Several of my UK-based associates have used Ipevo USB handsets in conjunction with Skype, some for several years. In fact, these have become a standard item in our packing for trade shows. Our booth has a Polycom Soundpoint IP430 and an Ipevo USB handset. The Polycom phone integrates with our US hosted IP-PBX (OnSIP) to provide calling to anyone and provides a handy, high-quality speakerphone. Alternatively, the Ipevo handsets supports free calling to those back at the home office that use Skype, including co-workers, friends & family.
Thus my prior experience with Ipevo products has been pretty good. I have found that they make affordable, simple audio interface devices for use with Skype or generic soft phones. In offering the X1N6 Internet Conference Station it appears that Ipevo is trying to take their game up to a new level.
I was also pleasantly surprised to find that Fry’s Electronics turned up as about the best source for the device. It was offered for $139.00 and shown as available at one of the three Houston area stores. Given that Fry’s has a very lenient return policy I decided to order the X1N6 online and arrange for in-store pickup. I would have 30 days to try the unit and still be able to return it if it should prove less than satisfactory.
The X1N6 is nicely packaged, resting in a formed paper cradle inside a simple box. It comes with limited printed documentation reflecting it’s utter simplicity. It was also accompanied by a USB-to-mini-USB cable for use connecting it to the host computer. There’s no power supply as the device is powered directly from the USB bus.
The X1N6 has only a few physical controls on the unit, the most obvious being a greenish volume control on one of the four “legs.” There’s also a simple on/off switch and a mute button along the side of one of the legs. The mini-usb connector is located underneath the unit, where there are reasonable provisions for passing the cable to/from the PC.
Finally, there’s traditional analog audio I/O for microphone and headset. With a PC type headset plugged into the X1N6 the device is really just acting as a USB audio interface for the headset.
When plugged into a host PC the X1N6 has a LED ring on the top that lights up in green when the device is active, changing to red when it’s muted.
The speaker is in the center of this ring, the microphones at the end of each of each leg, not unlike the Polycom SoundStation conference phones. This similarity gave me some hope that this device would be an effective solution for use with a group of 4-6 people around a small table.
On initial connection to my Windows XP Pro desktop the device was recognized and Skype asked if it should be used as the default audio device. It was also available to Eyebeam & PhonerLite, the other two soft phones that I use. Within the Windows Multimedia settings the device showed up as available for generic audio and voice applications.
Using the X1N6 just to listen to a podcasts I found that it sounded good. Moving on to making a test call I met with something of a surprise. I tweeted out an invite for someone to make a test call via SIP URI so that we might have a wideband call. @jdbalogh from Penn State University responded.
On this initial call the device worked, and I heard the other party just fine, but he reported that he was hearing a massive amount of noise. Since the device seemed to be generally working I thought that perhaps I had something about my windows XP desktop misconfigured.
From this initial call I moved on to making test calls between devices on my network, so that I could hear both ends of the call. The noise reported by @jdbalogh was readily apparent. I tried various tweaks to the Windows audio settings but nothing seemed to help.
Within the Eyebeam soft phone engaging the noise reduction and auto gain control settings helped somewhat. All that did was gate out the noise during quiet passages between words. The noise was still clearly evident underlying the active part of the conversation.
Audio capture directly into the editor from the X1N6, click above to see a full sized image.
To try and better understand the presence of this noise I used the X1N6 as an audio capture device directly recording into my audio editing program (Syntrillium’s Cool Edit Pro, now Adobe Audition.) The captured audio showed that the noise floor using the X1N6 measured -19 db and with a significant DC offset. The screen shot of the waveform editor shows this clearly. Note the audio level meter along the bottom of the screen shot.
Hunting around I found that Cool Edit Pro had a capture setting that would eliminate the DC offset. This resulted in a somewhat improved audio capture with the noise floor at -35 db.
In contrast a similar recording made using the Clear One Chat 50 showed the noise floor at -57 db. (see below)
Audio capture directly into the editor from the Chat 50, click above to see a full sized image.
To see the measured noise floor of the Chat 50 you’re going to need to click on the image above to see the full sized version. The waveform is simply not visible at the same vertical resolution as the the display showing the X1N6 noise. You can see that the audio level meters along the bottom indicate -57 db.
That’s a dramatic difference.
I performed one last experiment, which was to record the X1N6 output into Cool Edit Pro starting with the mute function engaged, toggling the mute off, then returning to muted. Here’s the resulting waveform.
Audio capture directly into the editor from the X1N6, toggling the mute function,
click above to see a full sized image.
A signal to noise ratio of 35 db is not even as good as analog FM radio, more typical of a bad AM radio signal. The impact of this noise on usability may not be obvious. To make that point I also recorded a brief voice sample. In making this recording I was sitting near the conference phone, as if I was having a meeting with 2-3 other people at a small table. You can hear the noise throughout the clip, but it’s most audible near the end.
Audio capture directly into the editor from the X1N6, voice sample,
click image above to see a full sized image.
Of course, you needn’t take my word for this. Here are MP3 versions of the sample recordings that correspond to the various waveform displays.
- Ipevo X1N6 raw recording, just ambient room noise
- Ipevo X1N6 recording, DC offset corrected, just ambient room noise
- Clear One Chat 50 raw recording, just ambient room noise
- Ipevo X1N6 recording, toggling the mute function on-off-on
- Ipevo X1N6 sample voice recording
All of this is very disheartening as it appears that the Ipevo X1N6 is so inherently noisy as to be completely unusable. On the basis of this finding I will be returning the device to the reseller for a refund.
I wonder if perhaps this one device is simply defective? Or is this generally indicative of the X1N6 ? It’s not possible to know without getting another device for evaluation. I don’t think that I’ll make that effort. Even so, it’d be good to hear from someone else who has used this device perhaps with better results.