Popular Post Nenon Posted March 27, 2019 Popular Post Share Posted March 27, 2019 This would be an attempt to extract some information from the “A novel way to massively improve the SQ of computer audio streaming” thread related to computer storage. People are trying different methods. Here are some of the options: 1. SSD drive 2. HDD drive 3. External HDD with external power supply 4. Optane memory 5. NAS accessed over the network 6. Run everything in ramroot Let’s discuss each of these. 1. SSD drive The general consensus is that SSD is noisy and should not be used. Powering by LPS helps but does not eliminate the noise. On 3/24/2019 at 3:50 PM, romaz said: SSD vs Optane Just for kicks, I decided to compare a 58GB Optane card against a Samsung 500GB 960 EVO NVMe SSD in the M.2 slot of the Asrock board. This was a very brief comparison because it didn't take long to realize how much more harsh the SSD sounded even with all of my isolation schemes in place. It's amazing how many commercial music server manufacturers continue to use SSD drives in their servers as if powering an SSD cleanly somehow addresses this harshness when it does not, at least not to my ears. I suppose you get used to the harshness over time but an SSD is about the worst thing I can imagine putting into a music server with the super fast NVMe drives sounding the harshest of all. As some may recall, in previous testing, I found the older, slower SATA II SSDs (especially the SLC variety) to sound less harsh then the newer, faster SATA III SSDs although the faster SATA III SSDs made music sound more alive and more immediate and so there was a trade off. It would appear that the Optane cards have the best of both worlds and so hats off to Larry for introducing us to the Optanes. I have read comments about how running AL in memory doesn't result in much improvement in SQ except for a slight improvement in smoothness. The point here isn't just running AL in memory for the sake of latency but also to be able to completely avoid using an SSD in the server. The Optane seems to be a nice compromise if capacity, low latency, and low noise are desired since Optane behaves more like RAM than an SSD. For those with a large Roon database who are looking for a brisk user experience with Roon, an Optane drive may be preferable to a USB stick for the Roon database. For sure, it would be preferable to an SSD. From a SQ standpoint, is an Optane drive preferable to having more RAM (16, 32, or even 64GB)? I'm not sure although according to Intel, a 58GB Optane drive only consumes 3.5w and so it would appear to draw much less current than RAM as a 3.3V device. On the other hand, many music servers use SSD including Innuos Statement. On 3/27/2019 at 2:39 AM, romaz said: Many fine companies are using SSDs and some of these companies feel that by cleanly powering these SSDs, these SSDs cease to become noisy. I find that to be untrue, however, not everyone is sensitive to the harshness of SSDs and so you have to ask yourself if it matters to you. To my ears, it's quite clear. 2. HDD Drive Some people say it's better than SSD, others say it's worse. One unavoidable problem is the additional vibrations if the HDD is inside the chassis. I will try to extract some posts on this topic later. On 3/27/2019 at 2:39 AM, romaz said: I personally find hard drives to be less noisy than SSDs but they are vibrating devices and so that's not ideal. HDDs are also not low latency drives and so for an OS, this is not ideal either. On 3/26/2019 at 8:21 PM, sandyk said: So are HDDs noisy as can be seen from the attached. The noise injected from an SSD back into the main PSU has faster rise and fall times, i.e. it is wider bandwidth noise but can be more readily filtered out using additional voltage regulation than a typical internal HDD which ALSO uses the +12V supply for it's motor, not just the +5V supply. SSDs powered from the motherboard are best reserved for processor intensive applications such as 4K Video editing and rendering where their faster speed is a great time saver. Incidentally, the OS SSD should also be powered with cleaner isolated and regulated +5V too, not just a Music SSD. Using additional external Linear PSUs to power an internal peripheral device may result in Earth related issues, not just audible hum. 3. External HDD with external power supply Keeps the HDD vibration out of your chassis and does not pollute your ATX power. On 3/26/2019 at 1:05 PM, ted_b said: Yes, we use a startech USB 3.1 card and a Startech USB 3.1 hdd enclosure, separately powered. The JCAT card is great too. https://www.amazon.com/gp/product/B00XLAZEFC/ref=oh_aui_search_asin_title?ie=UTF8&psc=1 Seems like the quality of cables and power supply makes a difference. On 3/26/2019 at 3:25 PM, elan120 said: