ValhallaRoom

[musicroom]

Originally Posted by Marando

Thanks Den!

I second that Den. This is one nice sound for a vocal buss.!

[cowudders]

Originally Posted by DTA

+1 for the Small Room!

Sean, I really hope you sell a zillion copies of this plugin! Here´s the reasons why ;-) :

- VRoom is just incredibly good value for an insane price! (So is shimmer!)
- Your communication with and incorporation of customers in development is perfect.
- You´re not only providing bug fixes in almost no time, but also are extending functionality and color with no extra fees constantly.
- I´m really curious and excited to see, what a brain like yours can do to the plugin world in the future ;-)

I always hesitated to write something like this, because with all the (justified) raving reviews about VRoom I just didn´t want to add another one... but, well... Now I just had to!! ;-)

Thank you, Sean.

what DTA wrote!

[jonmoore]

Of the options you gave I'd prefer:

- A variant of Narcissus, with a bit more echo density, and a slower attack.

But I'd also like to vote for Leonardo777's suggestion:

"a preset (or reverb) that sounds like the gold plate of my pcm91, or the very small rooms of the pcm96, the fist very lush and rich and the second very tight and kind of punchy"

Rich, lush small plate/room's that are perfect for short dense percussion treatments, especially handclaps...

JM
------------
Leftside Wobble's sounds on SoundCloud - Create, record and share your sounds for free

[borg64]

Originally Posted by seancostello

Nope. Ya gotta pick one.

I had a sneaking suspicion you'd say that.

My vote goes towards either Small Chamber or Small Room, but any one of the four would be the icing on an already well-frosted cake.

I agree with Hardtoe - truly one of the best value plugs I have ever bought.

[Trailmix]

Originally Posted by borg64

I had a sneaking suspicion you'd say that.

My vote goes towards either Small Chamber or Small Room, but any one of the four would be the icing on an already well-frosted cake.

I agree with Hardtoe - truly one of the best value plugs I have ever bought.

+1 for small room or smallchamber

[seancostello]

Originally Posted by borg64

I had a sneaking suspicion you'd say that.

This isn't out of any desire to be mean or parsimonious, so much as my recognizing a few limitations:

- There are only 3 reverb slots left in ValhallaRoom. When I designed the plugin, I figured out a trick (that I didn't know about when I created ValhallaShimmer): If I only use part of the range for a slider that selects different modes, this gives me room to expand in the future. Well, I'm coming close to using up all the room in VRoom. So I want to save a slot or two for when inspiration strikes. After this, I'll probably need to come up with a new "container" for algorithms that don't fit within the VRoom paradigm.

- Releasing more than one reverb mode means that much more development time and testing. I'm adding a mode because it makes me feel happier about releasing an update to VRoom (instead of boring bug fixes ), but I can't spend too much time on this - I gotta get UberMod out the door.

I have ideas for both small room/chambers, and a slightly "bigger"/"slower" Narcissus variant. People seem to like both, so the determining vote will probably be which algorithm works first and sounds better.

I'm pretty sure the Narcissus variant will sound good. Nostromo and Narcissus were created fairly quickly, and I feel that I stumbled upon a design paradigm that is easy to extend in a bunch of different dimensions.

There are some room ideas that would sound good and be lightweight without much research, that would basically be "a small version of Large Room, but Dark" or "a small version of Large Chamber, but Dark." I also want to explore a few options based on some research into a particular 1980's hardware reverb. The results might be fantastic, or might totally suck. This is how I figure out new stuff - lots of ideas seem great in my head, yet sound horrible when I implement them.

[Izzi]

Hey Sean,

Would really like to see a Eventide 2016 (Stereo Room) emulation in VRoom. Is this possible?

Thanks,
~Iz

[seancostello]

Originally Posted by Izzi

Hey Sean,

Would really like to see a Eventide 2016 (Stereo Room) emulation in VRoom. Is this possible?

I don't have an Eventide 2016. So that would make things difficult.

In addition, Eventide will be releasing native plugins of the 2016 in November, so they should have the 2016 emulation covered. I'd rather focus on original work. I may explore some 2016-ish ideas for the Small Room (I have some ideas about how things work), but I'd rather go for what sounds good, as opposed to directly emulating an algorithm from another developer.

Having said that, can you tell me what you like about the Stereo Room in the 2016? If you can describe the qualities (and/or point me to impulse responses), I might be able to create something that replicates the good parts of the sound without trying to clone things.

[Izzi]

Didn't know that Eventide was coming out with a Native version and "Emulation" wasn't the right choice of wording. What I meant was, something in that direction. There are no IR's out there that I'm aware of and trying to put how the unit sounds into words is difficult.

I do not have 2016 either but maybe you can get a vibe of the sound from this video:

Eventide - Reverb 2016 TDM Bundle

P.S.
I like Pensado's technique that used in the video!

