Introducing DSPi | A powerful, user friendly and open source DSP for less than a cup of coffee

[john61ct]

should now be a perfect clone of the Syn, complete with Presence control.

FYI, Syn has four separate "Modes", indicated by a pair of front LEDs

Process & Shape Mode (Both Lights On): Full processing mode. The surround sound channels are turned on, and the analog Width and Presence controls are actively shaping all channels, including the main outputs and headphones.

Stereo Mode (No Lights On): Standard 2-channel stereo output. The surround and center sound channels are turned off entirely, and all analog processing (Width and Presence) is bypassed.

Process Mode (Left Light Only): Surround sound channels are turned on, extracting ambiance for the center, surround, and subwoofer channels. The main left and right stereo channels remain completely clean and unprocessed.

Shape Mode (Right Light Only): Surround and center channels are turned off, but the analog Width and Presence controls are enabled to affect the main left and right channels (can only be accessed using the remote)

With music, user will need to choose the mode best suited to the content and how it was mixed, and of course their personal judgements.

 

[somebodyelse]

Is net

The UART and I2C interfaces talk to the same command orchestrator as the USB vendor interface, so all system commands are reachable via all interfaces. This is with the intentional exception of UART and I2C configuration commands, in order to prevent accidents in situations where USB is inaccessible.

You can read the full specification for both interfaces here.

Is WiFi/BLE control still on the roadmap for the W boards, or is it to be left to a separate board using DSPi-ESP32-Front-Panel or similar? Either way I'd like to suggest AES70/OCA support on the network side.

 

[Weeb Labs]

Is net

Is WiFi/BLE control still on the roadmap for the W boards, or is it to be left to a separate board using DSPi-ESP32-Front-Panel or similar? Either way I'd like to suggest AES70/OCA support on the network side.

BLE control might still make an appearance for the W boards but I'm afraid WiFi won't be supported. This is because the radio requires driver blobs and a networking stack, which would occupy far too much RAM to make sense. I think an external ESP32 is the most practical way to handle this.

I might be adding support for CAN under Control Interfaces, as it's quite well suited to this type of application.

 

[Wesseling Audio Lab]

Maybe it was mentioned but my outputs are setup I2S as shown on upper left panel but on lower screen it list them as SPDIF as shown in fist attachment
Also when I mute any output here

and unmute it will continue to display muted on the matrix mixer.. I guess mute on MM is sort of operating independently from PEQ/XO screen

Using DSPi Windows Console v1.1.5-beta5-hotfix and DSPi v1.1.5-beta5​

 

[Wesseling Audio Lab]

I'm interested in turning on the fan on class D amplifiers only at a high volume. it possible to use a Control Surfaces output and have it switch on/off with a prescribed volume level? Then I could use a 5v relay on the power line to the fan.

 

[MCH]

I'm interested in turning on the fan on class D amplifiers only at a high volume. it possible to use a Control Surfaces output and have it switch on/off with a prescribed volume level? Then I could use a 5v relay on the power line to the fan.

My recommendation to you is to use a pwm fan (=which allows speed control) and a temperature sensor in the amp connected to any random microcontroller. You can code something like "start spinning at low speed when temperature reaches x then increase speed gradually at increased temperatures". If well tuned, it is granted you will never ever hear it which I guess that's your valid concern.

 

[BURNART]

Is there a possibility to add frequency response transition point from stereo to upmixed? Like in Riven SRP? I need to go through posts before, because I just saw this thread and I'm wonder if it can be used in car.

 

[Wesseling Audio Lab]

My recommendation to you is to use a pwm fan (=which allows speed control) and a temperature sensor in the amp connected to any random microcontroller. You can code something like "start spinning at low speed when temperature reaches x then increase speed gradually at increased temperatures". If well tuned, it is granted you will never ever hear it which I guess that's your valid concern.

Control from actual temp is the best solution! thanks
Yeah, I don't want to hear the fan at all, or any noise in my room.

With some of these Class D amps from Ali that are measured here on ASR and elsewhere, the photo shows heat sink without fan and then when it arrives it has a fan.
Then I read about the issues with overheating at higher output. Anyway, fans are unplugged on my latest boards and no issue so far at my typical listening volume.

 

[robca]

Please be aware that running the pipeline at 96KHz will halve the available DSP instructions without an audible benefit.

To reinforce this: I have been peeking under the hood at what Genelec does, and their DSP (at least for the 8320A) runs at 48KHz. If it's good enough for the internal Genelec DSP, methinks it's good enough for most

 

[Blew]

To reinforce this: I have been peeking under the hood at what Genelec does, and their DSP (at least for the 8320A) runs at 48KHz. If it's good enough for the internal Genelec DSP, methinks it's good enough for most

It would depend on the quality of the reconstruction filter.

 

[Weeb Labs]

I'm interested in turning on the fan on class D amplifiers only at a high volume. it possible to use a Control Surfaces output and have it switch on/off with a prescribed volume level? Then I could use a 5v relay on the power line to the fan.

Yes. The LED and LED (Dimmable) components in Control Surfaces can be used for these purposes. The latter can produce a PWM signal with output level derived duty cycle. This could be used to control a fan.