On 3/26/2019 at 3:11 PM, auricgoldfinger said: Have you tried different USB cables? I wonder if they make an audible difference. I did. They do have some differences between the ones I have tried, but closer to subtle side. It is a toss up between iFI Mercury and JSSG360'd Lush, with JSSG360'd Lush leading slightly. On 3/26/2019 at 2:16 PM, elan120 said: In my case, power supply quality does make a difference, but I only compare stock wall wart with what I currently have on hand, which I end up with JS-2 on the enclosure, and LPS-1.2 on the PCIe card. On 3/26/2019 at 4:24 PM, numlog said: The USB adapter still uses a similar type of controller chip, as mentioned its possible to use an external supply with it easily but I would think the internal SATA has a more direct data path. If internal SATA supports a lot more inputs than what is needed disabling may be better but I would agree about keeping USB clean... both have potential pros and cons and even if 2 people heard the same thing they might not agree on which sounded best, the only way to really know is to try it out yourself . 4. Optane memory Very good feedback so far. The problem it has less memory than HDD, or SSD, or SAN, which makes it not very suitable for a huge collection of ripped CDs. Very good option for OS booting. It's also not supported on every motherboard. On 3/27/2019 at 2:39 AM, romaz said: Optane is great for the OS and for playback software like Roon or the Roon database. They're too small to be practical for storing all your music files but that doesn't really matter because with respect to noise, it is the OS drive that you have to worry about most since the OS or the playback software is constantly active, even during music playback. With music storage drives, once the file has been transferred to memory, these drives no longer have to be active and so I find it ok to put music files on large capacity hard drives since they're cheap but preferably on a NAS to keep any vibrating devices outside of the server chassis. Optane drives are designed to be able to function like an SSD but the tech they use is very different. They are not the same as SSDs and they have a different noise signature. 20 hours ago, romaz said: I have recently been educated that modern motherboards that have M.2 NVMe slots that aren't necessarily "Optane ready" (i.e. AMD boards) can still utilize an Optane card as a storage (or OS) drive but not as a cache drive, which is what Optane was originally designed for. This means that if your motherboard has an M.2 drive capable of NVMe, you are probably capable of using an Optane drive. It should be easy enough to try it. 5. NAS accessed over the network Good results reported and prefered method to some of the leading members in this forum. It probably makes sense to combine with good network cards and switch(es). On 3/27/2019 at 2:39 AM, romaz said: This is what I'm using and I'm finding that if you use the JCAT Femto Network card, the quality of the NAS becomes irrelevant since any incoming stream, whether it be from a NAS or from the internet (i.e. Qobuz or Tidal) gets cleaned up (reclocked) as it passes through the JCAT card. Without the JCAT card, I found that a file stored internally on an Optane card sounded a bit better (smoother) than a file streamed from my NAS. WIth the JCAT card cleanly powered by a Paul Hynes SR4, I can tell no difference at all among files originating from a NAS, streaming service, or locally from my Optane card (provided the same master is being used). I, too, have tried USB storage using compact flash media and a Lexar hub. With this hub powered by an LPS-1.2 and with this hub connected to my server via an IsoRegen powered by an LPS-1.2 and so a pretty elaborate and expensive setup just to store music files. As good as this sounds, I am unable to distinguish between files from this USB hub vs files from my NAS with the JCAT card in place. With the JCAT card, I am finding that the media storage device no longer matters. At all. 6. Run everything in ramroot Not even sure if this option should be here. It probably belongs to the AL thread. On 3/27/2019 at 2:39 AM, romaz said: Whether I ramroot or keep the OS on my Optane, I can't tell much difference at all. I am finding that Optane cards are very similar to RAM as far as noise and latency. Other interesting posts below. On 3/26/2019 at 11:12 PM, austinpop said: I had done a "Storage Location Study" a few months ago, in this post: https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?do=findComment&comment=900522 At the time, I had found that when the Roon Server was running on AudioLinux/Ramroot, the sound quality varied by music storage location as follows (from best to worst): in RAM, by placing music in the root directory, that gets loaded