Thanks,
~Iz

[seancostello]

Originally Posted by Izzi

Didn't know that Eventide was coming out with a Native version.

It was announced at AES. No official word at Eventide yet, but here's some info:

AESNY11: Eventide Introduces Native Plug-Ins, Reverb 2016 Stereo Room and Omnipressor announced in AU, VST and AAX formats -Sonicstate.com

And "Emulation" wasn't the right choice of words. What I meant was, something in that direction. There are no IR's out there that I'm aware of and trying to put how the unit sounds into words is difficult. Alas, I do not have 2016 either but maybe you can get a vibe of the sound from this video:

Eventide - Reverb 2016 TDM Bundle

Thanks,
~Iz

I'll take a look at the video. From my understanding of the algorithm, there is no modulation, so an impulse would capture the sound perfectly. I'll do a search for those again.

EDIT: I just started listening to the video. Wow, there is some SERIOUS chatter going on in the early part of the decay! I notice that Position is set fairly high (i.e. to the "back") during this part. This would be similar to setting "Depth" to a higher setting in VRoom.

I have a basic idea of how the algorithm works. Some of the aspects (or at least my theories about how things work) would not be reproducible in ValhallaRoom. The Early section in VRoom was designed to avoid certain artifacts, and those artifacts might be a big part of the 2016 sound. I plan on exploring these artifacts in future work, but I need a different plugin "container," with different parameters, in order to do so.

Other aspects of the Eventide algorithm might be worth exploring. Adding modulation to this type of algorithm might make things sound really dreamy. This is the nice part of not directly emulating old work: things that weren't done in the past, due to hardware limitations, can be added to modern designs.

[Izzi]

Thanks... I just saw the info regarding the Eventide release.

"Other aspects of the Eventide algorithm might be worth exploring. Adding modulation to this type of algorithm might make things sound really dreamy. This is the nice part of not directly emulating old work: things that weren't done in the past, due to hardware limitations, can be added to modern designs."

That's what I figured! Thanks again for making outstanding software and for being an outstanding guy!

I look forward your next plug-in... ValhallaÜberMod

All the best,
~Iz

[Deltones]

I would be happy with either the small room or the small chamber. You got the Alien's Universe pretty much covered.

[Den]

Originally Posted by Deltones

I would be happy with either the small room or the small chamber. You got the Alien's Universe pretty much covered.

Tuning music to 444 old tuning, can create healing effect on your music. 528Hz cures cancer.

http://web.mac.com/len15/MUSICAL_CUL..._Horowitz.html

[Den]

Originally Posted by musicroom

I second that Den. This is one nice sound for a vocal buss.!

I just finished mix on one album (POP/Rock) with this reverb algo.. On female vocal It really shine. Specially on ballads.

[henge]

Vroom already does the "small space that you don't notice but when you mute it the difference is dramatic" thing...and I'd love more of that!
The large spaces in Vroom are great but I mostly need small stuff for work.

[seancostello]

Originally Posted by Deltones

I would be happy with either the small room or the small chamber. You got the Alien's Universe pretty much covered.

I beg to differ. At least as far as names go, there is one other really good one.

[mossie23]

today, i compared valhalla room to some other reverbs i own, and i was quite surprised when i listened carefully to softube's tsar-1. posting this to see if other people hear the same.

what you hear is 2 claps with a pretty long reverb. the first clap is tsar-1, the second clap is valhalla room, with a setting that sounds roughly similar. the 'compressed' audiofile was compressed with ik's black 76 to bring the tail up (in a rather crude way).

what strikes me, is that it seems as if with tsar-1, you can hear combfilters at the very end of the tail. valhalla room has none of this, or just a really faint touch of it. i agree that the example is quite academic, but it brings out something i wouldn't expect in a $300+ reverb. isn't this what they call 'metallic artifacts'? i'm aware that tsar-1 with the highest density can be used for plates, and can thus sound metallic with longer reverbs. but with this example, the density was not even half of what it could be.

is anybody hearing the same as me? (you might need headphones)

[seancostello]

Originally Posted by mossie23

today, i compared valhalla room to some other reverbs i own, and i was quite surprised when i listened carefully to softube's tsar-1. posting this to see if other people hear the same.

what you hear is 2 claps with a pretty long reverb. the first clap is tsar-1, the second clap is valhalla room, with a setting that sounds roughly similar. the 'compressed' audiofile was compressed with ik's black 76 to bring the tail up (in a rather crude way).

what strikes me, is that it seems as if with tsar-1, you can hear combfilters at the very end of the tail. valhalla room has none of this, or just a really faint touch of it. i agree that the example is quite academic, but it brings out something i wouldn't expect in a $300+ reverb. isn't this what they call 'metallic artifacts'? i'm aware that tsar-1 with the highest density can be used for plates, and can thus sound metallic with longer reverbs. but with this example, the density was not even half of what it could be.

is anybody hearing the same as me? (you might need headphones)

I can hear this, too. Here's my take on things:

There are a few different ways of designing reverbs. One of the "classic" ways is to use building blocks called allpass delays. These are very useful building blocks, in that you can connect a bunch of them in series, and each one will increase the number of echoes. Put several of them in one or more feedback loops with filters and gain scalers, and you have a nice sounding reverb. Series allpasses as reverb tools date back to the earliest Manfred Schroeder papers of 1961 and 1962, with the "allpasses within a feedback loop" first explored by David Griesinger in the late 1970's.

