What are you using the G.O.S.T. Analyzer for ?

[dante]

For ScopeRise, I'd like a list of comments / applications for the new Analyser.

For example, not having used analyzers much before my first test will be to see if I can get a good visualisation of what my mastering chain is doing. So I will take the stereo outputs from STM to inputs a/b, then take the outputs from BrickMaster (last in chain) into inputs c/d. In the first instance I will just be running it over the next few weeks and observing. After Ive got to know it, I'm hoping it will reduce some of the trial and error of getting a mix to sound good on all 3 of studio monitors/car stereo and iPhone.

So I'm after comments about applications here firstly, and secondly, how this particular analyser potentially can achieve them. I realise its a new device, so initially it may be about expectations, with results to come later.

[wouterz]

I did a little latency test by measuring the phase shift of the waveform using the Analyzer.

Green = Direct output from Modular 4 (the catcher signal)
Red = Return signal from A16 Ultra (connected to Z-Link)
Blue = Return signal from Fireworx in bypass mode (connected to S/PDIF)

The waveform is 43.09 Hz
1000 ms / 43.09 Hz = 23.21 ms per cycle (the Analyzer is showing one full cycle)

The analyzer display is 1024 pixels wide

A16 Ultra Z-Link roundtrip
(23.21 ms / 1024 pixels) * 80 pixels = 1.81 ms roundtrip latency

S/PDIF roundtrip
(23,21 ms / 1024 pixels) * 28 pixels = 0.63 ms roundtrip latency

If I made a mistake in my calculations then I hope someone can correct me

Tip: The G.O.S.T. logo is a freeze button

[tgstgs]

the lighter points are measured values
whereas the darker points are interpolated to draw a line between the measured ones;

1 full periode is determined by the framesize of the display vs the framesize of the samplerate;
fp=sr/fs // 44100/1024=43.06640625 Hz;

time_in_milli_sek=samples/sr*1000;
1024/44100*1000=23.21995464852607709750566893424 ms frame you see at once
80/44100*1000=1.8140589569160997732426303854875 ms latency

yes absolutely correct but
i wouldnt have expected the source to be modified when bypassed;
also i do not see the dB raster is it PNGed away?

btw. when you write this code you would do a precalculate
1/sr*1000 only when the sr switches;
so you have just 1 multiply to do each time

welcome to my crazy world vibes

[wouterz]

also i do not see the dB raster is it PNGed away?

The dB raster is very dark on my screen, when I look at another angle to my screen I can see it but in normal use I don't.

[wouterz]

A16 Ultra Z-Link roundtrip
(23.21 ms / 1024 pixels) * 80 pixels = 1.81 ms roundtrip latency

The speed of sound is 340.29 meter / s
(340.29/1000) * 1.81 = 0.6159249 meter (= 2.02075 foot = 24.249 inch)

So in 1.81 ms sound can travel 0.62 meter through the air. My monitor speakers are further away than 0.62 meter so the biggest bottle neck in a Scope system (latency wise) is nature itself

[dante]

Some good application there wouterz.

So far I have noted 3 main applications :

1) Measuring latency
2) Mastering (comparing pre-master gain to post-master gain)
3) Studying the effect on signal shape of Modular components

Anyone else have any others ?

[jksuperstar]

For Mastering, I could see using an M/S device to get the mid and side content, then analyze the frequency of each. You can see how centered your bass frequencies are. Then one can look at the phase difference of left and right tracks to correct it. This helps you see if there's a reason your bass is thin before grabbing the EQ and turning it up.

[jhulk]

im going to use it

to see phase mismatch in scannable wavetables to remove clicks

and also aliasing frequencies in samples in the upper region so that i can add active filtering to remove the frequencies above nyquist

i would also like to add how fast the response in sending keys was

and they were delivered with in a 24 hour period

thanks s/c

[tgstgs]

if the wavetables are not bigger as 1024 samples you may display them via a historiesed catcher;
i do this with filterkernels for example;
send a click trough a filter historiesed the curve as well as catch it;
you see the impuls response in time domain and the frequency response in frequency domain;
for filters smaler blocksizes work better course the clickenergy spread over the bins is not that high;
also multiply y axis by 20;
please tell me if this works for wavetables as well;

good vibes

[jhulk]

i will

so i can send an impulse through a analog filter to capture its response that would be a cool feature

which would be good as you could then send the pulse at different frequency and res settings

this would make it easier for making a dsp model of an analog filter

my wavetables are 256 single samples by 128 cycles

i can de-phase the single cycles so they dont have opposing phases which causes the clicks

but i cant see the phase problems it is not until i try the wavetable in my sampler that i can here the phase change clicks

so im hoping with your plug i can see the phase change and then i can adjust it for that location

will leave you feedback on how i get on with it

im off from doing any work until the 7th jan so will be spending some time on this

which will save me lots of time

[Liquid EDGE]

What I'm intending to use it for as well as helping to check sub is center and to help see smooth mastering is done is this.

Helping with the mix.
You know when you get two or so sounds clashing. To help see and pinpoint where to cut or boost to clear way for each sound to stay consistent in the mix and clash no more.

Yes good ears can do this, but I really want to home in and get that exact nuance that will have the least detrimental effect to said sound to clear the way. Eq the sounds in such a way that it does not clash but sounds the same. Hope you get my wofflings.

[katano]

I'll use it for feedback!

[Mr Arkadin]

I'll use it for feedback!

[katano]

but hey, pssst. It's a secret

[dante]

but hey, pssst. It's a secret

Seriously, we now have :

1) Measuring latency.
2) Mastering (comparing pre-master gain to post-master gain).
3) Studying the effect on signal shape of Modular components.
4) Phase Mismatch correction.
5) Mastering (centering Sub Bass).

Any more ?

[Liquid EDGE]

6. Mixing. Comparing individual sounds in a mix of a piece of music to help with seeing where to make way and eq/clean up to make way and unclash sounds. Kicks and bass/sub, pads and melodies etc etc.