You are here: Home » Topic » Poor Sound Quality

Poor Sound Quality

This topic contains 17 replies, has 6 voices, and was last updated by  richdunlop 10 years, 9 months ago.

Viewing 15 posts - 1 through 15 (of 18 total)
  • Author
    Posts
  • #1169

    UK Phil
    Participant

    This post is just for information purposes. I bought a Linksys NSLU2 a while back and installed Unslung and Firefly on it and bought a Pinnacle (Roku) M1001 as a wireless player which I fed via the optical SPDIF output into a Denon AVR3802 surround sound amplifier. I was streaming FLAC files from the NSLU2 server. The problem was the sound quality was very poor, if I fed my CD player into the same optical input on the surround sound amplifier the quality was way better. This is a major disappointment, to the extent that I’ve returned the M1001 and given up on the project. What concerns me is how either Firefly or Roku are managing to trash what ought to be unadulterated data. This really ought to be like reading a data file where one would never expect to get errors. My understanding is that Firefly re-expands FLAC files to WAV format (so no data lost there) to stream over the wireless connection, if this isn’t the case then it’s a fundamental shortcoming of the whole setup and excludes it from any system that makes a pretence of being HiFi let alone audiophile.

    I guess my conclusion is buyer beware, I returned my M1001 and got a complete refund for it but still have the NSLU2, which is OK, but I would have been a very disappointed bunny indeed had I been stuck with the M1001 too. I would have considered it good money down the drain. I may try a Squeezebox3 with Slimserver running on a low power fanless PC at some point but I have to say I’m rather wary after this experience.

    #9495

    S80_UK
    Participant

    Hi,

    Sorry to hear of your disappointment. However, the Firefly forum is not the right place – since Firefly is a server, it is not itself responsible for any change to the quality of the sound. And since you were using FLAC, there should of course have been no difference.

    However, if you go to the Roku forums you will see that there has been considerable discussion relating to sound quality that some (but not all) users have experienced with the SoundBridge M1001. This is apparently due to the fact that the M1001 hardware cannot natively support the 44.1kHz sample rate of CD media (M500, M1000, M2000 don’t have this limitation). Work in investigating / resolving this with a possible improvement to the sample-rate convertor is still geing undertaken by Roku.

    Suggestion for Ron… When looking at support for configurable transcoding plug-ins to Firefly, wouldn’t it be cool to be able to decode FLAC to WAV and then also hook in a Sample Rate Convertor for the M1001 users. Not that you don’t have enough to do already… 🙂

    #9496

    rpedde
    Participant

    @s80_uk wrote:

    This is apparently due to the fact that the M1001 hardware cannot natively support the 44.1kHz sample rate of CD media (M500, M1000, M2000 don’t have this limitation). Work in investigating / resolving this with a possible improvement to the sample-rate convertor is still geing undertaken by Roku.

    Hmm.. didn’t know that. Are they using a fixed rate 48k dac or something?

    never noticed it, as I have a 500 and a 1000.

    Suggestion for Ron… When looking at support for configurable transcoding plug-ins to Firefly, wouldn’t it be cool to be able to decode FLAC to WAV and then also hook in a Sample Rate Convertor for the M1001 users. Not that you don’t have enough to do already… 🙂

    Gack!

    🙂

    #9497

    S80_UK
    Participant

    Hi Ron,

    Yes – the DAC clocking appears to be fixed to 48kHz on the M1001 – here’s the thread…

    http://forums.rokulabs.com/viewtopic.php?t=7009

    To save you reading all 14 pages, the M1001 uses an internal sample rate conversion algorithm. Depending on what they play, how they listen, and maybe on what they paid for their hi-fi, some folks can hear some low-level artefacts from this process. To present some kind of balance, there are many other users of the M1001 hardware that do not hear it. I have M1000 and M2000 and so I am also unable to hear it, but given how I listen I probably would if it was there.

    #9498

    rpedde
    Participant

    @s80_uk wrote:

    Yes – the DAC clocking appears to be fixed to 48kHz on the M1001 – here’s the thread…

    Oof. Yeah, I think I’d be pretty cheesed about that too. I don’t have golden ears, but I can still hear bad upsampling on cheap pc98 audio cards…. and if I could hear that on a piece of consumer audio hardware, I would also be unhappy.

    #9499

    gordo
    Participant

    This is a feature of the more recent hardware. m2000, m1000, and m500 do not upsample to 48 kHz but m1001 and soundbridge radio do.

    #9500

    UK Phil
    Participant

    Mmmm I guess you’re right this probably wasn’t the place to have made this post as it’s obviously a Roku problem, looks as though its ramifications are rather relevant to Firefly users though, as Ron said, he’d have been disappointed if he were in my situation. Thanks for the responses, I had wondered if it was an upsampling issue as I couldn’t think what else would have caused this problem, other than extreme jitter maybe. As I have the NSLU2 with Firefly on it I may well order an M1000 from Amazon and see if that over comes the problem. If I do I’ll let you know the result.

