[Michael Osadciw]

I figured this would be the best place to post this question, since many of us here love hifi...


Bare with me here...as I release my thoughts wildly here on this subject looking for a bit of clarification... I've read elsewhere a little bit about this subject, but am still slightly foggy...


How would you describe "noise floor"?


If I think I have this correct, would noise floor be referring to the level of noise created by virtually all components in a signal path? The lower the noise floor, the less audible some sort of noise is?


eg: in the original recording, whatever recording devices the artist uses, be it sound board, recording devices, the mixdown, etc...all would have a noise floor of some sort, would it not? How much of it captured in the recording would be determined by the components?


What about the resolution of recording? Is it safe to say that a native 16-bit recording would have a higher noise floor than that of a native 24-bit recording? (based on the capabilities of the technologies?)


That leads us to playback devices...hi fi components...claiming lower noise...(that would partially determine a noise floor, would it not?)


So audio components in a system would also determine a noise floor for in-home playback?


What about the choice in software? Is it also safe to say that if I had a 24bit PCM BD-audio disc of native 24-bit material, and played it back through a BD player's output, that with a well-designed component I could possibly experience music with a lower noise floor during playback than the same content on a 16bit CD played back through a CD player even if it is upsampled (because output precision is still 16bits?).


Hopefully I didn't sound confusing.

 

[Curt Palme]

To me, noise floor is what you hear (don't hear) when a system is on. In other words, there is a difference that can be measured between the hiss/hum/noise and the program audio. That is the signal to noise ratio, the noise floor is the 'noise'


All components, including a CD can contribute to the noise floor of the system (I'll argue that interconnects can't, but someone here will argue that point). Interconnects can contribute to the noise/hum in a system I guess, but only if they are incorrectly connected, or the wrong cable is used.


A lot of noise issues can be attributed to ground loops and signal level mismatches between equipment.


I personally don't see how 16 or 24 bit conversion can contribute to the noise floor, after all, zero volume is when all bits are at 0, so it shouldn't matter whether there are 16 or 24 of them.

I also don't think that software can contribute to the noise floor, although bad a/d and d/a converters sure can.

 

[oneobgyn]



Quote:
Originally Posted by Curt Palme /forum/post/15450255


To me, noise floor is what you hear (don't hear) when a system is on. In other words, there is a difference that can be measured between the hiss/hum/noise and the program audio. That is the signal to noise ratio, the noise floor is the 'noise'


All components, including a CD can contribute to the noise floor of the system (I'll argue that interconnects can't, but someone here will argue that point). Interconnects can contribute to the noise/hum in a system I guess, but only if they are incorrectly connected, or the wrong cable is used.


A lot of noise issues can be attributed to ground loops and signal level mismatches between equipment.


I personally don't see how 16 or 24 bit conversion can contribute to the noise floor, after all, zero volume is when all bits are at 0, so it shouldn't matter whether there are 16 or 24 of them.

I also don't think that software can contribute to the noise floor, although bad a/d and d/a converters sure can.

Well stated Curt....kudos

 

[QueueCumber]

 Noise Floor

 

[QueueCumber]

If you take the most general definition from that page and are considering the noise floor in terms of a listening room and/or recording space, you should also consider environmental noise along with noise internal to the electrical components and their signal carriers.

 

[Michael Osadciw]

ok, cool for the components...i get it.


...so regarding software then, a 16-bit CD won't have a higher noise floor than 24bit BD? For some reason I thought that with a 24bit disc, the noise ended up being a lot lower.

 

[Peter Nielsen]



Quote:
Originally Posted by Curt Palme /forum/post/15450255


I also don't think that software can contribute to the noise floor, although bad a/d and d/a converters sure can.

Bad software can do horrible things like outputting ones where there should be only zeroes.


Silence (all zeroes) can easily become noise (random ones) because of some "ingenious" DSP algorithm employed.

 

[Dennis Erskine]



Quote:
I thought that with a 24bit disc, the noise ended up being a lot lower.

To suggest that 24-bit has an inherently lower noise floor than 16-bit would be an incorrect assumption. As Curt noted, what happens during the A/D and D/A conversions is where the difference can occur. Most of the published work in this area carefully uses terms like "perceptual". The audible reduction depends upon how truncation, requantization and dither are handled. I am far from being current in this area (mid to late seventies). If you really want to dig into it, you can find several documents in the AES library. If you can find an old Meridian 618 or 518 manual, I believe you'll find a fair overview of the process.


OTOH, while various processing techniques can be used to reduce noise during the A/D or D/A process, many of the techniques can reduce the noise to levels much lower than the noise floor in consumer (and many pro) equipment and installations. Not that that is a bad thing, just an observation.


As a thought, don't confuse S/N ratio with "noise floor".

 

[NIN74]

Noisefloor on a 24 bit recording can be lower than on a 16 bit recording, depending on many things. Coded silence on a CD and a DVD-A can be instresting to compare, if one have a playbacksystem that have a very big s/n.

 

[Michael Osadciw]

well the reason behind my question is because I've got a software title on BD PCM 24/96 and CD PCM 16/44.1 (yes, there is actually a Blu-ray disc out there with only 2-ch audio on it, which is Nine Inch Nails: Ghosts I-IV). Comparing the two, volume matched as best as I could, I found that the "noise" that hissed around in the background of the music was significantly lower..."quieter" on the BD than it was on the CD. To me it gave me the impression of greater dynamic range as well...my perception of everything about the CD was altered.


So, I assumed that it was the PCM24/96 of the BD that contributed to this lower noise than the 16-bit CD system. I understand what is being said that bad DACs can contribute to higher noise, but in this case, I'm also going to assume that the DACs in my Meridian 508.24 CD player are much better than what's inside of my Panasonic BDP-10A BD player. I listened using the balanced outputs of the CD player and unbalanced from the Panasonic. FYI the rest of the system is a Ayre K-1xe, 2 Theta Dreadnaughts run as monoblocks for stereo audio, and Dunlavy SC-IV/A speakers.


So...I'm puzzled...what is contributing to this lower noise on the BD? (FYI, the BD player could not, in my opinion, create the same sense of presence as the CD did through the Meridian...even though I thought the BD sounded a little less abrasive and more weighty in several areas.)


Hmmmmm.....anyone?

 

[Michael Osadciw]

...could it be that this is part of the problem?

http://brianstagg.co.uk/p_t_a_clipressed/


could the CD be more compressed? check out the vinyl and CD comparisons at the bottom of this page. if the CD has greater compression applied to it than the Blu-ray, could the higher noise level I'm hearing on the CD be attributed to this? (assuming that the BD was mastered as the vinyl has been in this example)


I'm just tossing things out here for discussion...

 

[AndreYew]

I would have guessed more level compression on the CD as well. Depending on how they compressed the CD, the noise floor can be more audible. Your level matching could also be off.


--Andre

 

[NIN74]

I could be different masters. I had the Ghost ultra delux boxset but I sold it so I cannot compare now.

 

[Michael Osadciw]

aaawww why would you do that? I'm not letting go of my autographed box...

 

[Michael Osadciw]

I just went to listen to the Depeche Mode song, "Precious" ...I have the CD, never really gave it a good listen. Anyways, I can see where the guy in this website is coming from...when listening you can hear faint crackles consistently throughout the song...I'm assuming that's the sound of peaking... I hear that quite a bit in a lot of music...I do tend to listen to a lot of music that is loud, so I guess that is no surprise...