I could not find the AP software. But doesn't it run just on AP hardware?
Anyway, thanks to forum member JLO, here is a link to software that can do a similar trick.
http://www.audiosignal.co.uk/Resources/adddistortion.zip
Link is the the add distorsion files, zipped. It's from here:
http://www.audiosignal.co.uk/freeware.html
Anyway, thanks to forum member JLO, here is a link to software that can do a similar trick.
http://www.audiosignal.co.uk/Resources/adddistortion.zip
Link is the the add distorsion files, zipped. It's from here:
http://www.audiosignal.co.uk/freeware.html
(for Windows only) You can generate arbitrary multitone sets (any audio frequencies, each adjustable in size and phase) using PRAXIS software -- it can be downloaded free and used in a "free" (ie as in free beer) mode that supports this.
http://libinst.com/praxis_downloads.htm . The sets can be written out as WAV files.
Here's how to generate them (by the way, it's lots easier than it appears from looking at all these instructions):
* After installing, run PRAXIS and close the "what's new" and "Script launcher" forms that appear at start. You might have to confirm what soundcards and Windows Mixers are in the machine, just do what it says.
* To select the sample rate you want to use, find the "Levels Form" (you may have to get there via the "View" menu of the "MAIN FORM"). In the Levels Form, use its "Setup->D/A&A/D" menu to choose a soundcard that can support the rate and configure it for that rate. You can typically choose all the usual rates and though the soundcard might not really play the full bandwidth, the WAV files will contain them all.
* Locate the (MAIN FORM) on the screen, and go to the area labeled "stimulus". Click the drop-down box there and select the stimulus type "Fixed Tones (1ch asynch)". It appears greyed-out, but don't worry about that, it's only because the program can't directly play the tones out in Free/Demo mode, but you can still use it to generate wave files
* Click the button in the Stimulus area labeled "Config" (or "hide Config") so that a little form labeled "(Stim)" appears. In that form, click the button labeled "Edit Tones (Ch1)". A "Waveform Component Editor" will appear.
* The big white area in the middle of the Waveform Component Editor has a grid. Click in the area of the grid you want to modify: Type should be "Sine", choose frequency, amplitude (use 1 for each tone in the types of multitone sets that have been discussed here), and phase -- ignore "duty cycle" for sine tones. The actual modification should be entered in the up/down control at the top.
* to add another tone, click in the last row (empty until clicked) and modify one of its cells. To remove a tone, click somewhere in its row and then click the button "Remove" just below and to the left of the grid.
* alternative: TO GENERATE LOG OR LINEARLY SPACES SETS EASILY: Remove any waveforms currently defined, and click "Auto-Generate Wave sets" and a new form appears. Set the Start Frequency and something higher than the highest frequency you expect, choose whether you want to the frequencies to step linearly or be at multiples (log) of the prior frequency. Then choose the step size or multiplier where shown. You can do similar things with amplitude and phase, but will probably leave them set with start and limit set the same. Lastly, choose the number of sinewaves ("Max # of waveforms to add") and then "Auto-Generate Waveforms". Look through the grid and make sure it's what you wanted (else remove the waveforms and go back and Auto-Generate again).
* When you have all your tones defined, you have to make it so that their amplitudes all sum up to no more than 1.0. To do that, click "Auto Peak Set". and then (probably) "All Same".
* Now, to make the WAV file, click the "Make WAV file" button. Define how many samples you want to generate (how long the tones should play), what resolution, and whether a stereo or mono file, then "Ok". A form appears where you can specify where you want to save the file and what to call it. Do it and click "save".
* You have your file, ALL DONE! Exit the Waveform Editor by clicking on its "Done" button. Find the WAV file and play it the normal ways or burn it to CD with a Nero or Sonic if you have them.
___
some other things you can do:
*IF you want to save the recipe (what's in the grid) for making this waveform set, you can use the "Save to File" button (and "Load to File" to use it later).
* You can also generate stereo files with a different multitone set in each channel, for that start with Stimulus type ""Fixed Tones (2ch asynch)".
* If you want, you can use PRAXIS to look at the waveforms you generated (in time domain), just go to the form that looks like a graph ("Primary Plot") and use its "File->Open" menu to look for WAV files and then load the one you saved. I'll let you figure out how to adjust the display formatting.
* When you have a WAV file loaded into the Primary Plot, you can listen to what it contains -- use it's "File->Listen" menu.
* you can also filter and edit the file in the Primary Plot (via the Main Form's "PostProcess", apply FFTS, etc., if you want.
http://libinst.com/praxis_downloads.htm . The sets can be written out as WAV files.
Here's how to generate them (by the way, it's lots easier than it appears from looking at all these instructions):
* After installing, run PRAXIS and close the "what's new" and "Script launcher" forms that appear at start. You might have to confirm what soundcards and Windows Mixers are in the machine, just do what it says.
* To select the sample rate you want to use, find the "Levels Form" (you may have to get there via the "View" menu of the "MAIN FORM"). In the Levels Form, use its "Setup->D/A&A/D" menu to choose a soundcard that can support the rate and configure it for that rate. You can typically choose all the usual rates and though the soundcard might not really play the full bandwidth, the WAV files will contain them all.
