[ObiHuang]

Quote:

Originally Posted by Brent Hutto Ding, ding, ding. Give the man a kewpie doll! If you've read threads here reviewing portable music players, under "Sound Quality" they often go on at length about how fancy a software equalizer the player has. Or in the case of iPods, a lot of people will claim they are lacking in "Sound Quality" specifically because they don't have a configurable multi-band equalizer. You probably are thinking along the old-fashioned lines of "Sound Quality" meaning how close the player can get to the sound of the original source material with your particular headphone and/or amp. That's not a universal definition by any means.

Wait! I assure you there is an audible difference in the noise floor using the Rockbox firmware on very sensitive IEMs. I call it the "SSSSSSSSSSS" factor.

 

[omendelovitz]

Quote:

Originally Posted by ObiHuang Wait! I assure you there is an audible difference in the noise floor using the Rockbox firmware on very sensitive IEMs. I call it the "SSSSSSSSSSS" factor.

oh yeah, i forgot about that... and can confirm it. Never heard a noise like that from ANY other source I've owned - Apple OS, you SUCK!

 

[aos]

You're right, there IS a noise coming out of the iPod when paused and with max volume! I just tried it with my HD25-1, hard-pressed against my ears. It's not very noticeable in my office, but it IS there. That is rather interesting. I'd still say it some kind of signal processing that Apple does and Rockbox doesn't. I recall iPod used to measure very well in RMAA and had very little noise; it was 2G though, not the 5G that I have. Is it possible that 2G has better sound quality than 5G? I suppose it is. Anyway, I'm going to hook it up to my E-MU 1212m and measure it when i get home, as well as to my oscilloscope.

 

[omendelovitz]

Quote:

Originally Posted by aos Anyway, I'm going to hook it up to my E-MU 1212m and measure it when i get home, as well as to my oscilloscope.

Would you mind measuring the FR with some kind of audio analyzer program and post the graph? I'm so curious to see how Apple has 'tweaked' their sound currently. (4g and older had a significant <100Hz roll-off from the HO, especially noticeable as the headphone impedance dropped)

 

[aos]

I measured the iPod's headphone out with RMAA here. Line out (using dock) is a bit better in some respects.

Using oscilloscope, one can see that iPod generates about 15mV peak to peak broadband noise, at maximum volume (I did not take a picture of my oscilloscope screen, but I do have screen capture of RMAA running in calibration mode, showing the noise floor). This decreases as you change volume, and it gets below my oscilloscope resolution (about 2mV) when the volume is close to zero (you can see some spikes that seem to be 150kHz but they aren't large and worrysome). This implies to me that the cause is in digital domain and not with its amplifier. Perhaps it's playing digital zero when paused so what you hear and measure is 16-bit noise floor. If that is the case, RockBox probably either plays no signal when paused or mutes the DAC chip "properly". None of this would ave any effect on sound quality and noise while actually playing, however. If there is a difference in sound quality of playback with RockBox, the noise you can hear when paused isn't the cause.

For reference, iHP-120 generates about 10mV noise on its headphone out (screenshot of RMAA in real time), and this one seems to have a strong 1.5MHz component, way above audible range. It does not change with volume.

 

[omendelovitz]

nice work, aos. I would agree that the noise floor has no relation to playback audio quality, however you seem (at least it seems that way to me) to have shown that software and its implementation does affect the sound that the hardware outputs. Hence the fact that rockbox can sound better than Apple OS now seems more probable, n'est pas?

 

[aos]

It was never impossible but it's always unlikely. Improvement in sound quality usually comes from improving the analog amplifier or improving the DAC chip. You most definitely can change the sound while it's still digital though. I wouldn't call it improving though, I'd call it fixing bugs.

An iPod playing losless music (wav, alac) with equalizer turned off (and sound check turned off and whatever else there may possibly be turned off) should in theory apply no processing on the signal. If it does, and RockBox removes it, then perceived sound quality could improve.

Here's one example how this could happen. Say for some reason designer decided not to use volume control built into DAC chip, but implemented their own, in software. Let's say that DAC only accepts 16-bit signal. If you're now playing back music at any volume other than max, you'd be losing dynamic range in proportion to attenuation, as you will be losing precision (it's similar to dividing two whole numbers and ignoring whatever you get after decimal point). So if someone else comes along, removes your volume control software DSP program and instead uses registers on DAC chip to attenuate, you will not lose dynamic range any more (or not as much, in 16-bit at any rate). I doubt this is what's happening, but it is one possible scenario where a software change could affect sound quality. I would still call that bug fixing, though.

 

[omendelovitz]

I completely agree with you - I believe what we have here is a potAto/potato scenario. I think you are using the hardware potential as the baseline, whereas I was using the AOS (no pun intended) as the baseline. Suffice it to say that it seems as if Rockbox uses hardware to process the signal and that Apple processes the signal "differently". Either way, what I have found is that using my setup w/ the iPod LO and HO, both sound better (better resolution and imaging) on RB, and that imaging is the main improvement when using LO vs. HO on RB.

Cheers!

 

[wanderman]

the h3xx and the h1xx sound exactly the same. I think they are the best line out solutions because they are on the player body no need for subpack all you need is an interconnect and an amp. Optical + Lossless + dac + amp is all well and good but I use my mp3 more than 4 hours a day. The line out's on the h1xx and h3xx have next to no hiss compared to the headphone out on those same players. Rockbox works (nearly)flawlessly on them.

The optical+dac+amp transportable abomination is overrated imo. If I have enough room to bring all that would just my notebook. The vaio has a bigger screen, longer battery with wifi, and a larger hdd.