It's up to the mixer (ie; the composer in this case) to figure out how to control that sound and make it work for your composition if you choose to do so. I'd rather have the more raw and natural recording than an over-processed, awkwardly panned mess to deal with.
It wouldn't matter. Even if they did it front to back as you suggest, you will still get a thinner sound and more direct sound of the first chair and player sitting behind them. We aren't talking about a large ensemble here. It's between 3-6 players for those divisi sections. Regardless of configuration, you are going to get a very discernible sound in the positioning. That's how sound works in these environments. I'm sure you know this already though.sure, or they could have done what I assume other folks do and sample half the desks in a front to back configuration and not strictly just the desks closest to the close mics and closest to the center of the stereo field of the tree?
It wouldn't matter. Even if they did it front to back as you suggest, you will still get a thinner sound and more direct sound of the first chair and player sitting behind them. We aren't talking about a large ensemble here. It's between 3-6 players for those divisi sections. Regardless of configuration, you are going to get a very discernible sound in the positioning. That's how sound works in these environments. I'm sure you know this already though.
I honestly haven't heard any phasing issues either. Then again, there have been claims that Afflatus sounds "dated".Concerning Stereo Image "issues."
I haven't heard an audio examples of this by those complaining. With that said, I'm making an assumption.
There is an audio stereo difference between the Full Section and the divisi section when split. I've explained why in previous posts on other threads for Afflatus. To keep it simple, players in a room move air. The more players moving air, the more reflections in the room we get which gives us an audible spacial awareness in the stereo field. Now, cut those players in half. Less players means less air which means less reflections which means..yep..you guessed..less audible spacial awareness. This means we get a more "direct" sound from the smaller divisi sections.
I had speculated in previous posts about this with a confirmation from Strezov Sampling. Their divisi sections were simply taking the front half desks of each section and recording them. This means we lose those rear players and their reflections which help fill out the stereo field. Because they recorded the front half, we get a more direct sound given where the mics are placed. Couple that concept with the fact that they most likely recorded the divisi sections in a different session/day, we are going to have some subtle discrepancies in the mix but the stereo field "issue" you are hearing isn't an issue. That's just how sound works in a recording environment. It's up to the mixer (ie; the composer in this case) to figure out how to control that sound and make it work for your composition if you choose to do so. I'd rather have the more raw and natural recording than an over-processed, awkwardly panned mess to deal with.
In terms of phasing. Again. I haven't heard examples of this and when I tested this library for a few weeks before reviewing it, I didn't notice anything inherently disturbing to my ears concerning this. That doesn't mean it's not there of course but if you own this library and have these issues, make a recording of it and send that in a support ticket, so George and his team can fix it. As he's said early. They have a 30 day window to get an instrument update to NI. I don't know if this is an issue that needs to be included in that time frame but at least they will have a good example of those issues and will fix them in due time.
It's ended.@Strezov for all us last-minute people here, when exactly does sale end?
That pretty much sums it up. There have been times with other libraries that I thought I heard stereo image problems as well, when what I was hearing were reflections from the other side of the hall when using far mics - also, I noticed with libraries like Chamber Strings, the decca tree won't share the same specific image as just the rigs, ambient mics or close mics - that's because the decca tree configuration wasn't originally designed to present a faithful stereo image of a section, but was rather designed to get a full sound that jumped out at the listener.Concerning Stereo Image "issues."
I haven't heard an audio examples of this by those complaining. With that said, I'm making an assumption.
There is an audio stereo difference between the Full Section and the divisi section when split. I've explained why in previous posts on other threads for Afflatus. To keep it simple, players in a room move air. The more players moving air, the more reflections in the room we get which gives us an audible spacial awareness in the stereo field. Now, cut those players in half. Less players means less air which means less reflections which means..yep..you guessed..less audible spacial awareness. This means we get a more "direct" sound from the smaller divisi sections.
