REW signal generator alternative

[grid_stopper]

I am building a box to function as a standalone test tone generator, it will be centred around a raspberry pi +a 7" TFT screen with a more robust line amp attached to its audio output so I can drive heavier loads. My initial plan was to run REW on the pi and use a rotary encoder and a couple buttons for navigation, however it doesn't look like the individual modules in REW cant be run standalone, and the function generator window can not be resized, which makes it unfortunately not practical for this specific use case.

What I am wondering is if there are any similar function generators that are

A. similarly functioned(shows values in dbfs, voltage, dbu, dbv ect ect, does white+pink noise, can be calibrated)
B. works on Linux

Ive found a few signal generators for Linux but they are pretty bare-bones.
Alternatively, if anyone knows how I can make REW work for my purposes I am of course open to that as well.

Hopefully there is something out there that works!

 

[grid_stopper]

ok so it turns out REW has an API that allows it to be launched without the GUI which seems like a good starting point, but i believe(and someone please correct me if im incorrect...) that I would need to be connected to wifi. anyhow, utilizing it is definitely above my paygrade at this point. more research is required if i want to go the REW route.

 

[Tubetec]

I use a decomissioned Iphone as a balanced signal source ,
I prepare the test tone wav files in Reaper first ,
making 2 mono, out of phase tracks
this creates a balanced signal at the headphone jack , it will output around 0db into low loads .


I found a third party lightening to minijack line out cable cheap
It will put out around +6db max .
The onboard volume control provides attenuation of the test signal in a precise accurate and repeatable way ,
your just using the media player to play the prepared files , so no special software or Apple account is required .
It takes a little time initially to create your test tones and calibrate the levels but once you have that done you have a very handy portable high quality signal source with a touch screen .

Alternatively there are 3rd party generator apps for the iphone .

 

[grid_stopper]

I use a decomissioned Iphone as a balanced signal source ,
I prepare the test tone wav files in Reaper first ,
making 2 mono, out of phase tracks
this creates a balanced signal at the headphone jack , it will output around 0db into low loads .


I found a third party lightening to minijack line out cable cheap
It will put out around +6db max .
The onboard volume control provides attenuation of the test signal in a precise accurate and repeatable way ,
your just using the media player to play the prepared files , so no special software or Apple account is required .
It takes a little time initially to create your test tones and calibrate the levels but once you have that done you have a very handy portable high quality signal source with a touch screen .

Alternatively there are 3rd party generator apps for the iphone .

this is actually very similar to what ive been doing, however i am not only a glutton for punishment, but I am trying to learn some new skills and this is the project i have chosen to do so on.

As it stands all the other signal generators that are linux based that ive seen dont have a lot of the bells and whistles that REW has.

 

[cyrano]

Zynthian open source synth project?

https://zynthian.org/

 

[molke]

You could take a look at the ‘ancient’ software by Fons Adriaensen, especially Jaaa (sine and white noise), Japa (pink and white noise), and Jnoisemeter:
https://kokkinizita.linuxaudio.org/linuxaudio/index.html
Unfortunately, the GUI is probably not really suitable for small screens, and I don’t know if it can be operated with custom controls.

 

[Tubetec]

I found this and thought it would be worth a mention here since its available for linux also
https://github.com/koord-live/KoordASIO
Its based on FlexASIO , which I had trouble setting up before ,
This program provides an installer and GUI for the various Flexasio dependencies
Seems to work well enough with Windows ,REW and my laptops internal soundcard ,

After its installed you can run Flexasio test and port audio device detection from the commandline and it lists the sample rates and bit levels available on the device . Its much more configurable than ASIO4ALL , where you often end up stuck with 44.1khz 16 bit because it doesnt configure the drivers properly .
I had no problems using my internal soundcard at upto 96khz 24bit ,

 

[mrchapp]

I don't know how comfortable you'd be around creating a GUI app, but GStreamer might work here. audiotestsrc is an element that produces the following signals:
* Sine
* Square
* Saw
* Triangle
* Silence
* White uniform noise
* Pink noise
* Sine table
* Periodic Ticks
* White Gaussian noise
* Red (brownian) noise
* Blue noise
* Violet noise

A quick prototype to see if it works for your audio output can be tested with this:
gst-launch-1.0 audiotestsrc wave=pink-noise volume=0.8 ! pulsesink

(Ctrl-c to stop. You can switch pulsesink with pipewiresink or alsasink, depending on your setup.)

Also, if you do prepare sample files offline, you can play them back with:
gst-launch-1.0 filesrc location="sample.wav" ! wavparse ! pulsesink

GStreamer has a whole framework and there's plenty of sample UI apps using it, so it shouldn't be hard to whip up something for this purpose.