Allpass delays are not without their artifacts. In theory, they have a "flat" frequency response, in that all frequencies are passed through unattenuated. In reality, certain frequencies ring out longer than others. There is a particular metallic artifact that can be heard in the late decay of some reverbs that rely on heavily on allpass delays. To my ears, this is what I am hearing in the decay of TSAR-1.

At some point during the development of ValhallaRoom, I made it one of my goals to avoid these allpass delay artifacts. I was testing things with a particularly dry snare sample, and all of my allpass-based reverbs were sounding horrible with that sample, so I switched over to different techniques. The Early section was designed to get high echo densities, without using any allpass delays. About half the reverb modes in VRoom are allpass free, and the others use allpasses in conjunction with other design techniques to minimize the metallic coloration.

So, does this mean that allpasses are bad? Not at all! To me, they impart a distinctive coloration on the signal, but there are plenty of uses for that coloration. TSAR-1 reminds me of an extended version of the original 224 Plate algorithm - dense, somewhat colored, a touch of left-right motion in the decay, but with a beautiful sheen. TSAR sounds different than VRoom, and that is a good thing.

I should also add, that if you ran that same clap through ValhallaShimmer, you would hear a TON of metallic artifacts. It is part and parcel of how Shimmer makes its sound.

In the future, I want to explore more of the "vintage" algorithm space, complete with allpass colorations and the other artifacts associated with the older hardware boxes. These won't fit into ValhallaRoom, as trying to shoehorn artifacts into a plugin that was designed around avoiding those artifacts just won't work. So it will probably be another plugin, somewhere down the road.

[mossie23]

thanks for listening and taking the time to explain. just to be sure, i don't think i found a flaw in tsar-1. i like the way it works, and the ringing can actually sweeten melodic material. i was just curious and this struck my ears as a possibility to learn something about reverb.

you could be right about the algorithm they use. softube claim they actually fade seamlessly between a few algorithms, so there might be more to it than just the sound i got from it with this setting. but overall tsar-1 might be vintage inspired.

[Andy_bt]

Thanks a lot Sean, great answer as always!
Really looking forward to your forthcoming creations.
A.

[seancostello]

In another thread (one about Cytomic releasing AAX plugins), the following question was asked:

Originally Posted by Airon

Transition to AAX: A Real Programmer's Perspective - Avid Audio Forums

Sean, are you considering doing AAX DSP at all ? One of the Avid engineers said that each chip has 16 MB of RAM. How many instances of say, Valhalla Room would you be able to run just on RAM requirements of the DSP ?

I've been thinking about AAX a lot, and here are my current thoughts:

At some point in 2012, I will be releasing the Valhalla DSP plugins in AAX-Native. I am not currently planning to release them for AAX-DSP.

The reason for releasing AAX-Native plugins is simple: I have to, if I want my plugins to run in Pro Tools 11 and onward. For the time being, RTAS still runs in PT10 (and PT 7/8/9, all of which are being used by Valhalla DSP customers), so until Pro Tools 11 comes out, RTAS will work just as well. AAX-Native will probably be critical in order for Pro Tools to transition to 64 bits.

There are a bunch of reasons why I am currently not planning on AAX-DSP releases of my plugins:

- HUGE increase in development costs. Right now, I can design everything with a MacBook Pro, a Mac Mini (which runs Win 7), an MBox2 Pro, an iLok2, and some speakers, headphones, controllers, etc. For my day-to-day development work, everything fits into a backpack. AAX-Native can be developed with the same toolkit, with the addition of having Pro Tools 10 installed for testing. AAX-DSP would require a TI development kit, which isn't that expensive, a PT10 HDX system, which IS that expensive, and a new MacPro or some other system that can host the PCIe cards.

- Unknown increase in development time. The new TI DSP based HDX cards can theoretically run the same C/C++ code as the native plugins. However, being able to run the code, and being able to run the code IN REAL TIME are two separate things. Developers are reporting that they have to do extensive optimizations that are specific to the TI chip and the memory subsystem of the HDX cards, in order to get things running efficiently. I've done these sort of optimizations on a closely related TI chip, and it wasn't fun. I would estimate it would take a few months to figure out how to get things running and optimized for HDX, and I can't spare that time right now.

