New from FXpansion - D.CAM Dynamics [A suite of Four Dynamics Plugins]

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
Post Reply New Topic
RELATED
PRODUCTS

Post

paterpeter wrote:BTW: a BIG FAT THANKS to the FX pansion support to react so positively to the special request I had.
Hm... so that's the way to get it at an introductory price for existing FXpansion customers. I whish that was official instead.

Post

Angus_FX wrote:antithesist - generally no. A select few may sell it as downloads at some point in the future, but it's not a channel-wide boxed product.
Thanks!

Post

andy_cytomic wrote:
Breakpete wrote:
Breakpete wrote:Are these plugins zero latency?
Finally got around to download the demo, so I'll answer my own question:

I did a quick test of BusComp, ChanComp and EnvShaper in Renoise, and found they had 2 - 3 samples of latency using the default quality setting (2x oversampling).
Perhaps someone from FXpansion give a some more accurate information, but I believe are using a non-linear phase oversampler with around 4.18 samples of delay at DC which increases to around 70 samples at the highest frequency. The only way to compensate for this latency in your DAW is to run all tracks / buses through the same number of non-linear phase oversampling filters as each other, which has to be done manually.

Here is the plot of the amplitude response (top graph) and latency (bottom graph called "group delay") for the x2 filters in Synth Squad for a base sample rate of 44.1 kHz:

Image
Can we hear from fxpansion on this issue? For the record, I'm in no way affiliated with Cytomic, not even as a customer. Tally-ho.

Post

The plot looks about right. As you can see, within audible frequency range the delay is near-constant and <10smps (0.25ms, <10cm/4in air delay). Below 10K, it's 5smp +/- 1smp (~0.1ms, 5cm/2in air delay).

As I outlined previously, there is one case where it's liable to cause issues (namely, HF phase cancellation), and that's when running the comp in parallel in a DAW mixer with no additional processing in either path. (Once you throw an EQ in either signal path, all bets are off for phase integrity anyhow.)

For this reason we recommend you use the comps' in built mixers if you want to do parallel compression. The CrossComp has additional filters, which by design are not linear-phase, and that will have a much more marked impact on audio quality if you do parallel compression via the DAW mixer rather than using the built-in parallel knob.

If people are running in to other audio quality issues as a result of the oversampling, we're open to improvements (an 'Off' mode or a filter with a different set of trade-offs - likely to be higher latency and CPU load & some degree of pre ringing), but I have to say that our testers seemed pretty happy with the design as-is.
This account is dormant, I am no longer employed by FXpansion / ROLI.

Find me on LinkedIn or elsewhere if you need to get in touch.

Post

Thanks, I must have missed the previous explanation. I'll try to go back and find it. I guess I haven't thought about phase shift in terms of samples much.

I like to be able to control over-sampling when at all possible, but awareness of it and filter types is probably as good. I have at least one plug-in that must be also be over-sampling with non-linear phase filters like this.

Is your over-sampling scheme consistent for all sample rates? Some over-sample by 2SR at 44.1 and 48, but not at 88.2 and above, for example.

Post

Didn't go in to much depth previously - see reply to Breakpete above.

DCAM Dynamics & Etch oversampling' scheme is switchable 2x/4x, and doesn't vary by host sample rate.
This account is dormant, I am no longer employed by FXpansion / ROLI.

Find me on LinkedIn or elsewhere if you need to get in touch.

Post

will cpu hunger be adressed in the first update?
Finally!

Post

cyphersuit wrote:will cpu hunger be adressed in the first update?
As its guardian, feeding your CPU will be your responsibility. Just give it a little Artic Silver every now and again and you should be cool.

Seriously though, the CPU load is what it is. The sort of component modeling involved is CPU intensive by nature. These plugins are not CPU intensive simply because FX pushed them out the door half finished and optimising seemed like such a chore. This sort of product was not released in 2002 to run on PIIIs - they would have melted.

Post

Angus_FX wrote: If people are running in to other audio quality issues as a result of the oversampling, we're open to improvements (an 'Off' mode or a filter with a different set of trade-offs - likely to be higher latency and CPU load & some degree of pre ringing), but I have to say that our testers seemed pretty happy with the design as-is.
Please consider off option for oversampling! I know audio quality is under everyone's lips but I'd happily trade some quality that I couldn't hear within the mix anyway for cpu/latency...

Post

Did anybody else notice that the Channel Comp uses slightly more CPU on the 12:1 ratio than on any other setting?
Has anybody ever really been far even as decided to use even go want to do look more like?

Post

Was looking into picking this up, but then noticed this potential hazard w/ Logic regarding latency issues. Does anybody know about this?

http://www.fxpansion.com/forum/viewtopic.php?t=7578152

Post

Phase47 wrote:Was looking into picking this up, but then noticed this potential hazard w/ Logic regarding latency issues. Does anybody know about this?

http://www.fxpansion.com/forum/viewtopic.php?t=7578152
In other words, proceed with confidence, or, wait for bug fix?

Post

Phase47 wrote:
Phase47 wrote:Was looking into picking this up, but then noticed this potential hazard w/ Logic regarding latency issues. Does anybody know about this?

http://www.fxpansion.com/forum/viewtopic.php?t=7578152
In other words, proceed with confidence, or, wait for bug fix?
7 months later and it's still not fixed, if that's FXpansion's response to an "important" issue I'd say move on to something else.
"What embecile composed this list :/"

Post Reply

Return to “Effects”