I’ve requested access to the Beta programme, but pending that, I’ll summarise some of my findings here.
For about a month and half now, I own a Naim Uniti Star, firmware version 3.5.1.4972 (latest stable version at the time of writing). I’ve experienced many hangs, crashes and reboots. That I have so many may be related to the fact that I’m ripping CD’s a lot (more than 100 per week), but that’s just conjecture on my part. Anyway, on to some of my observations. All of which to me indicate software (firmware) problems; based on my experience in the field I would guess some kind of resource leak or concurrency problem.
Note: the on-screen messages I ‘quote’ involve a bit of hand-waving, as my Star is set to Dutch, and some messages disappear from the display too quickly for me to run up to it and read them properly.
The Star may spontaneously reboot for no apparent reason. I’ve yet to find a reproducible set of circumstances, though. Mostly it’ll reboot while I’m ripping, but that may be because I’m ripping a large part of the time I’m using the Star (busy ripping my entire CD collection). It may restart completely (blank screen, clicking power supplies, full reboot), or it may just reboot some subsystem. Example of the latter is that I was streaming a Tidal track while ripping a CD. It rebooted a subsystem (a click [from a power supply?]) could be heard, but the track continued playing. I got a “Welcome. Looking for paired remotes.” message, after which it spat out the CD is was ripping, followed by a “Failed to rip CD due to an unknown error. Error 518”.
The scary part after that, and other crashes I’ve experienced, was that I then could no longer switch the Star off! Neither through the remote, nor through the front-panel buttons. In some cases I was still able to connect to it through the Naim app, and switch it to stand-by through the Settings menu. And when in stand-by, I was sometimes able to long-press (a second or 3) the front-panel power button to switch it off (i.e. ‘real’ power-off, not stand-by, although still connected to the mains, and the front-panel power button is still lit, so let’s call it ‘cold’ stand-by as opposed to the ‘warm’ stand-by after a ‘soft’ power-off). After which I could switch it back on using the front-panel power button (not through the remote, obviously, because it was now in ‘cold’ stand-by), but after it booted up (yada yada yada, plenty of time to admire the Naim logo on the display), I still could not switch it off using either the remote or the front-panel power button. Yikes. The only thing that could get it back into a usable state was unplugging it. Something that can’t be good for the machine (and possibly also not for my speakers, as unplugging it while ‘live’ gives a quite audible “BOOMPrrrr” on them). And in my set-up, it’s also not easy, as the wall socket the Naim is plugged into is for all practical means and purposes unreachable (behind a cabinet), and the connector into the power input of the Star takes quite some force to pull out. [Which, to me, in itself is a good thing, as it makes for a tight connection, but it also makes me hesitant to unplug & plug back in too many times, for fear of wearing things out.]
But I’ve also had it reboot after I was playing an album through Tidal, paused the playback and put the Star on stand-by, came back after some time, switched the Star on and resumed playing. It played the first few seconds of the track, then went silent, then managed to show the cover image on the display, and only then rebooted.
Another problem I’ve experienced was while ripping a stack of CD’s, it would fail to rip a new CD that I put in, spitting it out with the message “CD contains no tracks”. After that, it would refuse to rip any CD, including ones that I had successfully ripped just minutes before. The only way to get it out of that state was to do a power-off (long press of the front-panel power button; which by the way I only found out was possible through a post somewhere on this forum; do people no longer write proper user manuals these days I wonder?). After booting up, I could successfully rip the CD it had spat out before.
Another annoying thing is that some of the ripped tracks have ‘hiccups’ (I would call them FLAC encoding errors, but the FLAC files themselves are OK). In one case the Star would (consistently, reproducibly) play until the ‘hiccup point’ in the track (which wasn’t even halfway in the track itself), then skip to the next track. Playing that same track with e.g. VLC had an audible hiccup but continued playing the rest of the track. And none of these were tracks from CD’s with ripping errors. After re-ripping these CD’s, the hiccups were gone. I realised too late I probably should have saved these FLAC files. So when this morning I had another one, I did save it. This is the audio spectrum at the ‘hiccup’ point:
As you can see, just before the 2:06 point in the track, there’s complete silence. An analysis with flac -a
around that point gives me (sorry, nerd alert ):
frame=0 offset=402 bits=136008496 blocksize=909 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=FIXED order=1 residual_type=RICE partition_order=0
warmup[0]=-2303
parameter[0]=10
subframe=1 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=0
warmup[0]=-3994
warmup[1]=-3762
parameter[0]=10
frame=1 offset=17001464 bits=28576 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=FIXED order=1 residual_type=RICE partition_order=0
warmup[0]=2912
parameter[0]=10
subframe=1 wasted_bits=0 type=FIXED order=1 residual_type=RICE partition_order=0
warmup[0]=-758
parameter[0]=10
frame=2 offset=17005036 bits=5464 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=FIXED order=1 residual_type=RICE partition_order=3
warmup[0]=-2172
parameter[0]=9
parameter[1]=0
parameter[2]=0
parameter[3]=0
parameter[4]=0
parameter[5]=0
parameter[6]=0
parameter[7]=0
subframe=1 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=3
warmup[0]=197
warmup[1]=860
parameter[0]=9
parameter[1]=0
parameter[2]=0
parameter[3]=0
parameter[4]=0
parameter[5]=0
parameter[6]=0
parameter[7]=0
frame=3 offset=17005719 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=4 offset=17005735 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=5 offset=17005751 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=6 offset=17005767 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=7 offset=17005783 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=8 offset=17005799 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=9 offset=17005815 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=10 offset=17005831 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=11 offset=17005847 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=12 offset=17005863 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=13 offset=17005879 bits=128 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=CONSTANT value=0
subframe=1 wasted_bits=0 type=CONSTANT value=0
frame=14 offset=17005895 bits=18696 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=3
warmup[0]=0
warmup[1]=0
parameter[0]=0
parameter[1]=0
parameter[2]=8
parameter[3]=9
parameter[4]=9
parameter[5]=9
parameter[6]=9
parameter[7]=8
subframe=1 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=3
warmup[0]=0
warmup[1]=0
parameter[0]=0
parameter[1]=0
parameter[2]=7
parameter[3]=8
parameter[4]=8
parameter[5]=8
parameter[6]=8
parameter[7]=8
frame=15 offset=17008232 bits=24744 blocksize=1152 sample_rate=44100 channels=2 channel_assignment=INDEPENDENT
subframe=0 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=1
warmup[0]=-1758
warmup[1]=-1846
parameter[0]=8
parameter[1]=9
subframe=1 wasted_bits=0 type=FIXED order=2 residual_type=RICE partition_order=3
warmup[0]=-570
warmup[1]=-734
parameter[0]=8
parameter[1]=8
parameter[2]=8
parameter[3]=8
parameter[4]=8
parameter[5]=10
parameter[6]=9
parameter[7]=9
As you can see, frame 1 is OK, frame 2 is the start of the encoding error, frames 3 through 13 are completely silent (type=CONSTANT value=0
), frame 14 is ramping up, and from frame 15 it’s OK again.
Since the Beta forum is not open to the public (unfortunately) I have no idea whether any of these problems a. have been reported previously and b. are being addressed by the current beta firmware.
Oh, and before anyone asks, yes, I did try a factory reset. That’s no joy, as that loses not just all your settings, but also all of your favourites and radio stations (and rebuilds the media library, which takes quite long if you have a lot of music). So that’s a hoop I hope I don’t have to jump through again.
To end on a high note, when it works, the Star sounds absolutely magnificent. My hopes are on the software team delivering a version that fixes most (if not all) of these.