Jump to content

Unveil - 'unreverberator'


nhaudio

Recommended Posts

Hello Denis,

thanks for the reply! I can imagine it to be a bit enervating to have all kind of requests coming in while the priority is developing major versions of the application, and I understand most parameters are too complex for convential naming. Standard parameter for Adaption would be super neat, though ;-) I'll try the visual approach as suggested. I'm sure looking forward to future incarnations of the software.

Link to comment
Share on other sites

  • Replies 70
  • Created
  • Last Reply

Top Posters In This Topic

Enervating? If I came across like I found your feedback enervating: that definitely wasn't my intention at all, it's just the nerd in me showing ;-)

Actually, we love feedback of any type -- after all, we make these tools for the people using them, and feedback can only help improve them, which is a win-win thing.

As to future incarnations, we've got some cool stuff for UNVEIL as well as some more advanced audio processing products in the pipeline....</tease>

By the way....is there any other "nah, that can't be done ....but I'd LIKE to be able to do that" type process you guys are missing in your arsenal? We've got some stuff in the making that we think qualifies for that description, but if there's anything in particular that would enhance your workflow, or even change it for the better, it can't hurt to let us know ;-)

Cheers,

Denis

Link to comment
Share on other sites

Hello Denis,

you didn't come over as enervated at all, I guess I was just overcautious in my wordings ::)

What would be super cool, and please excuse me if it sounds lame or crazy, cause i have not the slightest idea of how such a thing could be achieved, but lately I have been thinking of a noise reduction system that 100% focuses on vocals/speech. Like, in... take a noise print (say, voice print) of clean speech, speaker x. Then compare speaker x print to a "noise alone" print. Then be able to remove noise (adaptively) from a contaminated speech recording, where the system itself would know what not to touch (as it knows the voice print of speaker x). Even if we don't have a noise print, we could still tell the system to remove anything that is not speaker x (to some degree). Like a reversal of the Noise reduction systems we have at the moment. I'm not sure if this is understandable. Basically, I mean an advanced noise reduction algorithm that could work by comparing several prints (remove this, but leave that alone), as opposed to the standard algorithms where we have only one noise print (remove this, but do it blindly). Of course, it would be neat if the system would be able to analyze different selections of different input files and store them in a database, and then also be able to process them in a user defined matrix, as in: voiceprint1 compared to noiseprint2 would result in a combined analysis file (leave this alone but remove that), that then could be used to do the final noise reduction on the actual file, namely the contaminated recording. Maybe the system itself would need to "know" the human voice in all its shapes and shades of expression, or have an abstract "understanding" of the human voice. That's what I think would be the neuronal part. So, one would need all the basics of a super neat noise reduction system with all the standard parameters, plus the advanced (neuronal?) features aforementioned. I think what I basically mean is a cocktailparty effect noise reduction, where we can tell the system to concentrate on "this signal alone" and ignore the other stuff. Or at least, to some degree. Of course, matters are much more complex than that and it's all very far over my head... - I'm drifting away here, but I always wondered if something like that wold be possible some day.

If it sounds like a good idea, I can send my PP adress. :ph34r:

:lol:

Link to comment
Share on other sites

I would really like to see a shitty producer filter. My answering machine is great, but doesn't work very well on set

:)

The 'improve script' plugin.

Post has been asking for that for many years now. If only ....

I had that joke for 20 years in post! "What we need is the digital content improvement filter!"

Hehe, well our implementation of these will also include a causality inversion function, which causes directors to consider audio the most important aspect of the film and allocate more time & resources to that department. We're still having some issues with the code at this point, though ;-

Funnily, PITCHMAP has a similar kind of effect for composers. When the director comes in to listen to the results of the only orchestral recording session that was in the budget, and goes "Oh, that's AWESOME, it came out just like we wanted....but can we please change that chord sequence to something more like...like...Ryuchi Sakamoto versus Motorhead, you know....yeah, that's what we need, can we hear that please?".....you can now just play some new target chords to implement just that, hopefully demonstrating that it's all fine as-is *grin*

Link to comment
Share on other sites

Could you work substantially stronger magic if we provided your system with two tracks -- one of our best effort that would contain dialog plus unwanted background, plus a second track recorded 2-3 feet from first mic that has primarily the unwanted background?

