16 -> 24 bit

I can't cite sources but it's easy to understand if you think of video instead. Let's say you have a 1-bit video. Each pixel on your screen is either black or white. With 2 bits, each pixel is either white, light grey, drak grey, or black. It'll look better. 3 bits give you eight shades of grey (including black and white), 8 bits give you 256 shades of grey, 16 bits gives you 65,536 shades of grey, and 24 bits gives you 16,777,216 shades of grey. How many shades of grey your eye can distinguish then becomes the question. With sound, 24 bits should give you far more detail than 16 bits. Whether your ear can hear the difference I can't say. Personally I spend my money on soft foam earplugs (to protect my hearing when I'm out and about) and acoustic foam panels (to dampen echos) in my living room rather than buying expensive stereo equipment. --Tdkehoe 16:09, 14 February 2006 (UTC)

Hello. I wasn't requesting an explanation for myself. I already have an "almost sure" opinion on the point. What I was asking was that thearticle should have such explanation. In particular, if you search for this at Hydrogen Audio, you'll find the opposite opinion than is stated in the text. So as to support the flagrant bias in the text, one really should add enough explanation and sources. --Hdante 17:49, 13 February 2006 (UTC)

Well - there's some truth in this sentence. I really doubt that 16 vs 24 bitdepth has any influence on sound quality (contrary to samplerate). It does however have influence on dynamics, i.e. on 24 bit resolution you can code music that has greater difference between quietest and loudest sound. Many of 16-bit recordings need to be processed (compressed acoustically) to preserve optimum sound quality in whole volume range. The best example of this is "Dark Side of the Moon" compressed on CD and full range on SACD///

That's not really accurate. The dynamic range will be essentially the same: the amplitude will still go from 0%-100% of whatever your amplifier is pumping out. The difference is the precision with which the waveform is reproduced: a higher bit rate will reduce quantization error. It's similar to viewing a picture in 8-bit color vs. 24- or 32-bit color. On a good system, this is very definitely audible. In fact, there are distortion effects that do the opposite of this--reduce the bit depth in order to make the sound more grungy.

Something missing from this article is a discussion of the additional equipment needed for the higher quality audio to be of any use. Most home receivers and speakers weren't built to handle the increased bandwidth, and true, a lot of people don't care and won't need to spend all that money on an audiophile system. (Although the most recent edits are a bit ridiculous and not NPOV.)Torc 22:11, 21 July 2006 (UTC)

I searched Hydrogen Audio and found 25 pages of discussions about 24-bit audio. Could you give the specific topic title? Was this someone's opinion, or a published study? I've never heard 24-bit audio but I work developing voice audio equipment using 14-bit, 16KHz chips in portable battery-powered devices, and my voice sounds great. Then I use a 16-bit, 44KHz rack-mounted processor and my voice sounds far better. That's just my voice in headphones. I have no doubt that for orchestral sound on decent speakers most people could hear a difference between 16 and 24 bits.--Tdkehoe 16:09, 14 February 2006 (UTC)

Well, your mistake is comparing 16KHz vs 44KHz (which is a huge diference). For recording and all intermediate steps is necesary to use 24 bits or more. After finishing, you can go to 16bits/44KHz; that is 96db SNR or 96db of dynamic range, more than most equipments. Each bit is 6db more. BUT once you reach your noise floor, each more bit is useless. In very high quality equipment you can expect between 96 and 108db (16-18 bits), anything above 18bits is under background noise and there is no point having more digits than you can accurately mesure. —Preceding unsigned comment added by 201.231.97.156 (talk) 05:22, 9 November 2007 (UTC)

Hello. You may look at the Scientific/R&D forum, "Why 24bit/48kHz/96kHz/, If 16bit/44.1kHz is good enough?". The forum is about the members opinions, so it's not suitable for Wikipedia. --Hdante 16:48, 14 February 2006 (UTC)