DDMF Plugindoctor

[Compyfox]

May I ask why you want to "analyze" this range?


EDIT:
By the way - plugins that are wrapped via jBridge (32bit -> 64bit) are highly glitched in terms of analysis. The Delta analysis is about 5dB higher on the plot, and it has no flat line, but it looks like as if it was "pulses split into octaves". Same goes for the Dynamics (Att/Rel) view - here the readout is jumping around like mad.

Maybe something to look out for with the next fix

[MogwaiBoy]

May I ask why you want to "analyze" this range?

To see DC offset and subsonic noise.

[plexuss]

I updated the plots in my original post using 1.02 that fixed the phase problem. The post is a comparison of some SSL Bus Compressor software clones using PluginDoctor. The post is here: viewtopic.php?f=6&t=499635&p=7007415#p7007415

[Compyfox]

May I ask why you want to "analyze" this range?

To see DC offset and subsonic noise.

There are no stupid questions - but... why?

DC Offset is one thing, although that can be somewhat seen with the Oscilloscope and a shift in dB (up or down) and/or offset of the sine wave.

But subsonic noise? Guess I'll always learn something new every day...

I updated the plots in my original post using 1.02 that fixed the phase problem. The post is a comparison of some SSL Bus Compressor software clones using PluginDoctor. The post is here: viewtopic.php?f=6&t=499635&p=7007415#p7007415

Questions:
1) why is the Att/Rel plot for Acoustica N4 TimP Bus Pre like it is

2) why do you measure 50Hz as THD+N (usually, that's either 997Hz or 1kHz), especially if you show the "Intermodulation" in the plot directly below it? You're (hopefully) aware that the strength of harmonics can vary greatly at different frequencies?

[plexuss]

I updated the plots in my original post using 1.02 that fixed the phase problem. The post is a comparison of some SSL Bus Compressor software clones using PluginDoctor. The post is here: viewtopic.php?f=6&t=499635&p=7007415#p7007415

Questions:
1) why is the Att/Rel plot for Acoustica N4 TimP Bus Pre like it is

2) why do you measure 50Hz as THD+N (usually, that's either 997Hz or 1kHz), especially if you show the "Intermodulation" in the plot directly below it? You're (hopefully) aware that the strength of harmonics can vary greatly at different frequencies?

Hey Compy. I do not know why the AR plot is like that. It's irrelevant because the Pre has not dynamics. I think it can be considered and artifact of doing an envelope test on code that doesn't do dynamics. As for the 50Hz, I am using all the defaults. My feeling about it is, it shows the hamonics generated over a wider band. PluginDoctor is a visualization tool more than a measurement tool and so measuring standards are less important than informative visualizations. Thats my take anyway and so the 50Hz default achieves that.

[Compyfox]

Hey Compyfox. ... As for the 50Hz, I am using all the defaults. My feeling about it is, it shows the hamonics generated over a wider band.

Two things:
1) zooming is a thing
2) different harmonic response on different frequencies - hard to see with the UI and the "lower max sine peak" than expected - but happening

PluginDoctor is a visualization tool more than a measurement tool and so measuring standards are less important than informative visualizations. Thats my take anyway and so the 50Hz default achieves that.

Clearly not advertised as such. PluginDoctor is presented as "measurement tool" - and standard for decades at this point, is 1kHz for THD+N (some even consider 997Hz to be valid). For 50Hz, you have the "intermodulation mode"

[plexuss]

Hey Compyfox. ... As for the 50Hz, I am using all the defaults. My feeling about it is, it shows the hamonics generated over a wider band.

Two things:
1) zooming is a thing
2) different harmonic response on different frequencies - hard to see with the UI and the "lower max sine peak" than expected - but happening

PluginDoctor is a visualization tool more than a measurement tool and so measuring standards are less important than informative visualizations. Thats my take anyway and so the 50Hz default achieves that.

Clearly not advertised as such. PluginDoctor is presented as "measurement tool" - and standard for decades at this point, is 1kHz for THD+N (some even consider 997Hz to be valid). For 50Hz, you have the "intermodulation mode"

Thanks for the feedback. I look forward to seeing your comparisons using PluginDoctor.

[docdued]

I don't think there is an official standard for THD+N, one would always have to specify at which frequency the measurement was made. Anyway, I've created two new version today, we are now at v1.0.4 including a default settings option, increased frequency range in the lowend, Waves VST shell support and higher test frequency resolution (still quantized to get an integer number of oscillations in the test buffer).

