
The Sound
Contents
Let’s get right to the point: we hear absolutely no difference between the Samsung SSD and the Zzyzx Femto SSD. We really tried to hear something, but we did not succeed.
Perhaps it is due to the fact that we are playing with a Mutec MC3? Or that the Alpha PC is not good enough (no linear power supply)? Or maybe our ears are not good enough. Who knows.
RAM buffer
What could also be possible is that every piece of decent software takes in the data and puts it into RAM (Random Access Memory). That RAM – the fast memory of the PC, NOT the SSD or hard disk! – is used by the software to buffer a few seconds. Some software load the entire song in memory.
So in practice there is no playing from a disk at all. And timing is totally unimportant: the PC reads in “bursts” to the RAM. This is also easy to see in the task manager of the PC or Mac: you can see when you load a track, that the sofware briefly reads from the disk and puts it in the RAM.
With a streamer playing from a network, the same thing is done: the streamer first puts the data in the buffer (also to the streamer’s RAM) and starts playing. You will see that if you disconnect the network connection, the streamer just keeps playing.
Buffering the errors?
“But…” audiophiles will say…. “the errors can be buffered to RAM!”.
No: it can’t, because – as with networking and a copy to for example, an external usb-drive – error correction is built in. When a PC copies the data from the SSD to the RAM, checks are done – practically in real time – for data integrity. If that didn’t happen, there would be loads of data corruption!
For your information, there are also numerous error correction algorithms in a processor. Without those corrections, PCs would crash every minute.
“OK…. so no data errors. But that audio SD has less noise…”.
Yes: less noise is always good. However, the question is: is that one SSD the real culprit in the PC? Look at the environment in which this SSD is operating. You place it in an M.2 slot (basically pci-express) of a motherboard. A pcb packed with voltage regulators and a central processor with now billions of transistors. A processor that can draw tens of amperes under peak load and does its work at GHzs. We estimate that it generates slightly more noise than an m.2 ssd.
Our opinion
While this m.2 SSD certainly generates less noise, – in our opinion – the noise of an m.2 SSD is a drop in the ocean.
Thereby, it is totally irrelevant, since the data from the SSD is simply buffered in RAM. That RAM is not powered by a nice, clean power supply AND does not have a separate low phase noise Crystek clock.
In short: any “advantages” that might be there are killed within nanoseconds of pressing Play. Or are we now giving away an idea for the next product…. audiophile RAM?”











Great article. It’s a pity that the English comment regarding Revelation SSD and better RAM for audio is not shown here because I completely agree with the 1st comment: did you miss something?
There have been too many good results with the Revelation SSD + external power supply music files for at least 2 years to ignore. I would think that something else is going on with your set or with the (resolution) of the Alpha Audio computer if this does not improve anything audibly. (Or the mutec or your power supply or whatever?)
Since even longer ago it has been known, especially for DIY servers, that other RAM such as Apacer industrial wide temperature can also provide a huge improvement (if the Alpha Audio computer still uses standard RAM, definitely try this).
Julie OS could also be put on a smaller (eg 16 Gb) Optane M.2 memory for even better sound (if there is an M.2 slot left, definitely try!) Innuos used (only now) in the latest server SLC Apacer memory for the OS.
The music SSD fed with BV the JCAT Nano or an (even) better power supply should also give a better sound.
To make RAM sound even better, the power supply of the computer must also be of a higher level (the very best BV. Sean Jacobs). If the motherboard has 2 connections, definitely try with 12V 4pin and the ATX 24 pin with separate power supply. Also if the power supply is distributed, eg independent motherboard, SSD for OS and SSD for music, it also sounds better.
I’m certainly not saying here that you’re not doing well, so much and fast is being done in audio, but I recommend trying even more if you can’t hear a difference with this Revelation SSD or (Apacer) RAM.
There are simply too many users with very good results after upgrading memory (be it music SSD, RAM and OS memory). Also not even that expensive compared to other audio improvements. But as always not every system sounds the same and it can be more or less audible either ; YMMV.
I am very curious about results! Good luck!
Boy, are you late!
There IS Audiophile RAM on the market. For several years now. Apacer’s industrial grade RAM is touted as being superior to other types. I’m not smart enough to know how it’s supposed to work, but cruise over to Audiophile Style to get in on the conversation.
Neither am I smart enough to figure out if that powered SSD card works. I just know I almost went into cardiac arrest at the sight of the price.
Only God knows why we do what we do, and He’s apparently keeping it to Himself…because gods just wanna have fun. They just wanna. 🙂
Kan je navraag doen hoe de fabrikant deze SSD geïmplementeerd wil zien?