24bit vs 16bit, the myth exploded!
Jul 4, 2020 at 11:19 AM Post #5,671 of 7,175
Doesn't DSD have softer attacks because of the representation?
No, DSD does not have any characteristics that may affect the signal being recorded other than noise, which is pretty inaudible (even for DSD64 {-120 dBFS [20 bits]}). DSD does not sound different than PCM.
 
Jul 4, 2020 at 3:13 PM Post #5,672 of 7,175
No, DSD does not have any characteristics that may affect the signal being recorded other than noise, which is pretty inaudible (even for DSD64 {-120 dBFS [20 bits]}). DSD does not sound different than PCM.

Okay so much higher sample rate allows exactly same signal as PCM, except for the possible noise issue.
 
Jul 4, 2020 at 3:17 PM Post #5,673 of 7,175
except for the possible noise issue.
Noise even for undithered Redbook is pretty much inaudible. Noise is a non-issue since the CD became mainstream. PCM and DSD are designed to do the same thing, and they both do it correctly without any major deviations (only being noise a really minor difference at a technical level, but I already explained that).
 
Jul 5, 2020 at 3:46 PM Post #5,675 of 7,175
Thanks guys. I also want to know something else similar to this topic: Is FPGA better than chips and provide a lower latency and better sound, or it it, once again, placebo? Asking in regard of Hugo 2 vs Pro iDSD

FPGAs are used by boutique DAC manufacturers because their production volumes are so low that ASICs are financially unviable. A well designed ASIC will outperform a programmed FPGA for any given job - if not in performance then in power consumption.

In the case of Chord I've noticed just enough latency to make production work a touch awkward at times, but it really is just a bit. If you're just listening then there's no problem at all.

FPGAs are not some secret HiFI sauce despite how hi end DAC producers market them - they're just a cheaper way for them to "make" their own DAC chips. That said that doesn't mean that FPGAs are bad - in the right hands they can sound damned good.
 
Jul 6, 2020 at 4:52 AM Post #5,676 of 7,175
.Another clarification, sorry!
If we used a recording setup comprising of just one or two mics and recorded them in PCM and DSD, without editing, mixing or mastering, no one could tell the difference in a listening test. This is a contrived test though, that as far as I'm aware NEVER exists in commercial classical recordings. Commercial classical recordings typically use a minimum of 3-4 mics and sometimes more 50, so they have to be mixed, plus they're virtually always mastered and almost always edited. So, virtually all 1 bit DSD recordings are converted, still no audible difference though. The only exceptions I know of are a few recordings that have been live mixed in the analogue domain and haven't be edited or mastered.
Ik this doesn't happen. I was making a point that if we did the setup, it would be no difference in the sound.
 
Jul 7, 2020 at 6:25 AM Post #5,678 of 7,175
Okay so much higher sample rate allows exactly same signal as PCM, except for the possible noise issue.

Sort of! DSD does have a great deal more noise than say 16/44 PCM. However, given the far larger audio bandwidth of DSD/SACD, all that noise can be spread over a far wider band. You can easily see this in a spectrogram, the dither noise typically reaches it's peak around 25kHz and continues throughout the rest of the spectrum. This amount of ultra-sonic noise can potentially cause an IMD (inter-modulation distortion) issue with downstream equipment, amps and/or headphones/speakers, which is why Sony implemented a 50kHz analogue filter in it's SACD players and REQUIRED an analogue filter at 30kHz - 50kHz in it's licence agreement for third party manufacturers. This fact obviously makes a nonsense of audiophile claim that SACD contained important/useful information up to 100kHz.

In practice, as far as audibility is concerned, SACD (1bit DSD) and CD are pretty much identical when it comes to noise. In the critical hearing band 1bit DSD provides roughly 120dB of dynamic range and 16/44 (that also has noise-shaped dither applied), in the critical hearing band provides roughly .... 120dB of dynamic range!!

[1] Is FPGA better than chips and provide a lower latency and better sound, or it it, once again, placebo?
[2] Asking in regard of Hugo 2 vs Pro iDSD

1. How does lower latency provide better sound? Latency just refers to the delay between pressing play and the audio playing, or in recording situations it refers the time difference between the analogue input signal (say from a mic pre-amp) and the analogue output signal after passing through the AD/DA (digital conversion) loop. Typically, this delay/time difference is just a few milli-secs and to put that into perspective, the blink of an eye is typically around 300ms! As far as "better sound" is concerned, it's all relatively simple math, it doesn't matter what type of chip you use (ASIC or FPGA), they can all do the math required for audibly transparent/perfect reconstruction (conversion back to analogue).

2. Chord tout it's FPGA in terms of it's speed/ability to run the more demanding math of their bespoke reconstruction filter, that has more "taps" than conventional reconstruction filters. All very impressive as far as audiophile marketing is concerned but conventional reconstruction filters are already audibly perfect, so any audible difference cannot be anything other than placebo!

The results of Mark Waldrep's "HD Audio Challenge" are live.

