[rb2013]

  I should note not all SSD's are the same speed wise.  The inexpensive PNY 120 SSD I bought may actually be slower then a totl HDD:
 
http://www.thessdreview.com/our-reviews/pny-cs1311-ssd-review-120gb480gb/5/

 
My HDD was a WD Black 7600rpm 2TB - so close in speed to the 1TB WDD Blue shown above.  Note the PNY 120GB CS1311 speed.  I probably should have - and may down the road move to a larger faster SSD.  Although I have to say the boot times have dropped significantly, even with the 120GB.
 
But I believe this is what has caused the improved SQ with the PNY SSD vs the WD HDD:

 
Power demands on the SATA MB bus orders of magnitude less then the WD HDD.  Less power supply load means less PS noise for both the drive and the MB/CPU.  Not to mention the lack of motor vibration, read head servo noise modulation, ambient noise, etc...

What is interesting in this test - the older generation PNY CS2 211 240GB (373MB/S) )and 480GB (403MB/S) significantly outperform the newer generation PNY CS1311 480GB (210MB/S) and 120GB (96MB/S). 

 

[rb2013]

It appears that these SSD drives come in SLC versions - but like the SLC MicroSD and SLC USB Flashdrives are very expensive: Intel 32GB SLC SSD - $337
 
https://www.amazon.com/Intel-X25-SATA-Drive-SSDSA2SH032G101/dp/B001RLG0GY
 
Is there a SQ difference?
 
Reading that PC Mag review the newer gen PNY use a small SLC cache on the drive then the slower TLC -
 
So for medium sized files the transfer speeds are blazing fast - then drop off as the SLC buffer becomes saturated.  The test was done with 30GB movie files - but they show the results before saturation:
 

Finally, we wanted to see how performance was in a real world use when transferring large files to the SSDs. As we know, most TLC NAND based SSDs utilize a SLC caching algorithm to help improve performance, especially writes. This is needed due to the inherent slowness of TLC NAND. For this test we are going to simply stress write performance by transferring over a 30GB folder of movies off of one SSD to this one and time how long it takes. Once complete we can calculate the average speed.

 
 

During our real world file transfer test we can see that while benchmarks display high sequential write speeds within their testing file size of just a few GB, when testing more than that these TLC SSDs will slow during transfers. For the most part, the 120GB model is second to last at an average write speed of 96MB/s, however, the 480GB model was able to maintain above 200MB/s during testing.
To build upon this test we also looked to see where the speeds leveled off to by using HD Tune Pro.

120GB Left(top)/480GB Right(bottom)
 

 


 
 
So it looks like the 120GB SLC cache hits saturation at about 3GB and the 240GB at 35GB approx.   Both probably large enough for audio files without slow down.
 
So it seems there is little need for a SLC.

 

[somestranger26]

  I should note not all SSD's are the same speed wise.  The inexpensive PNY 120 SSD I bought may actually be slower then a totl HDD:

Any plans to try an M2 NVMe drive instead of a SATA drive, to see if there's any difference in sound quality?
 
Like this one: http://www.samsung.com/us/computing/memory-storage/solid-state-drives/ssd-960-evo-m-2-250gb-mz-v6e250bw/
 
It has a 13GB SLC cache to improve the speeds drastically over a regular TLC drive.
 

 

 
The USB sticks appear as regular Windows drives.  So yes they could be used to boot WIN10 - every one of the newer PCs has a USB stick boot option in the BIOS.
 
All that would be needed would be a copy of the C: drive (with the correct OS install) to one of the USB stick drives - then just set the BIOS to boot off that drive.

In order for this to work, you would need to do a full drive clone, not simply copying the C: drive. If you do this, ideally you would use a source and target drive of the same size, or with a larger target; downsizing while cloning has the potential for OS corruption.
 
The C drive is just the main partition on the drive, but there are others like MSR, EFI, GPT, etc. where the bootloader info and such are stored. The partition scheme depends on whether you're using an MBR or GPT partition table, but either way it's not as simple as copying the files on one partition to another.

 

[Superdad]

  Any plans to try an M2 NVMe drive instead of a SATA drive, to see if there's any difference in sound quality?
 

 
Yes, getting rid of active SATA interfaces has long been a sonic plus for me.  PCIe flash drives are a great way to go.

 

[rb2013]

  Any plans to try an M2 NVMe drive instead of a SATA drive, to see if there's any difference in sound quality?
 
Like this one: http://www.samsung.com/us/computing/memory-storage/solid-state-drives/ssd-960-evo-m-2-250gb-mz-v6e250bw/
 
It has a 13GB SLC cache to improve the speeds drastically over a regular TLC drive.
 

