This is a review and detailed measurements of the Emotiva RMC-1 AV Surround Sound Processor (for home theater). It is on kind loan from a member. The RMC-1 costs US $4,999 from the company direct. It is part of a new generation of high-end processors sporting 16+ channels of audio.
I like the serious, ready to go in an AV rack look of the RMC-1:
Despite the massive size, it is not too heavy to move around. The blue LED (OLED?) is a good choice for a dark theater.
The back panel sports the connectivity of a modern processor, leaving behind all the legacy connectors (component, composite video, etc.):
The good news unfortunately stops once you go past the mechanicals and look of the unit. Boot time is long and even waking up from standby is slow. Cycling inputs through inputs can be painfully slow as the unit seemingly tries to lock onto unused inputs. Just about every change produces a "Please wait" logo on the right. I don't mind things taking a bit longer but not being to navigate the UI makes no sense to me.
The remote control is a substantial metal box. It makes the buttons easy to find but it has no backlight. How can you use this in a dark theater??? Best to invest in a control system or unified remote than use the included one.
The owner had installed version 1.7 which was the last official version. That version proved very buggy with serious performance problems to boot. A little bird told me there is a version 1.8 that is a "total rewrite" and gave me a link to install. That did improve things including a few serious performance problems. He also passed on the message that the company wanted to talk to me before I publish the review. So I fired off an email to them last Friday and sat on the review until now (late Monday). No response back. This being an expensive and large unit that is taking a lot of space in my lab, I decided to proceed regardless. If the company has feedback, they are welcome to share it now.
There are still some operational bugs such as Reference Stereo not producing audio most of the time. I would have to switch back and forth to direct or even restart the unit to get it to produce audio.
Another disappointing thing is the OSD (on-screen display). Instead of using the entire display and show the unit status at a glace, it just replicates the two line menu. Really? Talk about taking a poor shortcut here. The menu as is, can be tedious to use. For example to activate a mode, you don't just push the center/enter button. Instead you have to move to another suboption to select a checklist. Never seen anything like this.
Since everyone uses these processors with HDMI input, I tried to use that output from my PC and yet again failed to get any output as with the last high-end processor I tested. I went through three video outputs from my PC (Nvidia, native Intel and a cheap USB to HDMI desktop extender) and three different monitors. All monitors would show video directly from the PC but once routed through RMC-1, nothing would come out. Windows would detect that RMC-1 was connected and was the "display type" but as noted, I could not get any video output. I am wondering if latest versions of HDMI chipset in these processors has broken compatibility with PC ecosystem somehow. I will keep trying to diagnose this problem. For now, all my testing is performed using AES/EBU digital input (balanced version of S/PDIF).
EDIT: An updated Version 1.9 firmware fixed some core issues that were dragging performance down. Please see here: https://www.audiosciencereview.com/...a-rmc-1-av-processor-review.11673/post-344170
AV Processor Audio Measurements
As usual, our goal here is to check the main audio pipeline to see if it is well engineered. It starts by feeding the unit a 1 kHz tone at 24 bits and see how low the distortion+noise is. The RMC-1 uses a high performance DAC from AKM:
Here is the specs for AKM4490:
This is what I get when tested in Reference Stereo mode which bypasses all DSP including bass management:
SINAD in this display is the same as THD+N expressed in dB and positive. This means that the AKM4490 datasheet's SINAD is 112 dB yet we are just getting 92 dB out of RMC-1. 20 dB of noise and distortion is left on the table due to suboptimal implementation.
Switching to Direct mode which leaves bass processing active reduces performance even more:
And it is not just harmonic distortion and noise. Notice the dual spikes around 300 Hz. What the heck are these? Their frequency changes with sample rate so it has something to do with that. But what? Notice that they exist in both Direct and Reference Stereo.
Here is just the FFT by itself, this time with 48 kHz sampling (which is very common in video content):
This is a seriously broken implementation of AKM4490. Since most people want to use bass management, you are losing nearly 27 dB of performance. Engineers often kill themselves to squeeze 3 dB better performance out a design so 27 dB is huge.