I had speculated in previous posts about this with a confirmation from Strezov Sampling. Their divisi sections were simply taking the front half desks of each section and recording them. This means we lose those rear players and their reflections which help fill out the stereo field. Because they recorded the front half, we get a more direct sound given where the mics are placed. Couple that concept with the fact that they most likely recorded the divisi sections in a different session/day, we are going to have some subtle discrepancies in the mix but the stereo field "issue" you are hearing isn't an issue. That's just how sound works in a recording environment. It's up to the mixer (ie; the composer in this case) to figure out how to control that sound and make it work for your composition if you choose to do so. I'd rather have the more raw and natural recording than an over-processed, awkwardly panned mess to deal with.
In terms of phasing. Again. I haven't heard examples of this and when I tested this library for a few weeks before reviewing it, I didn't notice anything inherently disturbing to my ears concerning this. That doesn't mean it's not there of course but if you own this library and have these issues, make a recording of it and send that in a support ticket, so George and his team can fix it. As he's said early. They have a 30 day window to get an instrument update to NI. I don't know if this is an issue that needs to be included in that time frame but at least they will have a good example of those issues and will fix them in due time.
Concerning Stereo Image "issues."
I haven't heard an audio examples of this by those complaining. With that said, I'm making an assumption.
There is an audio stereo difference between the Full Section and the divisi section when split. I've explained why in previous posts on other threads for Afflatus. To keep it simple, players in a room move air. The more players moving air, the more reflections in the room we get which gives us an audible spacial awareness in the stereo field. Now, cut those players in half. Less players means less air which means less reflections which means..yep..you guessed..less audible spacial awareness. This means we get a more "direct" sound from the smaller divisi sections.
I had speculated in previous posts about this with a confirmation from Strezov Sampling. Their divisi sections were simply taking the front half desks of each section and recording them. This means we lose those rear players and their reflections which help fill out the stereo field. Because they recorded the front half, we get a more direct sound given where the mics are placed. Couple that concept with the fact that they most likely recorded the divisi sections in a different session/day, we are going to have some subtle discrepancies in the mix but the stereo field "issue" you are hearing isn't an issue. That's just how sound works in a recording environment. It's up to the mixer (ie; the composer in this case) to figure out how to control that sound and make it work for your composition if you choose to do so. I'd rather have the more raw and natural recording than an over-processed, awkwardly panned mess to deal with.
In terms of phasing. Again. I haven't heard examples of this and when I tested this library for a few weeks before reviewing it, I didn't notice anything inherently disturbing to my ears concerning this. That doesn't mean it's not there of course but if you own this library and have these issues, make a recording of it and send that in a support ticket, so George and his team can fix it. As he's said early. They have a 30 day window to get an instrument update to NI. I don't know if this is an issue that needs to be included in that time frame but at least they will have a good example of those issues and will fix them in due time.
Having played Adagio and Anthology I can say for certain that those libraries suffered stereo imaging fairly badly, although when compared to Afflatus's imaging it's night and day. If you would like, I can post raw examples of each mic position in Afflatus with a quick riff for each section and you can pinpoint exactly which part is causing the issue you hear. That we if we do find anything it can be brought to Strezov's attention quickly.Obviously I don't own the library, so I can't tell if what you are describing is actually what is causing the stereo imaging problems. I can just say that it causes an unnatural sound and is an annoyance when listening to the music created. 8DIO had some issues with some of their string libraries that reminds me of this and this was not due to section sizes changing or anything like that. So until it is proved that this is only a problem when switching between sections, we don't really know, do we?
Regarding phasing I just think it is inherent to the recorded sound I am afraid. And maybe phasing is not the right technical term to describe what is going on but it is the closest label I can put on it.
If you guys don't hear these issues, good for you, it won't bother you.
Funny, I was just going to say something along these lines. But I opted not to because I have no experience with this in-depth of sampling, and I have no official training in acoustics. But it just seemed logical. I was thinking it may be more in the lines of an auditory illusion. Perhaps even certain frequencies are more prominent at causing this issue. I suppose this phenomenon could happen live as well. I guess I'm thinking about how high/med/low frequencies may reflect off surfaces differently, and travel differently. So I think in a sample library, any of these imperfections in acoustics are going to be amplified because you are isolating a certain perspective with a mic.That pretty much sums it up. There have been times with other libraries that I thought I heard stereo image problems as well, when what I was hearing were reflections from the other side of the hall when using far mics - also, I noticed with libraries like Chamber Strings, the decca tree won't share the same specific image as just the rigs, ambient mics or close mics - that's because the decca tree configuration wasn't originally designed to present a faithful stereo image of a section, but was rather designed to get a full sound that jumped out at the listener.
