference class performance.
There is essentially no performance loss turning on room correction.
View attachment 373830
Subjective Review
You cannot power on the Altitude 32 with the IR remote control. You have three options
1 - push the physical power button
2 - use Wake on LAN
3 - use 12V trigger
There is no HDMI-CEC support either, so you cannot power on the unit that way. It sounds crazy, but it's true. Only the Altitude 16 has IR remote power on. Being a hobbyist with an Altitude 32 is like buying a 2006-era Maserati Coupe or Quattroporte with its sand-casted naturally aspirated Ferrari V8 engine and an exotic F1 transmission system. Just because you can get a deal on the hardware doesn't mean that you get a deal on supercar maintenance. It would seem that the vast majority of Trinnov Altitude 32 owners run Control4/Creston-style systems and the need to power it on with a remote control or HDMI CEC is not a deal breaker. Even something like eARC/ARC can only be found in the very latest HDMI hardware revision. Love the AppleTV remote for controlling volume? Too bad. With all Trinnov's, you have zero HDMI-CEC support so you have to use the physical knob, the smartphone app, the web interface, or the IR remote. You do get volume control with 0.1 dB precision though. Interestingly, if I press the power button on my AppleTV, it is able to send the HDMI Power On command through the Trinnov to power on my display.
The Altitude 32, of course, sounds great. I do think it's superior to the HTP-1 with Dirac/DLBC in my room, although I do take advantage of the 3D remapping feature since my speakers are not in an ideal position. I get more tweaking and control with the Altitude 32 than I do with Dirac, so I may have tuned it to my preference over time. For example, my top-middle speakers only go down to 120 Hz. Most setups would just crossover to the LFE channel, which then gets distributed to all of my speakers. The problem is that I can localize the audio in the 80-120 Hz range. With Trinnov, I can use all of my subwoofers for the LFE content, but for bass management of my rear and height speakers, I can pick the nearest sub to individual channel to maximize the localization of the upper mid bass above 80 Hz. I could just spec speakers that have better bass response, of course. 3D remapping is a feature that I first experienced with my Sherwood R972 and it basically helps to reposition your speakers virtually into better positions. JBL/Harman aren't big believers of this and they actually remove this feature from the JBL SDP-75 but all I know is that it works for my setup. It's likely that most owners of the JBL SDP-75 also have JBL Synthesis or Revel speakers in a dedicated home theater room where the speaker locations are in the proper places.
The end result is that I really feel like all of the speakers disappear and it's as if the
mix is better. I've not had a big issue with the bass integration, although my understanding is that Trinnov has made several improvements to the algorithm over time so while I'm not getting the sophistication of Dirac Live Bass Control, I'm not getting the basic version that used to ship with Trinnov. Trinnov Waveforming is a future upgrade to the Altitude line which provides active control of the room acoustics, similar to Dirac ART but with added complexity and sophistication by requiring at least 4 subwoofers.
What about the analog audio SINAD?
It's only 90 dB! The second DAC revision tested by Amir got to 100 dB SINAD. The 3rd version with an ESS Sabre DAC should be above 100 dB, but if you're chasing SINAD, it seems like buying a bunch of external DACs is a good option too. (
Note: The rising DC in these graphs is an artifact of the E1DA Cosmos ADC)
View attachment 373835View attachment 373836
View attachment 373837
View attachment 373839View attachment 373840
Even though I
could use fancier DACs fed by the digital out of the Altitude 32, I'm satisfied running all active analog speakers with my 90 dB SINAD DAC and using AES/EBU for my subwoofer. Looking at the 1 kHz SINAD, we can see the noise is pretty low and the second and third harmonics are close to -105 dB. If you think about a real-world home environment and then movies that are mixed for 85 dB averages and 105 dB peaks, it's very likely that I never run into audible issues even at reference level. Coming from the Monolith HTP-1 and before that, the Yamaha CX-A5100, I've actually taken two steps back in SINAD performance and yet I'm happier with the end result. The difference is that going from the Yamaha CX-A5100 to the Monolith HTP-1 to the Altitude 32 has meant several leaps forward in room correction, and that ultimately has been the biggest improvement. Going over all the Optimizer settings would take hours, but one impressive feature is the
Excursion Curve feature, which lets you adjust the level of correction above and below the Schröeder frequency with greater granularity than the all-or-none approach of other EQ systems.
Reference Class
"Reference Class" or "SOTA" (State of the Art) gets thrown around a lot by audio enthusiasts. What we have in the Altitude 32 is arguably a reference "consumer" cinema platform for multichannel object audio, at least when it comes to the digital processing. By AVR standards, it's "vintage" -- it came out a decade ago in 2014! Ten years ago, a 55" OLED TV was $15,000. The analog stage of this older unit isn't up to par with what can be achieved with the latest technology, or currently shipping Trinnov units, but it's easy enough to buy external DACs if you have some combination of room and amplifier gain where higher SINAD matters.
In terms of audio quality, diminishing returns occur once you get to your first 5.1.2 Atmos AVR. The HTP-1 probably gives you 85-90% of the performance at a fraction of the cost and I bet the Sony STR-AN1000 receiver gives you 85-90% the performance of the HTP-1. We're well past the point of diminishing returns with the Trinnov Altitude 32, but it is nice to know that it is possible to have a processor with room correction that has no penalty in digital performance. In time, this level of performance should trickle down.
Edit: Thanks to Die Zwei pointing out that the Altitude 32 has 16 AES/EBU outputs, so you cannot run full digital if you go beyond 16 channels. Having the upgraded DACs can make a difference if your downstream amplifier and speakers benefit from the lower noise of the newer DACS.
