No. The raw BER goes up in NAND flash as a function of the number of program / erase cycles and also over time since last written (retention).
Good commercial / enterprise SSDs are very complicated and your oversimplification is not helpful and based on half-baked knowledge. The specifics of how this works is very drive dependent as well, but no. Honestly, as a user, you don't even have any clue if the data you just wrote to a consumer SSD is even stored in the MLC/TLC/QLC flash. Some drives have large amounts of SLC cache, some drives have large DRAM caches. In the case of the SLC cache there is no hard and fast rule about when that data will be committed to the slower flash because it's already non-volatile. Further, all controllers have very advanced wear leveling that monitor the RBER of the drive (and possibly every page individually) as it ages.
If things worked as you claimed, then maximum drive IOPS would be affected by power quality, which is demonstrably false.
What drive do you use? It's probably TLC flash unless it's older or very expensive. Better avoid QLC drives like Intel 660p like the plague then if you think the amount of error correction the controller performs affects your sound. What a joke.
If you really believe this, you should buy an Intel Optane 905p SSD. Optane is 3D XPoint (chalcogenide phase change memory), not NAND. It has a much lower inherent bit error rate than NAND and at least an order of magnitude better endurance.
Last edited:
Thanks Bob, yeah it's looking like it's 'let's gang up on Dan day' lol, it's water off a duck's back to me while the negative ego derived comments are borne of ignorance, belief systems and as you say it, stupidity.
I say you have good ears and call it how you hear it, keep trusting your ears.
.
I spoke about standard thumb drive plugged into various stand alone systems with USB input socket.
You are saying joke when you don't know the test conditions.
Fake Bybees? You are being critical of John Curl.A made-up one-liner story you tell, fake as fake Bybees
My apologies, I don't usually read all of your delusional ramblings.
And the controller of a USB flash drive still does wear leveling based on bit error rate and most of what I said is still accurate. You are clueless.
If you think this matters - which it does not - then don't use a USB flash drive which have one step above junk-grade flash.
Is there not some player made by a delusional audiophile that reads the entire file into memory before playback?
Last edited:
You're right. I am definitely wasting my time if nothing else.
Whatever you hear or think you hear, a 30 second Wikipedia level understanding of flash memory isn't the root cause.
This BS is tiresome and discourages any new people with decent reasoning skills from participating.
Flash memory is prone to write and read errors, this is known, this is expected and this is why error correction is designed in to the system.
Multiple TDs exhibited changes between recorded files and multiple subjects identified the changes.
The reasons for the subjective changes may be open to conjecture, the fact of such subjective changes is not open to conjecture despite your protestations.
Multiple TDs exhibited changes between recorded files and multiple subjects identified the changes.
The reasons for the subjective changes may be open to conjecture, the fact of such subjective changes is not open to conjecture despite your protestations.
Yes and via Squeezbox, you cant tell which duplicated albums is the MQA one because there is no indication via that interface. I promised myself to never play a MQA file through my system.
//
Back to old habits, Ovi?
The most useful advice for you was to try listening more. Look at the numbers is all you know how to do.
You quoted me above but you actually edited the quote. A sin, whatever the reason.
//
That you put it in a location where it does not matter, in other words, overspecced the part.
Use it for AVCC in a ESS DAC, or the comparable voltage reference pin on an AKM DAC, or any other location with poor power supply immunity, and your statement is patently untrue.
It was at least in part a question. Is there a player that reads the file from disk entirely before starting playback? It's something I can imagine an extremely paranoid person would do. I thought I've heard of some player that did this, but I don't want to misstate the facts.
I did not insult any makers of digital audio players, except those that are idiots. I do not know of any examples that allocate a buffer for the entire file. If any player reads the entire file into memory for playback purposes only, the author(s) deserve my scorn.
Well, it's not even a big sin these days with the amount of RAM available on a PC... but it is arguably bad practice anyway aside from the asinine reasoning that the storage device will not have to be accessed during playback.
Last edited:
Dan, when you hear a difference between 2 bit perfect copies of the same file, then one explanation would be that you have supernatural hearing, unmeasurable and unprovable to the norms. A waste of time and effort to pursue here.
I'll bite on this. I am 100% sure there is a measurable difference if you use a 7805 for VREF / AVCC, but how sure are you it's audible?
It might be, but I would not bet my life on it. The Rohm 7805 that AKM uses has a noise spec of 40 uV RMS from 10 Hz - 100 kHz. Not state-of-the-art, but appears to be lower than TI's LM317 if I read the datasheets correctly. PSRR is a bit higher in the audio band, too. Not amazing compared to new regulators, but I bet I could feed that to the VREF of an AK4499 and normal humans would fail to identify it in a blind test.
Agreed. Maybe he has X-Men mutant power of hearing.
Last edited:
Perhaps you are correct, I haven't followed the matter closely for far too long. However digital playback are not only PC based, it is going into smaller or embedded platforms like the Raspberry Pi with multi channel filtering, equalization and crossover function.
There you go - see attachment.
A file player (Amarra) by the same company that created mastering software that up until recently was used in production of about 90% of all digital releases on the market (CD and High Res) since 1989. Player uses the same playback "engine" as the mastering software. Modified (encrypted) version of this player is produced on the fly by the mastering software and is sent to client for playback of final DDP file to get their approval.
Now go and call Sonicstudio and tell them that you know better.
Ignorance is indeed a bliss.
Attachments
Chris, afaik ESS recommends the use of high end opamps such as the AD797 to generated a stable and noiseless reference voltage. I am not aware of AKM recommending a cheap bit like a 7805 for their reference voltage, so if you have a link, I am interested. What I am aware of is for example the voltage regulator for the AK4497 https://www.akm.com/content/dam/doc...dac/ak4497eq/ak4497eq-en-evaluationmanual.pdf. At the end of the document, it shows quite an elaborate schematic for a voltage regulator based on an AD817.
- Status
- Not open for further replies.