KVR Audio

andy-cytomic · Post by **andy-cytomic** » Thu Jun 22, 2017 3:50 pm

meloco_go wrote:
andy-cytomic wrote:News update: The Glue v1.3.17 is up. This version fixes a problem with bug with the new oversampling system. Click on the little (+) button in the download section and remember to double check in a new project that the plugin is authorised if you haven't done so in a while:

https://cytomic.com/glue
Great! Thank you!
I wonder with new oversampling options: since there are control voltage path and audio path in this compression, maybe it would make sense to have separate OS options for both? Could be a waste of CPU and a PITA to get the timing right, but maybe it is a way to get the correct behavior of diode (as you have shown in Scream) and still have full phase coherency when used as the parallel processor.

The same implicit solver is used in The Scream as in The Glue to for the diodes, I've just added non-linear capacitance to the diode modelling in The Scream, but otherwise it's the same math. With linear phase oversampling everything is already time aligned properly, there is no gain to be made in terms of time alignment from splitting the gain reduction calculation from changing the gain.

meloco_go · Post by **meloco_go** » Fri Jun 23, 2017 11:11 pm

andy-cytomic wrote:The same implicit solver is used in The Scream as in The Glue to for the diodes, I've just added non-linear capacitance to the diode modelling in The Scream, but otherwise it's the same math. With linear phase oversampling everything is already time aligned properly, there is no gain to be made in terms of time alignment from splitting the gain reduction calculation from changing the gain.

I thought about the example you have shown in The Scream thread, where, due to the pre-ringing of linear phase oversampling there was a situation where it resulted in waveform differences when compared to hardware behavior. You have shown how minimum phase oversampling solves this. But I don't know if similar can be applied to the Glue.

PS just upgraded to latest betas -- all seem to work great now! Thank you for the update!

MogwaiBoy · Post by **MogwaiBoy** » Sun Jun 25, 2017 6:52 am

Let us know when the Beta goes legit : )

andy-cytomic · Post by **andy-cytomic** » Fri Jul 07, 2017 12:12 pm

News: The Glue v1.3.18 is ready for download. This build fixes a minor possible bug with the new oversampler in linear phase mode, and also improves the nulling ability of the ovrsampler in all modes. It's a recommended update for everyone.

The Glue v1.3.18 (7 July 2017)
https://cytomic.com/glue

New: reduced oversampler latency to use the lowest possible to support current realtime and render oversampling amounts
New: added advanced setting to allow dynamic updating of latency when switching between realtime and render – note: very few hosts support this, so this setting has to be manually changed in the settings.xml config file
Fixed: uninitialised variable in linear phase oversampler

Everglide · Post by **Everglide** » Fri Jul 07, 2017 1:50 pm

Are you considering making a modelled preamp? I'm currently testing the Acoustica Audio Gold with various Neve preamps. It gives a dimension and richness I can't reproduce with other preamps or non-linearity EQs. I'm using the Glue to compress once the signal is amplified with the Gold pre's but it taxes my CPU. I'm hoping you could sprinkle your algorithmic dust on a Cytomic preamp!

meloco_go · Post by **meloco_go** » Fri Jul 07, 2017 2:47 pm

Everglide wrote:Are you considering making a modelled preamp? I'm currently testing the Acoustica Audio Gold with various Neve preamps. It gives a dimension and richness I can't reproduce with other preamps or non-linearity EQs. I'm using the Glue to compress once the signal is amplified with the Gold pre's but it taxes my CPU. I'm hoping you could sprinkle your algorithmic dust on a Cytomic preamp!

If you own The Drop you can try playing with it as a general "color" device. If you engage HPF you can set the frequency very low, so it won't really affect audio bandwidth much, and some of the circuits provide plenty of character.

gas pump · Post by **gas pump** » Fri Jul 07, 2017 5:22 pm

Agree or disagree?

https://www.youtube.com/watch?v=FbbW-rh7SWY

HcDoom · Post by **HcDoom** » Fri Jul 07, 2017 9:28 pm

Disagree...judging by his comments and manners I believe he comes from very rich family who told him: do what ever you want, we will pay. Than he said: I want to be an audio engineer (but no talent) and parents paid for seminars and bought him everything he needs. He compares The Glue which is definitely one of the best sw compressors ever created with best aa technology and he said it sucks, than he compares it with FF ProC2 on clean setting...and uses same measures on two totally different compressors. Not to mention wrong gain staging and wrong level outputs. Wont comment any further.

MogwaiBoy · Post by **MogwaiBoy** » Fri Jul 07, 2017 9:30 pm

Interesting premise - going to have to check out his other videos.