My hypothesis on this would be these following possibilities - the mics when mixed together create a very full image which isn't in line with classical recording and/or the lush sections contain two sample sets to facilitate divisi switching, so the layering could be causing what some would perceive as blur or over-saturation. Not sure though unless I put up examples of everything to test these ideas.Funny, I was just going to say something along these lines. But I opted not to because I have no experience with this in-depth of sampling, and I have no official training in acoustics. But it just seemed logical. I was thinking it may be more in the lines of an auditory illusion. Perhaps even certain frequencies are more prominent at causing this issue. I suppose this phenomenon could happen live as well. I guess I'm thinking about how high/med/low frequencies may reflect off surfaces differently, and travel differently. So I think in a sample library, any of these imperfections in acoustics are going to be amplified because you are isolating a certain perspective with a mic.
It's one thing if they didn't keep the mics in the same position when recording each RR or Velocity. But I am assuming that all of the RR and Velocity layers for a specific articulation were recorded in the same session, in a relatively small window of time. So I wouldn't think that that would be the cause. If there are stereo imaging issues, I'd be really curious of the cause. It may indeed be due to mic choice and positioning of the mics, or room acoustics, or a combination of things.
All of that being said, I have no issues with the sound. If I notice something glaring I will certainly say something.
I am only hearing jumping around in the 2nd and 3rd examples of each articulation/patch. It doesn't seem to be the main frequencies, but rather sub-harmonics and harmonics of the main notes. But overall there still seems to be a consistency of the behavior of the sound. So it still sounds natural to me. It seems believable to me, like it's how it may be heard live. I feel like we may be getting into some subjective territory. I've never seen a sample library analyzed in this way. I wonder if we did examples like this with other libraries if we'd see the same or similar things. The only way to truly test this is to do double-blind tests.My hypothesis on this would be these following possibilities - the mics when mixed together create a very full image which isn't in line with classical recording and/or the lush sections contain two sample sets to facilitate divisi switching, so the layering could be causing what some would perceive as blur or over-saturation. Not sure though unless I put up examples of everything to test these ideas.
Yes, we would in fact see similar things, just with different anomalies depending on the space the strings were sampled in - for example Air Studios London for Chamber Strings. You'll notice that some libraries have more tight and distant images (classical) whereas the more cinematic-style libraries will have a closer/mid sound that's fuller and wider, which may cause more stereo perception and "jumps". Close mics tend to be similar barring mic choice and post processing.I am only hearing jumping around in the 2nd and 3rd examples of each articulation/patch. It doesn't seem to be the main frequencies, but rather sub-harmonics and harmonics of the main notes. But overall there still seems to be a consistency of the behavior of the sound. So it still sounds natural to me. It seems believable to me, like it's how it may be heard live. I feel like we may be getting into some subjective territory. I've never seen a sample library analyzed in this way. I wonder if we did examples like this with other libraries if we'd see the same or similar things. The only way to truly test this is to do double-blind tests.
The following WAVs contain a legato riff played for each section's close, decca and hall mics for Chamber Violins I & II, Chamber Violas, Chamber Celli and Lush Basses (there are only Lush Basses for legato). You'll notice that the close mics place the sections in traditional seating, whereas the decca mics contain a more centred and full stereo image, while you'll hear the reflections from the walls in the hall mics, giving them a wider image. If anyone would like examples made with different patches or mic mixes, feel free to request and I'll post them as well.
https://www.dropbox.com/sh/a281nvelkod9jzi/AACIBVQEEJQofQ-xqNwZ_cBpa?dl=0
The following WAVs contain a legato riff played for each section's close, decca and hall mics for Chamber Violins I & II, Chamber Violas, Chamber Celli and Lush Basses (there are only Lush Basses for legato). You'll notice that the close mics place the sections in traditional seating, whereas the decca mics contain a more centred and full stereo image, while you'll hear the reflections from the walls in the hall mics, giving them a wider image. If anyone would like examples made with different patches or mic mixes, feel free to request and I'll post them as well.
https://www.dropbox.com/sh/a281nvelkod9jzi/AACIBVQEEJQofQ-xqNwZ_cBpa?dl=0
I know this is a Berlin Strings vs Afflatus thread, but since CSS has been brought up multiple times, I just wanted to post a quick demonstration comparing the legato.