into RAM with the ramroot boot option. on an Optane SSD partition on my NAS. I just tried this experiment with my new server topology, with the bridged JCAT Net Card Femto, which has so improved my Tidal/Qobuz streaming. I ran the same experiment, and this time the results are indeed different: music in RAM still sounds best music from my NAS is a close second, and beats out... ...music on the Optane! Please note: neither RAM nor Optane are practical for a > 1TB library like mine, but I'm treating Optane in this study to represent DAS (direct-attached storage). The fact that NAS as storage location is now beating out DAS is an exciting development, which really makes the JCAT card an excellent value. I can just use a NAS for music storage, and the JCAT card really elevates the SQ pulling from this source. Now, to be fair, I have not tried other DAS options like the Startech USB 3.1 enclosure with HDD option. But to be honest, I like the simplicity of just letting my live on my NAS. Maybe some day I'll be motivated to try the HDD option, but I will more likely experiment with an LPS on my NAS to see if it even makes a noticeable difference. The fact that music files in RAM still sounds better suggests that Euphony are on to something with their file caching feature. To my knowledge, that isn't something AL does. Perhaps this explains why Roy found he liked Euphony on the Roon Server. odelay, 87mpi and numlog 1 2 Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 27, 2019 Author Share Posted March 27, 2019 I think we need to dive a little deeper into the storage problem and try to understand what is causing the harshness or other differences people are hearing before deciding which media is best. A file is a file. It would be identical no matter if it is stored on a SSD, HDD, USB stick, NAS, or something else. So what makes one storage media worse than another? My guess would be it's the retrieval process of the file. Some storage medias generate more noise than others. SSD must be falling in that category. And I am not ignoring sandyk's post or saying that HDD generates more noise than SSDs or vice versa... I am also not saying that the noise id the main problem, just suggesting that the problem is probably somehow related to the file retrieval process. If the only problem is only during the file retrieval process, we "should" be able to copy a file from let's say a SSD drive to the RAM, shutdown the SATA controller and play the file. And if there was any harshness, it should disappear, right? Unless of course there are other factors here. And I have no doubts there are... But this should not be a difficult test from someone to do with AL in ramroot. Another test we could do is copy the same file to RAM multiple times from different sources and compare. I would gladly do that if I manage to replicate the harshness issue. Bought a Samsung SSD to try it. Also, I wonder if writing a file to a noisy storage device has any effect. And I don't mean effects on the music you are playing while writing the files but later, when you play that same file. Does it make a difference if I RIP my CDs to a SSD, or HDD, or USB flash drive? Locically, the process of writing a file should not matter. And just the retrieval of the file should matter... I can't see any logical explanation how the noise could be attached to a file, because it was stored on a SSD drive for example. You do a md5 checksum on that file and a file that was stored on a USB stick (or even memory), it would be exactly the same. So the only way I can explain the difference in sound would be something related to the retrieval process while playing the file. But who knows, people still can't explain why two cables with the same technical specs (i.e. resistance, capacitance, etc.) sound different. And probably 95% of the world thinks we are crazy and this is a placebo effect. But let's not talk about cables here. My point was that in the digital audio world things are probably exponentially more complex. R1200CL 1 Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 27, 2019 Author Share Posted March 27, 2019 @davide256 - do you mean 1 TB? Still trying to understand your post, and it would make more sense if you meant 1 TB, not 1 GB. Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 27, 2019 Author Share Posted March 27, 2019 16 minutes ago, sandyk said: If you have say a 3M long USB-A -to USB-B cable , try plugging a USB memory stick into that and copy the ripped file from SSD to there, then see if you can hear the degradation with the USB cable. Interesting.... So, are you saying that if you copy the file from the SSD via the 3M long USB cable to a memory stick, and copy the file from the memory stick back to the SSD (as a second copy of the original file) and you play both files from the SSD, you hear a difference? I did not think there would be difference, but if people are telling me there is I would certainly give it a try. It's an easy test to make. But playing one file from my SSD and the other from the memory stick via the long USB cable would be an extremely unfair comparison. So I would rather see the copied file back on the SSD and played from there. And, yes my system is very resolving, and I can easily hear differences on my speakers... actually my speakers (system) are more resolving than my HD800 headphones. Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 27, 2019 Author Share Posted March 27, 2019 I am absolutely sure there will be difference if you play one file from the SSD and the other from a 3M USB memory card. I don't even need to test that to believe it. However, copying a file through a long USB cable to a memory card and back to the SSD and hearing degradation when played from the SSD would be a real eye opener. Not that I don't believe it. Those days are past. I might be a little sceptical at the moment, but I have seen a lot of things in our high end audio world that don't make any sense but make an audible difference. That would also mean that if we care about high quality digital, we should never download tracks from places like HDTracks, unless they ensure us they use good power, cables, storage, vibration isolation, and everything else that could potentially make a difference. And even then, what about the ISPs this traffic is going through?!? None of them are using audiophile switches/routers and linear power supplies. We would need a "master" FLAC/WAV/AIFF/whatever delivered on our special media, copied directly from the master server. Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 30, 2019 Author Share Posted March 30, 2019 2 hours ago, jabbr said: 5. NAS I am 100% confident that employing a NAS running ZFS with a good RAM buffer, and connected by a good fiberoptic Ethernet switch, that assuming the NAS is reasonably decent, that there will not be SQ effects of the hard drives, nor will the PSU supplying the NAS be audible. Indeed I have compared 10Gbe Intel x520 cards with special Solarflare NICs as well as Connextx-4 100 Gbe Mellanox NICs and these on the NAS "sound" the same. (do you know what the jitter specs on a 100Gbe NIC have to be in order to hit that eye diagram? ) http://forms.solarflare.com/Media/Default/PDFs/Platforms/Solarflare_SFN7322F_10GbE_Adapter_Brief.pdf https://www.mellanox.com/related-docs/prod_adapter_cards/PB_ConnectX-4_EN_Card.pdf https://www.intel.com/content/www/us/en/ethernet-products/converged-network-adapters/ethernet-x520-server-adapters-brief.html http://uk.tek.com/dl/65W_28494_1_LR_Letter.pdf https://www.evaluationengineering.com/test-issues-techniques/article/13007997/100-gbs-physicallayer-testing-tips-and-tricks It seems like a lot of people agree with that here. Especially those who use JCAT network cards and/or audiophile switches. I think that option goes to the top of the list and might be the best (sounding) option for large music collections. @jabbr - any recommendations for a decent NAS? I believe @austinpop uses a 4 bay Synology NAS. From what I see so far, people tend to prefer (in that order): 1. NAS. JCAT Net / audiophile switches make that option even better. 2 - a. An external USB HDD with external power supply. Adding an external USB card could be an improvement. 2 - b. USB Tower / SDXC cards / LPS 3. HDD 4. SSD Of course there are as many opinions as people, so not to start a debate, especially knowing that a few people think SSD is better than HDD. But that seems to be the general consensus so far. I did not add Optane as it's not available in Terabyte sizes yet. 87mpi 1 Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted March 31, 2019 Author Share Posted March 31, 2019 3 minutes ago, jabbr said: I don't use audiophile NICs or audiophile switches because frankly mine are much much better. Solarflare and Nexus? Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted April 17, 2019 Author Share Posted April 17, 2019 Has anyone done any experiments treating the SSD drive with EMI absorber? I am wondering if that makes any difference in SQ. Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Nenon Posted April 27, 2019 Author Share Posted April 27, 2019 48 minutes ago, moriez said: @Nenon any progress or update? I was patiently waiting for you guys to finish arguing :). But it would be nice to go back on topic. I have been waiting for various parts, so I can perform different tests. Most orders are arriving next week, some will take longer. But I will post any updates worth mentioning. Industry disclosure: https://chicagohifi.com Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now