Marlene's Musings: Rockbox

You´re about to read the fourth out of nine reviews. You don´t need to read them all, just pick the unit that seems to be the most interesting to you. In the weeks to follow I´ll review the following units: Kenwood DP-5090, Pioneer DV-610, Sony CDP-470, Pioneer BDP-140, iRiver IMP-550, Sony MZ-R 55, Sony MZ-R 37, Sony NW-A 1000 and the Sansa Clip+. Stay tuned and enjoy the new review: the Sansa Clip+! I also implore you to read the article describing my rigorous testing methodology before you actually start reading this one.

Sansa Clip+

EDIT 14.01.2013: Several people on the Sansa forum were doubting my measurment results. As it turned out they were right. I´ve updated this article accordingly.

One shouldn´t believe what all others are saying, really not. For years I and countless others have been lead to believe that Rockbox is an awesome alternate firmware created to improve certain portable media players. As you will read in this now updated review of the Sansa Clip+ it in reality isn´t what everyone thinks it is. For comparison I´ll leave the original text of this review online for everyone to read. When I reviewed this unit in October 2012 I had just installed the newest Rockbox firmware for the first time. I did not use it in the two years prior because the original Clip+ firmware by Sansa was working fine and without flaw. My decision to finally install it was based on numerous positive remarks about Rockbox and as I wanted my Clip+ to be at its best I thought that I had a valid argument in favor of using it. Because of that I tested this tiny player with 48 kHz material only when instead I should have used 44.1 kHz material. You´ll see why:

You know how often I´ve talked about the Sansa Clip+ being a decent player in almost every review I´ve written for this blog? I´m actually very ashamed of myself because of that, also because I trusted NwAvGuy's measurment and conclusions when he reviewed it. Well, I have to be realistic: his player was much newer than mine which now is over two years old. It´s likely that it has aged during this short time... which does render it unreliable and a waste of money. Using lossless FLAC files it also should have an advantage over, say, the MZ-R 55 which works with lossy ATRAC compression. But there you have it: bad hardware destroys the best possible theoretical advantage. Just for your information, I did my measurments two times: the first with volume at maximum output (0 dB), the second with -3 dB reduced gain. You´ll see why:

Comparison, FLAC 44.1 kHz, original firmware vs. Rockbox

Comparison, FLAC 48 kHz, original firmware vs. Rockbox

Frequency response, FLAC 44.1 kHz, original firmware vs. Rockbox

Frequency response, FLAC 48 kHz, original firmware vs. Rockbox

Intermodulation distortion, FLAC 44.1 kHz, original firmware vs. Rockbox

Intermodulation distortion, FLAC 48 kHz, original firmware vs. Rockbox

As you can see on the pictures above 48 kHz performs just horrible when using Rockbox. Intermodulation distortions and frequency response deviations are so strong that they WILL be audible. With 44.1 kHz material it´s - sort of - the other way round, Rockbox performs slightly better. Now it´s also evident that the Clip+ actually performs very well, it certainly doesn´t show age related imperfections as I was boldly stating in my original review! It escapes me why so many people programming Rockbox get something basic as playback of 48 kHz so badly wrong. This means that I´ll have to update the conclusion to this and the other articles accordingly, it also means that I won´t use Rockbox with 48 kHz material. Since now everything has changed I have to do my listening tests again, using 44.1 and Rockbox.

Stage impression is a bit more compact than the reference but very stable and more holographic. Everything in the center is pronounced slightly and more forward while at the same time instruments don´t wander around or change their size. There´s one obstacle though: the Clip+ fails to render reverberation convincingly, recording venues sound drier than usual. While dynamics are very good they are too aggressive and a tad contrived. This aggressivness doesn´t extend to timing and snap though; mids and treble are a bit slower while bass is a bit faster than the reference. Furthermore, the Clip+ pronounces upper bass and diminishes deep bass, which - when combined - helps punch and snap but occasionally hurts stage impression. Treble is pronounced as well; that should help resolution & detail but instead it adds to its aggresiveness and dynamic impression, music sounds a bit more strident and 'glassy' as a result. One thing however is crystal clear: compared to my original review it sounds much, much better. It´s almost as good as the very balanced and somewhat boring sounding Sony NW-A 1000; I assume that the sound of the Clip+ will appeal to more people. Recommended - but please remember to use 44.1 kHz material only.

The sharp high frequency dropoff will be audible. It isn´t a misinterpretation of RMAA, I confirmed the results with iZotopes Ozone 4 and its EQ Matching function. The awful amount of intermodulation distortion will be audible too, it reaches alarming -6 dB at 20 kHz. You might understand that I didn´t do a jitter test, it would have been pointless anyway. With distortions as severe as these jitter related artifacts will be obscured completely. Less gain on the output lowers distortions somewhat, the high frequency drop off however stays. It´s no wonder that this player has a reputation of having a very strong and deep bass: when treble gets reduced the perception of deeper frequencies will of course be different! BTW, these are not the first measurments I have done - when I got it two years ago I measured it to compare it to my Creative Zen Mosaic and back then it still had an extremely flat frequency response and less distortions. Which means only one thing: the parts used on its mainboard have aged during the course of only two years. How does it sound? Well, considering the distortions and the high frequency dropoff it still sounds ok - surprisingly so. One of its sonic advantages always were its explosive dynamics... they are nice if you´re able to ignore their artificial character. I´ve complained numerous times about a flat and constricted stage that was not even able to render reverb convincingly. The stage is its weakest aspect; it sounds nothing like the reference because it´s way too compact and very flat. Crispness always seems forced upon the sound and "faked" (distortions can leave that impression if applied improperly - if you don´t know what I´m talking about try an Exciter). The Sansa Clip+ combines a grungy signature with dryness, grainyness and an aggressive, ostensibly analytical sound. It was better when it was new but it has aged considerably during a short amount of time, therefore I can´t recommend it. I´m appalled of myself that I´ve spent money for it, that I´ve trusted it, that I´ve used it.

Rockbox, 44.1 kHz:

Sonic Balance: