Login

Nebula & THD plots

Officially Licensed 3rd Party Developer Libraries
Free 3rd Party Programs

Nebula & THD plots

Postby yr » Fri Sep 09, 2011 11:07 am

In the "CDSM" thread, Michael mentions that the VST plugin analyser only shows the lowest dynamic levels of a Nebula preset. This morning I made a small test using NAT, and created a preset with only the last repetition being manipulated (we are talking -43.5dBFS peak level). As reported by Michael, only the last (and manipulated) repetition could be observed with the VPA.

The big question is, why are 3rd party developers using VPA plots on websites/manuals, and is it done correctly. For instance, I can get the same THD plot in VPA as shown on a console manual (tested with a -3dBFS signal). If I understand it correctly, Nebula is not going to use the lowest dynamic layer with such a loud signal.

Anyways, I'd like to hear users/developers take on this issue. Perhaps there should be some kind of guide-lines as to how presets/libraries are presented.
Reuven | post-production & sound-design | scenography |
website | nebula presets
User avatar
yr
Beta Tester
Beta Tester
 
Posts: 438
Joined: Mon Mar 29, 2010 9:04 am
Location: Amsterdam

Re: Nebula & THD plots

Postby yr » Fri Sep 09, 2011 8:34 pm

ok, I think I figured it out (I hope). Did some more tests with NAT/analyzer(s). What happens is that the vst plugin analyzer always shows the frequency response of the lowest (weakest) layer. That means that the frequency plots that you see in library manuals/websites are probably not representing the actual response at nominal levels but rather the "weakest" preset layer. Probably not a big deal since there wont be a huge variation between the 2 in most cases.

When it comes to the harmonic structure, the vst analyzer is actually showing the maximum THD levels that a preset generates, and not the distortion generated by the "weakest" layer (or base level distortion). That is why you can expect similar results from the vst analyzer and other "real-time" analyzers used for checking presets in your daw.
Reuven | post-production & sound-design | scenography |
website | nebula presets
User avatar
yr
Beta Tester
Beta Tester
 
Posts: 438
Joined: Mon Mar 29, 2010 9:04 am
Location: Amsterdam

Re: Nebula & THD plots

Postby Cupwise » Fri Sep 09, 2011 9:21 pm

how would vst analyser know where the lowest or highest sampled dynamic was, to show the results at that point?

one program could be made to have samples going down to -15dbfs, and another could go down to -40dbfs. how would the analyser know to show you the frequency response at those points, when you switch from one program to the other?

just me personally, but i don't feel that the harmonic graphs are all that useful anyway. i mean, harmonics ARE there if the program has them, and looking at them on a graph probably won't give most people a great idea of what it is going to sound like. i'm sure there are people who can get use out of them (harmonic graphs), but i feel that the distortion content is something that you have to hear, and get familiar with through experience using that effect/program, to know when you might need it. just my opinion.
Cupwise
Expert
Expert
 
Posts: 773
Joined: Tue Nov 23, 2010 2:03 am

Re: Nebula & THD plots

Postby yr » Fri Sep 09, 2011 11:10 pm

I have no idea how it happens (perhaps Giancarlo does) but I'm quite positive it does. Not only was it reported by Michael (CDSM) yesterday, but I tested it several times today. It's actually very easy to do: make a preset with the lowest repetition processed differently then the one above it. Open it in the vst analyzer and you will only see a frequency response that corresponds to the lowest repetition.

As for the comments about the analyzer(s)- I agree with you. But I think when 3rd party developers use this kind of graphs in manuals/web-sites, it should be clear what they are actually showing...

