Explaining the Graphs in Noise reduction and Reverb Reduction

Post questions and find answers on the latest SpectraLayers release here.
Post Reply
1953mafg
New Member
Posts: 13
Joined: Wed Jul 29, 2020 4:22 pm
Contact:

Explaining the Graphs in Noise reduction and Reverb Reduction

Post by 1953mafg »

Hi,

I was hoping someone could explain what we are seeing [x and y axis and the graphed line(s)] in the graphs that are shown in the Noise Reduction and Reverb Reduction Processing.

Thanks,

Mike

NoiseRed.JPG
(26.61 KiB) Not downloaded yet
RevbRed.JPG
(26.21 KiB) Not downloaded yet

Robin Lobel
Moderator
Posts: 379
Joined: Wed Jul 03, 2019 3:39 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Robin Lobel »

In the noise graph, X is frequency and Y is power. The white curve corresponds to the noise print you registered earlier, and the red curve corresponds to the threshold (based on that noise print) that you defined with the parameter Tolerance. If your signal goes below that red curve, it's considered noise and gets cut from your recording.

In the reverb graph, X is time and Y is power. If you click the Analyze button above, you'll see a white curve which shows you the temporal shape of your reverb (how long it is, and how it fades over time). Anything below the red line will be cut. You can adjust the red line with the parameters. It can takes some adjustments though, the white curve is an approximation, not a ground truth analysis of the reverb. So try to remain as conservative as possible to not cut too much from your signal, which could yield artefacts. The same goes for the noise reduction process.
Robin Lobel - SpectraLayers creator and lead developer

1953mafg
New Member
Posts: 13
Joined: Wed Jul 29, 2020 4:22 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by 1953mafg »

Robin,

Thanks - that's very helpful - it would be good to cut and paste that into the Operation Manual for future editions.

Mike

User avatar
Nspace
New Member
Posts: 27
Joined: Fri Jul 12, 2019 6:12 am
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Nspace »

Robin Lobel wrote:
Thu Aug 13, 2020 6:48 pm
In the noise graph, X is frequency and Y is power. The white curve corresponds to the noise print you registered earlier, and the red curve corresponds to the threshold (based on that noise print) that you defined with the parameter Tolerance. If your signal goes below that red curve, it's considered noise and gets cut from your recording.

In the reverb graph, X is time and Y is power. If you click the Analyze button above, you'll see a white curve which shows you the temporal shape of your reverb (how long it is, and how it fades over time). Anything below the red line will be cut. You can adjust the red line with the parameters. It can takes some adjustments though, the white curve is an approximation, not a ground truth analysis of the reverb. So try to remain as conservative as possible to not cut too much from your signal, which could yield artefacts. The same goes for the noise reduction process.
+1 Thanks Robin for this and Thanks also Mike for the specific question.
1953mafg wrote:
Thu Aug 13, 2020 7:32 pm
it would be good to cut and paste that into the Operation Manual for future editions.
This thread is already added to my "0 SL7 Usage" folder.

Robin Lobel
Moderator
Posts: 379
Joined: Wed Jul 03, 2019 3:39 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Robin Lobel »

Will do.
SL7's documentation has been rewrote from scratch, so if you feel some areas lack explanations or details, please mention it so it can be clarified.
Robin Lobel - SpectraLayers creator and lead developer

1953mafg
New Member
Posts: 13
Joined: Wed Jul 29, 2020 4:22 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by 1953mafg »

Robin Lobel wrote:
Thu Aug 13, 2020 8:51 pm
SL7's documentation has been rewrote from scratch, so if you feel some areas lack explanations or details, please mention it so it can be clarified.
Robin,

Maybe you could consider adding more detail on the import and export of the Time Markers and the Spectral Markers, i.e., the format that the file would need to be in to import data from another program.

I wrote a small program to import labels from Audacity and convert them to Spectralayers and RX7 format so that I could quickly and easily find locations that I had marked in the audio files I'm working on in Audacity. It took a while to get the format correct so that Spectralayers would recognize it.

Thanks,

Mike

Puma0382
Senior Member
Posts: 2013
Joined: Wed Dec 15, 2010 6:11 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Puma0382 »

1953mafg wrote:
Fri Aug 14, 2020 2:04 pm
Maybe you could consider adding more detail on the import and export of the Time Markers and the Spectral Markers, i.e., the format that the file would need to be in to import data from another program.

I wrote a small program to import labels from Audacity and convert them to Spectralayers and RX7 format so that I could quickly and easily find locations that I had marked in the audio files I'm working on in Audacity. It took a while to get the format correct so that Spectralayers would recognize it.
Off topic now I know, but...
... begs the question, how does WaveLab cope with sharing markers to-and-from SL7..? Maybe Robin (and PG - WaveLab developer) could look into that...
System 1:- Win10 64bit, Gigabyte H81M m/board, Intel i7 4790 3.6Ghz, 16Gb RAM, NVIDIA GTX 750 Ti, 2 x 22" HD monitors; Steinberg UR44; Cubase Pro v10.5.20, WaveLab Pro v10.0.40, Studio One v4.6.1, Dorico Elements v3.5.10, Addictive Drums 2, Komplete 12, StylusRMX

System 2:- Win10 32bit, Q6600 2.4 Ghz, 4Gb RAM, NVIDIA GeForce 8400 GS, Delta 1010LT; Cubase Pro v8.0.40, WaveLab Pro v9.1.0, Komplete10, StylusRMX

Robin Lobel
Moderator
Posts: 379
Joined: Wed Jul 03, 2019 3:39 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Robin Lobel »

FYI the .wav format can store markers, it's part of the format specifications. So most audio softwares with markers support should be able to read and write markers from/to .wav files. SL, WL and RX can read/write markers from/to .wav files for instance. Unfortunately Audacity does not seem to support this.
Robin Lobel - SpectraLayers creator and lead developer

Puma0382
Senior Member
Posts: 2013
Joined: Wed Dec 15, 2010 6:11 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by Puma0382 »

Robin Lobel wrote:
Fri Aug 14, 2020 4:59 pm
FYI the .wav format can store markers, it's part of the format specifications. So most audio softwares with markers support should be able to read and write markers from/to .wav files. SL, WL and RX can read/write markers from/to .wav files for instance. Unfortunately Audacity does not seem to support this.
Ok - thanks; was being a bit naive there - I thought maybe this was talking about something more 'specialist'... ;)
System 1:- Win10 64bit, Gigabyte H81M m/board, Intel i7 4790 3.6Ghz, 16Gb RAM, NVIDIA GTX 750 Ti, 2 x 22" HD monitors; Steinberg UR44; Cubase Pro v10.5.20, WaveLab Pro v10.0.40, Studio One v4.6.1, Dorico Elements v3.5.10, Addictive Drums 2, Komplete 12, StylusRMX

System 2:- Win10 32bit, Q6600 2.4 Ghz, 4Gb RAM, NVIDIA GeForce 8400 GS, Delta 1010LT; Cubase Pro v8.0.40, WaveLab Pro v9.1.0, Komplete10, StylusRMX

JohnfromSonar
Junior Member
Posts: 160
Joined: Fri Nov 24, 2017 11:14 pm
Contact:

Re: Explaining the Graphs in Noise reduction and Reverb Reduction

Post by JohnfromSonar »

Robin I can't recall ever being more impressed by a software developer. Please keep doing what you are doing.

Post Reply

Return to “SpectraLayers 7”

Who is online

Users browsing this forum: No registered users and 1 guest