Camilla SQ

[michaelagiles]

Good Evening! I recently tried to implement an eq solution using Camilla. I have a tri-amped TL system, using 4th order XO filters and had attempted to increase the bass extension, using camilla. the bass was indeed extended and sounded very good but I was very disappointed in the overall sound quality. The bass improvement was definitely NOT worth the degradation.

As suggested I set up the "FLAT" first and then the desired bass correction. Even the "FLAT" setting was unacceptable. After the enthusiasm expressed for Camilla her, I had expected better.

I am using RPI4 w/ 7.3.0, and audiophilediyer's AD1868 hat.

Has anyone else had this experience? Am I missing something something obvious?

Thanks!
MG

[Tim Curtis]

What "audiophilediyer's AD1868 hat" are you using and can you provide a link?

[Miss Sissy Princess]

Is this what you have?  

https://audiophilediyer.com/ad1868-r-2r-dac-pi-hat/

I have a tri-amped TL system, using 4th order XO filters...

It's generally easier to get help if you spell things out like "crossover" and "transmission line speaker" (I think that's what you meant by "TL").  I'm not sure what problems you are trying to solve with tri-amping and fourth order crossover filters.  Nor do I know if the fourth order filters are analog or DSP in nature, implemented at line level or post-amplifier within the speakers.

The bass improvement was definitely NOT worth the degradation.

As suggested I set up the "FLAT" first and then the desired bass correction. Even the "FLAT" setting was unacceptable. After the enthusiasm expressed for Camilla her, I had expected better.

Do you have any measurements showing that there was actual degradation, versus perceived degradation?

[michaelagiles]

Is this what you have?  

https://audiophilediyer.com/ad1868-r-2r-dac-pi-hat/

I have a tri-amped TL system, using 4th order XO filters...

It's generally easier to get help if you spell things out like "crossover" and "transmission line speaker" (I think that's what you meant by "TL").  I'm not sure what problems you are trying to solve with tri-amping and fourth order crossover filters.  Nor do I know if the fourth order filters are analog or DSP in nature, implemented at line level or post-amplifier within the speakers.

The bass improvement was definitely NOT worth the degradation.

As suggested I set up the "FLAT" first and then the desired bass correction. Even the "FLAT" setting was unacceptable. After the enthusiasm expressed for Camilla her, I had expected better.

Do you have any measurements showing that there was actual degradation, versus perceived degradation?

Good morning! Yes - this is the hat I am using.
It is an analogue XO. And my first foray into DSP.
I do not have any way to
 measure distortion etc. Nor am I trying to solve any problems! This is a system I have been building/refining for 20 odd years and - I am very happy with it! Just looking for a little better bass extension and thought this might be the tailor-able solution that would work
I have been using moode for several years and again , I am more than happy with it! And the AD1868 hat has been a revelation!
Thanks for your response!

MG

[TheOldPresbyope]

@michaelagiles

This line from one of your posts intrigues me

As suggested I set up the "FLAT" first and then the desired bass correction. Even the "FLAT" setting was unacceptable. After the enthusiasm expressed for Camilla her, I had expected better.

I have little experience with CamillaDSP but I don't recall noticing degradation in sound quality when I tried the FLAT setting. Is it possible that you are experiencing clipping because of excessive gain somewhere in the audio chain? 

Unfortunately I'm just musing in my hotel room with no moOde player within sight (or hearing) so I can't experiment with settings ATM.

Regards,
Kent

[the_bertrum]

Unfortunately I'm just musing in my hotel room with no moOde player within sight (or hearing) so I can't experiment with settings ATM.

Ouch, what purgatory!

[TheOldPresbyope]

Unfortunately I'm just musing in my hotel room with no moOde player within sight (or hearing) so I can't experiment with settings ATM.

Ouch, what purgatory!

Then there's the atrocious unprotected WiFi offering. Thank the gods for VPNs, but with or without one the throughput is abysmal compared to what I enjoyed at home.

I can do better using my phone as an AP but that eats up my data plan.

Still, I should have a stand-alone moOde player/headphones rig running with my ripped CD collection RSN. If I can find the SSD that is

Regards,
Kent

[Miss Sissy Princess]

Thanks for the additional information.

I do not mean any disrespect, but we are all prone to hearing differences where none exist -- especially us audiophiles (it's no badge of shame to members of the general public to say 'yeah, those amps sound the same').  It's established science at this point, with numerous studies showing that to be the case.  There are many reasons for it, including evolutionary biology, but that's wandering way off of the primary topic of your thread.

What I recommend is first establishing whether you can hear a difference in a blind test with Camilla in or out of the signal path with Camilla set flat.  Do you have a friend or family member who could help you conduct an impromptu blind listening test?

Assuming you show an ability to detect the difference in a blind test, then some more exploration of Camilla settings would be in order.  It's quite complex and I am sure that others here are more experienced and proficient with it than I am.

