PS: To compensate for inter channel differences (phase and amplitude), proceed as follows:
Right after the first characterization run (thus before you have saved the result), connect the notch filter to the other channel (and don't forget to adjust the channel modes as well!) and do a second run. Now the average of amplitude and phase is automatically calculated and you can save the result as outlined in the previous comment.
Right after the first characterization run (thus before you have saved the result), connect the notch filter to the other channel (and don't forget to adjust the channel modes as well!) and do a second run. Now the average of amplitude and phase is automatically calculated and you can save the result as outlined in the previous comment.
Last edited:
Hi Braca,
As for notch filter characterization, I'm not sure if Diana.V1.54.9-PW.zip is correct and up to data. Normally, during characterization, one should get waveforms as shown in fig.1. But the latest version shows something different, see fig. 2, which is wrong. Clearly I was fiddling with some settings and forget to put it back. Sorry for the inconvenience. ASSAP I will put the correct version on my website.
Cheers, E.
As for notch filter characterization, I'm not sure if Diana.V1.54.9-PW.zip is correct and up to data. Normally, during characterization, one should get waveforms as shown in fig.1. But the latest version shows something different, see fig. 2, which is wrong. Clearly I was fiddling with some settings and forget to put it back. Sorry for the inconvenience. ASSAP I will put the correct version on my website.
Cheers, E.
Attachments
Notch filter stuff
To all,
I've uploaded version V1.49.10. Now the waveforms are correct again when characterizing a notch filter.
The waveform at the ref. channel should be a sine plus a superimposed sawtooth with amplitude of 0.2 times the amplitude of the fundamental.
NB: Make sure that the fundamental frequency exactly coincidences with the notch frequency. Then the amplitude of the test signal should be at a minimum.
As for compensation of inter channel differences, please forget what I've written in post #761. It doesn't work that way (any longer?). Maybe it was old feature and later on deleted, or the other way around: a new feature, though not yet implemented. Sorry, I can't remember which of the two it was. Anyhow, this feature isn't that important.
Cheers, E.
To all,
I've uploaded version V1.49.10. Now the waveforms are correct again when characterizing a notch filter.
The waveform at the ref. channel should be a sine plus a superimposed sawtooth with amplitude of 0.2 times the amplitude of the fundamental.
NB: Make sure that the fundamental frequency exactly coincidences with the notch frequency. Then the amplitude of the test signal should be at a minimum.
As for compensation of inter channel differences, please forget what I've written in post #761. It doesn't work that way (any longer?). Maybe it was old feature and later on deleted, or the other way around: a new feature, though not yet implemented. Sorry, I can't remember which of the two it was. Anyhow, this feature isn't that important.
Cheers, E.
Notch filter characterization, extended procedure.
Anyhow, this is the correct procedure:
1. Connect the sound card in loopback mode. One channel directly (the reference channel) and the other channel via the notch filter (the test channel) .
2. In the Distortion menu select which channel is used as reference.
3. Select in the Option menu 'Characterize notch filter'.
4. Now run the first measurement for say 100 records by pressing the Distortion button.
5. When finished, don't save the result yet. Instead, exchange the channels, i.e. connect the notch filter to the other channel
6. In the Distortion menu set the reference channel accordingly.
7. Start the second run by pressing the Continue button.
8. When finished, go to File->Save Notch Filter and save the result as usual
The difference with the simple procedure as outlined in post #760 is that inter channels imbalances, in terms of gain and phase, are averaged out and errors are hopefully reduced to zero.
NB: No interpolation has been applied to compensate for mismatches (if any) between harmonics as saved in the notch file and harmonics of subsequent THD measurements. So ensure to use the correct and same fundamental frequency, which matches the notch filter itself, the characterization file and THD measurements.
Cheers, E.
Well, it does work, but I simply forgot how to do it (by using the continue button).
Anyhow, this is the correct procedure:
1. Connect the sound card in loopback mode. One channel directly (the reference channel) and the other channel via the notch filter (the test channel) .
2. In the Distortion menu select which channel is used as reference.
3. Select in the Option menu 'Characterize notch filter'.
4. Now run the first measurement for say 100 records by pressing the Distortion button.
5. When finished, don't save the result yet. Instead, exchange the channels, i.e. connect the notch filter to the other channel
6. In the Distortion menu set the reference channel accordingly.
7. Start the second run by pressing the Continue button.
8. When finished, go to File->Save Notch Filter and save the result as usual
The difference with the simple procedure as outlined in post #760 is that inter channels imbalances, in terms of gain and phase, are averaged out and errors are hopefully reduced to zero.
NB: No interpolation has been applied to compensate for mismatches (if any) between harmonics as saved in the notch file and harmonics of subsequent THD measurements. So ensure to use the correct and same fundamental frequency, which matches the notch filter itself, the characterization file and THD measurements.
Cheers, E.
Last edited:
Notch filter implementation
And her some stuff about the notch filter itself: Schematic, construction, simulated response and measured response.
(Note the ugly XLR connectors of the Lynx L22
)
Cheers, E.
And her some stuff about the notch filter itself: Schematic, construction, simulated response and measured response.
(Note the ugly XLR connectors of the Lynx L22
)Cheers, E.
Attachments
Hi David,
I agree, totally! BTW, don't worry, I'm using them too (from Rean, that are the best, at least in my opinion).
Cheers, E.
Attachments
REAN is owned by Neutrik and does maintain very high standards. Its more focused on the higher volume stuff.
I find the connector morass in audio really frustrating. I have more cables and adapters than I can manage and still end up making new ones all the time. I have instruments with XLR's BNC's and banana jacks for the same thing. No easy solution. Mobile stuff has really pushed this issue. It also makes finding a decent ground reference really hard.
for general purpose I go thru dual banana jax, xlr to dual banana, bnc to dual banana adapters, clip leads from dual banana. Just make sure to leave the pin 1 connector from the xlr long enough to plug in anywhere for adaptive grounding.
I love the idea of mini xlr, but nobody uses it in the pro audio world, hence....
+1 for Rean, Neutrik bought them years ago (15-20?) and they market (slightly) less expensive connectors thru them. They make my go to RCA cable end connector of choice. (which is never if I had a choice!)
Cheers
Alan
I love the idea of mini xlr, but nobody uses it in the pro audio world, hence....
+1 for Rean, Neutrik bought them years ago (15-20?) and they market (slightly) less expensive connectors thru them. They make my go to RCA cable end connector of choice. (which is never if I had a choice!)
Cheers
Alan
for a T&M environment I can agree that XLR's are bulkier than needed. I am not so convinced that in the pro stage environment they are not the correct connector to keep using. In that regard, I don't think you'll see any change over to the mini. Some of the wireless mics and lavaliere mics have gone to the mini as that is a great choice in that situation, on the body to be hidden. And way better a secure connection than most of the alternatives.
and the stage business doesn't like to change from a proven technology, too many things already go wrong on a regular basis.
Cheers
Alan
and the stage business doesn't like to change from a proven technology, too many things already go wrong on a regular basis.
Cheers
Alan
Applying a notch filter
Hi Braca,
Please tell me about your findings. In case you prefer to create the notch filter file by other means, it might be helpful to know how this file is read by DiAna. I hope you understand the listing below, written in the C language.
Cheers, E
Hi Braca,
Please tell me about your findings. In case you prefer to create the notch filter file by other means, it might be helpful to know how this file is read by DiAna. I hope you understand the listing below, written in the C language.
Cheers, E