 

[kboe]

To reinforce this: I have been peeking under the hood at what Genelec does, and their DSP (at least for the 8320A) runs at 48KHz. If it's good enough for the internal Genelec DSP, methinks it's good enough for most

I would prefer high bit-depth over high sample rate at any time. High sample rate increases upper frequency limit which no human being can hear. High bit-depth decreases rounding errors and therefore quantisation noise which IMHO is much more relevant especially in this case where volume control is done before DA conversion.

 

[TOMSON]

How should I set the I2S bus as an input circuit? I have a PCM1808 module. I am attaching photos of the circuit. The PCM5102 output circuits are working properly.
The input voltage is 5V, and the regulator is 3.3V.

 

[Jumpybah]

Come devo configurare il bus I2S come circuito di ingresso? Ho un modulo PCM1808. Allego le foto del circuito. I circuiti di uscita del PCM5102 funzionano correttamente.
La tensione di ingresso è di 5 V e quella del regolatore è di 3,3 V.

 

[TOMSON]

With one I2S input pair I can only select GPIO1, when I switch the input to I2S the CPU load on Pico 2 is 100% on USB, it works OK

 

[Weeb Labs]

I am happy to announce the release of DSPi Firmware v1.1.5 Beta 6, along with Console for Windows and Console for macOS.

This release introduces several new functions, a minor redesign and fixes some longstanding bugs.

Here are the most significant changes:

• RP2350: Stereo Upmixer enables a center channel and surround channels to be derived from any stereo input. It includes Sinner and Logician engines that work similarly to the Schitt Syn and Dolby ProLogic II upmixers.
• RP2350/2040: All output clocks now remain active through flash writes and system state changes, with brief audio fades. This eliminates all pops and clicks while changing settings and effectively renders External Mute Control redundant.
• RP2350/2040: IR remote control input is now much faster, enabling very rapid button presses and repeats to be registered.
• RP2350/2040: Control Surfaces now includes nouns for Stereo Upmixer, Psychoacoustic Bass, output delay in milliseconds and preset reload.
• Console: Full system states can now be imported and exported. Worried about losing your presets during major firmware updates? Export them!
• Console: Dashboard cards correctly follow renamed channels and pair correctly when a channel is disabled.
• Console for macOS: The parameter copy/paste function can now be activated via Cmd+C and Cmd+V keyboard shortcuts. It operates on the selected channel. Coming to Windows soon.
• Console for Windows: Response graph pills have been retired in favor of the sidebar pills, which now serve exactly the same function. Click to enable/disable. Coming to macOS soon.
• Console for Windows: Linkwitz Transform filter values can now be edited correctly and must be explicitly applied in order to take effect. Switching to another filter type discards the values for safety.

We have now reached the final beta and next will be the final v1.1.5 release. I can't wait to show you some of the fancy things like software VCXO that we have been playing with for v1.1.6!

The usual disclaimer: Please bear in mind that this is still a beta release and while great care has been taken to thoroughly test all functionality, some bugs may remain. Should you encounter a bug or unexpected behavior, please don't hesitate to make a report so that it can be squashed! You can do that either here, on GitHub or via the Discord server (many helpful people there).

Finally and as always, DSPi is a not-for-profit open source project funded entirely by donations and out of pocket. If you wish to support the project, you can do that via Ko-fi or Patreon.

 

[gadget-man]

Five betas in one month! I can barely keep up with testing each version—and you’re developing during this time! I’m still thrilled about this project. GREAT JOB!

I noticed one (minor) bug in the beta6 console (macOS Tahoe 26.5.2):
> In “Control Surfaces,” the values on the second level can’t be selected; they just flash briefly. It’s very similar to what happened in beta2 in the “/AutoEQ/Favorite Profiles/” menu.

I wish you continuing success with this project!

 

[mike_008]

Could Upmixer be made more universal? Meaning not limited to LR as input channels. In a home cinema environment one could make additional Atmos ceiling speakers (4->6 configuration). Or some additional surround speakers.

 

[Weeb Labs]

Could Upmixer be made more universal? Meaning not limited to LR as input channels. In a home cinema environment one could make additional Atmos ceiling speakers (4->6 configuration). Or some additional surround speakers.

This should actually be possible already. You can feed the upmixer a stereo pair of surround channels to produce two more (and correctly process the initial pair).

Direct support for this function is absolutely feasible and will likely make an appearance down the road.

Five betas in one month! I can barely keep up with testing each version—and you’re developing during this time! I’m still thrilled about this project. GREAT JOB!

I noticed one (minor) bug in the beta6 console (macOS Tahoe 26.5.2):
> In “Control Surfaces,” the values on the second level can’t be selected; they just flash briefly. It’s very similar to what happened in beta2 in the “/AutoEQ/Favorite Profiles/” menu.

I wish you continuing success with this project!

Thank you for the report! I don’t have access to Tahoe at the moment but everything seems to work correctly under Sonoma. I will investigate this evening.

In other news, I will soon be pushing USBtrx to the supporter channels. It is a more specialized firmware whose sole job is to offer audio input and output via USB, with various configurations from 10-in to 4-in/6-out, 6-in/4-out, 10-out and others. Supported interfaces will be ADAT, TDM8, I2S and SPDIF.

Several users have asked about this tool and I suspect many might find it useful.

 

[mike_008]

One maybe stupid question. Started playing yesterday with latest version. My use case would be for one pico2:
8ch i2s(in) to ADATout AND 8ch ADATin to 8ch i2s(out). Is this possible? Because it seems 8 i2s(out) channels are actually always mirrored to ADATout.