Anyway, going with the better number of the two, here is where the unit ranks:
And within 18 processors and AVRs tested so far:
I know, you are thinking the output level is too high so let's look at a new test that shows us the distortion+noise relative to output level:
This is a head scratcher. The Reference Stereo mode (green) improves with increasing output level to 4 volts. The Direct mode works backward, besting the Reference Stereo at lower output levels but then declining at full 4 volts. Some kind of gain scaling must be behind this odd behavior. Either way, our 4 volt output is not stressing the unit. It can't produce better than 92 dB SINAD no matter which mode or output level you choose.
You want to get depressed? Check out what happens when we measure intermodulation distortion relative to input level (digital sample values):
Sigh. We start with excellent, low noise performance rivaling our good desktop DAC reference (Topping DX3 Pro ~$250). But then at just -35 dB, distortion takes over and the RMC-1 loses to a little phone dongle from Razor which retails for US $25. Why, why, why???
Hey, maybe it handles jitter well. Let's find out:
Nope. Nothing other than main spike at 12 kHz should be there. Instead we have an array of sidebands indicating very clear, and correlated jitter component. As you see, the jitter remains whether I used AES or Optical input.
There are some bright spots like dynamic range:
Linearity could that too except that just sitting there, it changes its performance from near excellent to a step below:
Even filter performance is odd:
What is with that high level of ripple? I could not find a filter mode in AKM4490 that has this kind of response.
Turns out we are not done depressing ourselves. Check out THD+N versus frequency:
A $99 DAC is in dashed blue. Forget that we don't match that curve. Notice how distortion rises with lowering of frequency. Best DACs have no frequency dependency. Those that do have rising distortion with frequency as you see in Blue. But the RMC-1 has the opposite problem, typically indicative of poor power supply design. In Direct mode distortion rises to whopping 0.1% THD+N! That is a 25 dB drop from our reference 1 kHz tone we used in the dashboard!!!
Let's dig into this further:
We see that distortion rises with level (lower worse). Compare this with the graph I post earlier that was at 1 kHz. The variation there was much smaller.
We can see the frequency dependence by looking at the full spectrum:
Notice how the harmonic peaks of the 20 Hz (in red) are much higher than those of 10 kHz tone (in blue).
Conclusions
This is a stunning failure for a high-end AV Processor. Usability is not great, and bugs are easy to hit on. The worse news by far though is the measured performance of the pipeline from digital input to analog output. Despite using a premium DAC chip, the measurements paint a very ugly picture. This is a hardware design problem and not fixable with firmware updates and such.
Coming on the heals of another failed high-end processor, I am starting to think there is not anyone performing these basic performance checks before the design is released to the public. And not enough expertise is applied to key analog (and gain management) subsystem.
Even the DSP processing seems faulty. Why lose so much performance because you turn on the DSP? IN this day age, there should be no issue whatsoever to use high resolution internal representation to avoid screwing up 24 bit audio samples. Now, in a $300 AVR maybe that is a stretch but in a $5,000 processor?
Really, these companies would be in better shape if they bought a $1 DAC that has a SINAD of 95 dB and get that kind of performance. The $9 Apple dongle does that. Why put in an expensive DAC chip and incur the huge hit to the BOM (bill of material) if you are not remotely able to get its performance out the door from the XLR connectors?
If no one is left in the industry, ask me to test for heaven's sake. I take no joy in writing these reviews. Both companies know who have built these high-end processors know where to find me. I am not looking for work but please, one of you needs to build a decent high-end processor that we can be proud of. Don't let this industry sink backward anymore. With video processing no longer needed or included, let's get audio right.
Anyway, it should be obvious that I cannot recommend the Emotiva RMC-1. Buy it for features if you can live with its bugs. But don't think you have a high-end Audio product.
------------
As always, questions, comments, recommendations, etc. are welcome.
