Hi William,

A very unfamiliar style of music for me. Nevertheless, the composition is obviouosly very-well done. You gave plenty of room for each group of instruments, and they compliment each other naturally. Besides the praise the piece certainly deserves, I'd like to comment on the mix.

1) When I listen on headphones, the separation between left and right is very obvious, while not much is going on in the center. Seems like it's caused by panning rather than by anything else. It's ok in the speakers, though.

2) Some drums sounded a little boxy to me, this is especially apparent in the end. Nothing a little EQ-ing couldn't fix.

All the best,

Crusoe.

Hey Oceanview,

VSL samples are stored in large files, which is really nice for SSDs that have a fast connection bus.

One could buy two NVMe 2Tb each (maybe even three, if the motherboard supports it), so that's 4-6 Tb of fast internal storage. If you can afford that, and that's enough in terms of size, go for it and don't look back. But don't ditch that external drive, by all means :)

I share samples on my PC between a few NVMe m2 drives and an external SSD via SATA III connection. Heres's some stats:

/dev/nvme0n1 (this is the NVMe one):
 Timing cached reads:   26930 MB in  2.00 seconds = 13482.85 MB/sec
 Timing buffered disk reads: 8884 MB in  3.00 seconds = 2960.68 MB/sec

/dev/sda (this is SSD Samsung 860 Evo 1TB via SATA III)
 Timing cached reads:   26346 MB in  2.00 seconds = 13189.70 MB/sec
 Timing buffered disk reads: 1622 MB in  3.00 seconds = 540.34 MB/sec

As you can see, in my particular case NVMe is very far ahead. Can't tell how close this is to the real-world usage, though.

Cheers,

Crusoe.

Originally Posted by: Oceanview

Hello VSL,

But it seems this is now not the only factor for the Synchron libraries. Do you have any recommendations regarding drives, chips, speeds, etc?

Hi Oceanview, I think VSL stated this a few times, you do need a decent solid-state drive for the sample storage, too. I can confirm this first-hand. If you search the keywords, you'll find a few discussions about this subject. I don't yet own Synchron Strings Pro, so I can't claim anything about them. But SYS, Pianos, pretty much everything using Synchron Player does apply.

Cheers,

Crusoe.

Thanks MMKA for the detailed response. All I really care is the final result in the end, however it is achieved. Between real legato of VI libraries and emulated one of SYS, I'm in favor of real legato, personally. Apparently, things have changed given what you say about overlapping notes.

The observation about better crossfading is valuable, too. It was a similar experience for me with SYS. I simply had to avoid XF.

Cheers,

Crusoe.

A question for the people who already bought the new library. One much debated topic of the first SYS release was "natural legato", or rather its absence, depending on how you look at it. Can anyone please tell if it's the same in SYS Pro, or different ? For me personally this would be an important part of the buying decision. Maybe for the others, too.

I'm listening to the demos; couldn't so far find one that could make the answer obvious. Also, apologies if I missed the question already answered elsewhere.

Cheers,

Crusoe.

Macker's comment is pretty much what I would say myself. Could be not even the release and reverb settings, but the articulations. Sharp note endings contribute so much to the playfulness of this piece.

Cheers,

Crusoe.

I don't have MIRx, does it allow setting dry/wet ratio for individual instruments? If yes, then certainly do it; instruments in the lower register get a little more reverb, the ones in higher register less. But don't follow this blindly, of course, it should sound right to you, of all things. I personally thought your basses are too wet, actually.

And the bigger picture: try reducing overall wetness of both components a little, and compensate that with a short reverb, as in my first message. Also, why such a big venue in MIRx? Try a smaller one, it should blend better, in my opinion.

Also, try to increase pre-delay gradually, maybe that willl help, too.

Good luck,

Crusoe.

Hi Fastel,

On the musical side of things, you did a pretty good job. Plenty of dynamics, and the most of the time the balance between the vocal and the orchestra is good.  Maybe in those louder parts you could let the vocals breathe just a little more.

