Magenta 3 this friday

[pablo1980]

I do find Magenta sound a bit harsh (I think I am not the only one)

In my opinion, The sound is inferior to newer acqua releases or similar model third party libraries.

Do you think that Magenta will be updated to the new technology and improve a bit its sound?

[cjperez]

Just Curious.. Any idea when Magenta 4 with core 8 or 9 will be out? (not that i have a problem with Magenta 3)

Giancarlo, any info on this?...

[giancarlo]

yes it will be core9. Not sure about the release date.
Not sure we'll add more internal gain steps, there was a project. Anyway NOT this year.

[markgalup]

Just wanted to resurrect this topic - any updates on when we might be so fortunate as to see Magenta4? I think this could be one of your most popular Acquas if you nailed it with your current standards

Best,
MG

[Chicoria]

Come on, folks. It's been a while Magenta 3 is out and everybody knows it has some issues. For me the biggest one is the 18KHz not being able to function as a shelf eq. Something very weird happens if you try to boost it when in shelve mode. Anyway at least a bug fix should've come soon.

I am an enthusiast of Acustica Acqua and Nebula but I'm kind of unhappy seeing you from Acustica to put out a lot of new stuff but unable to solve issues with older plugins and your loyal costumers.

Please, give us some update about that.

Cheers
Chico

[davidgary73]

Magenta 4 > around 8th September product update (G said Magenta4 will be a revolution product update)

[kenbobplus3db]

Quoting from my earlier post: (http://www.acustica-audio.com/phpBB3/vi ... 27#p117868)

These are my top three Magenta revisions for v4:

1) Shelving EQ available on all 4 bands. Part of the magic of the modeled device is the ability to stack shelving equalizers on top of each other. (A 3Khz shelving boost with a 16hz shelving cut becomes a very wide bandpass, for example). This can't be done on Magenta 3.

2) Control + click needs to reset band gain to 0 not the middle of the boost (+ 6 db , +16 or so on the input trim relative to the default setup). Its not a lot of fun when you're tired, working fast, and you control+click your tweeters into oblivion.

4) A global band-gain scaling control running from 0.5 to 2.0. Think of this as a more/less control. After you've got your eq setup, and want a bit more of everything, run the gain multiplier to some factor > 1. Perhaps your client loved the direction of your eq, but wants everything dialed back a touch. Then just pull your scaling factor back to some value < 1.

[I'd like this to be a standard feature on all acqua EQs]

*New*
6) Please take a look at the way you're sample rate converting your vector files. If you down-sample to 44k and then upsample to 88k, then does that not take the high frequency content out of your 88k vector set? My thinking is that the anti-aliasing removes the content at 44k, but then it doesn't show up in the upsampled 88k convolution. Likewise, it seems like the 21-24k content is similarly missing when working on 48k audio.

[This may be the start of another topic]

[heraldo_jones]

G said Ivory is the latest product for September and looks like Lemon will be the one for October so I guess Magenta4 will be around Nov/Dec. Hope to see the Vari-Mu and Slam!, also I liked the Elop once I tried it. And the Voxbox for the channelstrip would be killer ... my wet dreams

[realhomerecording]

6) Please take a look at the way you're sample rate converting your vector files. If you down-sample to 44k and then upsample to 88k, then does that not take the high frequency content out of your 88k vector set? My thinking is that the anti-aliasing removes the content at 44k, but then it doesn't show up in the upsampled 88k convolution. Likewise, it seems like the 21-24k content is similarly missing when working on 48k audio.

[This may be the start of another topic]

Where did this information come from? Yes, of course one should always downsample from the highest resolution source. Upsampling doesn't add any new useful data. Unless there is some other rationale for doing things that way...

[kenbobplus3db]

Where did this information come from? Yes, of course one should always downsample from the highest resolution source. Upsampling doesn't add any new useful data. Unless there is some other rationale for doing things that way...

This from the purple manual:

"1.1. Sampling Process
The sampling process has been performed by Acustica. The unit was sampled with mastering quality converters, using a method which takes considerably more time than the normal sampling standard. This method is of benefit to the entire audio spectrum. Two sample rates are provided with PURPLE; the native sample frequency was 96 kHz. The 44.1kHz, 48 kHz, 88.2 kHz frequencies were derived from the native one by a down-sampling and up-sampling process. This method avoids any negative sample rate conversion (SRC) artifacts when matching project
different sample rates and also helps in project loading times."

I assume this is the procedure for the alternate sample rate vector packs on all acquas-- but I'd love to be proven wrong. It makes for some odd implications. I've included a spectrogram for the pearl2 compressor at 88khz-- can you tell at which point I've activated the preamplifier?

[markgalup]

Where did this information come from? Yes, of course one should always downsample from the highest resolution source. Upsampling doesn't add any new useful data. Unless there is some other rationale for doing things that way...

This from the purple manual:

"1.1. Sampling Process
The sampling process has been performed by Acustica. The unit was sampled with mastering quality converters, using a method which takes considerably more time than the normal sampling standard. This method is of benefit to the entire audio spectrum. Two sample rates are provided with PURPLE; the native sample frequency was 96 kHz. The 44.1kHz, 48 kHz, 88.2 kHz frequencies were derived from the native one by a down-sampling and up-sampling process. This method avoids any negative sample rate conversion (SRC) artifacts when matching project
different sample rates and also helps in project loading times."

I assume this is the procedure for the alternate sample rate vector packs on all acquas-- but I'd love to be proven wrong. It makes for some odd implications. I've included a spectrogram for the pearl2 compressor at 88khz-- can you tell at which point I've activated the preamplifier?

This has gotten me thinking since I've read it. Should probably start it's own topic to get some more chatter going on, I'd love to know the general thoughts. I hate to miss part of the spectrum even though I'm working at a higher resolution than 44.1k, ya know...

MG

[RJHollins]

are you working higher than 96K ?

[markgalup]

Hehe, touche, no I'm not. My computer would punch me in the face if I tried my mix sessions at 96k. Thus, I mix at 48k

MG

[mathias]

I assume this is the procedure for the alternate sample rate vector packs on all acquas-- but I'd love to be proven wrong. It makes for some odd implications. I've included a spectrogram for the pearl2 compressor at 88khz-- can you tell at which point I've activated the preamplifier?

in this case (pearl2 comp), the 48-88 vectors for the preamp are missing.
request them at support.

with the correct samplerate-vectorfiles, you don't get the bandstopfilter of a lower samplerate.
i don't know anything about the conversion-procedure described in the magenta-manual, but i know for sure, that it does not lead to bandstop-filters that are lower than half the samplerate, as it should be.
i guess that conclusion is not correct.
with all those samplerate-packs i have tested it was never the case (or it was a bug).

again, ask support for providing the missing sampleratevectors (48-88) for pearl2 comp.

EDIT: inbetween use this workaround:
change <DEFAULTRATEKHZ> 44 </DEFAULTRATEKHZ>
to <DEFAULTRATEKHZ> 96 </DEFAULTRATEKHZ>
in pearl2comp.xml

this will convert the samplerate from 96 to 88 on loading for the preampvector and circumvent the bandstop-filter your picture is showing.

[markgalup]

mathias, thanks for the input. That definitely makes sense.


Best,
Mark