This is a review and detailed measurements of the Allo DigiOne Raspberry Pi digital audio (S/PDIF) interface. It retails for USD $99 and is on kind loan from Allo. Unlike its higher-end brother (DigiOne Signature), this is a single board implementation at half the price.
When combined with a Raspberry Pi, and a suitable "audio operating system" (e.g. Volumio used in this review), you have a $135 networked S/PDIF interface. Hook the output of this board to your favorite DAC and you automatically add networking to it, allowing you to place the computer server elsewhere and stream content to your audio system.
Testing digital audio outputs is quite different than our typical analog measurements. Here, we are interested in how pure the digital output is as to produce good performance out of any DAC regardless of how good its S/PDIF interface is. I took a shot at this when testing the Allo Digione Signature but I was not too happy with them as the results were rather hard to interpret. Here, I have taken a different shot at it, hopefully making it easier to understand. So let's get into it.
Measurements
Given the success of my analog measurements using a "dashboard" view, I created a similar performance dashboard. The stimulus signal is "j-test." The J-test is a square wave at 1/4 of sampling rate (12 kHz here since our sample rate is 48 kHz). In addition, the low order bit is toggled + and - one causing all the bits to flip at once (e.g. 9999 to 10000 and back). The toggling of all the bits is designed to aggravate/accentuate jitter over S/PDIF cable. Ideally, we would see no trace of it if the system is resilient to what content is being transmitted. In other words, the S/PDIF signal needs to be independent of the audio being played to perform its role as an ideal "digital transport."
To have a reference, I tested my Audio Precision APx555 in its loopback mode with its unbalanced (S/PDIF) digital interface:
Focusing on top right (FFT), we see the spectrum of noise and spurious tones going as far as 300+ kHz. We see a noise spectrum starting at 10 picoseconds at low frequencies and dropping. We have a few spikes above 30 kHz but their amplitude is exceptionally low at just 6 picoseconds.
The time domain (oscilloscope) view on the left reflects the same with the waveform seemingly being random with no distinct patterns (good).
RMS level of jitter is 73 picoseconds. Peak is highly variable and falls in 200 pico seconds or so.
A cool but odd feature of APx555 is to treat the digital stream as "audio" and perform classical measurements on it. Here, I am having it output THD+N equivalent in the form of SINAD in dB. As in analog audio, the higher, the better. Yes, 16 dB is not a lot but we are talking about digital sample recovery so it is fine.
Next, to get really calibrated, I thought I test the Topping D10 as a S/PDIF interface bridge from USB:
Starting with FFT on top right, we immediately see the 250 Hz square wave in J-test becoming visible as jitter. 250 Hz is a square wave so when decomposed into frequencies, we see it and all of its odd haromonics (250, 750, 1250, etc). So we clearly have data dependency in the S/PDIF output of the Topping D10.
We see the same in scope display on the left where the sine wave pattern of the 250 Hz harmonics are clearly visible. The waveform does not look random as it did with APx555 output. RMS jitter has risen to 126 picoseconds and peak is nearly double at 400 ps. Jitter SINAD as dropped to 11 dB.
Now let's see how Digione does:
We get an output that is darn near APx555. No sign of the 250 Hz harmonics as we saw in Topping D10. RMS jitter is 72 ps. SINAD is a dB or so lower at 14 (this value jumps around fair bit).
Scope display on top left shows a much more random sequence again which is good.
And here is the same measurements for Allo Digione Signature:
This is a tiny bit better. We are hitting the best case performance of Audio Precision APx555 here so not much more room to move. I lost those results but a higher resolution FFT showed the signature to have lower noise than non-signature DigiOne.
Conclusions
Allo is one of the few audio companies that has excellent engineering skills and importantly, verifies its designs using measurements. So it is no wonder that even the non-signature DigiOne produces such textbook performance (i.e. essential match to Audio Precision APx555). At $99 it is a bargain way to get networked S/PDIF functionality.
Given the willingness of Allo to participate in online audio communities, I can't find any faults with the company or the product.