Cupwise wrote:how would vst analyser know where the lowest or highest sampled dynamic was, to show the results at that point?

one program could be made to have samples going down to -15dbfs, and another could go down to -40dbfs. how would the analyser know to show you the frequency response at those points, when you switch from one program to the other?

just me personally, but i don't feel that the harmonic graphs are all that useful anyway. i mean, harmonics ARE there if the program has them, and looking at them on a graph probably won't give most people a great idea of what it is going to sound like. i'm sure there are people who can get use out of them (harmonic graphs), but i feel that the distortion content is something that you have to hear, and get familiar with through experience using that effect/program, to know when you might need it. just my opinion.
Reuven | post-production & sound-design | scenography |
website | nebula presets
User avatar
yr
Beta Tester
Beta Tester
 
Posts: 438
Joined: Mon Mar 29, 2010 9:04 am
Location: Amsterdam

Re: Nebula & THD plots

Postby audioanal » Fri Sep 09, 2011 11:21 pm

VPA has very low resolution.
The best way to test Nebula and other plugins is inserting in your daw a tone generator as a first instance, then Nebula and after them Voxengo SPAN 2 at the highest resolution.
If you want to test distortion just generate a sine at different freq. If want to test phase and freq response just generate pink noise.
This test is the most accurate, I think.
An example of a great test:
http://www.hispasonic.com/blogs/analisi ... 737sp/2737

I'll make some tests to compare the developers plots in VPA and the same programs in SPAN and I'll upload here the results when I get some time.
audioanal
User Level II
User Level II
 
Posts: 22
Joined: Thu Dec 16, 2010 12:21 am

Re: Nebula & THD plots

Postby Cupwise » Fri Sep 09, 2011 11:35 pm

yeah i agree about using noise for frequency responses, with span. i've been doing that for a while, even when sampling stuff to see what i'm going to get.

i've also decided to not include distortion plots in my stuff anymore, but that's me.

i still don't see how vst analyser could know what level to send into nebula to get the lowest dynamic sample.. have you tried making programs with different amounts of dynamic steps and different distances between each step, and using your test you described to see if in those cases, it still always shows the lowest sample?
Cupwise
Expert
Expert
 
Posts: 773
Joined: Tue Nov 23, 2010 2:03 am

Re: Nebula & THD plots

Postby yr » Sat Sep 10, 2011 8:02 am

I also created a preset with the first and second steps having different levels of distortion (flat frequency) and another step in the middle is eq'd (rest of the steps untouched). VPA shows a flat frequency plot (last step) and max THD. I don't think VPA knows which level to use- it automatically uses the last (weakest) layer. I don't know why or how..

I used both Span and Nugen to compare to the VPA. The harmonic structure is similar with some small variations. The frequency response of the "real-time" analyzers depends on the input level but reflects the "proper" dynamic level. In that sense they are better then the VPA.

I should add it's possible some developers use the appropriate layer for creating the plots.
Reuven | post-production & sound-design | scenography |
website | nebula presets
User avatar
yr
Beta Tester
Beta Tester
 
Posts: 438
Joined: Mon Mar 29, 2010 9:04 am
Location: Amsterdam

Re: Nebula & THD plots

Postby audioanal » Sat Sep 10, 2011 3:52 pm

I'm coming.
Some comparisons beetween Span and VPA
N**e 1081 console + Eq from AlexB

neve 1081 Hm.jpg
neve 1081 Hm.jpg (118.41 KiB) Viewed 1041 times

neve 1081 Hm 2.jpg
neve 1081 Hm 2.jpg (142.11 KiB) Viewed 1041 times
audioanal
User Level II
User Level II
 
Posts: 22
Joined: Thu Dec 16, 2010 12:21 am

Re: Nebula & THD plots

Postby audioanal » Sat Sep 10, 2011 3:55 pm

very very different curve in both analyzers
Attachments
neve 1081 HS 2.jpg
neve 1081 HS 2.jpg (150.88 KiB) Viewed 1039 times
neve 1081 HS.jpg
neve 1081 HS.jpg (129.75 KiB) Viewed 1039 times
audioanal
User Level II
User Level II
 
Posts: 22
Joined: Thu Dec 16, 2010 12:21 am

Re: Nebula & THD plots

Postby audioanal » Sat Sep 10, 2011 3:58 pm

more
Attachments
neve 1081 HS 4.jpg
neve 1081 HS 4.jpg (140.4 KiB) Viewed 1036 times
neve 1081 HS 3.jpg
neve 1081 HS 3.jpg (124.45 KiB) Viewed 1036 times
audioanal
User Level II
User Level II
 
Posts: 22
Joined: Thu Dec 16, 2010 12:21 am

Next

Return to 3rd party libraries

Who is online

Users browsing this forum: No registered users and 4 guests