The RMAA (RightMark Audio Analyzer) Source and Audio Device Measurement Thread [Overview of Measurements in Post #3, Tutorial in Post #2]

[pila405]

In your measurements it goes to -1.5db at 20KHz, while only at -0.15 on Ohm Image (both NL). How come?

 

[Samueru Sama]

  In your measurements it goes to -1.5db at 20KHz, while only at -0.15 on Ohm Image (both NL). How come?

Because it has a series inductor.
 

 
Edit: nvm, thought you where asking why high frequencies drop with a load. Chris already stated that is because of the aliasing filter of his PC.

 

[HiFiChris]

  In your measurements it goes to -1.5db at 20KHz, while only at -0.15 on Ohm Image (both NL). How come?

 

 
[...]

Fwiw: I think it's obvious by looking at a couple of my graphs, however for those who haven't noticed it: My sound card has got a steep low pass filter [starting] around 17-18 kHz, so obviously the number right at 20 kHz is not the actual deviation - my results are "only" accurate up to 18 kHz.
 

 

[pila405]

Oh, ok. Wasn't sure because it did pretty good job with the Shinrico SHD5

 

[HiFiChris]

  Oh, ok. Wasn't sure because it did pretty good job with the Shinrico SHD5

 
But as you can see with the SHD5, the FR is also suddenly steeply "cut off" above ~ 18 kHz as this is where my soundcard seems to have a steep lowpass filter. In reality, the FR curve would just linearly extend above 20 kHz without starting to drop around 18 kHz.

It really is no big deal though, as besides that, it measures flat, and this part of the frequency response is quite irrelevant anyway, along with that when realising that my soundcard will have a lowpass kicking in around 18 kHz, one can easily imagine how the FR would continue above it until 20 kHz.
 

 

[earfonia]

  TUTORIAL:
 
 
Requirements:

In theory, now one could start measuring right now, but it is recommended to check whether everything that is necessary is available:

Hardware:

computer (running Windows), audio input (-> soundcard/interface), 3.5 mm (male) to 3.5 mm (male) cable, 3.5 mm stereo to 2x 3.5 mm stereo y-split, DAP (, headphone amplifier in case you want to amplify the signal with hardware instead of with Audacity)

Software:

Audacity, Right Mark Audio Analyzer (RMAA)
 
Realisation:

Before beginning with the actual measurements, it is recommended to test whether the used equipment (soundcard/audio-interface) records signals linearly and therefore is good for making measurements
 
........
 
 
Congratulations - you can now perform simple measurements with RMAA.

 
Thanks a lot for this!
 
Btw, I'm looking for the 'verdict criteria', the last column in the summary table where it states the verdict, poor, average, very good, etc.
 
For example:

 
Do you have the criteria for those values? I'm curious what is considered poor, average, good, very good, or excellent.
Have been searching for those criteria, but couldn't find it so far.

 

[HiFiChris]

 
Btw, I'm looking for the 'verdict criteria', the last column in the summary table where it states the verdict, poor, average, very good, etc.
 
For example:

 
Do you have the criteria for those values? I'm curious what is considered poor, average, good, very good, or excellent.
Have been searching for those criteria, but couldn't find it so far.

 
Reasonable question.

All the manual gives is "For those unfamiliar with measured technical parameters a verbal mark is provided.".

Obviously the developers have have defined the range of what measured value translates to what verdict. "Excellent" for the dynamic range could (for example) be anything better than 94.99 dB (A), whereas anything between 90 to 94.99 could be "Very good".

I guess we will never know for sure unless someone pings RMAA's programmers, looks into the source code or approximates them by performing a large number of tests.
 

Anyway, the numbers and "verbal remark" are only reliable if the same soundcard, levels during the measuring process and loads are used, because every soundcard will perform differently and also show different results for the parameters that exceed the simple frequency response and frequency deviation measurement.

 

[earfonia]

I guess we will never know for sure unless someone pings RMAA's programmers, looks into the source code or approximates them by performing a large number of tests.
 
Anyway, the numbers and "verbal remark" are only reliable if the same soundcard, levels during the measuring process and loads are used, because every soundcard will perform differently and also show different results for the parameters that exceed the simple frequency response and frequency deviation measurement.

 
Ok. Probably if I decided to buy the Pro version I will ask them this question.
 
I guess with some recommendation on the level of measurement, load, and good audio interface, RMAA can be useful. I will do more comparison on how close RMAA result compared to my QA401. I currently only have HRT LineStreamer+ for ADC, hopefully, good enough.
Need some study to do comparable measurement setup.

 

[castleofargh]

RMAA into the average input will look very sad compared to what you can do with a QA401. knowing that the value you measured isn't the ADC itself is one big reassuring information ^_^. if I measured more than half a device a year I would definitely get one.

 

[hakuzen]

Walnut V2

DAP/Amp.

DAP, unloaded, and loaded with Pai Audio MR3 and other multi-BA: (note different scale to fit response to MR3): DAP, loaded with other kind of phones (dynamic drivers):



DAP LO (line out) vs PO (phones out): DAP LO, amplified by Topping NX1a (very low output impedance):




Amp mode, unloaded and loaded:
Zishan (DAC mode) LO + Walnut + **** UEs (1DD+1BA, 16Ω, 110dB, officially)



They are unusable for sensitive multi-BA phones.

Unloaded curve is far from being flat. Some people say they remind you 80's players; it must be due to the premature roll-off at both sides, lows and highs, together with a slight mid centric bump. Energetic, but I guess congested as well. Unloaded curve from line out shows less roll-off both sides, specially highs roll-off.
It sounds like 80's players, I must agree.

Measures (@ 1KHz, 0dBFS), using DMM (Zishan as source for amplifier mode, 1.874Vrms in-signal):
Output impedance: phones out, 100Ω!!!; line out: >3KΩ.
Max. output (not clipping / "low" distortion): phones out, 3.32Vrms (@64Ω); line out: 0.33Vrms; amplifier mode (+10.7dB, aprox.), 3.67Vrms (@64Ω).

Noise with MR3 is awful; it's like loud metallic machines; it's no hiss, it's some kind of leaking from power current.


Strange excessive high out impedance, low line out power, and high power (for the price) up to 3.67Vrms.

The amplifier, while powerful, is almost unusable. First, the bass roll-off is enormous [Edit: .. in the case of D14; no bass roll-off unloaded, when amplifying Zishan].
Second, the steps in volume potentiometer; usable volumes are located in a 45º arc of the wheel. Using a 1.874Vrms in-signal, it starts at 7.30h, 0.18 at 8h, 0.4 at 8.30h; 3 at 9h, 5 at 10h, 5.8 at 11h, and 6.1 to 6.4 from 12h to 5h (Vrms); so suitable volumes are from 8h to 9:30h; very difficult for volume fine adjusting.


Edit: added hiss info
Edit2: added more info about amplifier mode
Edit3: corrected max Vrms no clipping/distortion, added more multi-BA phones
Edit4: corrected potentiometer "hours"

 

[HiFiChris]

  Walnut V2
 
DAP/Amp.
 
[...]

 
That's... just plain awful!

Thanks, I've added it to the third post.

 

[hakuzen]

GAD mini A1
 
Amplifier (used Zishan as source, 1.874Vrms and 0.137Vrms lineout); made by Ghost Audio DYI (i guess).
 
Unloaded, and loaded with Pai Audio MR3 and iBasso IT03:                                             Unloaded, and loaded with other kind of phones:
 
   
 
Measures (@ 1KHz, 0dBFS), using DMM (Zishan as source, 1.874Vrms):
Output impedance: 15Ω.
Max. output unloaded: 6.18Vrms (measured near 5.7Vrms before it starts distorting a bit, at 64Ω).
 
Our critical phones don't like that output impedance. It even affects a bit to some low impedance dynamic drivers.
Unloaded curve is very flat and extended.
 
Hiss with MR3 is low. 9h-11h, minimum; 11h-3h, low hiss; from 3h, noticeable (you can't listen to MR3 beyond 12h!).
But while you are touching the volume potentiometer, a loud noise is generated.
 
Volume steps: 0.18 at 9h; 0.8 at 10h; then, +1Vrms each hour (2.8 at 12h, 5.8 at 3h); 6.18 at 5h (Vrms).
 
Powerful amplifier (battery supplied), with very nice response except for sensitive multi-BA iems.
 
Edit: forgot to mention that left and right are inverted in my unit, don't know why.

 

[hakuzen]

Zishan DSD AK4495 (stock version)

DAP/DAC.

Unloaded, in modes "Sharp roll-off" (FR), "Short roll-off" (SR), and "Super Slow roll-off" (SSR):



Unloaded, phones out, line out, DAC modes "Sharp roll off" and "Short Delay Sharp roll-off" (phones out):


Open image in a new tab to get detail; used this format because all responses are very similar and wanted to show the difference between modes "Sharp roll-off" and "Short Delay Sharp roll-off" (yellow one, you can distinguish its corners). I haven't investigated deeper in such difference, nor its relation with ASIO configuration for the DAC (by the way, it was recognized by the computer -Win7- and by ASIO4ALL automatically). Difference between "Slow roll-off" and "Short Delay Slow roll-off" modes in DAC mode, are analogue.