    Phil.

    #9501

    richdunlop
    Participant

    @UK Phil wrote:

    I may well order an M1000 from Amazon and see if that over comes the problem.

    If you’re ordering an M1000 make sure it is an M1000 (connectors under the end caps) before you buy. Roku originally marketed the M1001 as an M1000 when it first came out so you might get a nasty surprise if you aren’t careful. Your best avenue for an original M1000 purchase may be Roku direct or ebay.

    #9502

    UK Phil
    Participant

    Thanks Rich I had wondered if it was quite that straight forward, Roku have made a real screw up here haven’t they? That Squeezebox is sounding ever more attractive.

    #9503

    richdunlop
    Participant

    @UK Phil wrote:

    Thanks Rich I had wondered if it was quite that straight forward, Roku have made a real screw up here haven’t they? That Squeezebox is sounding ever more attractive.

    I’ve been a little surprised with some of Roku’s hardware design decisions such as the upsampling and lack of ethernet on the radio. Hopefully they’ll take account of the feedback they’ve had.

    For me squeeze isn’t the solution though. If you buy a Squeezebox you have to use Slimserver and I don’t like being locked in to things, no matter how capable. The beauty of the Roku is that it supports so many protocols – DAAP, UPNP, RSP. You have lots of flexibility on the server side in terms of both hardware and software. The Roku / Firefly / NAS combo really is killer for music…

    I’m sure that you could pick up an original M1000 on ebay or you may be able to get a refurb M1000 direct from Roku.

    #9504

    S80_UK
    Participant

    @richdunlop wrote:

    I’m sure that you could pick up an original M1000 on ebay or you may be able to get a refurb M1000 direct from Roku.

    Yep – that’s how I got started (a used M1000 from a UK seller on Ebay). Loads of fun (and pleasurable expense) ever since. 😀

    #9505

    CCRDude
    Participant

    Too bad the M1000 doesn’t allow WPA for Wifi if I’m not mistaken?

    #9506

    S80_UK
    Participant

    @ccrdude wrote:

    Too bad the M1000 doesn’t allow WPA for Wifi if I’m not mistaken?

    Correct – the older platforms which use the compact flash WiFi solution are WEP only. Possible solutions are…
    1) use a wired connection
    2) used a bridge / client mode access point wired to the SB (has the benefit also of allowing 802.11g rather than 802.11b or mixed mode, if that’s an issue, and often provides better WiFi signal range).
    3) use 128 bit WEP on the basis that it’s good enough for domestic needs. (I know it’s hackable if someone wants to try – but there are easier targets out there in the form of unprotected networks set up by non-techy users who buy their routers at the supermarket along with the groceries.)

    #9507

    CCRDude
    Participant

    3) People who would be wardriving would go otherwhere anyway, but the network is connected to the office as well, and that one has too many unscrupulous enemies. I wouldn’t even go for WEP with SSID hidden and MAC restrictions.

    2) Too bad I’ve got kind of an ecological mind and wouldn’t want the additional “useless” power consumption 😉

    1) Would probably the best for me… if I had known about this before buying, damn 🙄 😉

    I currently have two M1001, and a neighbour as well. We’ve planed some experiments for this… recording a LP both in 44,1 kHz and 48 kHz to be able to do some comparison of exactly the same signal on the same hardware. I know audiophiles don’t like these simple 1 vs. 1 comparisons a lot, but for us it’ll be sufficient and we could really compare the same signal in unconverted quality (due to the analogue source).

    #9508

    S80_UK
    Participant

    @ccrdude wrote:

    3) People who would be wardriving would go otherwhere anyway, but the network is connected to the office as well, and that one has too many unscrupulous enemies. I wouldn’t even go for WEP with SSID hidden and MAC restrictions.

    I use all three of those – and VPN on top for the office stuff but from a different PC.

    @ccrdude wrote:

    2) Too bad I’ve got kind of an ecological mind and wouldn’t want the additional “useless” power consumption 😉

    I used a Netgear gaming adaptor for a bridge for a while (but there are cheaper options). It takes about 3 watts, or half what my NSLU2 takes.

    @ccrdude wrote:

    … if I had known about this before buying, damn 🙄 😉

    The story of my life! 🙄

    @ccrdude wrote:

    I currently have two M1001, and a neighbour as well. We’ve planed some experiments for this… recording a LP both in 44,1 kHz and 48 kHz to be able to do some comparison of exactly the same signal on the same hardware. I know audiophiles don’t like these simple 1 vs. 1 comparisons a lot, but for us it’ll be sufficient and we could really compare the same signal in unconverted quality (due to the analogue source).

    Should be fun – at least you will know where you stand. Hopefully you will find that you don’t need to spend any more money. 😀

Viewing 15 posts - 1 through 15 (of 18 total)

You must be logged in to reply to this topic.