OhGizmo!: OhGizmo! » Archive » Kalista CD Player Takes on Vinyl
Joakim Arfvidsson
· 4 years ago
This is for the stupid kind of audiophile.
And technically, this is ridiculous. To reproduce vinyl sound, a CD player does not have to be better. It just has to be worse, and add in whatever stuff the vinyl process does. Like scrapings, bending, distorsion, uneven frequency response.
I've been wondering for a while, could I make money by targeting an overprices product at stupid people? That's what this segment of the audio business is doing.
Connor
· 4 years ago
Actually, it's not that ridiculous. You can upsample the data on the CD and interpolate it several different ways to get a richer sound from a 16/44.1 recording. Also, to get that vinyl sound you could do some subtle saturation and dynamics on the upsampled digital signal. However this could all be done in software on a reasonably spec'd PC so there's really very little justification for such a steep price tag. CDs aren't better quality than vinyl. In-fact, it's arguably impossible to make a comparison on quality since they use entirely different sound-replication models. However I think the digital -vs- analog argument is beyond the scope of this thread.
Todd
· 4 years ago
I think it is within the scope here since the ultimate goal of high-end cd playback is to emulate vinyl (eluded to above). That said, I've heard an $8000 cd player A/B'd with a $350 turntable and the contrast is stark. The only cd playback system I've heard the approaches analog sound is the dCS stack (about $32,000 for the 3-box system).
Josh Ray
· 4 years ago
Hey guys, this is Josh Ray over at SonicFlare. To clear up some confusion, vinyl is almost always better. If the mastering on the LP is terrible and the CD is great then, hey, CD will beat it. But all being equal, vinyl gives CDs the smackdown.
I recently posted an article on the "jitter" issue that gives CDs their nastiness:
Also, the real question isn't if vinyl is better than CDs, but if computer audio is. The article above also explains why computers audio, done right, can murder CDs by totally bypassing jitter and allowing for extremely advanced software EQ, upsampling and room correction.
Expect to see high priced HDD servers that put all these features into an audiophile-approved package. I've heard a few of these systems and, let me just say, they're amazing...plus, you don't have to get off your lazy ass to change the CD or LP.
Joakim Arfvidsson
· 4 years ago
Problem with CD as I see it is that they shot for "good enough, while minimizing bitrate" when designing it. Theoretical signal processing (I'm in the field) proves that 44.1 kHz is sufficient to reproduce frequencies up to half that exactly, given samples of infinite resolution.
I'm not positive that 16 bits per sample is good enough to approximate "infinite", especially when you want to have some dynamic range. With 24 (giving a resolution that is 256 times better), we're fine.
For the frequency cap of 22.05 kHz, that's more than I can hear. But I would go for safe and welcome the 96 kHz sampling rate of the new formats.
The jitter problem is bad design on the part of some player manufacturers. As noted, it's not even expensive to solve. It is trivial [for a conscious electrical engineer such as myself] to design something with lower jitter and sway than a turntable could achieve. The solution would include digital buffering.
Now that the signal is perfectly well defined according to theoretical results, the last question is DA conversion. There are ways to go wrong there, and this is the only interesting area.
It seems like a completely fucktarded way to have a low jitter, having a good clock to regulate speed of the disc and all this mechanical stuff, when they could use that same clock to pour stuff out of a digital buffer.
All Comments
And technically, this is ridiculous. To reproduce vinyl sound, a CD player does not have to be better. It just has to be worse, and add in whatever stuff the vinyl process does. Like scrapings, bending, distorsion, uneven frequency response.
I've been wondering for a while, could I make money by targeting an overprices product at stupid people? That's what this segment of the audio business is doing.
You can upsample the data on the CD and interpolate it several different ways to get a richer sound from a 16/44.1 recording. Also, to get that vinyl sound you could do some subtle saturation and dynamics on the upsampled digital signal. However this could all be done in software on a reasonably spec'd PC so there's really very little justification for such a steep price tag.
CDs aren't better quality than vinyl. In-fact, it's arguably impossible to make a comparison on quality since they use entirely different sound-replication models. However I think the digital -vs- analog argument is beyond the scope of this thread.
I recently posted an article on the "jitter" issue that gives CDs their nastiness:
http://www.sonicflare.com/archives/computer-aud...
Also, the real question isn't if vinyl is better than CDs, but if computer audio is. The article above also explains why computers audio, done right, can murder CDs by totally bypassing jitter and allowing for extremely advanced software EQ, upsampling and room correction.
Expect to see high priced HDD servers that put all these features into an audiophile-approved package. I've heard a few of these systems and, let me just say, they're amazing...plus, you don't have to get off your lazy ass to change the CD or LP.
I'm not positive that 16 bits per sample is good enough to approximate "infinite", especially when you want to have some dynamic range. With 24 (giving a resolution that is 256 times better), we're fine.
For the frequency cap of 22.05 kHz, that's more than I can hear. But I would go for safe and welcome the 96 kHz sampling rate of the new formats.
The jitter problem is bad design on the part of some player manufacturers. As noted, it's not even expensive to solve. It is trivial [for a conscious electrical engineer such as myself] to design something with lower jitter and sway than a turntable could achieve. The solution would include digital buffering.
Now that the signal is perfectly well defined according to theoretical results, the last question is DA conversion. There are ways to go wrong there, and this is the only interesting area.
It seems like a completely fucktarded way to have a low jitter, having a good clock to regulate speed of the disc and all this mechanical stuff, when they could use that same clock to pour stuff out of a digital buffer.