ARTA

[DcibeL]

Having worked with ARTA for some time, there's only a few things that I wish would be changed. I'm not sure if the developer checks this forum, maybe these thoughts would be better suited to a direct email?

LIMP - when completing lead compensation, I short my leads and press the RLC button, which provides values in mOhm and uH, which I then have to move some decimal places to enter this in the lead compensation as Ohms and nH, so it would be great if the units for lead compensation were changed to mOhm and uH.

LIMP - when calculating lossy inductor model, it would be great to see a plot of the modeled impedance to verify that it has a good match to the measured impedance.

LIMP - I fail to understand the need to re-calibrate if the FFT size is changed. The channel difference remains unchanged, but sure enough the results will be out of whack without a re-calibration.

STEPS - The distortion as a % screen is great as it will normalize the response result, so the value is simply % of the fundamental. However, I would prefer to view this data as dB values as well, so -40,-60,-80dB instead of 1%, 0.1%, 0.01%. My thought would be to change the name of this screen to "Normalized Distortion" which a checkbox to change units from dB to %.

ARTA - A plot of the window shape would be great to visualize the different windowing types.

ARTA - I fail to understand why t=0 isn't at bin 300.

ARTA - I would very much prefer if the arrow keys moved the cursor by one bin at a time, rather than simply scrolling the display. For fine cursor placement I always end up zooming in which wouldn't be needed with a simple fine cursor control like using arrow keys.

I think that's all I have for now.

[iMat]

Hi DcibL
I've checked forum but sometimes I expect that other members give their own insights. It helps me to better setup ARTA procedures to user needs.

Let start with your requests and views,

LIMP is offering you to check TSP model - what you need to do after the tsp and inductance estimation is done is to close dialog box and click F3 key. You will get target curve (red one) calculated from estimated parameters..

LIMP calibration is in some cases very critical, especially when you measure capacitance. Even a small error can change angle from 90 to 90.xx value, and the result will be interpreted as inductive. I am trying to fix this for next version.

Plotting of window shape in time domain is not of great help. The fact is that exponential smoothing slightly moves poles and zeros of transfer function, but time function does not reveal that shift. It is better to compare frequency response with different windows.

Reference time zero in ARTA is arbitrary set to sample position 300. There are few reasons for this, the main one is that we works with digitally filtered AD/DA systems that exhibit pre-ringing (which virtually appears before time zero).

Movement of cursor (or marker if exists) can be obtained by clicking keys Shift+cursor.

Ivo

[DcibeL]

Hi Ivo, thanks a lot for responding, this is a great help!

Originally Posted by iMat

LIMP is offering you to check TSP model - what you need to do after the tsp and inductance estimation is done is to close dialog box and click F3 key. You will get target curve (red one) calculated from estimated parameters..

Perfect! I didn't find this feature documented in the manual, perhaps it could be added to the "view" dropdown menu for easy access.

Originally Posted by iMat

LIMP calibration is in some cases very critical, especially when you measure capacitance. Even a small error can change angle from 90 to 90.xx value, and the result will be interpreted as inductive. I am trying to fix this for next version.

It's not a big deal, more of a curiosity as I thought the only meaningful result of the calibration is the channel difference.

Originally Posted by iMat

Reference time zero in ARTA is arbitrary set to sample position 300. There are few reasons for this, the main one is that we works with digitally filtered AD/DA systems that exhibit pre-ringing (which virtually appears before time zero).

The confusion comes to a new user who has set up my mic and expecting time of flight of something like 3ms, but when I place my cursor near the start of the impulse it shows 6ms. The reference impulse is not shown, and I eventually found from another user that it is located at sample 300, and pressing the home key will locate the cursor at sample 300, and I have to measure from sample 300 to the measured impulse to determine time of flight. This was not clear to me when reading through the help file trying to learn the software.

I still think setting t=0 at 300 sample is a good idea, or at least a feature to set t=0 to the cursor location so that it can be arbitrarily set by the user for whatever special case scenario exists.

Originally Posted by iMat

Movement of cursor (or marker if exists) can be obtained by clicking keys Shift+cursor.

Perfect, thank you!

[JMB]

Hi Ivo,
Just wondering if it might be possible to add the following:
1.ability to show a legend on the graphs as an option for all of the software
2. the ability to show different colors for mulitple overlay/target curves in LIMP to more closely evaluate the different T/S models
3. Larger FFT option in LIMP?
Thanks,
Jay

Sorry to repeat this but I think it got missed a few pages back.

[DcibeL]

Ivo,
I was working with ARTA again today and have a couple simple requests:

- Load FRD or ZMA directly as overlay. Currently in LIMP to load an overlay it requires a saved .lim file, it would be great to be able to load .zma (space separated) or .csv (comma separated) data directly.

While on the LIMP topic, I am curious what information is included in a .lim file that isn't include in a .zma? The .zma is plain text readable by most anything, and the .lim file appears to be binary format that is only readable by LIMP so I am curious what benefit there is to saving the .lim file over the plain text.

Similarly in ARTA - load overlay requires .pir, .sfo or .ovf file, again it would be great to be able to load .frd or .csv plain text files directly as an overlay.

[ente]

Hi Dcibel,

what is the reason for your question, what do you want to do with the additional overlays?
If you want to compare measurements, you can also use "load target curve". There you can load zma and lim files.

Regards
Heinrich