Getting great results running DD in manual mode and automating on a clip by clip basis here.. Very easy to tweak and get good results

Then maybe and instance in auto mode with very minimal reduction on the dx chain..

As for dereverb, haven't been able to get any remarkable results there yet, but everything else is great for my needs.

Be aware that sometimes noise masks problems with the recorded dialog. I've definitely had instances where listening to the dialog the first time it just sounds noisy, then after denoising it sounds 'clean' but with artifacts. But then going back to the original what sounded like artifacts was there all along, just covered by noise. Of course, hitting the denoiser hard will make the voice crappy, but in some cases it's more a matter of unveiling defects in my experience.

Nothing "remarkable" here yet either, but definitely "enough" for dialog to be brought "forward" enough to satisfy clients. It's a bit tricky to set though, but careful denoising will help perceived pumping I find.

That pic is hilarious! The reference made me laugh, thanks.

That's a good point there. However the (only) material I have tested it on yet was MKH50 going into SD 788T. So I doubt that this was in the signal. Maybe pushing too hard, maybe a not ideal situation, but multiband expansion for me quite often seems to win over these FFT (or in this case probably combined) noise-print style algorithms. But in this case it was marginally. Then again, on my brutally revealing speakers, it is "there".

How does a Cedar work? Does anybody know? Very fine multiband expansion?

I'll keep ABing in the near future of course

There is now RX 3.01 out, where Dialogue Denoiser runs almost twice as fast (without any quality loss). Please check it out.

Downloaded. I'll check it out. Thanks!

Cool!

I don´t see the optimization of Diadenoiser on the featurelist. You should absolutely mention it there.

Whoops!

Seems to be a bug in 3.01 where dia-denoiser and dereverb won´t recognize the ilok when in RTAS or AS. Standalone and the others seem fine.

PT 10.3.7
Mac os 10.8.4
newest ilok drivers

It's in the feature list of the Advanced version.

I´m an idiot.

I downloaded RX3, not RX3 advanced.

Move along, nothing to see here.

Yep, just opened a current post project with a number of DD's running, and I gained back 10% CPU in Pro Tools 10 HD. Thanks for the Friday Freebie!

i picked up RX3 (non-advanced) to remove some micro ear bleed on a new show coming in december ... very impressed with the Spectral Repair ... if all goes well this will save a network show ... and I have not read any manual - just a few youtube videos over the last year ... did not really have a need for it up until now and now i feel like i wish i had this type Repair Kit all along ... thank you Izotope ... can anyone compare this to the Sony Version ?? if they are complimentary I may pick that up as well, but if they both do relatively the same I will save some cash for another tool .. thank you john

What's micro ear bleed?

micro ear's are IFB's that are RF .. without the need for an over the ear Clear Tube Loop - fits in the ear ... crank them up to loud or they are loose and you can get bleed into mic's ..

The DeClipper is brilliant!

Trying to decide if RX3 Adv is worth it... it seems like it gains a ton of features over the RX3, which is probably why it costs so much. I can't afford Adv in any way right now, but I feel like if I was to get RX3, it would have to be Adv...

Redjacket on here is selling advanced copies for five bills. Look him up.

Sent from my SAMSUNG-SGH-I337

I used DD on some audio recorded on a beach and mic'ed too far away. Typical nightmare audio. It was a quick no-time-for-the-manual fix... and... amazing results.

+ 1 for a makeup gain (maybe auto-makeup gain too). And maybe a Presence knob for after processing like Unveil has, I lost some highs that had to be eq'd back in.

Agreed.

Sent from my SAMSUNG-SGH-I337

Since we are in feature-request-mode:

I would love to see a gain function in Audio suite mode.
Sometimes it is easier to lift what you want out of the noisefloor than to reduce the unwanted sutff, and in stand alone mode RX has this function. Would be a great addition to spectral repair.

Oh .... i need to learn how to do this ...

"lift what you want out of the noisefloor than to reduce the unwanted sutff"

I disagree with this methodology.
I tend to view out as much noise that I can without effecting the quality of the dialog.
Reason being, so you don't hear things pumping when raising quiet lines or end of sentences that drop off. Also, you might have mixed a noisy scene, leaving lots of noise in, because theres music to help hide it.
Then, after playback, someone decides they don't want music there, and you have to remix the entire scene.


I also find it simply being lazy.

This deserves being repeated. In type of programming I've done this has often been the case, and having to deal with it on the spot when music has been pulled out is just wasting the client's time (when attending the mix playback).

maybe i am thinking about this process described the wrong way ... i agree i would be cleaning up the surrounding noise First, as my process is ... but pulling say a word out of context in a noisy program might be helpful ... i do not do episodic - mostly i get a few things here and there that are unusable at best ... and sometimes just need to not touch anything but this one element that needs to poke out a bit ... in that case it could be helpful to know how to do it ..

but again i agree as you say clean the surrounding area / pushing the noise away from the wanted elements ...

i hope i explained that well ...

the earpiece bleed i am about to start dealing with is the producer yelling into the ear of the talent on set / and then leaving his finger on the Momentary button too long and you hear the director calling cameras as well .. and he was loud ... EQ'ing the 3k bleed touches the dialogue to much IMHO, but the Spectral does a nice job of attenuating the bleed to a usable degree ...

mind you, no producer should be talking to the talent that much during a taping but it happened .... and I used to Production Mix "Change of Heart" years ago and the producer fed every single line to the Host through his IFB ... i had a fostex speaker in the booth and would listen to this direct IFB feed - brilliant it was .. as the host could listen and repeat while speaking - he was an animal at it ... and no one knew the better ... the production i am speaking of cleaning was not as smooth by any means ...

cheers

john

How does the noise reduction sound compared to rx2. I really didn't like it at all ( personal taste as I know lots of you like it) if it sounds the same i wont upgrade at this point, might save for a cedar instead

Sent from my Nexus 4

Over use of De-noiser units including Cedar make your dialog sound worse than with the noise there..
I've been hearing lots of overuse on TV and even a motion picture .. Please back off on the noise reduction.

Sent from my Nexus 4

It's a new world out there. They know anything can be fixed in the mix these days. I have video editors place room tone for me to use (thanks) which contains the producer(s) talking to the host and others on set without an earpiece (but not thanks) all the time. Apparently they can't hear it in the edit suite and think all is well. Nobody ever tells the producer of course.

It's ok, my last room tone was preroll.... lol

Sent from my SAMSUNG-SGH-I337

I don't disagree.
It's knowing how to use and not abuse your tools.

I think the big problem is that work people want to use a tool like CEDAR to do the whole thing in one go. And it doesn't work that way. I have heard of people using 3 instances of CEDAR ONE on their dialog tracks.

It just doesn't work that way.
But it's not hard to get clean dialog, using a variety of processes at the same time.

I recently saw a movie, outside scene, nighttime. And even though there were additional crickets cut, I could hear the little surge of additional production crickets on each line. That should have been notched out. It's not that hard. I also don't want to hear a chorus of production crickets all over my dialog either.

I approach my TV dialog, the way I would a feature. Clean, full. Every camera angle, every shift, every character,every word, has to match as much as possible. Anything less is lazy.