I always had a lot of respect for Mark Waldrep, both in terms of the quality of his recordings and his generally "no nonsense", factual approach to marketing. There was one exception though, I did disagree with him about his statements of the audible benefits of high-res audio. As he states, " I was convinced that high-resolution recording — real HD-Audio — would be perceptible." and "I was among the strongest advocates for this new and exciting “upgrade” to audio reproduction." However, a few years ago he (reliably/controlled) tested HD vs CD and kudos to him, he actually publicly admitted he couldn't distinguish them.

Now he's gone a step further, tested hundreds of others and states: "The hundreds of people that have participated in the second round of the HD-Audio Survey, have confirmed the results of the previous project. It is no longer possible to claim that “hi-res audio” is an important next step in the evolution of audio. HD-Audio is completely unnecessary for the reproduction of hi-fidelity." !

G
 
Jul 7, 2020 at 10:15 AM Post #5,679 of 7,175
Now he's gone a step further...."It is no longer possible to claim that “hi-res audio” is an important next step in the evolution of audio. HD-Audio is completely unnecessary for the reproduction of hi-fidelity." !

G
What you want to bet...this will have very little impact on anything. As much as I agree with his findings, this is like trying to stamp out a false religious belief. People believe because they want to, facts be damned, and this one is reinforced by "higher and more is better", even if it isn't really.
 
Jul 7, 2020 at 8:05 PM Post #5,680 of 7,175
I used to love the sound of CDs and chuckle to myself at people that listened to vinyl. One day I had the chance to sit and listen to a quality 24bit recording. As I listened, my mind became open to the idea that perhaps there was something better than redbook. There was no positive bias towards analog / 24bit for me (in fact quite the opposite) but my position shifted. I have certainly heard recordings that are indistinguishable 16bit / 24bit. I have also heard drastic differences in quality of the 16bit / 24bit sound. Perhaps that is due to the care which goes into the preparation. I do know that highly compressed music causes me listening fatigue. That may be a factor for others as well.

What I do not understand why some wish to deprive others of the joy of listening to something that sounds better to them. If one listens and doesn't hear something better - great, move on. You saved some money for the future. No one is forcing anyone to listen to 24bit. I listen to what I like because I enjoy it. Others may find a "premium product" increases their satisfaction level. Perhaps there are certain people who have better hearing capabilities than the majority of the population which are more sensitive to 16bit / 24bit differences. I just don't understand why this has to be ideological polarizing. I don't think it would hurt me to let others be wrong in this case. My advice is let the people enjoy the music how they choose.

Peace
 
Jul 7, 2020 at 8:08 PM Post #5,681 of 7,175
I have also heard drastic differences in quality of the 16bit / 24bit sound.

I'm interested... Which specific recordings and what process did you use to compare them?
 
Jul 7, 2020 at 8:28 PM Post #5,682 of 7,175
I used to love the sound of CDs and chuckle to myself at people that listened to vinyl. One day I had the chance to sit and listen to a quality 24bit recording. As I listened, my mind became open to the idea that perhaps there was something better than redbook. There was no positive bias towards analog / 24bit for me (in fact quite the opposite) but my position shifted. I have certainly heard recordings that are indistinguishable 16bit / 24bit. I have also heard drastic differences in quality of the 16bit / 24bit sound. Perhaps that is due to the care which goes into the preparation. I do know that highly compressed music causes me listening fatigue. That may be a factor for others as well.

What I do not understand why some wish to deprive others of the joy of listening to something that sounds better to them. If one listens and doesn't hear something better - great, move on. You saved some money for the future. No one is forcing anyone to listen to 24bit. I listen to what I like because I enjoy it. Others may find a "premium product" increases their satisfaction level. Perhaps there are certain people who have better hearing capabilities than the majority of the population which are more sensitive to 16bit / 24bit differences. I just don't understand why this has to be ideological polarizing. I don't think it would hurt me to let others be wrong in this case. My advice is let the people enjoy the music how they choose.

Peace
Well thats fine on the other head-fi sub forums, but this is sound science. Sound science is just that the science of the audio.
 
Jul 7, 2020 at 8:38 PM Post #5,683 of 7,175
My advice is let the people enjoy the music how they choose
The point of knowing and discussing science in an audio forum is to get a better grasp of the inner workings of our equipment. The point is not only to enjoy, is also to learn and improve.
 
Jul 7, 2020 at 8:42 PM Post #5,684 of 7,175
If I find I like the taste of rat poison in my morning coffee, I hope someone deprives me of the joy by giving me a clue!
 
Jul 7, 2020 at 8:44 PM Post #5,685 of 7,175
I'm interested... Which specific recordings and what process did you use to compare them?
There are many examples listed in this thread
Well that's fine on the other head-fi sub forums, but this is sound science. Sound science is just that the science of the audio.
Science gave us the flat earth and many fallacies in history. It is not perfect and only explains a very small fraction of the knowledge of the universe. Of course it is important, but it is also limited. It's also important to understand why there are outliers rather than dismissing them.
 

Users who are viewing this thread

Back
Top