I plan to set up my listening environment at home using an active software based crossover running on a PC, that outputs the audio signal on a multi-channel audio interface to multiple amplifiers. While there are countless audio interfaces on the market, most of them are stereo only. The Gigaport EX from ESI Audio features an eight channel DAC suitable for 7.1 audio, that seemed a good fit for what I have in mind. In this short review, I present my findings regarding the audio quality of the Gigaport EX interface. I already own an interface from ESI, the U24 XL, which I reviewed earlier and I was curious whether ESI has improved some of the shortcomings I found in the U24 XL.
For the measurement, the Gigaport EX was connected to my notebook using the USB cable that was shipped along with the interface.
The Gigaport EX was tested using REW 5.19 running on Debian buster together with the U24 XL, which recorded the output of the Gigaport EX. The analog audio input of the U24 XL does not have measurement instrument quality, but is not too bad and suitable for some basic measurement of the Gigaport EX output.
Here is what the noise floor of the line out looks like with the Gigaport EX set to 48kHz mode:
I also tested the Gigaport EX in 96kHz mode:
Regardless how I set the Gigaport EX, the noise floor is pretty high and shows a lot of spikes in the spectrum. First I though my measurement setup was wrong and I was seeing some weird artifacts, that are not there in reality. I changed my mind once I connected my headphone to the output. There is really a terrible squeaking noise at the output of the Gigaport EX, which is clearly audible and highly disturbing.
When I did some research prior to buying the Gigaport EX, I found many users complaining about noise issues. Noise an hum are common issues in larger installations when using single ended signal transmissions so I didn't worry too much whether this might affect me. My measurement setup was all floating with no earth connection whatsoever so the noise really stems from inside the Gigaport EX. What strikes me here is that the spikes in the spectrum start at exactly 1kHz and almost all higher frequency spikes are straight multiples of 1kHz.
I have no idea how the noise is generated and how it gets to the audio output at such alarming levels, but would speculate this could originate in the switching power supplies, maybe from some inter-modulation of the different boost converters. ESI has demonstrated earlier in the U24 XL that they are not really competent in power supply design so it would not surprise me if that was the issue here again.
The good news is that the eight line outputs do not suffer any frequency response issues. This should be self-evident, but seems worth mentioning in light of the headphone output frequency response.
As advertised, the Gigaport EX runs plug and play on Linux. Just plug it in and it works right away with all channels available. It also registers as analog input, but this is not available of course.
Here are some photos of the PCB assembly, which are probably only relevant to tech minded people like I am. Note that once you click on any of the photos, very large full size images will be loaded and this may take some time due to limited bandwidth of my web server connection.
The PCB has components on both sides packing a lot of electronics on a tiny PCB. Here are some details and my opinion about the circuitry:
Underneath the grey block lower left side on top side of the PCB is a micro controller (Bravo SA9128). I assume that the grey thing on top of the micro controller is a ferrite patch for EMI suppression. The DAC is the AK4358VQ from AKM. There are plenty of NJM4580 operational amplifiers for the analog audio section. The power supply was somehow improved compared to the U24 XL and now features a dedicated boost converter for the negative supply. This means that the analog supply is now true symmetric +/-5V. There is also an extra dedicated +5V supply for the digital part, which is a sign of good will. However, there are no filters or any other measures to lower noise on the analog supplies, which is really sad. Calling the output of a boost converter an analog supply is legal, but a poor engineering decision in my opinion.
The build quality of the assembly is a bit subpar in my opinion. There are plenty of residues on the assembly. Some of the residues are soldering flux and some are of unknown origin to me. Some components look like they were hand soldered by less skilled operators. This is all okay for consumer electronics and nothing to worry about, but the U24 XL was built with slightly better quality in my opinion.
The Gigaport EX is advertised as "professional 24 bit and 192kHz USB audio interface with 8 separate high quality analog output channels". In my opinion this product does not really target professional users because those would expect balanced XLR interface instead of unbalanced RCA connectors. While 24 bit and 192kHz sound great on paper, the old 24 bit and 48kHz U24 XL easily outperforms the much newer Gigaport EX. For me the extreme noise always present on the analog outputs is totally unacceptable. Also, the lack of bass at the headphone outputs really make me doubt that the Gigaport EX is a well engineered product. From a German company I would expect much better engineering. Just like with the U24 XL, a few electronic components worth a few cents each and some days in the lab could have resulted in a much better product. I'm going to return mine and will be looking for a better multi channel DAC.
Of course, I am not the first person who reviewed the Gigaport EX.
I found a pretty detailed review including measurements on ProSound.IXBT.com. The measurements were fine and the conclusion quite positive. I wonder about the very different findings and don't know why I experienced noise issues and ProSound did not.
The big retailer Thomann has three customer reviews and two out of those three reviewers complain about noise issues. What one of the reviewer describes fits my findings regarding the multiplies of 1kHz spikes:
I can confirm it is noisy. High frequency noise is present. I measured with other sound card connected to same PC noises as 8, 16, 24, 32, 40 kHz frequency peaks about 25 dB higher than general noise floor.