Bricasti M7

[jpchartrand]

You could always host it on a free cloud service like Mediafire - much faster than torrents

+1

[SWAN]

You could always host it on a free cloud service like Mediafire - much faster than torrents

dont think you can get 20+Gb free tho which is prob what we are looking at...

[jeffbdavis]

dont think you can get 20+Gb free tho which is prob what we are looking at...

Exactly. Plus, they're kind of a pain to use. Like only being able to download one piece at a time or taking forever because they limit bandwidth to unpaid users...

[CoolColJ]

Mediafire has 50Gb free I believe, and has no download limits, and you can parallel download with download managers

I have gotten up to 5+ MB/s on it

[richie43]

I still vote for torrents. It will only be slow if nobody seeds. And since it IS a legitimate free download, I think it is fitting to use a medium that is notorious for crack-ware for a good free product. If enough people seed, there is no better way to do this in regards to bandwidth AND ethics or "vibe". It's worth a try, for sure. And if people end up being curmudgeons and not seed, then that would be the time to investigate other methods. Just my opinion.

[vicnestE]

Torrents are great for this kind of download.
Much easier and quicker than download one file after another.

And thank Michael and Henry for the unselfish deeds.

[Definity]

Yeh course I understand H no probs thanks for the reply


I'm all for MediaFire, because I can just imagine people constantly asking for people to seed after the initial first seeding, this could go on for years especially when new people sign up.

surely we could put it into one big rar/zip file and upload it.

[Cupwise]

hi henry!
i'm just posing a thought i've had about reverbs for a while now. you've clearly been putting a lot of time/thought/work into this thing, so you may have already considered this, but it's something i've never seen mentioned anywhere so i figured i'd float it past you.

i noticed with my own reverbs, that when it comes to dynamics, there's something about the way that nebula works which i think keeps it from perfection. the thing i'm talking about is the fact that you only have one envelope follower, which is controlling which dynamic step's impulse is played at any given time. i noticed that with really long reverb tails, and with the default nebula reverb release time of 1ms, what happens is this- if you have a drum hit that's average level is at around -5dbfs, nebula will begin playing the impulse sampled at -5db. you can clearly hear this with my tube fm3 stuff where the higher steps begin to distort (and were recorded that way), and you can tell that it's the higher sampled levels you are getting, DURING the drum hit. BUT, the moment that the actual drum is finished, say you have a single kick drum and nothing after it, the 1ms release time means that the envelope follower almost instantly takes the impulse player down so that it's playing the quietest sampled dynamic step. that means that if you have a 10sec reverb, the sample at the level matching the input level is played only during the duration of that input, but the rest of the 10sec is played by the quietest sampled step, because now nothing else is going into nebula so the envelope follower takes the impulse/sample player down to the lowest one. in the real world, if a loud sound triggers a natural reverb, it will reverberate as the reverberation of a loud sound throughout it's duration. with default nebula reverb program settings it magically goes down to reverberating as if it were a really quiet sound, once there's no new input keeping that envelope follower up in the higher dynamic steps.

i could be wrong about this somehow but i did do a few tests back when i thought about it first and they seemed to confirm what i'm saying. so with my most recent reverb sets i adjusted them in a way that i thought made more sense, which is to increase the release times to match the length of the reverb. this way the tail stays playing the reverb sample for the dynamic step that it was triggered in. but there are flaws with that too, such as now if you have a loud sound and quiet sound after, the quiet sound will reverberate with the higher dynamic sampled reverberation. i think it's probably the best way though, and that scenario is MUCH better than the default, where all reverb tails (when there's no new input coming) will always be from the lowest sampled step.

maybe this has something to do with your perception of the nebula version not being quite as real as the bricasti? you play a clap or kick or single snare and the reverb tail you hear is all from the lowest sampled step, but the drum hit was much louder and so that isn't realistic, and the bricasti definitely isn't working that way. maybe try increasing release to match verb length, or maybe a little shy of it (if you haven't already thought of this). because there really isn't much point in having a full 30db sampled range at 1 sample per db, creating a huge filesize, if the tails you hear are always going to be using the one sample taken at -30db, and never any of the other 29.

sorry for the long post and i'm definitely not trying to grandstand or anything here, it's awesome what you and michael are doing with this thing. i've just felt that this issue should be discussed (maybe someone else thinks of a better solution?) for a while now, and you DID ask for suggestions . maybe you or michael have already thought of this issue, but so far all neb verbs i've seen use 1ms release.

also i agree with those who said that it working with free nebula might be more in accordance with bricasti's wishes (but ultimately it's up to you guys). in the end it might bring more new users that way anyway..

[Henry Olonga]

Thanks Tim,
Awesome info there. I think to be honest Michael may find it more useful as I am just doing the sampling aspect. You can of course come on board with the testing. The more input we have from a select few who know how to tweak Nebula like yourself, the better.