For a few years the Grateful Dead implemented an analog version of this with phase inversion and matched instrumentation mics. They sang close into one of the mics, which let them stand in front of their speaker array (another spectacular undertaking of its own).

Link to comment
Share on other sites

Could you work substantially stronger magic if we provided your system with two tracks -- one of our best effort that would contain dialog plus unwanted background, plus a second track recorded 2-3 feet from first mic that has primarily the unwanted background?

Well, for any intelligent process, more information will always be better, yes. But I believe that a good tool should not depend on having other material than that which is to be processed to do what it does --- as it would then only help you achieve results if this condition is met. So IMHO relying on additional information is not an ideal strategy, and kind of "the easy way out". It is of course always good if the tool *allows* you to *optionally* supply it with additional information, but the tool needs to *already* know a lot so that it will always be able to deliver at least satisfactory results right out of the box.

Cheers,

Denis

Link to comment
Share on other sites

(...) It is of course always good if the tool *allows* you to *optionally* supply it with additional information, but the tool needs to *already* know a lot so that it will always be able to deliver at least satisfactory results right out of the box.

Cheers,

Denis

i think that was what i was trying to say but took me half a page, ha - i think any additional analysis data would be great for surgical precision processing, but sometimes you don't have of such discrete extra snippets, and that's why it would be great to have the tool already "know" what is considered good vs. bad, let's say in a worst case scenario. would be really good to have a combination of standard and advanced processing options at hand: a standard noise removal tool approach, combined with an intelligent module which has an intrinsic knowledge of "good" signal's characteristics (but should be able to do additional learning). the different processing options should be able to be combined to get the best of both worlds (i.e., to mix the "standard" noise print approach and the amount of "brain" work). even if the tool would concentrate on the human voice alone, it would need a major bulk of "brains" for the tool itself and maybe an intelligent database, because the human voice can take on so many shapes... i mean, one would have to be able to tell the brain what to look for: a whispered voice will have totally different chracteristics from a singing voice, let alone to think of the many shapes the human voice comes in with even standard speech. looking forward to see what you guys will come up with. really inneressing.

Link to comment
Share on other sites

I have a question.

Altiverb allows post folks to use the clap of sticks to map reverb in order to add matching reverb to a dry ADR recording. Does this software allow a similar mapping to assist in removing/reducing undesired reverb in a production track? Or is everything manual?

Robert

Link to comment
Share on other sites

  • 3 months later...

Hey all!

Just a short heads-up that UNVEIL version 1.5 is coming November 15th, adding a bunch of stuff:

  • Now supports RTAS, AAX Native and VST on MacOS X and Windows in addition to AudioUnits
  • New preset management functionality accessible from within the plug-in GUI allows user presets to be used across all plug-in formats on all platforms
  • New multi-mono capability in Logic Pro 9.x to allow for individual settings for channels in multi-channel/surround scenarios
  • New authorization app for more streamlined user experience
  • Completely re-designed automation system for improved automation workflow
  • New factory presets
  • New option to enter values numerically by double-clicking controls

This is a free update for users of v1.0.x. We'll be showing it at AES in SF, Oct 26-29, booth 639, so if you're in the area it'd be great to meet you!

Cheers,

Denis

Link to comment
Share on other sites

  • 11 months later...

Have just been mucking around with Unveil on a particularly live room, and I'm blown away at how effective it is. 

 

There's not excuse for not utilizing all the tricks in our location recordists arsenal for minimizing reverb, but it's still great to know what's available.

Link to comment
Share on other sites

No, but it's a great tool for us posties. Ditto the unreverberator in RX3.

 

Now... just please don't tell DPs about it. Or else they'll be saying F#&k Sound even louder: "How cares if the boom isn't close? They can deverb it in post. I've got a friend with this great plugin..."

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...