Had tons of rain yesterday and winds. Woke up this morning with one of our massive 40+ foot trees had fallen across our drive away and onto poor neighbor's roof! Everyone is safe thankfully but I am needing good chunk of money to become happy again after dealing with this all day. So please donate what you can using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
I like the serious, ready to go in an AV rack look of the RMC-1:
Despite the massive size, it is not too heavy to move around. The blue LED (OLED?) is a good choice for a dark theater.
The back panel sports the connectivity of a modern processor, leaving behind all the legacy connectors (component, composite video, etc.):
The good news unfortunately stops once you go past the mechanicals and look of the unit. Boot time is long and even waking up from standby is slow. Cycling inputs through inputs can be painfully slow as the unit seemingly tries to lock onto unused inputs. Just about every change produces a "Please wait" logo on the right. I don't mind things taking a bit longer but not being to navigate the UI makes no sense to me.
The remote control is a substantial metal box. It makes the buttons easy to find but it has no backlight. How can you use this in a dark theater??? Best to invest in a control system or unified remote than use the included one.
The owner had installed version 1.7 which was the last official version. That version proved very buggy with serious performance problems to boot. A little bird told me there is a version 1.8 that is a "total rewrite" and gave me a link to install. That did improve things including a few serious performance problems. He also passed on the message that the company wanted to talk to me before I publish the review. So I fired off an email to them last Friday and sat on the review until now (late Monday). No response back. This being an expensive and large unit that is taking a lot of space in my lab, I decided to proceed regardless. If the company has feedback, they are welcome to share it now.
There are still some operational bugs such as Reference Stereo not producing audio most of the time. I would have to switch back and forth to direct or even restart the unit to get it to produce audio.
Another disappointing thing is the OSD (on-screen display). Instead of using the entire display and show the unit status at a glace, it just replicates the two line menu. Really? Talk about taking a poor shortcut here. The menu as is, can be tedious to use. For example to activate a mode, you don't just push the center/enter button. Instead you have to move to another suboption to select a checklist. Never seen anything like this.
Since everyone uses these processors with HDMI input, I tried to use that output from my PC and yet again failed to get any output as with the last high-end processor I tested. I went through three video outputs from my PC (Nvidia, native Intel and a cheap USB to HDMI desktop extender) and three different monitors. All monitors would show video directly from the PC but once routed through RMC-1, nothing would come out. Windows would detect that RMC-1 was connected and was the "display type" but as noted, I could not get any video output. I am wondering if latest versions of HDMI chipset in these processors has broken compatibility with PC ecosystem somehow. I will keep trying to diagnose this problem. For now, all my testing is performed using AES/EBU digital input (balanced version of S/PDIF).
EDIT: An updated Version 1.9 firmware fixed some core issues that were dragging performance down. Please see here: https://www.audiosciencereview.com/...a-rmc-1-av-processor-review.11673/post-344170
AV Processor Audio Measurements
As usual, our goal here is to check the main audio pipeline to see if it is well engineered. It starts by feeding the unit a 1 kHz tone at 24 bits and see how low the distortion+noise is. The RMC-1 uses a high performance DAC from AKM:
Here is the specs for AKM4490:
This is what I get when tested in Reference Stereo mode which bypasses all DSP including bass management:
SINAD in this display is the same as THD+N expressed in dB and positive. This means that the AKM4490 datasheet's SINAD is 112 dB yet we are just getting 92 dB out of RMC-1. 20 dB of noise and distortion is left on the table due to suboptimal implementation.
Switching to Direct mode which leaves bass processing active reduces performance even more:
And it is not just harmonic distortion and noise. Notice the dual spikes around 300 Hz. What the heck are these? Their frequency changes with sample rate so it has something to do with that. But what? Notice that they exist in both Direct and Reference Stereo.
Here is just the FFT by itself, this time with 48 kHz sampling (which is very common in video content):
This is a seriously broken implementation of AKM4490. Since most people want to use bass management, you are losing nearly 27 dB of performance. Engineers often kill themselves to squeeze 3 dB better performance out a design so 27 dB is huge.
