iZotope Iris

[TwoToneshuzz]

For instance, it'd be much more interesting if each marked area was treated as it's own key-group with it's own sends, envelopes and filters. It'd also be much more interesting if it did some stretching and tempo syncing together with the ability to force some marked zones as static, non-pitch shifting.

Cheers!
bManic

I like these ideas! From the early days after the Iris release.

It's sounds like it would be a heavy duty expansion of Iris's capablilty. I see this suggestion as something that builds on and expands Iris's core capability.

You would need to handle this with some extra tweaks to the idea. So that you could implement this by enabling it on a per case basis up to a number of say four per sample lane for a start.

[Brando]

I have my reasons for responding to this the way I did. So I rest easy about being called a fool by you.

Let's be fair - I said you were making a fool of yourself. As in, your behaviour, not your level of intelligence.
I simply asked you to go back and to re-read the statements in this very thread, as they were made by Whyterabbit, and by yourself.
Pot=Kettle.
And- really - lighten up. Threads are intended to be a way for people to express an opinion. I can't see anything in Whytrabbit's statements that were in the least derogatory, or even slightly negative - yet that's how you took them.

[TwoToneshuzz]

@ Brando

Okay Thanks

[TwoToneshuzz]

I just found another important difference of Alchemy and Iris. In Alchemy you could edit the existing harmonic content by deleting or multiplying the amplitude of existing harmonics or adding new harmonics with a certain amplitude (by selecting the drawing color between black and white). This allows a deep editing of an imported sample or allows to create a totally new sound from scratch.

It would have been nice if Iris would have contained a 3D spectrogram display like iZotope Ozone 5 does. Based on the price of Iris (especially the full price) this is not a thing a customer could not ask for.
http://www.izotope.com/products/audio/o ... bridge.asp

Ingo

This seems like a well defined workflow feature idea or two. Not sure if I'd need or want to do spectral editing like this seems kind of fussy, but it sounds possible to implement, first make a selection then access a gain control pop up on right click per selection.

3D spectrogram, is beautiful for sure, but i'd imagine in use it would take too much CPU power away from the actual business of the plug in. If it didn't have any CPU hit I'd use it. But no it's not high on my list.

[ajoe]

Iris reminds me of the Steinway piano sitting in my living room; it is a nice piano but could have been so much more.

The Steinway makes nice sounds but if you want to play more than one sound at a time, you have to hold down each and every key you want to hear. Want to play a chord, you need to figure out which notes go together. If the keys you want to play are too far apart, tough luck. If you want sustain, you have to use your foot to hold down a pedal. Jeesh!

But get this, that Steinway (probably a v. 1.0 model) has no support for external sounds, no effects, no filters, no LFOs, and ADSR is 100% manual.

I sure hope history doesn't repeat itself. Come on Iris, you don't want to be the next Steinway, do you?

[TwoToneshuzz]

Iris reminds me of the Steinway piano sitting in my living room; it is a nice piano but could have been so much more.

The Steinway makes nice sounds but if you want to play more than one sound at a time, you have to hold down each and every key you want to hear. Want to play a chord, you need to figure out which notes go together. If the keys you want to play are too far apart, tough luck. If you want sustain, you have to use your foot to hold down a pedal. Jeesh!

But get this, that Steinway (probably a v. 1.0 model) has no support for external sounds, no effects, no filters, no LFOs, and ADSR is 100% manual.

I sure hope history doesn't repeat itself. Come on Iris, you don't want to be the next Steinway, do you?

That's funny

Now the idea here is to go into a modus where anything goes. You could attach helium balloons to the back of the keys in the steinway to give them more lift!

