- Joined
- Dec 21, 2011
- Messages
- 4,991
So we already know MOTOTRBO subscriber audio AGC sucks. That was fixed in the other thread.
I have always noted the MOTOTRBO RMN5127C enhanced mic sounds very bassy and muffled on both analog and digital, when used with the XPR5000-series mobiles. I haven't tested on a 4000-series; I'll assume it's the same. I have a HMN4079D (XTL enhanced mic) which I used to do a comparison test with. The results indicated the XTL mic sounds much better than the RMN5127C. But why is that?
Internally, the mics are close enough to being the same. Both have a programmable microcontroller/USB chip.
(On black modular connector, A = ASTRO25/HMN4079D) T = TRBO/RMN5127C. Also note the 4079D does not support all button-functions of the 5550.)
The RMN5127C uses the ATMEL MEGA 16U2-AU. The HMN4079D uses the ATMEL AT43USB325E-AU. (Note: Supposedly the HMN4079 is now up to F-revision, so it's possible they're using the MEGA 16U2-AU). Both chips contain FLASH memory.
---
When the XPR5550 is first powered up, it looks for any IMPRES audio accessories connected to the radio. It appears when an IMPRES accessory is detected, the radio loads an audio profile (equalizer settings) from the mic itself. The radio then caches these parameters for the duration the radio is powered on.
I learned about the caching as I can power-up the radio with the HMN4079D attached, then disconnect it, and plug in the RMN5127C and the audio profile from the 4079D is retained. If I power cycle the radio, then it will load the RMN5127C profile, which sounds terrible.
For the heck of it, I also tried a non-IMPRES mic -- the HMN1090B (XTL potato mic). It's simply plug-and-play. No IMPRES settings are loaded. It sounds the best of all of them.
Attached are audio samples:
View attachment HMN1090B__RMN5127C_analog-test.mp3
View attachment HMN1090B__RMN5127C_digital-test.mp3
View attachment HMN4079D__RMN5127C__analog-tests.mp3
View attachment HMN4079D__RMN5127C__digital-tests.mp3
I'd really like to see Motorola evaluate the abysmal audio performance of the RMN5127C and come out with a fix. The mic is flashable, so there's no reason (besides perhaps custom hardware for flashing) it can't be updated to work properly. Motorola has issued firmware updates for IMPRES accessories in the past.
At the risk of sounding paranoid, I'm really starting to wonder if Motorola is purposely trying to sabotage the audio-quality of their MOTOTRBO products, so as to be inferior to that of their high-tier APX (P25) products. DMR and P25 Phase II use the AMBE+2 vocoder/codec at bitrate of 2450bps, with 1150bps for Forward Error Correction. (There are differences in the air interfaces, but at the codec level, they're identical). Why is the master mic gain set too high? Why does AGC perform so poorly? (Resolved in other thread).
There's absolutely no reason at all they should sound different. But it is apparent there's plenty of room for audio improvement in the MOTOTRBO product family. It's as if they haven't put any optimization or done any testing with the product.
I have always noted the MOTOTRBO RMN5127C enhanced mic sounds very bassy and muffled on both analog and digital, when used with the XPR5000-series mobiles. I haven't tested on a 4000-series; I'll assume it's the same. I have a HMN4079D (XTL enhanced mic) which I used to do a comparison test with. The results indicated the XTL mic sounds much better than the RMN5127C. But why is that?
Internally, the mics are close enough to being the same. Both have a programmable microcontroller/USB chip.
(On black modular connector, A = ASTRO25/HMN4079D) T = TRBO/RMN5127C. Also note the 4079D does not support all button-functions of the 5550.)
The RMN5127C uses the ATMEL MEGA 16U2-AU. The HMN4079D uses the ATMEL AT43USB325E-AU. (Note: Supposedly the HMN4079 is now up to F-revision, so it's possible they're using the MEGA 16U2-AU). Both chips contain FLASH memory.
---
When the XPR5550 is first powered up, it looks for any IMPRES audio accessories connected to the radio. It appears when an IMPRES accessory is detected, the radio loads an audio profile (equalizer settings) from the mic itself. The radio then caches these parameters for the duration the radio is powered on.
I learned about the caching as I can power-up the radio with the HMN4079D attached, then disconnect it, and plug in the RMN5127C and the audio profile from the 4079D is retained. If I power cycle the radio, then it will load the RMN5127C profile, which sounds terrible.
For the heck of it, I also tried a non-IMPRES mic -- the HMN1090B (XTL potato mic). It's simply plug-and-play. No IMPRES settings are loaded. It sounds the best of all of them.
Attached are audio samples:
View attachment HMN1090B__RMN5127C_analog-test.mp3
View attachment HMN1090B__RMN5127C_digital-test.mp3
View attachment HMN4079D__RMN5127C__analog-tests.mp3
View attachment HMN4079D__RMN5127C__digital-tests.mp3
I'd really like to see Motorola evaluate the abysmal audio performance of the RMN5127C and come out with a fix. The mic is flashable, so there's no reason (besides perhaps custom hardware for flashing) it can't be updated to work properly. Motorola has issued firmware updates for IMPRES accessories in the past.
At the risk of sounding paranoid, I'm really starting to wonder if Motorola is purposely trying to sabotage the audio-quality of their MOTOTRBO products, so as to be inferior to that of their high-tier APX (P25) products. DMR and P25 Phase II use the AMBE+2 vocoder/codec at bitrate of 2450bps, with 1150bps for Forward Error Correction. (There are differences in the air interfaces, but at the codec level, they're identical). Why is the master mic gain set too high? Why does AGC perform so poorly? (Resolved in other thread).
There's absolutely no reason at all they should sound different. But it is apparent there's plenty of room for audio improvement in the MOTOTRBO product family. It's as if they haven't put any optimization or done any testing with the product.