Anyway, going with the better number of the two, here is where the unit ranks:
And within 18 processors and AVRs tested so far:
I know, you are thinking the output level is too high so let's look at a new test that shows us the distortion+noise relative to output level:
This is a head scratcher. The Reference Stereo mode (green) improves with increasing output level to 4 volts. The Direct mode works backward, besting the Reference Stereo at lower output levels but then declining at full 4 volts. Some kind of gain scaling must be behind this odd behavior. Either way, our 4 volt output is not stressing the unit. It can't produce better than 92 dB SINAD no matter which mode or output level you choose.
You want to get depressed? Check out what happens when we measure intermodulation distortion relative to input level (digital sample values):
Sigh. We start with excellent, low noise performance rivaling our good desktop DAC reference (Topping DX3 Pro ~$250). But then at just -35 dB, distortion takes over and the RMC-1 loses to a little phone dongle from Razor which retails for US $25. Why, why, why???
Hey, maybe it handles jitter well. Let's find out:
Nope. Nothing other than main spike at 12 kHz should be there. Instead we have an array of sidebands indicating very clear, and correlated jitter component. As you see, the jitter remains whether I used AES or Optical input.
There are some bright spots like dynamic range:
Linearity could that too except that just sitting there, it changes its performance from near excellent to a step below:
Even filter performance is odd:
What is with that high level of ripple? I could not find a filter mode in AKM4490 that has this kind of response.
Turns out we are not done depressing ourselves. Check out THD+N versus frequency:
A $99 DAC is in dashed blue. Forget that we don't match that curve. Notice how distortion rises with lowering of frequency. Best DACs have no frequency dependency. Those that do have rising distortion with frequency as you see in Blue. But the RMC-1 has the opposite problem, typically indicative of poor power supply design. In Direct mode distortion rises to whopping 0.1% THD+N! That is a 25 dB drop from our reference 1 kHz tone we used in the dashboard!!!
Let's dig into this further:
We see that distortion rises with level (lower worse). Compare this with the graph I post earlier that was at 1 kHz. The variation there was much smaller.
We can see the frequency dependence by looking at the full spectrum:
Notice how the harmonic peaks of the 20 Hz (in red) are much higher than those of 10 kHz tone (in blue).
Conclusions
This is a stunning failure for a high-end AV Processor. Usability is not great, and bugs are easy to hit on. The worse news by far though is the measured performance of the pipeline from digital input to analog output. Despite using a premium DAC chip, the measurements paint a very ugly picture. This is a hardware design problem and not fixable with firmware updates and such.
Coming on the heals of another failed high-end processor, I am starting to think there is not anyone performing these basic performance checks before the design is released to the public. And not enough expertise is applied to key analog (and gain management) subsystem.
Even the DSP processing seems faulty. Why lose so much performance because you turn on the DSP? IN this day age, there should be no issue whatsoever to use high resolution internal representation to avoid screwing up 24 bit audio samples. Now, in a $300 AVR maybe that is a stretch but in a $5,000 processor?
Really, these companies would be in better shape if they bought a $1 DAC that has a SINAD of 95 dB and get that kind of performance. The $9 Apple dongle does that. Why put in an expensive DAC chip and incur the huge hit to the BOM (bill of material) if you are not remotely able to get its performance out the door from the XLR connectors?
If no one is left in the industry, ask me to test for heaven's sake. I take no joy in writing these reviews. Both companies know who have built these high-end processors know where to find me. I am not looking for work but please, one of you needs to build a decent high-end processor that we can be proud of. Don't let this industry sink backward anymore. With video processing no longer needed or included, let's get audio right.
Anyway, it should be obvious that I cannot recommend the Emotiva RMC-1. Buy it for features if you can live with its bugs. But don't think you have a high-end Audio product.
------------
As always, questions, comments, recommendations, etc. are welcome.
Had tons of rain yesterday and winds. Woke up this morning with one of our massive 40+ foot trees had fallen across our drive away and onto poor neighbor's roof! Everyone is safe thankfully but I am needing good chunk of money to become happy again after dealing with this all day. So please donate what you can using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
Last edited: