I'm and my friend needs an external USB sound card which among its direct purpose should be used with Millett Audio Interface and Cordell distortion magnifier (loopback mode speaking in programmer's term).
Soundcard Interface
CordellAudio.com - The Distortion Magnifier
This should work yet if anyone have any cons, please share.
Thanks in advance.
Soundcard Interface
CordellAudio.com - The Distortion Magnifier
This should work yet if anyone have any cons, please share.
Thanks in advance.
Sorry. I forget to mention in my previous post that we are going to use Aurora with AddOn B, which, according to datasheet have:
* 8x XLR analog output, can also be used unbalanced (RCA)
* 8x RCA analog input
* 8x XLR analog input
* 4x RCA digital input (S / P-DIF via coaxial cable)
* 4x TOSlink digital input (S / P-DIF via fiber optic cable)
The only question remains how well Aurora works in loopback mode. 2 pcs with "Add-on B" are quite expensive, almost 600 Euro.
* 8x XLR analog output, can also be used unbalanced (RCA)
* 8x RCA analog input
* 8x XLR analog input
* 4x RCA digital input (S / P-DIF via coaxial cable)
* 4x TOSlink digital input (S / P-DIF via fiber optic cable)
The only question remains how well Aurora works in loopback mode. 2 pcs with "Add-on B" are quite expensive, almost 600 Euro.
I suggest wait review from audiosciencereview.com
After in my opinion, even if I order one device with addonB, I will prefer something more modular. ADC DAC DSP mCPU separate.
After All-in-one it's good, faster to finish your project.
After in my opinion, even if I order one device with addonB, I will prefer something more modular. ADC DAC DSP mCPU separate.
After All-in-one it's good, faster to finish your project.
Last edited:
Has someone from ASR actually sent an Aurora to Amir to measure? I've seen people talking about it but I have yet to see someone either measure it or send it in for review.
I have some equipment that can be used for measurement of the ADC/DAC performance (Victor oscillator & 1kHz notch). If anyone's interested I can run a few tests. I breifly measured it in loopback before and it performed close to the specifications of the ICs (though the voltage output is smaller than the 4vrms standard Amir uses).
As for anything more substantial than pure specs, it mostly comes down to your specific implementation. Aurora is simply a hardware platform you can add whatever you want to, and the performance of the peripherals is dependent on your skill. The volume control for instance is a great way to introduce tons of modulation if you're not careful how you implement it.
The volume control on the ADAU1452 is a 10 bit ADC, meaning it can resolve ~3.2 mV of signal. The stock SigmaStudio volume configuration is just multiplying that straight into the audio signal with no Schmitt effect or quantization rejection, so if your particular volume control implementation has more than 1.6 mV of noise, you'll see modulation in your signal equivalent to the rise/decay time set in the SigmaStudio volume ramp. Also, if your particular volume input lands right near the transition from one 10 bit value to the next, you can get sometimes get modulation even with lower noise levels.
This was particularly a problem for me since I was using a 10 bit digital potentiometer for my configuration. The easy solution was just to truncate the 10 bit ADC input to 8 bits in SigmaStudio and set up my digital pot to avoid the transition points, but the "real" solution would be to add in quantization rejection and smooth the input with a cap in hardware.
This was particularly a problem for me since I was using a 10 bit digital potentiometer for my configuration. The easy solution was just to truncate the 10 bit ADC input to 8 bits in SigmaStudio and set up my digital pot to avoid the transition points, but the "real" solution would be to add in quantization rejection and smooth the input with a cap in hardware.
Raphael from Auverdion.de confirmed that Aurora DSP works in full-duplex mode.
This comes into play whenever you use the volume control input on the board and have volume control enabled in the control software. Input/output selection is irrelevant.
Thanks. Sorry for the dumb questions, would a standard analog pot (say 5k or 10k ohm) give any issues, and what does this modulation entail in terms of the effect on the sound? Is it just that the volume (attenuation) would step up and down randomly?
Last edited:
I've not extensively tried with analog potentiometers on my board yet, since that doesn't apply to my specific project. I tried a 5k linear pot at one point when debugging and noticed a slight decrease in SINAD performance (1-3 dB), but this should be taken with a grain of salt, as my testing methods at the time were focused more on speed than accuracy.
Looking at it from an EE perspective, lower resistance potentiometers with shorter wiring should have less overall noise (so long as the 3v3 supply can keep up with the current draw), but any continuous input signal has the potential of landing on a transition point, even with minimal noise. Lowering noise simply lowers the chance of encountering the problem.
Yes, the volume would go up and down (not discretely, as SigmaStudio implements continuous volume ramps) resulting in amplitude modulation (AM). This makes sidebands visible in an FFT, similar to the sidebands caused by jitter (effectively FM). This can be audible at lower volumes.
Oof. That's quite curious.
Overall, the volume control situation is a bit disappointing as I'm certainly not technically capable enough to fix it, but I have hope that someone will come up with a solution that can be readily passed along.
Just thinking out loud, is there a way to check the voltage ranges for each step/bit change? Perhaps a stepped attenuator with carefully-selected values could work in the meantime since each position it would land you right in the middle of each range?
Never mind, I don't think this would work. Too late to edit the post, unfortunately.
It would work, it would just be a lot of effort for the potentially marginal benefits it would offer. I would just ignore it unless it's an obvious problem like in an atypical use case like mine. I originally just brought it up as an example of performance changing in relation to specific user implementations.
Well the easy solution is to use the rotary encoder for volume control. Both control actually the same volume block. It is just a matter of which component you prefer. The rotary encoder does not have the voltage issue.
Although I am wondering a bit, wether it is really an issue. As far as I understood is sudermap applying an external voltage instead of connection a plain potentiometer.
Anyway I have now some free days and can have a look into this.
Raphael