[TwoToneshuzz]

i like it
i have some other spectral tools but iris is much quicker to use
for a 1.0.0 release the gui quirks i experienced are acceptable

some features i would want to see in a free 1.* update::
{a} choice of individual stereo outs for each of the four parts for routing in our daw. good enough for phatmatik pro good enough for iris
{b} possibility to save load & browse presets for the effects at least (sux redialing the same favourite settings all the time when creating new patches)
{c} possibility to save load & browse patches (sample, masks, settings etc) for each individual part. right now it only allows to switch samples per part in the preset browser (when browsing presets click on full browser to open that menu)
{d} some slight refinement of the selection tools especially the magic wand

i may have missed something though as i havent rtfm

These seem clear enough. More part of the pre and postprocessing side of the Iris tool set.

Except for the bit about the magic wand, I guess he means a sensitivity or threshhold control.

In {c} I wonder if he knew about the load sample but preserve selections feature..That sort of allows for some of what he is asking for here.

My way to use this would be to create selection templates in a preset and then import my fresh samples. This could be also done with an effects setup. It's clunky I know but would still work, especielly if you just want to try out various effects or selections on a bunch of different materiel..

[TwoToneshuzz]

Very nice synth, great interface!

The one feature I really miss coming from Cameleon/Alchemy is the ability to morph between the four soundsources.

Anyone know if you can "move" in more than one direction through the sound at the same time? (eg in Poseidon I can have different voices moving through the sound landscape at different rates and in different directions - very cool)

Morphing between the four soundsources. This comes up quite alot spectral morphing and such. I think though if you have access to actually drawing on the sonogram you are doing a type of hands on spectral morphing, and when using 3 different sammples you can do some effective crossfades in a very specific way that might in the end be more rewarding.

Morphing several files togther seems like a somewhat different approah. While Iris is about doing detailed very specific work. Morphing seems like a type of sends effect where sounds move through a black box processor that does realtime processes on the input and spits out an output. Here I'd guess you tweak settings in an educated way to achieve your desired results but your still basically flying blind compared to the way you would work in Iris.


Anybody into the technical aspects of Morphing between sound files in realtime? I'd love to here some details if there is anyone with deep knowledge on it checking out this post. Any ideas where in the Iris workflow morphing would fit in? Pre or post sonogram painting?

Perhaps if you did this with your selections in Iris as a type of post process you could tweek or meld/glue your sounds together in interesting ways.

aMused reference to the way you can move through a sample in more than one direction, there is only 2 directions on the time line as yet? anyways 2 at the same time sounds interesting indeed. Nothing to stop you from doing a version of that now either by having more than one instance of Iris playing or using the same sample in all three slots. But it would be CPU intensive. Each movement would require a sample voice..

So to implement this in one instance say for 3 times 2 directions Iris would require a healthy bump upwards in the polyphony otherwise it would only work in monophonic modus.

[TwoToneshuzz]

Just off the top of my head: How about saving snap shots of selections and then morphing between them via a env or lfo? How about a time sync grid overlay and snapping spectral selected regions to the grid, or sweeping freq ranges through the audio range modulated by an assignable LFO? How about feathering a selections bourders or smoothing? And then there is all the stuff you COULD have done with time stretching, multiple loops, modular routing, inter voice modulation, etc.

Some interesting things here.

Ability to make a snapshot of selections and then morphing 2 or more seems useful.
I'd setup different presets in two different instances of Iris in my host to acheive the same effect.

Snapping spectral selected regions to the grid. That would be nice too, for now I could load a sample file with pulses recorded at the subdivisions I needed and use that as a visual guide.

Sweeping freq ranges through the audio range modulated by an assignable Lfo? I do that down stream using a eq plug, like filterscape. Though I'm not not sure I've understood what he means here..Sweeping the selections so they move or just the output after the selections.

Feathering selection borders or smoothing? That would be a wonderful refinement.

This jasinski is into this I can see.. I like his ideas very much..

[TwoToneshuzz]

So this is like the long awaited instrument version of 'METASYNTH's image filter ???
4 engines ..but no ability to use separate midi chanels for each ..this means that when we want toplay each channel separately , we have to divide the keyzone into 4 zones ..meaning a limited range for each engine ., no separate outputs
2.no ability to control the volume of the selected/masked spectral content ..selections outputs the original file at full volume .., so maybe ad an opacity slider would come in handy , maybe use red , green for placement in stereo image .;like in metasynth
Limited range of tools , ability to rotate ( rectangular )horizontal/vertical selections into diagonal for rising/falling filter efects .
Also , since iris can analyze the spectral content of an audio file , why not include the ability to superimpose :apply the spectral image to a new file ...for vocoding effects .example ...ananlyze spectral content of voice ...apply this as a filter on a new file
)

To play one of the four "engines you could modulate the gain controls for each. Basically solo the one you want to play.

opacity slider sounds like a interesting idea or just a selection gain control on right click.

Stereo placement within each sample, would that require the ability to have each selection have it's own mix out. Sounds like it would be very expensive CPU wise. Maybe if you limited it to four outputs per sample this could be possible. Still quadrupling the number of voices that would need to be reserved for each instance.

Vocoding effects seem present in a lot of tools as an effect. Could this be done downstream in the audio signal flow? They could add this to the list of internal effects in Iris, but my preference would just to have 4 stereo outputs out of the plug and using external vocoders like Waldorfs Lector or for additional filter madeness then filterscape.

In the final analysis isn't all this extra functionality a double edge sword, a single instrument gets more powerful but also more unwieldy, opapque and complex...

[rj0]

Just off the top of my head: How about saving snap shots of selections and then morphing between them via a env or lfo? How about a time sync grid overlay and snapping spectral selected regions to the grid, or sweeping freq ranges through the audio range modulated by an assignable LFO? How about feathering a selections bourders or smoothing? And then there is all the stuff you COULD have done with time stretching, multiple loops, modular routing, inter voice modulation, etc.

Some interesting things here.

Ability to make a snapshot of selections and then morphing 2 or more seems useful.
I'd setup different presets in two different instances of Iris in my host to acheive the same effect.

Snapping spectral selected regions to the grid. That would be nice too, for now I could load a sample file with pulses recorded at the subdivisions I needed and use that as a visual guide.

Sweeping freq ranges though the audio range modulated by an assignable Lfo? I do that down stream using a eq plug, like filterscape. Though I'm not not sure I've understood what he means here..Sweeping the selections so they move or just the output after the selections.

Feathering selection borders or smoothing? That would be a wonderful refinement.

This jasinski is into this I can see.. I like his ideas very much..

+1

[TwoToneshuzz]

Another Track using Iris.

Sounds from Patchpool subsription set, and the iZotope libraries, Wood, Glass, Voice


https://www.box.com/s/weimnxtrzrlynvfshbft

The some of the Wood library sounds are useful to break up the soundscape with something crispy and in your face. Well done and something new!

[TwoToneshuzz]

I've gone through alot of the thread now for the 2nd time and it was such a pain to bring the different interesting points up on the the front page again so I pasted the posts into a document and now drop then here. I see this activity as a way to understand something about the expectations mostly power users have had to Iris and to consider where these viewpoints come from. Many have used Alchemy, Cube or Posiedon. At least Cube and Posiedon are much more expensive and very mature products. So these can give ideas to Feature improvemnets to later versions of Iris if there is a user base to support it.

I don't agree or disagree with any of these improvements I'm just looking them over and thought others might find it conveniet to have the remaining bits in besides the ones already posted in this one post here instead search through 40 pages of a thread.

The Quotes are attributed at the bottom of each segment of text.

renderful wrote:
… Along with time-stretching and other tempo-based features, I'd like to see tempo based selectors/brushes, so we can paint and/or select 1/4 triplets and things like that.

Excellent idea! Let's keep up the pressure on our friendly iZotope project manager for time stretch and tempo/beat based features in a 1.x upgrade (update)!

/Joachim

Just installed the demo. First impressions - I love the interface, especially how clean it is, the fact that you can remove all extra details and just focus on the spectral display, and that it can be resized even to almost full screen. That is something I asked for long ago for Poseidon (on the other hand Poseidon has a beautiful 4D waterfall display - something I would love Iris to have). Some of the presets are quite nice too but I would like to see more diversity. In terms of editing power it is intuitive but a bit limited in that you appear to be able to operate on the audio file by selecting and moving around various details but it doesn't look like you can draw in your own or manipulate the partials directly (unlike in Alchemy) Also the filters are very basic compared to Poseidon and the lack of multi point envelopes is a big limitation too. I do like the All view where you can manipulate the 4 audio files together but there doesn't seem to be any way to place them in any sort of temporal sequence, while their selections can be moved around the files themselves are all at the left edge. Poseidon is much better at manipulating the time domain than Iris - you can also do more complex things with the position of the play cursor in Poseidon. Seems fun to play with though.

aMused
----


While I'm finding it fun to play with, Iris definitely feels like a work in progress to me. Some FR's:

1. Separate filters for each section rather than one Global filter. This puts a real damper on filtering each sample separately. It also means that all the FX sends are PRE-filter unless you use the FX master mode, but then you lose the sends per sample. There should really be one filter per section with an envelope dedicated to each, with a pre/post selection for the sends.

2. Limited Modulation resources. More LFOs assignable to different parameters would be most welcome. It would also be nice to see LFOs affect shape positions and filtering intensity... maybe even groups of them...

3. Macro routines for drawing. For example, if you want to fade in drawn sections (cone shapes) it should be a macro allowing you to rubber-band select the start section to whatever point and execute a fade in from start to selection. It should also be possible to draw straight lines, and other geometrical shapes.

4. The ability to "feather" and create subtle fade in/out effects with the drawn shapes, anywhere and to whatever degree. Currently the method in point 3 is the only way to simulate this, but whatever the shape, whatever is drawn simply plays at full volume.

5. Save of filter patterns. I can see the reuse of certain patterns with different content, which currently means re-drawing them every time.

6. Import of graphic images as filter patterns, with full transform and editing capabilities. Of course

I really hope they don't take all our ideas and the put out a $999 version like someone else mentioned, although you have to admit, as much as it's infuriating, that it's not a bad commercial strategy...

Breeze


Things I'd like to see in a future version:

Waterfall display mode
Drawing, editing and filtering directly onto the spectral image as well as just selecting areas - in Cube 2 you can even draw on a comb filter
Multi point envelopes
More innovative filters
Image import
Spectral morphing and the ability to impose the spectral analysis of one sound onto another
Ability to change relative positions of samples in the "All" Window (so that they are in sequential relationships instead of just layers)


Amused

About suggestion :
+1 for the A/B switch to compare frequencies (could be extend to a A/B preset btw.)
I would suggest too more automation possibilities, I'm working on a interface with Usine to add physics movements to parameters and could be cool, in a general way, to have control (direct VST parameters or midi learn) for example on :

* Switch to each sample view.
* the 3 Invert selection buttons.
* the open mix window.
* the All button


This really help work with tactile screen (or by midi interface of course)


Nay-seven


I've experimented with the demo for about an hour now, and I like the ease with which one can drag a sample into the interface and the fast resynthesis.

I think I'm finding that I can't do a few basic things, however. Is there a way to:

1. Select a single freq, or band or freqs without dragging a selection tool? I'm used to using Spear and Morphine and other programs that isolate freq bands. You just click on a band to select it and in Spear, you can CTRL click to select or deselect the ones you want to work on.

2. See the exact freq or band of freqs that one is selecting? The Frequency ruler offers a broad view, and can be set to show increments of 5 hz but I can't find a way to see the exact freq.

