ACCEPTED: Creating cross-platform room correction for Kodi
#16
(2015-05-18, 01:01)RogerS Wrote: ...to include microphone calibration data in AESource, but make sure while you are coding that inclusion afterwards will be possible....

My Plan after GSoC is to integrate Python (SciPy, NumPy & Matplotlib) into my addon Wink Then anyone can write small scripts to design filters.
Latest news about AudioDSP and my libraries is available on Twitter.

Developers can follow me on Github.
Reply
#17
(2015-05-18, 22:48)wisler Wrote:
(2015-05-18, 01:01)RogerS Wrote: ...to include microphone calibration data in AESource, but make sure while you are coding that inclusion afterwards will be possible....

My Plan after GSoC is to integrate Python (SciPy, NumPy & Matplotlib) into my addon Wink Then anyone can write small scripts to design filters.

This would be awesome Smile
Not only for your project but for everyone Big Grin

Thanks
Reply
#18
Hello Wisler,

This is a neat feature, and it may be great for a situation where you have a room without a receiver, but some things aren't adding up to me. I don't understand why I would want to offload this task to Kodi when it seems as though I would get greater benefit out of using my receiver for the same task.

1. Once this feature is working in Kodi, for users that have true AV receivers, would you recommend users use this feature, or use the software built into their receiver?
2. What are the hardware environments you feel will benefit from this the best?
3. How do you plan on replacing the AMP that is part of your receiver?
4. It's my understanding that the sound processing on a receiver is much better than those used in sound cards (especially THX enabled receivers), have you done any testing on this?

Help me understand why Kodi is trying to compete with receivers for this feature.

EDIT: After thinking about it a bit, it seems as though this feature would be great for people who have a computer speaker system, like the Logitech Z5500 or Z906, where the control console isn't a fully featured receiver. This would provide a receiver like feature to these types of computer speaker systems, they wouldn't otherwise have. I consider that to be very useful, but I still don't see how this feature would be used to replace a functioning AV receiver.
Reply
#19
I'd say there are arguments on both sides:

Dedicated hardware in receiver = most likely (well, possibly) quicker, less load on the HTPC, but also a fixed implementation. On the room correction side of things you can often adjust relative levels, but there's not much you can do about the curves used etc. once calculated by e.g. Audyssey. There may be some presets but they adjustment is pretty crude typically between those.

Software = more load, but potentially much greater flex-and-tweak-ability.

You could set up a system with, e.g. powered speakers (typically great quality & value for money as the integrated amp is perfectly matched to the speaker hardware) - or one with just a simple amp(s) and speaker and no receiver. Dedicated power amps can be a lot better at the amp bit than integrated receiver/amps (apparently).

Also, with software, you could create different rules for different scenarios basically and more easily switch between them, possibly even automatically based on e.g. time of day or type of content, rather than have to futz about with the usually fairly awful AVR interfaces/remotes.

I'd say there is plenty of coolness potential here Smile
Addons I wrote &/or maintain:
OzWeather (Australian BOM weather) | Check Previous Episode | Playback Resumer | Unpause Jumpback | XSqueezeDisplay | (Legacy - XSqueeze & XZen)
Sorry, no help w/out a *full debug log*.
Reply
#20
From my perspective, I've wanted DRC for a long time although I have a receiver. The reason is quite simple - it's old, pre-HDMI but sounds fantastic and to improve on it I'd have to spend £x00 where x > 500, money I don't have (washing machine broke yesterday!). Also would give the possibility of doing things that receivers just don't do e.g. active crossovers and allow future audio formats to be supported without junking existing kit.
Reply
#21
I can already read it in the newspaper. "Kodi stops people from wearing dirty clothes" Wink
AppleTV4/iPhone/iPod/iPad: HowTo find debug logs and everything else which the devs like so much: click here
HowTo setup NFS for Kodi: NFS (wiki)
HowTo configure avahi (zeroconf): Avahi_Zeroconf (wiki)
READ THE IOS FAQ!: iOS FAQ (wiki)
Reply
#22
I concede! Here is my .02 cents if anyone is interested.

1. The feature is great for scenarios with no receiver, or a receiver\boxed system which does not have this option.
2. If you have an expensive receiver you will probably want to continue using your receiver for this feature.
3. If "like OP" you are interested in getting rid of your receiver this feature helps bridge the gap between the features you would loose by dumping your receiver. (I still believe that current mid range and up receivers provide a wealth of benefits to sound processing that Kodi is unlikely to fully replace) but everyone has a different use case, and this feature will certainly help a lot of people.

Additional Questions:

1. Will you be able to enter the distance of each speaker from the viewer, into kodi?
2. Will this feature be able to detect if a speakers polarity is incorrect?
Reply
#23
(2015-05-19, 17:44)enen92 Wrote: This would be awesome Smile
Not only for your project but for everyone Big Grin
Yeah this is not only a project for me, it's also a project for the community Blush


(2015-05-20, 02:04)guyonphone Wrote: ...I don't understand why I would want to offload this task to Kodi when it seems as though I would get greater benefit out of using my receiver for the same task....
One sentence: you get more flexibility (that's what bossanova808 wrote Blush)

(2015-05-20, 02:04)guyonphone Wrote: 1. Once this feature is working in Kodi, for users that have true AV receivers, would you recommend users use this feature, or use the software built into their receiver?
2. What are the hardware environments you feel will benefit from this the best?
3. How do you plan on replacing the AMP that is part of your receiver?
4. It's my understanding that the sound processing on a receiver is much better than those used in sound cards (especially THX enabled receivers), have you done any testing on this?
1. Hard to say before it is not finish.
2. & 3. I plan to use a dedicated AMP without any signal processing capabilities and do all the signal processing stuff at the Kodi side.
4. It depends on the used sound card or DAC, no at the moment I didn't tested any hardware. The AudioDSP feature isn't finished and I don't have enough money to buy some test hardware. But I think we can get a great music experience with a very good DAC or sound card.