[Compyfox]

I'll take a look at v1.04 once I feel better.

Still... what's the main problem to not get access to 1kHz, but rather offset ones?
And will the other tool suggestions be ported?

What about the offset of signal strength (sine) in the THD+N plot? (signal setup to 0dBFS, UI/plot shows -6dB max peak for the fundamental)



Other than that... look at spec sheets of devices from the last 30+ years. All of them were measured at the nowadays universally accepted 1kHz. Pity that C.Budde isn't around anymore - I think he could have easily explained why 1kHz is considered "a standard".

Thanks for the feedback. I look forward to seeing your comparisons using PluginDoctor.

Won't happen, as I don't have these tools. Sorry.

[Zaphod (giancarlo)]

I updated the plots in my original post using 1.02 that fixed the phase problem. The post is a comparison of some SSL Bus Compressor software clones using PluginDoctor. The post is here: viewtopic.php?f=6&t=499635&p=7007415#p7007415

Questions:
1) why is the Att/Rel plot for Acoustica N4 TimP Bus Pre like it is

2) why do you measure 50Hz as THD+N (usually, that's either 997Hz or 1kHz), especially if you show the "Intermodulation" in the plot directly below it? You're (hopefully) aware that the strength of harmonics can vary greatly at different frequencies?

Hey Compy. I do not know why the AR plot is like that. It's irrelevant because the Pre has not dynamics. I think it can be considered and artifact of doing an envelope test on code that doesn't do dynamics. As for the 50Hz, I am using all the defaults. My feeling about it is, it shows the hamonics generated over a wider band. PluginDoctor is a visualization tool more than a measurement tool and so measuring standards are less important than informative visualizations. Thats my take anyway and so the 50Hz default achieves that.

preamps can be "dynamic"
Due to harmonic distortion they can be like expanders. Since harmonic distortion in nebula is frequency and program dependant they can be like exciters, or feature a multiband behaviour. When you overload they can act like compressors, multiband ones (diamond lift, crimson, primestudio charleston). Normally 1kHz is the standard for the harmonic distortion plot, but they prefer close values in order to show aliasing in a better way. Anyway if the plugin has a frequency dependant harmonic distortion, a volterra analysis could be better, or using different values for the sine. Practically also the representation used in izotope rx is very good, because you plot distortion at varying sines and a single plot gives a lot of information about how distortion or aliasing changes

[Compyfox]

Oh, I'm fully aware that pre's and tape machines can act as "compressors" (especially if you're slightly overdriving the hotspots).

What I (currently, might be a knot in my brain) do not understand however, is the att/rel behavior. At least in this form. Shouldn't there be barely anything, or rather "slow" att/rel readouts? This is what's puzzling me here.



Arguing about the 1kHz tone as "standard" - completely different topic

[docdued]

It's "standard" because it's what most people did. Someone probably started to use 1000 Hz just because it's such a nice, even number and it's also somewhat in the middle of the hearing range. But there's nothing inherent in a 1 kHz signal that brings out hidden artefacts or something like that better than, say, 900 Hz or 1100 Hz.

[Compyfox]

Like I said, Christian Budde is sadly not around this place anymore. But he'd definitely have something to say about that. My personal spin: it's 1kHz because how (easy?) it can be built in hardware form.


In the plugin realm however, a "crooked" frequency can(!) easily show aliasing/"artifacts" rather than a static/stable 1kHz. Hence why it is recommended to also measure at 997Hz, or even with a sine sweep. If you really want to look that deep.

So on one end, it's okay that your engine uses slightly offset frequencies - although the handling is not ideal IMO. On the other, I'd still love to have static/well known 200Hz, 500Hz, 1kHz, 10kHz for measurements as well.


Then again... I'm currently more than covered in this area with different tools.

[DeathByGuitar]

I'm most likely gonna buy this plugin in the next few days. One thing I don't understand though is why the website doesn't display the current version number or changelog for updates. Would surely be useful information to have.

[plexuss]

I'm most likely gonna buy this plugin in the next few days. One thing I don't understand though is why the website doesn't display the current version number or changelog for updates. Would surely be useful information to have.

Just a little detail: its not a plugin. its an app. it loads a plugin and then analyzes it.