 

[ruffrecords]

One problem with software generated waveforms is that it is hard to generate low distortion low level signals for testing mic pres for example. A -50dBu output has a resolution of only 46dB on a 16 bit system so the distortion will be at least 0.5%. For this reason I feed computer generated outputs via a test box which includes switched attenuators of 40dB and 60dB so a full 16bit resolution signal is produced at low level.

Cheers

Ian

 

[kags]

One problem with software generated waveforms is that it is hard to generate low distortion low level signals for testing mic pres for example. A -50dBu output has a resolution of only 46dB on a 16 bit system so the distortion will be at least 0.5%. For this reason I feed computer generated outputs via a test box which includes switched attenuators of 40dB and 60dB so a full 16bit resolution signal is produced at low level.

Cheers

Ian

Is the volume control on most USB interfaces controlling the DAC or an analog attenuator? I've assumed analog on the less expensive, 2-channel units, but maybe I'm wrong and need to set up some attenuators!

 

[ruffrecords]

Is the volume control on most USB interfaces controlling the DAC or an analog attenuator? I've assumed analog on the less expensive, 2-channel units, but maybe I'm wrong and need to set up some attenuators!

I am pretty certain they are analogue but its going to be hard to use them for 40dB or more attenuation. One tiny movement and you calibration is lost. Fixed attenuators are repeatable.

Cheers

Ian

 

[Tubetec]

The signal is digitally attenuated when you reduce the level in software , so yes it will get bit munched

I found high sample rate test tones here ,
https://www.audiocheck.net/testtones_highdefinitionaudio.php

I dont see any reason not to be using 192khz for test tones .

 

[kags]

i did some research and I see that the Focusrite 2i2 that I use for testing has an analog output volume control. I adjust the output level on my scope each time, and I keep the test tone at something like -12 dBFS, so good resolution.

 

[cyrano]

Get yourself an old RME FF400. Digital input and output control to 0.1 dB precision. PGA2500 as mic preamp is wonderful for measurements. 100% repeatability. Headphone out goes down to DC if needed.

I've pondered a few times over DIY measurement stuff over the years. Even built something that worked, years ago. These days, I hardly need anything besides that FF400.

 

[JohnRoberts]

One problem with software generated waveforms is that it is hard to generate low distortion low level signals for testing mic pres for example. A -50dBu output has a resolution of only 46dB on a 16 bit system so the distortion will be at least 0.5%. For this reason I feed computer generated outputs via a test box which includes switched attenuators of 40dB and 60dB so a full 16bit resolution signal is produced at low level.

Cheers

Ian

For testing mic preamps it is useful to pad down line level (or as hot as you got) signals down -60dB or so.

At the same time you can control the impedance presented to the mic preamp input from the pad to mimic typical mic source impedance (150-200 ohm).

JR

 

[kags]

For testing mic preamps it is useful to pad down line level (or as hot as you got) signals down -60dB or so.

At the same time you can control the impedance presented to the mic preamp input from the pad to mimic typical mic source impedance (150-200 ohm).

JR

I got used to this setup from SSLs that provide oscillator patch points for 0 dBu and -60 dBu. Today, only some of those oscillators are still in working order. I just refurbished an old HP 209 signal generator from our tech shop deep storage. Handy unit if low distortion isn’t necessary.

 

[JohnRoberts]

I got used to this setup from SSLs that provide oscillator patch points for 0 dBu and -60 dBu. Today, only some of those oscillators are still in working order. I just refurbished an old HP 209 signal generator from our tech shop deep storage. Handy unit if low distortion isn’t necessary.

Line up oscillators built into console master sections were optimized for amplitude and frequency stability, not low distortion. It was popular to use function generator ICs that typically converted triangle waves to sine waves using non-linear diode break circuits.

These were generally used for printing reference level and/or slate tones on master tapes, zeroing VU meters, and the like.

JR

 

[ruffrecords]

I got used to this setup from SSLs that provide oscillator patch points for 0 dBu and -60 dBu. Today, only some of those oscillators are still in working order. I just refurbished an old HP 209 signal generator from our tech shop deep storage. Handy unit if low distortion isn’t necessary.

My old Ferrograph test set has a millivoltmeter with an input attenuator switch which gives it a range from -60dBu to +40dBu in 10dB steps. You don't see that too often these days.

Cheers

Ian

 

[ricardo]

My old Ferrograph test set has a millivoltmeter with an input attenuator switch which gives it a range from -60dBu to +40dBu in 10dB steps. You don't see that too often these days.

You got a working RTS-2 ??!! I'm jealous, Ian.

I need an tiny Allen set screw to fix mine. Then I'll gut it and install SOTA circuitry using the switches, meter & caps.

 

[ruffrecords]

You got a working RTS-2 ??!! I'm jealous, Ian.

Yes, it is a beauty. I had to replace the meter movement but it works fine.

I need an tiny Allen set screw to fix mine. Then I'll gut it and install SOTA circuitry using the switches, meter & caps.

The switches are certainly industrial grade!

Cheers

Ian