So you made a totally new crossover? Or made a crossover in a new slot?
And didn’t change any wiring or anything else?
weird.
Any ideas why/how that happened?
After I sent miniDSP help desk/tech support a screenshot of inverted impulse responses from digi input, but not inverted from the analog, tech support conceded they were unable to help further, and forwarded it onto their engineers to investigate.
I’m yet to get a formal response from miniDSP engineers.
Ugh, it seems pretty much all DSP systems are plagued with either feature or usability issues. It is definitely the future of our hobby - what with embedded DSP on amplifier modules and widespread availability of the knowledge to implement things. However, if we are restricted to basic crossover functions and weird issues like this continue to be a problem my plans of building a measurement/simulation rig with a DSP system are just not going to happen.
FWIW, old school LspCAD has a simulator in it that allows you to use both channels of an amplifier to do measurements with. It outputs the crossover for each driver on individual channels. Very useful feature.
I changed from asymmetric XO slopes to much more evenly matched or symmetric slopes and now I cannot tell when the inversion buttons are pushed upon any of the speakers 3 way XO's, so the phase issue is one of the users creation and the "fix" is in-fact the inversion button.
I'll edit this post to include the specific changes made later tonight when I am in front of the DSP unit . . .
Oops - Sorry - got distracted with shop tooling upgrades - I'll post the XO settings soon . . .
OK - I just opened the DSP panel and it wanted to sync the settings so I did (even though I had saved the settings last night before closing the app) and now the inversion problem is back in spades.
It effects the mids and tweeters where it was only the mids in the past.
Gotta invert the left Mid and left tweeter to get the normal balance between the channels otherwise it sounds all funky with none of the mix centered.
Maybe I should email STING and other artists and ask them how their stereo master mix was done in the studio during production to see if this DSP is working correctly? (Hey, Sting; Did you guys put the bass drum and vocals in the center or did you push them out of phase so they sound like they are coming from 2 different mic sources about 25 feet apart? - Oh no mate, we were on two different continents and the latency was a right real bitch but we streamed that stuff right to the master track just like that - what is your problem with my CD, bloke?).
Nothing is repeatable/stable - I give up.
It sounds good now but it just isn't right at all . . .
Looks like I goofed-up some settings concerning the delays on the channels and got the right channels out of sync with the left . . . still experimenting after correcting it . . . So, for now - I am a moron. Disregard my July 19th post, for now.
I could certainly suggest a few GUI enhancements/improvements to increase usability of the MiniDSP interface at this point.
On the "Output" configuration GUI/Page:
Place a red vertical bar between channel 4 & 5 to aid in visual separation of the left and right channel configs
Any channel header labeled "Unused" would have the entire channel-vertical settings-bar grayed-out and all DSP functions Bypassed for that channel
Extend the "Channel Link" function to the "Delay" settings as an option (although I understand the flexibility offered by the default GUI in this regard - it's just not flexibility that 99% of us need).
Have the FILE NAME appear in the upper banner for the PRESET LOADED.
Well since that time, I did check my work. Like I say, I thought I was going around the bend.
Usually this thing is user error. It's common.
To continue, I took a break from the miniDSP 2x4HD. I loaded the crossover and biquads settings etc by hand into nanoDigi, and using external D/A converters, everything is hunky dory for my 3 way.
I'll be developing another 2 way in the coming weeks, and will try the 2x4HD again. We'll see if I can replicate the problem. I hope it will be resolved, and turn out to be user error. Operator dependant problem.
Comments
Changed the XO's tonight and the phase issue disappeared - no inversion required and the sound is really awesome.
So you made a totally new crossover? Or made a crossover in a new slot?
And didn’t change any wiring or anything else?
weird.
Any ideas why/how that happened?
After I sent miniDSP help desk/tech support a screenshot of inverted impulse responses from digi input, but not inverted from the analog, tech support conceded they were unable to help further, and forwarded it onto their engineers to investigate.
I’m yet to get a formal response from miniDSP engineers.
Ugh, it seems pretty much all DSP systems are plagued with either feature or usability issues. It is definitely the future of our hobby - what with embedded DSP on amplifier modules and widespread availability of the knowledge to implement things. However, if we are restricted to basic crossover functions and weird issues like this continue to be a problem my plans of building a measurement/simulation rig with a DSP system are just not going to happen.
FWIW, old school LspCAD has a simulator in it that allows you to use both channels of an amplifier to do measurements with. It outputs the crossover for each driver on individual channels. Very useful feature.
I changed from asymmetric XO slopes to much more evenly matched or symmetric slopes and now I cannot tell when the inversion buttons are pushed upon any of the speakers 3 way XO's, so the phase issue is one of the users creation and the "fix" is in-fact the inversion button.
I'll edit this post to include the specific changes made later tonight when I am in front of the DSP unit . . .
Oops - Sorry - got distracted with shop tooling upgrades - I'll post the XO settings soon . . .
OK - I just opened the DSP panel and it wanted to sync the settings so I did (even though I had saved the settings last night before closing the app) and now the inversion problem is back in spades.
It effects the mids and tweeters where it was only the mids in the past.
Gotta invert the left Mid and left tweeter to get the normal balance between the channels otherwise it sounds all funky with none of the mix centered.
Maybe I should email STING and other artists and ask them how their stereo master mix was done in the studio during production to see if this DSP is working correctly? (Hey, Sting; Did you guys put the bass drum and vocals in the center or did you push them out of phase so they sound like they are coming from 2 different mic sources about 25 feet apart? - Oh no mate, we were on two different continents and the latency was a right real bitch but we streamed that stuff right to the master track just like that - what is your problem with my CD, bloke?).
Nothing is repeatable/stable - I give up.
It sounds good now but it just isn't right at all . . .
(Sigh)
Updated findings in above post . . .
Looks like I goofed-up some settings concerning the delays on the channels and got the right channels out of sync with the left . . . still experimenting after correcting it . . . So, for now - I am a moron. Disregard my July 19th post, for now.
I could certainly suggest a few GUI enhancements/improvements to increase usability of the MiniDSP interface at this point.
On the "Output" configuration GUI/Page:
Place a red vertical bar between channel 4 & 5 to aid in visual separation of the left and right channel configs
Any channel header labeled "Unused" would have the entire channel-vertical settings-bar grayed-out and all DSP functions Bypassed for that channel
Extend the "Channel Link" function to the "Delay" settings as an option (although I understand the flexibility offered by the default GUI in this regard - it's just not flexibility that 99% of us need).
That's all for now . . .
Minor epiphany - compare the source mix to the DSP mix using headphones.
I have been chasing my tail . . .
TKTRAN - check you work.
Well since that time, I did check my work. Like I say, I thought I was going around the bend.
Usually this thing is user error. It's common.
To continue, I took a break from the miniDSP 2x4HD. I loaded the crossover and biquads settings etc by hand into nanoDigi, and using external D/A converters, everything is hunky dory for my 3 way.
I'll be developing another 2 way in the coming weeks, and will try the 2x4HD again. We'll see if I can replicate the problem. I hope it will be resolved, and turn out to be user error. Operator dependant problem.
what DACs are you using for nano digi?
D03k