Right off the bat, Afflatus is MUCH easier to play. There's just no way around it. It took me several takes to get the hang of the CSS, while Afflatus was a walk in the park.
The first clip is the Violins 1 patch from CSS in Con Sordino, followed by the Scene d'Amour patch.
The second clip is the standard Violins 1 patch from CSS, followed by the Lush Strings patch from Afflatus.
I also wanted to note that no reverb has been added. As you can hear, Afflatus is much closer/dryer sounding (which I greatly prefer). CSS is more lush, but I feel that can get in the way of detail at times.
Personally, I'd take Afflatus.
In all your examples, first one is fine. But the second two (decca and hall) have phase issues. You can hear it instantly. Third one (Hall) is worse than the second (decca) though. It's the same for all four instruments.
Thanks for the analysis. I'm wondering though, is it possible to get a "false positive" with layered instruments? I mean, you should be able to tell by just dropping the waveform into your DAW if it is out of phase. I should be able to record a piece with these patches and it will show that it is out of phase when I mix down the track. I shouldn't need a meter to tell me that. So if a meter is telling me that, but the actual visualization of the waveform isn't, could it just be the meter is getting a false positive based on multiple layers of the same instruments playing? Some level of phasing is natural when layering certain instruments. It seems string instruments would be a prime candidate for that sort of phenomenon.In all your examples, first one is fine. But the second two (decca and hall) have phase issues. You can hear it instantly. Third one (Hall) is worse than the second (decca) though. It's the same for all four instruments.
With both decca and hall the minute it plays, you can hear how ultra wide it sounds and the middle sounds like it's sucked away, or a void (at least that's how I always hear phase issues).
I then dropped both into the daw and put both Nugen visualizer on it and the waves dorrough meter and both the phase light on the dorrough meter lit up for decca and tree, and the polar display in visualizer put the spikes in the out of phase area, confirming what my ears told me. But you don't need meters to tell, you can hear it instantly because it's pretty strong.
A lot of people, if they don't have engineering experience often mistake phase issues for wide stereo and actually (especially for dance music) messing with the phase is used to make ultra wide dance tracks that sound like they're wider than the speakers. It's not great to do but people do it anyway. But for me, it drives me nuts because it sounds like I'm listening on headphones but with my speakers - I hear the sound in my left and right ears but the middle of my head feels like it's in an anechoic chamber.
As to whether it's baked into the samples or happened in mixing or editing is anyone's guess. It's very possible no one checked for phase issues during the recording session or it was so hectic no one heard it. It could happen a lot of ways - it could just be that the multiple mics for the decca and hall were set up incorrectly or it's possible one of the mics (especially if it's older) or an xlr were wired incorrectly with pin 3 hot which would cause phase issues with that mic. Vintage mics especially can be wired with either pin 2 or 3 hot (there was no standard back then) and so it used to be de rigueur to test mics for phase before recording to be sure.
The good news is that this should be all fixable as long as the samples are all recorded on their own tracks, as they most likely are. If the mics are set up badly or incorrectly, it could just be a matter of shifting some of the mics a few milliseconds forward or back. If it's a pin 3 hot problem, then the phase needs to be flipped for that mic's track. Of course if the mic setups were submixed to stereo for recording then the library is screwed. But I highly doubt that because these days with tracks being so plentiful, there's no reason to not record everything on its own track. Submixing is a 50s/60s thing.
The bad news is that they have to go back and basically do this for all the recordings and all the samples so it could be very time consuming. Especially if they submixed everything to stereo before they edited. Then they'd have to fix the phase issues, submix again and re-edit all over again.
Anyway, that's my take on the whole thing.