If you get everything set properly to minimize or eliminate any degradation (with the help of said experts), and you can still hear said degradation in a blind test, then maybe that's as far as you can go with Camilla.

If you get Camilla to a transparent state, then you can move on to experimenting with EQ and DSP to improve bass extension.

I understand the power that Camilla has, but I don't have the T&M gear to really fully utilize it (I'd probably make things worse rather than better).  Did you try to accomplish what you wanted using the simple parametric and graphic EQ rather than Camilla?  

As I am sure that you know, at some point, you run into limitations on bass extension imposed by speaker drivers, cabinet design, amplifiers, and the listening room (none of which I know much about in your system).  That's why I own speakers large enough to be buried in (a thought that has probably occurred to my wife on multiple occasions), each of which includes a 15" subwoofer, 15" passive radiator, and a 10" woofer.  As the car nuts say, there's no replacement for displacement.

[michaelagiles]

Thanks for the additional information.

I do not mean any disrespect, but we are all prone to hearing differences where none exist -- especially us audiophiles (it's no badge of shame to members of the general public to say 'yeah, those amps sound the same').  It's established science at this point, with numerous studies showing that to be the case.  There are many reasons for it, including evolutionary biology, but that's wandering way off of the primary topic of your thread.

What I recommend is first establishing whether you can hear a difference in a blind test with Camilla in or out of the signal path with Camilla set flat.  Do you have a friend or family member who could help you conduct an impromptu blind listening test?

Assuming you show an ability to detect the difference in a blind test, then some more exploration of Camilla settings would be in order.  It's quite complex and I am sure that others here are more experienced and proficient with it than I am.

If you get everything set properly to minimize or eliminate any degradation (with the help of said experts), and you can still hear said degradation in a blind test, then maybe that's as far as you can go with Camilla.

If you get Camilla to a transparent state, then you can move on to experimenting with EQ and DSP to improve bass extension.

I understand the power that Camilla has, but I don't have the T&M gear to really fully utilize it (I'd probably make things worse rather than better).  Did you try to accomplish what you wanted using the simple parametric and graphic EQ rather than Camilla?  

As I am sure that you know, at some point, you run into limitations on bass extension imposed by speaker drivers, cabinet design, amplifiers, and the listening room (none of which I know much about in your system).  That's why I own speakers large enough to be buried in (a thought that has probably occurred to my wife on multiple occasions), each of which includes a 15" subwoofer, 15" passive radiator, and a 10" woofer.  As the car nuts say, there's no replacement for displacement.

Thanks all for your input!
Firstly - there is a facility for Camilla to count clipping occurrences and I checked it after about 1/2 hr of play - 5 incidents - so some but certainly not enough to account for what I heard.

re:perceived "improvements" vs only "changes" - yes - I have done this to myself several times (MANY times!) and in fact did so this time. I was enjoying the new bass sound  and it was only on getting accustomed to it and listening to the rest of the music that I heard things I didn't like. It was then that I went back to the FLAT setting to try to find the source of the perceived degradation. And discovered that the process of diverting to another chain of process seemed to be the source. Whether there was any other filter processing involved or not. In my view, testing/measurement is a tool to help us discover a more enjoyable experience, or why we are NOT enjoying the experience as we feel we should. And sometimes this does point us to quirks/anomalies in our own perceptions and preconceptions.

Yes - I have tried both the Parametric and Graphic EQ and found them also not to my taste in this system. In fact I use the para eq in my less revealing bedroom system and find it helpful there.

I appreciate your scientific rigour (Canadian spelling eh!) but really, as Tim says regularly - "Enjoy the music!". I expect that there might be a setting - or series of settings - that will produce a more transparent result and when I have a chance will try to dig a little deeper into this. But for now, I am enjoying the music!
Thanks Again!
MG

Kent - thankfully I am no longer in a position to have to suffer this fate. Try to have fun! A portable system might be the ticket!

[michaelagiles]

i kinda wanna revive this thread

things i noticed so far:

1. process priority should be increased like the tweaks for mpd, it helped with the "nasty" overall sound (its not that bad but certainly noticable) i didnt found a way yet to make this permanently, unfortualy i dont have much linux skills

also increasing chunksize helped with sq, tho i found that as soon you increase priority you get pretty much the same effect

2. making a dither filter in the end of the chain helped

3. i noticed that inverting the phase on the start of the chain isnt a good idea, its better todo it at the end (right before dither), i have to revert the phase for both channels because apparaently early topping d10 models have inverted phase

4. rate_adjust should be enabled (i set it to 1second instead of 10)

5. watch out with filters, they have an effect on phase, try to keep it as little as possible (you can check with the graphs)

i think those are the things i discovered so far that did the most

Thanks Ghoostknight! I will try these when I get a chance. Bigger fish to fry at the moment! A lightning strike took out a power supply, a power amp and fried 2 mid/tweeters! Youch!
MG