Needless to say, I highly recommend the Allo DigiOne to build a networked digital audio interface for streaming applications.
-----
If you like this review, please consider donating funds using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
When combined with a Raspberry Pi, and a suitable "audio operating system" (e.g. Volumio used in this review), you have a $135 networked S/PDIF interface. Hook the output of this board to your favorite DAC and you automatically add networking to it, allowing you to place the computer server elsewhere and stream content to your audio system.
Testing digital audio outputs is quite different than our typical analog measurements. Here, we are interested in how pure the digital output is as to produce good performance out of any DAC regardless of how good its S/PDIF interface is. I took a shot at this when testing the Allo Digione Signature but I was not too happy with them as the results were rather hard to interpret. Here, I have taken a different shot at it, hopefully making it easier to understand. So let's get into it.
Measurements
Given the success of my analog measurements using a "dashboard" view, I created a similar performance dashboard. The stimulus signal is "j-test." The J-test is a square wave at 1/4 of sampling rate (12 kHz here since our sample rate is 48 kHz). In addition, the low order bit is toggled + and - one causing all the bits to flip at once (e.g. 9999 to 10000 and back). The toggling of all the bits is designed to aggravate/accentuate jitter over S/PDIF cable. Ideally, we would see no trace of it if the system is resilient to what content is being transmitted. In other words, the S/PDIF signal needs to be independent of the audio being played to perform its role as an ideal "digital transport."
To have a reference, I tested my Audio Precision APx555 in its loopback mode with its unbalanced (S/PDIF) digital interface:
Focusing on top right (FFT), we see the spectrum of noise and spurious tones going as far as 300+ kHz. We see a noise spectrum starting at 10 picoseconds at low frequencies and dropping. We have a few spikes above 30 kHz but their amplitude is exceptionally low at just 6 picoseconds.
The time domain (oscilloscope) view on the left reflects the same with the waveform seemingly being random with no distinct patterns (good).
RMS level of jitter is 73 picoseconds. Peak is highly variable and falls in 200 pico seconds or so.
A cool but odd feature of APx555 is to treat the digital stream as "audio" and perform classical measurements on it. Here, I am having it output THD+N equivalent in the form of SINAD in dB. As in analog audio, the higher, the better. Yes, 16 dB is not a lot but we are talking about digital sample recovery so it is fine.
Next, to get really calibrated, I thought I test the Topping D10 as a S/PDIF interface bridge from USB:
Starting with FFT on top right, we immediately see the 250 Hz square wave in J-test becoming visible as jitter. 250 Hz is a square wave so when decomposed into frequencies, we see it and all of its odd haromonics (250, 750, 1250, etc). So we clearly have data dependency in the S/PDIF output of the Topping D10.
We see the same in scope display on the left where the sine wave pattern of the 250 Hz harmonics are clearly visible. The waveform does not look random as it did with APx555 output. RMS jitter has risen to 126 picoseconds and peak is nearly double at 400 ps. Jitter SINAD as dropped to 11 dB.
Now let's see how Digione does:
We get an output that is darn near APx555. No sign of the 250 Hz harmonics as we saw in Topping D10. RMS jitter is 72 ps. SINAD is a dB or so lower at 14 (this value jumps around fair bit).
Scope display on top left shows a much more random sequence again which is good.
And here is the same measurements for Allo Digione Signature:
This is a tiny bit better. We are hitting the best case performance of Audio Precision APx555 here so not much more room to move. I lost those results but a higher resolution FFT showed the signature to have lower noise than non-signature DigiOne.
Conclusions
Allo is one of the few audio companies that has excellent engineering skills and importantly, verifies its designs using measurements. So it is no wonder that even the non-signature DigiOne produces such textbook performance (i.e. essential match to Audio Precision APx555). At $99 it is a bargain way to get networked S/PDIF functionality.
Given the willingness of Allo to participate in online audio communities, I can't find any faults with the company or the product.
Needless to say, I highly recommend the Allo DigiOne to build a networked digital audio interface for streaming applications.
-----
If you like this review, please consider donating funds using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