3. See actual amplitude of a selected band?

4. Change selected freqs? Shift the pitch or change the amp envelope, etc, of just the selected section? (EDIT: I understand that one can select a band of freqs, detune them, save that as a new sample, load it, remove the band from the original resythesis, and then play both sounds by using Mix. But that seems a bit complicated if the goal is just to hear how a slight change in pitch affects the overall sound.)
Last edited by Jake Jackson on Sun Apr 29, 2012 2:07 pm; edited 1 time in total


Jake Jackson


Same here. I'm not a synth collector, in fact: I only have a few. But carefully chosen ones.
None has given me the same feel as Iris has. It's easy to use, has a near infinte range of possibilities and allows much more control tha I thought it had when I first tried it out.

What I would like to see added is the ability to create your own brushes: small black and white png (gif has to be paid for)in the shape you want. White doesn't change, black is used as the brush. Add some anti-aliasing and even the sky is no limit anymore.
And if that is not possible, or too difficult ( I'm not a sound engineer) perhaps individual width and height, so circles could become elipses. And also squares /rectangles. Dragging them is ok, but I'd love to be able to draw repeating sets of a same width/height.

Iris does not have to enter competition with Alchemy or Harmor. It has to stay "simple". No need to import images ( I know, I mentioned that a while ago), only brushes we can shape ourselves.

The ability to use the filters on a full-blown track/wav could be another offspin of RX2, without the real repair tools. If they create it, for the same price as Iris, I'm interested.
----
Cubase 6.5 64bit | Windows7 | Selected set of VST&VSTi

ErikH

I am enjoying my Iris license very much, but the program is definitely crying out for some more advanced features.

The first and foremost feature that I think Iris needs is the ability to assign a depth to regions the mask drawn over the spectrogram so that the mask operates as more than just a logical mask. This would enable more natural-sounding transitions between sounds and also more natural simulated filter sweeps. I understand that enabling a drawing capability like this would add a level of technical depth to the program that is very complicated.

Better drawing tools would probably be necessary to implement depth maps, for example the ability to draw and adjust the size of shapes, draw and connect paths, assign gradients or blends to regions, have brushes with different shapes and spreads, etc.

Time stretching was also mentioned, and I agree that it would be nice to be able to do that. I have also mentioned in this thread that I would like to be able to set a point in time where the motion through the spectrogram is held (during the sustain portion of an envelope) at a fixed time like a single cycle waveform.

Various import and export functions would also be useful, such as import and export of images instead of sounds, direct export of wave forms at a fixed note, etc.

I am not sure if this is the appropriate place to suggest a laundry list of other features, but I can think of many others.
----
------------------------------------------
Gribs

[whyterabbyt]

stage 1 : attack folk just for making FRs
stage 2 : comment/respond to all FR's suggested
stage 3 : start making own FR's
stage 4 : self-appointment as collator/editor of everyone else's FR's

i think i get it now.

[Shabdahbriah]

stage 1 : attack folk just for making FRs
stage 2 : comment/respond to all FR's suggested
stage 3 : start making own FR's
stage 4 : self-appointment as collator/editor of everyone else's FR's

i think i get it now.

Yeah... IMHO ~ This was one of those exercises in the "receptivity" (to new ideas in general) and some basic/classic psychology of "innovative/innovation" cognizance dynamics, where-in one becomes the "innovator" (proponent/et al) after appropriate indoctrination; e.g. a form or manner of (essentially) 'peer-pressure'*

*among other things: the challenge of juggling (for lack of a better term) resisting/embracing the persistence of any intrigue or interest in limitations, which might cause one to 'doubt' vs. the praise in the innovations current status AND its potential by others, which might cause one 'doubt' about their (own) attitude regarding limitations, vs. self-limitations (conditioning).

(I just pulled all of that out of my ass, 'cause I wanted to play-on 'loquaciousness')

So, it's kinda like a spontaneous bowel movement.

But then, one can see the 'picture' on the puzzle-box lid... and yet not grasp it's 'elements' to any great degree... until you open the box, and set the lid aside.

and begin to 'connect-the-dots' ~ yourself.

... fascinating.