Modes FR (loaded with Pai Audio MR3 and iBasso IT03), SR (w/MR3), and SSR (w/MR3); and loaded with other kind of phones:






"Turbo" mode (high gain) results i've tried, are practically identical to these; no extra differences related to impedance along frequencies.

Measures (0dBFS), using DMM:
Output impedance: phones out, 0.2Ω (@1KHz), 0.5Ω (@120Hz), 0.9Ω (@60Hz), due to output coupling caps; line out: 450Ω (@1KHz).
Max. output (not clipping / "low" distortion):
- phones out: high gain, 2.74Vrms (@20Ω), 3.64Vrms (@33Ω), 3.75Vrms (@64Ω, or unloaded); low gain, 0.276Vrms
- line out, fixed: high gain, 1.885Vrms; low gain, 0.138Vrms
Large power difference between normal and "Turbo" mode. Normal for most IEMs and some cans, high gain when difficult to drive phones.

Hiss with MR3 is low. 9h-11h, minimum; 11h-2h, low hiss; from 2h, noticeable (anyway, you can't listen music with MR3 beyond 12h, too loud!).
While you are touching the volume potentiometer, some noise is generated.

Everything is alright with this DAP/DAC, unloaded flat extended response, even impedance variation for critical IEMs is acceptable, except for the coupling capacitors, responsible of the high pass filter and so the bass roll-off when using every phone i've used (it seems proportional to the phones impedance: lower impedance, higher bass roll-off). It's not a big roll-off anyway (with a 16Ω load, -1dB at 40Hz, -1.75 at 30Hz, -3dB at 20Hz; with 8Ω, it goes very worse, though). It uses two Elna 470uF, 10V, caps. Solder skilled people could try to replace the caps, if they fit into the Zishan: I think 6mm(diameter) x 24mm(length) or 8mm x 15mm, should fit (stock ones seem to be 6.3 x 12, aprox.).

Edit: used stock Zishan, LME49720NA as LPF, but also Muses02 as LPF; of course, this didn't affect the coupling caps action in lows, and i didn't notice any difference in the obtained curves.
Edit: added further info.

 

[macky112]

hi all,
 
I just tried RMAA for the first time, and am not sure if I did it correctly, so comments/suggestion welcomed.
 
all DAPs has EQ turned off, all DAPs volume set around 80%, laptop mic-in volume set to 100% and pre-gain at 0%.
 
I dont have a 3.5mm Y-splitter, so I am only able to get the unloaded chart.
 
Walnut at 50% volume (12 o clock position, with 6 o clock being off), laptop mic-in volume adjusted to 20% due to clipping
 

 
Walnut at 25% volume (9 o clock position), does this chart illustrate channel imbalance at this volume?
 

 
Fiio X3ii
 

 
AGPTek M20 (Benjie S5)
 

 
Ruizu X02 (AGPTek A02)
 

 
here is the info table for all 4 above DAPs
 

 

[Samueru Sama]

  hi all,
 
I just tried RMAA for the first time, and am not sure if I did it correctly, so comments/suggestion welcomed.
 
all DAPs has EQ turned off, all DAPs volume set around 80%, laptop mic-in volume set to 100% and pre-gain at 0%.
 
I dont have a 3.5mm Y-splitter, so I am only able to get the unloaded chart.
 
Walnut at 50% volume (12 o clock position, with 6 o clock being off), laptop mic-in volume adjusted to 20% due to clipping
 

 
Walnut at 25% volume (9 o clock position), does this chart illustrate channel imbalance at this volume?
 

 
Fiio X3ii
 

 
AGPTek M20 (Benjie S5)
 

 
Ruizu X02 (AGPTek A02)
 

 
here is the info table for all 4 above DAPs
 

 
Unloaded measurements don't say much about the real performance of the device. The only thing that can be said of those is that the walnut sucks.
 
And the channel imbalance is because of the potentiometer of the walnut. Usually you want it to be under 1dB from its max output voltage (the potentiometer all the way up) to -40 dB in reference to the max output voltage, you can do this with the calibration signal of rightmark, since it reports the amplitude of both channels, however its likely you will saturate your line in with the walnut at full volume, and really the thing already has other problems of major concern. Like a 100 Ohm output impedance.