- HDX latency advantages not necessary for reverb/delay effects. The big selling point of HDX is that you can run things with a RIDICULOUSLY low latency: 4 samples. That is amazing. For EQs and compressors, this would greatly help out in live tracking. For a reverb on a send, it is far less essential. Predelays of 10 milliseconds or more are commonly used for reverbs, which translates to latencies of over 400 samples. Even the shortest predelay on a reverb will result in the output signal being delayed by several milliseconds, in order to realistically simulate a room of any size (a rough rule of thumb is that a distance of 1 foot equals 1 millisecond of delay). Plate reverbs can have delays around 2 milliseconds for the highest frequencies, but plates are usually used in conjunction with a predelay of 15 to 30 milliseconds. In short, any latency can be "folded" into the predelay time, so spending a bunch of energy to get a reverb or delay running at a super-short latency seems like overkill.

For larger developers with existing TDM and RTAS plugins, AAX-DSP would be an excellent format. For this small developer, the increased cost and development time is daunting, and the improved latency of the AAX-DSP plugins seem wasted on delay based effects that are used as sends.

[scottwilson]

Thanks for the down-to-earth explanation of the differences in AAX Native and DSP.

Makes a lot of sense... So I wonder if after the optimizations for TI if the same code will still run on x64 without having #ifdefs every other line of code.

Looking forward to getting some ubermod!

-s

[theOnlyMoment]

it's rough that small independent developers need to purchase an entire pro tools hd rig just in order to develop a plugin for it. It's quite an understandable reason as to why you ccouldn't

[markham]

Originally Posted by theOnlyMoment

it's rough that small independent developers need to purchase an entire pro tools hd rig just in order to develop a plugin for it. It's quite an understandable reason as to why you ccouldn't

I think this is one of the main reasons why so many of the most exciting plugs of the last several years have not been coded for TDM.

[mossie23]

could anybody please explain to me what the differences are between 'echo density', 'modal density' and 'reflection density'? i see this terms pop up here every now and then, but can't find a place where they are explained (outside of technical papers).

[seancostello]

Originally Posted by mossie23

could anybody please explain to me what the differences are between 'echo density', 'modal density' and 'reflection density'? i see this terms pop up here every now and then, but can't find a place where they are explained (outside of technical papers).

"Echo density" and "reflection density" are the same thing: the number of reflections per second. In a real room, and in most good algorithmic reverbs, the echo density builds with time to a fairly high level. Different rooms and algorithms will have different initial echo/reflection densities, and will build at different rates.

"Modal density" is the number of resonances per Hertz. Higher modal densities can result in a smoother decay. A real room will have a modal density that increases with the square of frequency, while most digital reverbs have a fixed average modal density over all frequencies.

Bigger rooms have a higher modal density, which means that there are probably less resonances hanging out by themselves and causing trouble. Think about a bathroom versus a concert hall - the bathroom will have a few obvious, really loud resonances, while a concert hall is super smooth.

For algorithmic reverbs, the more delay memory used in the algorithm, the higher the modal density. You can get a really dense reverb (in terms of echo density) using 50 msec of delay, but the thing will sound totally metallic if you turn the decay time up. Generally speaking, you want to have 1 second or more of delay memory in order to get smooth tails. The memory has to be filled up with echos/reflections, in order to not sound grainy, and that takes cycles.

[mossie23]

thanks sean! added it to the file "good_things_that_sean_said_about_reverb.doc"

[syncussion]

Originally Posted by kozlikha

Upload Files, MP3, Music. Free File Hosting To Share Files

Stereo Wide Hall used on both Piano & Strings.

Sean, we are all hooked on your gorgeous ValhallaRoom

Downloaded the wav, but it's all distorted?

[kozlikha]

Originally Posted by syncussion

Downloaded the wav, but it's all distorted?

Just noticed it. For some reason it was RC tube from SSD that was causing distortion

[Beatworld]

Slightly off topic but not entirely....

Very proud to announce Blue Umbrella's debut album, "Ancient Freedom Songs".

Blue Umbrella | Ancient Freedom Songs

This album has been a year in the making.
During that time many different ITB reverbs have been used.
All the usual suspects were tried......

In the end, Valhalla Room was used for all ITB reverb duties.

There were 6 separate aux reverb sends used, each with Room.
Vox, snare, sax, guitar 1, guitar 2, percussion.
All aux sends used Dark algos, mainly Dark Room and Dark Chamber.
Some keys and strings tracks used Room as an insert.

We used Room because we love the space it creates without being too obvious. Subtle, smooth and wide.

Thanks Sean, love your work.

Here in Perth the CD Launch is Friday night at Steve's in Nedlands.