M1001 noise

RokuMike · Tue Jan 23, 2007 1:27 pm

Yes, as already posted, we have reproduced it using some of the sample files that people here graciously provided.

andyinstal · Wed Jan 24, 2007 11:49 am

Mine is going on Ebay tonight for anyone interested. We have been griping for months (year?) about this issue and now they are finally admitting that there is a problem? Too little, too late. Good luck to the rest of you.

beun · Wed Jan 31, 2007 1:31 pm

As a relatively new user I can now also report that the noise is there on my new M1001. Very obvious on soft passages like the beginning of Alan Parsons "I robot" and the Art of Noise "(Two fingers of) Love"

The Highlander · Wed Jan 31, 2007 1:54 pm

Just one suggestion:
While waiting for a third party to solve the problem of the
resampling algorithm, simply handle the stream as a
multichannel surround stream and pass the data
without any modification of sampling rate
or output level directly to the digital outputs.
This could be an options in the settings of the SB,
at least this a solution for wav-Files.
In my opinion this is the only solution for a bit-perfect
streaming of audio content!

grommet · Wed Jan 31, 2007 2:12 pm

The Highlander, if it wasn't clear in the numerous threads here... the M1001, unlike the older M500, M2000 & M1000 models, can only output 48 Khz as designed. This is the reason the algorithm exists in the first place. So, unless you have 48 kHz content being served from your server, it needs to upsample.

The Highlander · Wed Jan 31, 2007 2:21 pm

Hi grommet.
How is the option for passing through
multichannel streams working if its not
possible to bypass the upsampling?

grommet · Wed Jan 31, 2007 2:28 pm

That option (in my opinion) is named incorrectly. It just twiddles a bit on the digital output.... a PCM valid/invalid bit.

dinkeldorf · Thu Feb 01, 2007 9:59 am

OK can the fix be applied to existing 1001s?

guysivey · Tue Feb 06, 2007 10:58 am

For the record, I can hear the noise on the Pink Floyd track mentioned, on my 1001.

I tried creating 2 wav files of silence at 44 and 48 k to see what happens, but both of them were (nearly) silent. Strangely there was some residual noise when the files started, but the same for both files.

I then created 2 wav files of very quiet music and then faded them in three times to get varying levels. No difference in sound from the 1001.

Maybe it depends on the nature of the original wav file?

tps · Tue Feb 06, 2007 4:28 pm

My experiments seem to indicate that when the M1001 plays a file with "digital zero" at 44.1 KHz, the M1001's digital output indeed puts out 48 KHz digital zero. However, playing a 44.1 KHz file of dither, then the noise components appear on the M1001's 48 KHz digital output. I used random dither because it is about the lowest level noise above digital zero in the digital audio world. Most modern CDs use some form of dither to take the hard quantization "edge" off of extremely low level audio signals.

Dan Smart · Tue Feb 06, 2007 4:39 pm

dinkeldorf wrote:
OK can the fix be applied to existing 1001s?

There is as yet no known fix.

Dan

cinnamonsigh · Sun Feb 11, 2007 4:40 pm

Excuse my ignorance for I have not had the time to read this entire thread, but does anyone know if this noise problem also shows up in the Pinnacle Soundbridge M1001?

And what about the Pinnacle Soundbridge Home Music?

HarryDymond · Sun Feb 11, 2007 4:55 pm

cinnamonsigh wrote:
Excuse my ignorance for I have not had the time to read this entire thread, but does anyone know if this noise problem also shows up in the Pinnacle Soundbridge M1001?

And what about the Pinnacle Soundbridge Home Music?

That would be a "yes" on both counts, with the provisos that:

1.) Not everyone can hear it.
2.) I believe (please someone correct me if I am wrong) that those who can hear it only hear it with certain 44.1 kHz tracks, not all 44.1 kHz tracks.
3.) You can avoid the problem if you re-sample to 48 kHz on your computer. To do this without sacrificing quality, you will need to re-rip your CDs with the appropriate settings, unless your library was ripped to a lossless format. (In which case you can perform the upsample process on your computer without loss of quality, it'll just take a while.)

The Highlander · Mon Feb 12, 2007 1:31 am

HarryDymond wrote:
3.) You can avoid the problem if you re-sample to 48 kHz on your computer. To do this without sacrificing quality, you will need to re-rip your CDs with the appropriate settings, unless your library was ripped to a lossless format. (In which case you can perform the upsample process on your computer without loss of quality, it'll just take a while.)

Sorry, your last statement is not correct.
Upsampling from 44.1 kHz to 48 kHz is not a simple task.
Because of the ratio of 147:160 the samples are calculated values
and this means the result is an approximation to the original signal.
For the same reason it is not reversible and therefore it is not lossless.

dinkeldorf · Mon Feb 12, 2007 8:35 am

OOPS what I meant was "don't assume a fix could be applied to existing units".

Dan Smart wrote:
dinkeldorf wrote:
OK can the fix be applied to existing 1001s?

There is as yet no known fix.

Dan

M1001 noise

Who is online