Vienna Symphonic Library releases Steinway D Synchron Stage

[tfort]

Funny timing. AppleInsider reviewed external SSDs today, and liked the Samsung X5:

https://appleinsider.com/articles/18/12/12/comparing-the-samsung-x5-thunderbolt-3-ssd-versus-the-gtdrive-mobile-ssd-glyph-raid-atom-ssd

 

edit: comments reference the OWC Express 4M2 NVMe/M.2 enclosure and the Akitio Thunder3 Quad Mini 2.5" SSD enclosure.

[Mark Schmieder]

Ah, but the X5 is $1200. It's the T5 I'm looking at, which is almost 1/3 that price.

 

Thanks for the review pointer though; I will read it quickly and then make a snap decision, as it's almost 3pm and that's the shipping cutoff date for anything that has any hope of getting here before the holiday week.

[Mark Schmieder]

Looks like the Glyph is the way to go. I had written it off earlier when I saw it listed as Atom, thinking it was only compatible with Atom-based computers.

 

At any rate, there are a lot of confusing variants at Amazon, with some being slimmer but no difference in title, and a three-to-one or almost four-to-one premium for the 2 TB model vs. the 1 TB version -- but only for the slim one (the wider one, as shown in the review as well, is cheaper).

 

These ship directly from Amazon vs. third-party sellers though, so maybe as long as I get the order in by tonight, I can still receive it by 22 DEC.

 

Update: weirdly, the thinner models are half the speed of the wider models, but the latter use two SSD's in RAID configuration so that simultaneous read/write is possible.

 

For sample libraries, only reads matter (except when installing the libraries, but that is a rare action).

 

Fortunately, in spite of how many redundant product pages Amazon has (some just due to colour choices), there is a nice detailed write-up at the bottom of each, so I should be able to quickly compare those to figure out why there are two variants at different widths and whether the RAID vs. non-RAID is the reason for the width difference.

 

Amazon publishes the date a product is introduced, which is also a big help, so I'll compare those to see if one is more recent than the other.

[Reezekeys]

I'm sure this library is bandwidth-intensive, especially when multiple mic positions are streaming, but 44.1 or 48K (and even 96K) streaming audio is not exactly warp speed transfer for a hard drive. For example, the tech specs for MOTU's 1248 interface says it can move 64 channels of 44.1/48K audio over USB*2*. I'm sure a USB3 connection to any decent SSD is capable of a lot more than that. I remember seeing the voice counter for my old NI NY Piano go at least into the 60s, years ago when I was using the first consumer 7200rpm notebook drive (a 60GB Toshiba IIRC) over a firewire 400 connection no issues. Up until now we've somehow been making do playing streamed virtual instruments without TB3 and NVMe drives. You think you really need these latest & greatest (and expensive) drives & interfaces? My $17 Inatek USB3 enclosure with UASP support is probably overkill for what I do. Not saying my experience should be the metric for others, but that one might want to be realistic about the tools needed for this particular job.

[Mark Schmieder]

The wide version of the Glyph Atom SSD is from spring 2015 and the slim one (more expensive, but slower) is from summer 2018. Makes you wonder why the brand-new review covers the older one, but it could be due to form factor and other concerns regarding intended market.

[Mark Schmieder]

OK, this is the one I just ordered:

 

https://www.amazon.com/Glyph-Silver-USB-C-Compatible-Thunderbolt/dp/B00Z14VFLI/ref=sr_1_47_sspa?ie=UTF8&qid=1544660151&sr=8-47-spons&keywords=glyph+raid+ssd&psc=1

 

New Egg has them for twice that price.

 

There was a 4 TB model listed at $1000 also, which is a good price as it scales linearly from the 2 TB model and most are charging way more. But the product info was contradictory and incomplete so I didn't dare take the chance, as Amazon can be very tricky for technology purchases.

[burningbusch]

For sample libraries, only reads matter (except when installing the libraries, but that is a rare action).

 

True, but there are sequential and random reads. In a library such as the VSL Steinway, you have a mix of both but it's the randoms that will kill the system (my guess). If, when you play a note 20 stereo samples are triggered, those scattered samples need to be found and read from the storage medium, in essence, at the same moment (obviously that's not possible but it needs to happen very quickly with no interruption). As the note is held and the initial small buffer is expended, theoretically, a sequential read takes place. The only problem is ALL the other samples need their expended buffers filled so it goes back to being random reads. The manufacturer specs and that AppleInsider shootout refer to sequential read/writes. Random read/writes can be a fraction of those speeds, but there are variables.

 

As VSL has provided no specs as to what is required with regards to the storage medium, it's anyone's guess as to what is needed for the library to perform as intended. Normally, that's not an issue but because this particular library seems at the extreme (4,000 samples per note), I absolutely think it's warranted.

 

Busch.

[Mark Schmieder]

That makes a lot of sense. Based on how this library is programmed, well over 200 samples can be played at a time, and that's where it started choking.

 

The drive that I just ordered, should perform at close to 800 Mbs, which is likely a factor of ten or more greater than my current HDD.

 

By the way, Sweetwater temporarily has this drive on sale as well, for the same price ($479), so I ordered from them instead. I forgot to mention that earlier.

[Nathanael_I]

I'm sure a USB3 connection to any decent SSD is capable of a lot more than that..... You think you really need these latest & greatest (and expensive) drives & interfaces? My $17 Inatek USB3 enclosure with UASP support is probably overkill for what I do. Not saying my experience should be the metric for others, but that one might want to be realistic about the tools needed for this particular job.

This.