The comparison with Pro-C 2 is actually a good one. The Fabfilter is in Clean mode so much less harmonics and noticably less pumping - the sound is clearer and more defined.. very true and respectful to the original source - no surprises there. The Cytomic isn't quite so pristine in response and that's a big part of its charm.

Both have their place!

I agree that "glue compression" is nothing special at all and infact all compression has a kind of "glue" effect because you are effectively reducing dynamics and increasing RMS power which is where the thrust (or "guts") of the music resides. This is why saturation also has a "glue" effect, although it combines frequency smearing too.

virtualpt · Post by **virtualpt** » Sat Jul 08, 2017 12:16 am

gas pump wrote:Agree or disagree?

Totally disagree. He seems to have rather an inflated sense of his own importance, combined with an extraordinary ability to sound like he doesn't really know what he is talking about. Hopefully he will grow up & be less of an arse in a few years

jpga · Post by **jpga** » Sat Jul 08, 2017 1:46 am

Everglide wrote:Are you considering making a modelled preamp? I'm currently testing the Acoustica Audio Gold with various Neve preamps. It gives a dimension and richness I can't reproduce with other preamps or non-linearity EQs. I'm using the Glue to compress once the signal is amplified with the Gold pre's but it taxes my CPU. I'm hoping you could sprinkle your algorithmic dust on a Cytomic preamp!

Please do a pre Andy.

All the best,,

BRBWaffles · Post by **BRBWaffles** » Sat Jul 08, 2017 3:06 am

gas pump wrote:Agree or disagree?

https://www.youtube.com/watch?v=FbbW-rh7SWY

This guy has absolutely no idea what he's talking about. Even if he did manage to to get two compressors to sound similar (which he didn't), that doesn't discredit "glue" as a perceptible sonic characteristic/phenomenon. I mean, in this test you could just as easily interpret his results as evidence that Pro-C 2 is also a good "glue compressor". Which doesn't discredit or falsify "glue" at all.

Personally, I don't care for a lot of audio engineer jargon like "warm", or "punch", or "glue". It's not that I'm skeptical as to whether or not the phenomena they generally aim to describe exist. It's that people often have disparate interpretations of what these terms even mean, and I feel it only adds to the ambiguity that naturally leads to YouTube videos like this. So, I try not to use them. I will not say "warm" when I mean mild saturation. I will not say "glue" when I mean cohesive. I will not say "punch" when I mean percussive. But, having said all that, this YouTuber is full of shit, and has no idea what he's doing.

andy-cytomic · Post by **andy-cytomic** » Sat Jul 08, 2017 3:15 am

gas pump wrote:Agree or disagree?

The Glue is an accurate model of the envelope follower and sidechain non-linaerities of a G series SSL Bus Compressor, as far as I know the Pro-C in clean mode is not. He hasn't even matched the levels very well, his comparison clearly shows the Pro-C has auto-gain on, so all he has showed is he prefers louder over quieter. So for me the idea behind the comparison is totally fair, but his execution is sloppy and so the conclusions are invalid.

aciddose · Post by **aciddose** » Sat Jul 08, 2017 8:00 pm

I think his conclusion is made due to bias and suffers from confirmation bias. He starts out advising that his opinion is "glue compression" as an effect is nonsense.

I agree 100% with this point of view because it is absolutely 100% true. "Glue compression" is not an effect that you can suddenly apply to your track to get the next hit.

The problem is he overlooks the fact that "glue compression" is not an effect but a technique much like the difference between using all analog gear vs. ITB and the effect this can have on your mix due to your methods and strategies in working with the issues associated with the platform.

For example:

In the analog world we're dealing with a lot of noise and other physical constraints. (Hanging yourself on a noose made of patch cables?)
ITB we have limited processing power, aliasing and to be honest a quite nasty UI (click click.)

This means the two platforms can produce much different results simply because of all the factors at play and how those things influence the actions you take while mixing and mastering.

So his conclusion is true in the very narrow context in which he sets it up; which is the whole reason he set it up that way.

Everglide · Post by **Everglide** » Sun Jul 09, 2017 1:06 am

meloco_go wrote:
Everglide wrote:Are you considering making a modelled preamp? I'm currently testing the Acoustica Audio Gold with various Neve preamps. It gives a dimension and richness I can't reproduce with other preamps or non-linearity EQs. I'm using the Glue to compress once the signal is amplified with the Gold pre's but it taxes my CPU. I'm hoping you could sprinkle your algorithmic dust on a Cytomic preamp!
If you own The Drop you can try playing with it as a general "color" device. If you engage HPF you can set the frequency very low, so it won't really affect audio bandwidth much, and some of the circuits provide plenty of character.

Thank you for the idea! I got fairly close with both HPF, LPF and EQ before The Drop. This was in fact cleaner and retained the detail compared with Gold

Cytomic 'The Glue' Compressor