To everyone else please do keep the ideas coming. Please give more information and ideas that pertain to sampling specifically. Once I begin - probably in the next couple of weeks, the format will be set in stone. So get your opinion in soon. It just helps me to figure out what the majority consensus is.


I am going to be a little queiter in the coming months with various projects so I will be less and less a presence on the board but I will pop in once in a while to check up and give status updates.

Take care guys.

H

[botus99]

First: I work at 44.1khz

Do 96khz for everyone. If we complicate things with different versions for different editions of Nebula, then some people who move up will have to delete the first set. This seems like a pretty big annoyance to the user, in addition to the annoyance of creating a whole different library in the first place!

DO 96khz FOR EVERYONE!!! I work in 44.1, but I don't mind the conversion, and 96khz is sufficient for high-end audio needs IMO.

Second: I love the ease of MediaFire and Torrents alike

But for the love of all that you consider holy, PLEASE do it with torrents! At least for the initial release this would make the most sense.

MediaFire will only let you upload files 200mb or smaller (on a free account), so the library would be separated into *does the math* 60 different RAR/ZIP/7Z files! That sounds fun right? 60 clicks away from the Bricasti... now 59.... 58... oh the joy!

Accompanying the initial release with a torrent, we would see more people getting it sooner with less clicking around to worry about. After the bulk of people get the release we can discuss other options. At that point, most people will have it, bandwidth will be less of an issue because less people will need to download it after the torrent subsides.

I really think this is a no brainer

[jeffbdavis]

Do 96khz for everyone. If we complicate things with different versions for different editions of Nebula, then some people who move up will have to delete the first set. This seems like a pretty big annoyance to the user, in addition to the annoyance of creating a whole different library in the first place!

That's a good point...

[Cupwise]

96khz seems an obvious call. but, if part of the idea were to help promote nebula itself, then a 44.1khz version might help with that. if you took just the general population of people hanging out at a place like kvr for example, id wager that MOST of them use 44.1khz. and they'd see the set only being in 96khz as a con, because it would take longer to not only download, more space to store, but more importantly load times would be hideous. (and does anyone know how ram use works here? if you load a 96khz program and neb converts to 44.1, i'm guessing you only use the ram of a 44.1khz program, but i'm not sure.)

as henry pointed out, it can be sampled at 96khz then other versions could be made in post. i'm really not even necessarily suggesting to henry that he do make a 44.1khz version, because it would be lots of extra work (yes, even just converting a finished set from one rate to another is lots of extra work) and that's his call if he wants to do it. i'm just mentioning some of what i see as things to consider.

[RJHollins]

I agree with Cupwise as to the library sample rates.

Understanding that this has multiple issue to consider.

I do enjoy working at the higher SRs [actually prefer 88.2 due to better conversion to 44.1] ... however, for more practical reasons, the majority of my work is at native 44.1 due to resource AND destination considerations.

The idea that the 'general music making population' would request 96k, I think, would be rather slim. Even in the 'Pro' community it may be a minority [no facts to back that ... just a considered consensus from readings in various audio forums].

I'm not debating the quality from higher rates ! Until I can add server farms and storage into the facility, practicality must be considered.

Should two versions be released, I would grab both ... but I would likely install 44.1 for day to day work.

hey ... just my 1/2 cent

[adam73]

i think a single 20gb zip file could be a great idea, but due to many different download speeds some people may wait for a long time to try, if each preset is between 200-300mb then maybe an upload of individual presets? also i'd like to say a huge thank you to
Michael and Henry for making this happen, very kind of you

[Tim Petherick]

I agree with Cupwise as to the library sample rates.

Understanding that this has multiple issue to consider.

I do enjoy working at the higher SRs [actually prefer 88.2 due to better conversion to 44.1] ... however, for more practical reasons, the majority of my work is at native 44.1 due to resource AND destination considerations.

The idea that the 'general music making population' would request 96k, I think, would be rather slim. Even in the 'Pro' community it may be a minority [no facts to back that ... just a considered consensus from readings in various audio forums].

I'm not debating the quality from higher rates ! Until I can add server farms and storage into the facility, practicality must be considered.

Should two versions be released, I would grab both ... but I would likely install 44.1 for day to day work.

hey ... just my 1/2 cent

I agree with the 88.2 sample rate thing.Not only do I think it's great because of 2:1 src ratio but because 88.2 sounds tight and smooth at the same time. I think that lavry white paper could be right!!If you go down to 48khz the punch is more like analog.
Dan lavry states that around 60khz is the ideal sampling rate.

My first Nebula release is going to be in 88.2 , i've been experimenting for weeks on this.Think it's gonna be good!!