Reezekeys makes great points. You don't need to overspend on this. Buy a Samsung or Crucial SSD at whatever size works. Put it in a simple USB3 enclosure. You will NOT get more performance for spending more, or using higher-bandwidth interfaces. USB3 will be around for 5 more years, easy. Once on an SSD you have done what you can do. I run terabytes of orchestral samples in my composing rig, all active at once. It is RAM and SSD's. The speed of either is not important.

[Mark Schmieder]

Vienna Instruments posted a bugfix to Synchron Pianos tonight, but I forget which issues got addressed -- it's in their forum talk. I just installed it, but worked a super-long day today and am too tired and brain-dead from math and physics all day, to do any music software testing. :-)

[vanderSchoot]

I don't want to spoil the party , but why do the VSL Synchron piano's sound so cold ?

I have Garritan CFX, The Hammersmith, CinePiano, Ravenscroft, Embertone Walker 1955, Art Vista super grand , my piano and more...

I can clearly hear the tremendous detail in the Synchron latests piano's, but i cannot get over the .... opposite of warm / organic tone.

 

It reminds me of a very accurate sampleset of an analogue synth ... accurate in every way but the 'analogue' presence so to speak.

 

But thanks for the info !!

Nice to read so many details about certain 'niche' products.

[Mark Schmieder]

Hmm, I feel the opposite. To me, this is the warmest yet. But only when I select the specific mics and other settings that I listed earlier. I like that we have so much control.

 

For the first time since Pianoteq, I feel like I'm sitting down at a real acoustic piano vs. triggering samples. Only with this one, it's a Steinway vs. an approximation that is a nice piano in its own right.

[Nathanael_I]

I suspect it is nothing more than the crystal clarity of the recording. The technical quality of the recording is one of the finest recorded piano sounds I've heard. It is shockingly clear, in phase, etc. The detail added by the close mics is borderline unreal.

[Mark Schmieder]

I have 64 GB of RAM on the way from OWC now as well, as there is a special at the moment and I also get a further rebate by sending in the original 8 GB.

 

For 32 GB the price was the same as Crucial and Kingston (I think OWC uses Samsung memory; they certainly don't manufacture their own).

 

At 64 GB, the other brands for some crazy reason bump the price by more than a factor of two, even though you could just buy two 32 GB sets and have exactly the same four boards.

 

I could not have considered this purchase up until Cyber Week (which is being extended), as the memory prices for iMac 18,3 DDR4 RAM were sky-high and invariant from when I bought the computer in July 2017 through to November 2018.

[Nathanael_I]

Congrats! Having lots of RAM is wonderful! Lots of room for wonderful samples! Add an SSD and that piano will sing!

[burningbusch]

Seems a controller with release velocity is pretty important with this one, no?

 

Busch.

[Mark Schmieder]

Ah, yes, release velocity: how can ANYONE live without it -- especially on acoustic piano work! What a difference in realism it makes, for phrasing and note-to-note continuity as well as detailed articulation rendering.

 

Off-hand, I can't remember which controllers support it. One of the reasons I sold my DSI prophet 12 is that it didn't. The Moog Voyager does, and ironically has the most complete MIDI spec of any synth I've ever owned. One would expect that more of a device from one of MIDI's inventors.

 

Of course, that's just the synth action keybeds, which maybe one wouldn't expect as much to support key off -- though I find it really helps with the Voyager patches a lot.

 

For weighted action keybeds, I don't know the current status.

 

Maybe someone else can publish a comparison chart specific to release velocity, for the main digital pianos from a half dozen manufacturers preferred on this forum.

 

As for the Synchron Piano specs regarding release velocity, my ears told me it was being used, but I didn't check the specs yet to make sure, and can't do that until home tonight (and probably not until this weekend). I think Vienna's stuff supports it though.

 

Given how many controllers and software products use zero as the release velocity, I always set those to 64 as that gives subtly better rendering for most sound sources. Not as good as variant release velocity -- especially triggered by a live player, who will be intuitively adjusting how fast they release notes based on aural feedback.

[burningbusch]

I based it on this video:

 

[video:youtube]

 

Busch.

[Dnsmo]

And did you notice what controller is being used to demo the D-274? A Lachnit MK-22, which retails for about $3,900 USD, or $4,500 for the "Studio" model. It's the board Friedrich Lachnit of Vienna developed around a Fatar TP/40W action. Among other things, he replaced the Fatar silicon contacts with a proprietary light sensor system that can transmit very high-res velocity, including aftertouch.

[burningbusch]

And did you notice what controller is being used to demo the D-274? A Lachnit MK-22, which retails for about $3,900 USD, or $4,500 for the "Studio" model. It's the board Friedrich Lachnit of Vienna developed around a Fatar TP/40W action. Among other things, he replaced the Fatar silicon contacts with a proprietary light sensor system that can transmit very high-res velocity, including aftertouch.

 

Good stuff. Didn't know that.

 

Busch.

[Nathanael_I]

And the wonderful thing is that controller doesn't have the spongy key feel of those contacts on the bottom. I haven't decided to want it yet, but that keyboard is a special one. It may still be that the Kawai MP11se has it beat with longer key geometry, but that's a comparison that would involve international travel. One wonders....

[Mark Schmieder]

I spent a few more hours tonight tweaking every possible setting, ending up close to where I started but with slightly different mix imaging and mild use of the surrounds for some light ambience.

 

I find the algorithmic reverb and the body size simulator, both smear the sound too much and diminish the articulateness.

 

I still personally prefer the tube mics to the condenser mics, as they are warmer and less prone to proximity buildup in the lows.

 

If anyone would like my preset, send me email -- I don't see a way to attach that 56 KB file here in the forum. I'll see if the FAQ's say how.

 

Update: The FAQ's say file attachments have been turned off in these forums. As the file is small though, anyone who wants it should be able to receive it via regular email, if they contact me with that info.