The spatial picture created by reverb could be improved, yes. It is actually not that bad. I can't however give a more detailed comment here, because you didn't mention reverb settings that you used.

One fairly standard trick that could work here is to add the same *short* reverb to both vocals and the orchestra. It won't make the result too washy, and the same time will provide the need spatial glue.

Cheers,

Crusoe

Hi William,

I re-listened to those parts I tried to describe as "hissing", and apparently it's a cumulative effect of the tremolo and violins on the left, which I didn't mention because my brain is too small. It's still very short-lived and now I'm thinking it would actually sounds the same in a real-world performance. At 2:30-2:32, for instance. Anyway, the performance is so good that this rather tiny thing is the biggest complaint I could think of.

Oh yes, thanks William and VSL for the walkthrough!

Cheers,
Crusoe.

Hi William,

Wow. This sounded more or less flawless to me. Which I don't remember hearing in the field of virtual orchestration. Maybe a couple of times. The first 2 minutes were 100% convincing and musical. Maybe, and I'm not 100% sure here, a few times you could use filtering to reduce hissing that is created by the tremolo strings. I reckon that hissing is created by the reverb. Very much impressed!

Cheers,

Crusoe.

Look what I've found:

CPU Performance vs. Real-Time Performance in Digital Audio Workstations (DAW)

Really nice explanation of the DAW performance subject, simple but without oversimplification. I'd sum up it like this: it's the same as with one's children. One has to invest time and sweat into their kids, no amount of "smart toys" can replace that. Hope you catch my drfit :)

Cheers,

Crusoe.

Holy ... Thanks for pointing this out :)

I've just had a look here:

https://www.cpu-monkey.com/en/compare_cpu-amd_epyc_7302-996-vs-amd_ryzen_threadripper_3970x-948

TR price is substantially lower, must be a different model pair that you looked at. Nevertheless, this page gives a pretty good idea of the comparative performance. 

All the best,

Crusoe.

Ok, then I guess you need not concern yourself with the thread number. I don't know how it works exactly, but it's more or less a resource management thing. Any CPU of the grade you are aiming at is not a bottleneck. Not even remotely. This CPU will sit 97% idle in the stratospehere, while mere mortals, memory chips, SSDs and the like are trying to pull of the mighty feat of moving around those 256Gb of data.

The shopping list looks fine to me. You might want to consider the newer Threadripper 3970X, although it has higher TDP, which is enough to substitute a fireplace, if your client needs one. Also, the audio interface could become a bottleneck, depending on what your client has right now. Samsungs are a good choice, btw, I'm not convinced by the performance of the new PCIe 4.0 modules. But this will probably change in the future.

Good luck :)

Crusoe.

Hi Trickytrick,

You are welcome. I wonder why you've selected a server CPU? This is not obvious to me at all. Also, I don't really understand the question about threads per instance. It's just a number of cores times two. Or you meant something else that I didn't quite grasp.

Cheers,

Crusoe.

Hi Maarten,

This is a beatiful piece, I liked how you used the strings to create the background for the melody, really beatiful. And the melody is catchy and likeable, you centered your piece around it, I only wish you'd develop it further. I wonder why you called this tune a jazz piece, I thought it's a pre-jazz tonal piece with maybe a few touches of jazz harmony.

The clarinet is very nice, too. Maybe with a few additional articulations you could make it even better.

A few technical criticisms:

- The claps and cajon are very dry; I believe adding some reverb could give them a better place in the mix.

- The bass is too quiet or at least saturated by the piano; you need to give it some more space in the mix.

- The instruments don't follow each other well rhythmically. Claps are out of sync with the piano, I think, but I couldn't figure this out 100%, maybe it was something else that confused me.

Cheers,

Crusoe.