 
In order for this to work, you would need to do a full drive clone, not simply copying the C: drive. If you do this, ideally you would use a source and target drive of the same size, or with a larger target; downsizing while cloning has the potential for OS corruption.
 
The C drive is just the main partition on the drive, but there are others like MSR, EFI, GPT, etc. where the bootloader info and such are stored. The partition scheme depends on whether you're using an MBR or GPT partition table, but either way it's not as simple as copying the files on one partition to another.

None at all - on a dedicated music server running Fidelizer Pro to strip processes to a bare min - that would fruitless overkill.  Running a typical redbook audio track of say 50MB the speed of a SATA SSD like the PNY 120GB is 446MB/s up to say 3GB.  So that audio file is using roughly 0.10 of the throughput capacity.  No need for a faster connection.  But for a bit more money you could use the 240GB PNY with 493MB/s throughput up to say 35GB - now you're down to like .0986. Even with a Hi Res 24/192 audio file with approx 250MB you still are no where near reaching a major of the max throughput.  Read the test clearly the WD HDD advantage is only on a 30GB video file - after the SLC cache reaches saturation.And that  was for the slower write speeds - read speeds (as in audio file playback) of SSD's are even faster.
 
Now more importantly is getting that SSD drive power source away from the MB SATA power feed.  See my post regarding powering the SSD with an ext LPS - for $6.
http://www.ebay.com/itm/221446206429?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
 
The PNY SSD drives come with a neat drive cloning software called ACRONIS True Image - it imaged a 1TB WD to the 120GB PNY SSD without issue (of course the C: drive was only using approx 60GB).

 

[rb2013]

   
Yes, getting rid of active SATA interfaces has long been a sonic plus for me.  PCIe flash drives are a great way to go.

I think the more important issue is moving the power source away from the PC MB - whether it's SATA or PCIe to an ext LPS:
 
Sound familar?

 
http://www.head-fi.org/t/829639/usb-strikes-back-watch-out-aoip-usb-ethernet-chain-beats-all-at-least-for-me/135#post_13131323
 

06-06-2015, 08:48 PM

#157​

​

Superdad

Music lover; Owner, UpTone Audio

​

Join Date​

Apr 2008

Location​

Mariposa, California

Posts​

4,272

Hi Kiwicol:
Glad you are able to appreciate the SD card. But suggesting that an SSD will thus sound better than a spinning drive is the wrong conclusion. Music storage/playback SQ differences are all about the active drive interfaces--their own and the computer's chipsets for the same--and the activity, noise, and power supply spikes they generate at playback time in the DAC-connected computer. SATA, FireWire, and Thunderbolt all generate far more "stuff" than SDXC, PCIe flash, or Ethernet attached storage.

I've swapped a HD and an SSD into the same OWC FireWire enclosure (powered by an LPS) and heard virtually no difference. But put either of those drives into a different enclosure (the quad-ported OWCs with eSATA, USB 3.0, etc. sound worse), and THAT is easy to hear.

​

 

 

[Adosero]

I remember having read somewhere that the extra heat caused by PCIe was a no go for audio but I didn't try it personally.

 

[johnjen]

My Mac has a PCIe bus SSD which is my boot volume, and while this isn't a ram disk there probably isn't much difference, performance wise.
 
JJ

 

[rb2013]

  I remember having read somewhere that the extra heat caused by PCIe was a no go for audio but I didn't try it personally.

I think the way these devices are powered - best externally with a low noise LPS - may be the critical factor.  Maybe going with an ext PCIe case/ SSD powered by a LPS.  But that gets even more complicated and expensive.  A SATA SSD powered externally by a LPS should do the trick.  Anyway I'll report what I find with my set-up in the next week so once that cable arrives.

 

[Mr Underhill]

***Whoops***
 
Been a plonker.
 
Having not seen quiet what I expected for my USB F1 interface via the microRendu I thought I would see if there was an updated Firmware available. I loaded up the drivers and found I was running Firmware v2.0. On the Kitsune site they had firmware 2.2 for the F1 and so I loaded it, it completed and .......appears to have turned my F1 into a neat black brick. The F1 powers but there is no communication going on.
 
Just in case you are tempted to try firmware from other than the manufacturers site, please learn from my sad experience.
 
Sigh.
 
On a brighter note I have:
 
1. Sold my Mutec3+USB;
2. Ordered a replacement F1;
3. Ordered an IFI Purifier 2.
 
I also bought a 128GB USB 2.0 stick and copied some files onto it, this I plugged in to my NAS to have a listen.
 