(2015-05-20, 02:04)guyonphone Wrote: I consider that to be very useful, but I still don't see how this feature would be used to replace a functioning AV receiver.
Because not everyone has a AV-Receiver. Another use case for e.g. could be a car or if you don't have enough space for a full AV-Receiver. On the other hand the room correction algorithms are all closed source and no one except the company, which developed the algorithms knows the inner workings. If the software part in Kodi for audio processing is more powerful you can spend more money for e.g. on better speaker or an AMP.
Also the mic from audassey seems very cheap to me and not every AV-Receiver supports other microphones and not every AV-Receiver supports to upload correction files for a microphone.
Don't understand me wrong I really like my Denon X2000 and it's room correction system. I use it every evening and I'm quite happy with my environment, but I really like to develop audio signal processing stuff and prove if I can beat the commercial systems Big Grin
I don't know if I can beat them, but if know one tries we don't know what is possible.

(2015-05-20, 04:34)bossanova808 Wrote: Also, with software, you could create different rules for different scenarios basically and more easily switch between them, possibly even automatically based on e.g. time of day or type of content, rather than have to futz about with the usually fairly awful AVR interfaces/remotes.

I'd say there is plenty of coolness potential here Smile
That's a perfection description and I can't say it better. Thanks.
Room correction is only one use case what is possible with the addon adsp.xconvolver. It's a efficient convolution engine and you can load filters to achieve this features (similiar like BruteFIR):
(2015-05-21, 12:45)MrRimmer Wrote: ... old [AV-Receivers], pre-HDMI but sounds fantastic ...
That's another great use case. Use old AMPs with new signal processing algorithms Wink

(2015-05-21, 19:58)guyonphone Wrote: ...
2. If you have an expensive receiver you will probably want to continue using your receiver for this feature.
3. If "like OP" you are interested in getting rid of your receiver this feature helps bridge the gap between the features you would loose by dumping your receiver. (I still believe that current mid range and up receivers provide a wealth of benefits to sound processing that Kodi is unlikely to fully replace) but everyone has a different use case, and this feature will certainly help a lot of people.
2. Or you can test which one sounds better Blush
3. I think no because AudioDSP is very young and we don't have enough Add-ons. I don't know if you are fimiliar with C/C++ code, but if you can read it. You should look into it and I think you would be impressed and how powerfull ActiveAE is. I think not all commercial AV-Receivers have this powerful engine. With AudioDSP the gap would be closed at one day.


(2015-05-21, 19:58)guyonphone Wrote: 1. Will you be able to enter the distance of each speaker from the viewer, into kodi?
2. Will this feature be able to detect if a speakers polarity is incorrect?
1. & 2. Hard to say at the moment and I don't know which features would be integrated during GSoC. But the first version can measure your room impulse response and can correct it. The two features which you mentioned can be integrated after GSoC.
Latest news about AudioDSP and my libraries is available on Twitter.

Developers can follow me on Github.
Reply
#24
Hello Wisler,

Thank you for your thoughtful reply. I have been fully swayed by the benefits of adding this feature. I am excited about it being added, and would like to thank you for taking the time and effort to create it. I was a bit concerned that my comments may be considered to be negative in their nature, however I viewed them as legitimate questions which were important to discuss, and I think it has spawned some good discussion. You guys have answered all my questions, and it is clear you have really thought this through. I unfortunately cannot read C\C++ (not a programmer), but i'm quite excited to see what you come up with, and I would be very happy to help test the feature when it's ready for testing!

Thanks again!
Reply
#25
@all
AV-receivers can do anything - in five years maybe even washing your underwear with ultrasonics while watching TV Wink - but nothing well. Sad
Whenever a new HDMI-standard is published, you have to buy new equipment (players, AV-receiver, TV/projector). I'm interested in using a semi-professional USB audio interface with PA-amplifiers and BiAmped RiPol-speakers. For a 5.2-setup I need 12 channels (5 BiAmp-Speakers + 2 Subwoofers). A DSP has to use a high-pass and a low-pass per BiAmp-Speaker and a parametric equalizer to apply linearity calibration data from a measurement microphone. The DSP also needs to correct the group phase of the speakers.

@Wisler
My current plan is to use the Room EQ Wizard with a microphone to measure linearity, phase and group phase of speakers, use System-Wide-DSP-Guide to convert the results for the LADSP-plugin of LADSPA which I want to use in ALSA. I wanted to use the Behringer U-Phoria UMC1820 USB audio interface, but unfortunately it seems to be vaporware.

Maybe you can contact the REW developers for help on your Kodi-DSP-support?
Reply
#26
I hope Im alive to see native microphone input support in Kodi via this and it works with Kodi Karaoke Smile

Great job Smile
Reply
#27
depends on your age Smile
Reply
#28
Just last week Roger Reed 5 months and a few days older than myself bought the farm you may have read about it
Reply
#29
(2015-07-22, 12:03)uNiversal Wrote: I hope Im alive to see native microphone input support in Kodi via this and it works with Kodi Karaoke Smile

Great job Smile

Thanks, but I only use the input to measure room impulses. So Karaoke is with this implementation not possible. We decided to implement the microphone input only in adsp.xconvolver.

Maybe one day we will see native microphone support.
Latest news about AudioDSP and my libraries is available on Twitter.

Developers can follow me on Github.
Reply
#30
SO I will never see it... OK
Reply

Logout Mark Read Team Forum Stats Members Help
ACCEPTED: Creating cross-platform room correction for Kodi3