Hi Dave,

Lovely piece, I liked it. I thought the bassoon remained clueless as to what the oboe is up to, it just went on happily, maybe suspecting something just a for few short moments. That's the power of harmony :)

Counterpoint was more or less flawless, and the interplay between the instruments was clever, as usual. You also used fairly straightforward harmony of the classical period; it was clear and very easy to follow.

Sometimes panning tricks worked for me, sometimes they didn't. At 1:42 oboe jumps half the stage in an instant, which doesn't sound as if it was intended (even with your explanation). And indeed, those stage settings blur the feeling of space that unites the trio. Dynamic panning of those proportions is very hard to pull off, in general, and I understand your focus was on fun. And torturing the mix settings is not a very fun thing to do, yes :)

All the best,

Crusoe.

Hi John,

3.1 is a very good option, I suppose. If you have enough space in your room, that is. You will hear plenty of bass with a subwoofer, but how helpful it will be in a small room (for mixing) is not a given. But there's no question, if you care about bass below 60Hz, you need a sub.

Foam, yes. Stuff from Universal Acoustics is reasonably priced and works well. It also doesn't smell when unpacked, which is nice. Pluto-1 is a good starter kit. Use it to diffuse direct reflections fro the speakers.

Bass traps - yes. Can't give the exact recommendation here. If you are adventurous, you could make them yourself following the advice here. Set them up in the corners, as many people suggest.

Acoustic measurements... maybe, but that's after the previous two points. You will get used to the new room, and after some time you will actually know what needs a correction (with some luck). Not precisely, of course, and that's where the measurements will come in handy. But the other way around... I'm not so sure. It's the same principle as with mixing in general - don't do something unless you know why you need to.

Cheers,

Crusoe.

Hi John,

it's a pretty big subject. There's plenty of information on the the web, YouTube included. How did I personally know that my room needs treatment? I turned on some talk show, very well recorded audio, and yet half of all the words were munched, drowned in a wash of the room echo. It's as simple as that.

Cheers,

Crusoe.

Hi John,

Aside from what was already said, include your room into the equation, too. If it's a small un-treated room, you will easily overwhelm it with 2 5'' monitors. Want more bass? Get a subwoofer, not more speakers. And treat your room, if it's not treated. The question of low frequencies is really orthogonal to the quantity of speakers.

Cheers,

Crusoe.

Merry Christmas to Dave, Becky and everyone who reads this!

Dave, Becky, indeed, your performance brought a smile on my face. Thanks for doing it.

Crusoe.

Hi Martin,

I'm not an expert on the subject by any means, so what follows is a bit of speculation.

Who doesn't want that, hehe. This is a pretty ambitious goal, I'd say, whatever choice you make in the end. And the problem is the 128 samples plus 0 issues combination, not the CPU power or anything else. The average performance will certainly improve with a newer PC, but the *worst*-case situation could stay. Audio performance is a function of many variables, and I therefore refrain from a direct advice.

I'd personally consider economic reasons above all else. You do have VE Pro, and you do a working set-up, and you have enough money for a reasonable new PC that you could upgrade memory-wise etc. as the time goes by. You'd also retain a working one-machine set-up, which is nice in my opinion, given all the hassles of a new PC. 

You can look at the LatencyMon measurements of my AMD 3900X here.

Merry Christmas,

Crusoe.

Here's the results of LatencyMon run on my system.

Cheers,

Crusoe.

Hi Ben,

sure, I will send them some time soon. I think I have LatencyMon installed.

Cheers,

Crusoe.

Hi Ben,

That's good to know. Without any extra knowledge of the specifics, one would think that the bias could be the case indeed (with Intel compiler around, and Intel's overall dominance not so long ago in the past).

Just to add a bit of info on 3900X, all 12 cores get their share of work under a heavy load, resulting in not as much of a load on a single core, which is noticeable (Cubase and VE Pro 6). It's also worth noting that I still stay on the X370 platform, upgraded from Ryzen 1800X (which was pretty good, too, especially given its price). Works as a charm.

Cheers,

Crusoe.