All the tracks via the USB appear to have slightly more detail, but also have more edge. The only track which sounds better is Signe from Clapton Unplugged, aided I suspect by the lack of vocals.
 
I have tracked down a site selling the Startech at a reasonable price and I am probably going to order one tonight.
 
M

 

[rb2013]

  ***Whoops***
 
Been a plonker.
 
Having not seen quiet what I expected for my USB F1 interface via the microRendu I thought I would see if there was an updated Firmware available. I loaded up the drivers and found I was running Firmware v2.0. On the Kitsune site they had firmware 2.2 for the F1 and so I loaded it, it completed and .......appears to have turned my F1 into a neat black brick. The F1 powers but there is no communication going on.
 
Just in case you are tempted to try firmware from other than the manufacturers site, please learn from my sad experience.
 
Sigh.
 
On a brighter note I have:
 
1. Sold my Mutec3+USB;
2. Ordered a replacement F1;
3. Ordered an IFI Purifier 2.
 
I also bought a 128GB USB 2.0 stick and copied some files onto it, this I plugged in to my NAS to have a listen.
 
All the tracks via the USB appear to have slightly more detail, but also have more edge. The only track which sounds better is Signe from Clapton Unplugged, aided I suspect by the lack of vocals.
 
I have tracked down a site selling the Startech at a reasonable price and I am probably going to order one tonight.
 
M

Oh I heard about that - I think that firmware update was for the SU-1 possibly?   I don't think the firmware on the F-1 is user updateable - but can be by the factory.  I'd contact Singxer directly about that.  If possible to salvage it would be worth it.
 
The Mutec is gone  - I suppose it was bettered by the F-1?  Not worth the cost to keep as a SPDIF reclocker?
 
As I mentioned the USB stick needs burnin (I know I get flamed here about this - but it's what I have experienced - they have crystal XO clocks and mico-processors like the F-1 - so why not?).  I found 100-150 hours things smoothed out.  Of course the real benefit comes with running from behind the GI wall of the CAT6 Ethernet UTP on the Startech REX.
 
Happy to hear I will have a positive or possibly negative confirmation on another system.
 
Speaking of salvaging a F-1 the Wolfson i2s to SPDIF board came in.  I hope to use it to resurrect my friends modded F-1. I will power the Wolfson SPDIF board with a separate 5VDC LPS.  It will an interesting test of this alternative SPDIF coax path in terms of SQ.
See below:
 

 

[mtruong34]

***Whoops***

I have tracked down a site selling the Startech at a reasonable price and I am probably going to order one tonight.

M

Care to share this site which may benefit interested others?

 

[taz23]

@rb2013, I remember that you will be comparing the Pro3z with the F-1 when you receive the Pro3z.  I wonder how is that coming along?
 
Looking forward to your impressions, and thanks for sharing your hard work with all the informative posts.

 

[Mr Underhill]

 
Oh I heard about that - I think that firmware update was for the SU-1 possibly?   I don't think the firmware on the F-1 is user updateable - but can be by the factory.  I'd contact Singxer directly about that.  If possible to salvage it would be worth it.
 
The Mutec is gone  - I suppose it was bettered by the F-1?  Not worth the cost to keep as a SPDIF reclocker?
 
As I mentioned the USB stick needs burnin (I know I get flamed here about this - but it's what I have experienced - they have crystal XO clocks and mico-processors like the F-1 - so why not?).  I found 100-150 hours things smoothed out.  Of course the real benefit comes with running from behind the GI wall of the CAT6 Ethernet UTP on the Startech REX.
 
Happy to hear I will have a positive or possibly negative confirmation on another system.
 
Speaking of salvaging a F-1 the Wolfson i2s to SPDIF board came in.  I hope to use it to resurrect my friends modded F-1. I will power the Wolfson SPDIF board with a separate 5VDC LPS.  It will an interesting test of this alternative SPDIF coax path in terms of SQ.
See below:
 

Hi Bob,
 
Thanks for the thoughts, I am hoping my error will help others avoid the same fate.
 
I did email Singxer. They got back to me a day later and advised me to just use firmware from their site ....no mention of a possible fix. I will watch your experiment with interest.
 
I will keep running the USB Stick .....but I was not surprised. The effect for me was akin to the effect of SMPS on sound quality. From my POV I am enjoying doing this one step at a time.
 
M

 

[Mr Underhill]

Care to share this site which may benefit interested others?

https://www.ballicom.co.uk/usb2g4lext2-startech-com-4-port-usb-2-0-over-lan-.p1009343.html?ref=PLA&gclid=CIfr9Le9rdECFeop0wodmpgLMw
 
£436 inc VAT