* Locate the (MAIN FORM) on the screen, and go to the area labeled "stimulus". Click the drop-down box there and select the stimulus type "Fixed Tones (1ch asynch)". It appears greyed-out, but don't worry about that, it's only because the program can't directly play the tones out in Free/Demo mode, but you can still use it to generate wave files
* Click the button in the Stimulus area labeled "Config" (or "hide Config") so that a little form labeled "(Stim)" appears. In that form, click the button labeled "Edit Tones (Ch1)". A "Waveform Component Editor" will appear.
* The big white area in the middle of the Waveform Component Editor has a grid. Click in the area of the grid you want to modify: Type should be "Sine", choose frequency, amplitude (use 1 for each tone in the types of multitone sets that have been discussed here), and phase -- ignore "duty cycle" for sine tones. The actual modification should be entered in the up/down control at the top.
* to add another tone, click in the last row (empty until clicked) and modify one of its cells. To remove a tone, click somewhere in its row and then click the button "Remove" just below and to the left of the grid.
* alternative: TO GENERATE LOG OR LINEARLY SPACES SETS EASILY: Remove any waveforms currently defined, and click "Auto-Generate Wave sets" and a new form appears. Set the Start Frequency and something higher than the highest frequency you expect, choose whether you want to the frequencies to step linearly or be at multiples (log) of the prior frequency. Then choose the step size or multiplier where shown. You can do similar things with amplitude and phase, but will probably leave them set with start and limit set the same. Lastly, choose the number of sinewaves ("Max # of waveforms to add") and then "Auto-Generate Waveforms". Look through the grid and make sure it's what you wanted (else remove the waveforms and go back and Auto-Generate again).
* When you have all your tones defined, you have to make it so that their amplitudes all sum up to no more than 1.0. To do that, click "Auto Peak Set". and then (probably) "All Same".
* Now, to make the WAV file, click the "Make WAV file" button. Define how many samples you want to generate (how long the tones should play), what resolution, and whether a stereo or mono file, then "Ok". A form appears where you can specify where you want to save the file and what to call it. Do it and click "save".
* You have your file, ALL DONE! Exit the Waveform Editor by clicking on its "Done" button. Find the WAV file and play it the normal ways or burn it to CD with a Nero or Sonic if you have them.
___
some other things you can do:
*IF you want to save the recipe (what's in the grid) for making this waveform set, you can use the "Save to File" button (and "Load to File" to use it later).
* You can also generate stereo files with a different multitone set in each channel, for that start with Stimulus type ""Fixed Tones (2ch asynch)".
* If you want, you can use PRAXIS to look at the waveforms you generated (in time domain), just go to the form that looks like a graph ("Primary Plot") and use its "File->Open" menu to look for WAV files and then load the one you saved. I'll let you figure out how to adjust the display formatting.
* When you have a WAV file loaded into the Primary Plot, you can listen to what it contains -- use it's "File->Listen" menu.
* you can also filter and edit the file in the Primary Plot (via the Main Form's "PostProcess", apply FFTS, etc., if you want.
I should have known ISO frequencies were just the standard 1/3 octave filter centers. AP adjusts them slightly to hit FFT bins which is what you will want to do unless you are using a continuous time analog spectrum analyser. The sidelobe leakage will contaminate the results otherwise. I wonder how good soundcards do in generating this?
Here's a link to John Risch's paper on frequency picking.
http://www.geocities.com/jonrisch/PhiSpectral1.htm
I plan on playing with this over the next few weeks, maybe try an inverse RIAA one with high crest factor and run it through a couple of preamps.
http://www.geocities.com/jonrisch/PhiSpectral1.htm
I plan on playing with this over the next few weeks, maybe try an inverse RIAA one with high crest factor and run it through a couple of preamps.
Decent ones, at least, do pretty well -- see the screenshot at http://libinst.com/PRAXIS.htm, which is from a nothing special USB soundcard putting out out two sets of 20 log-spaced tones each and recording them back in. Those aren't on FFT bins, though, but use a good windowing function (btw, the auto-waveform set generator has an option to put the tones onto the closest FFT-bin frequencies, which will let you work without windowing if done right).
(btw, the data in the shown plot wasn't made with data from the soundcard shown on the setup)
Certainly looks good for speakers etc. Depending on how you define DNR looks like 90-100dB. Some electronics is going to invariably be better than the op-amps in a sound card anyway.
You can use HOLMImpulse's IFFT to generate (close to perfectly) exact FFT-bin freqs in arbitrary mag/phase relationships. Only that it's not so easy to get the exact crest factor as wanted (in case crest factor is a concern).
- Klaus
- Klaus
Thanks for link, Scott. Good reading.
But with the interest in a "better" test signal, such as multitone, comes the question: "How do we interpret the results?" This isn't going to be easy.
Most of these multitone signals seem to be seeking a result that is closer to our subjective impressions than the old, simpler tests. But how do we know what the measurement results mean?
For some help we might look back to the work of Matti Otala in the 70's and Sakai and others in the 60s for a good starting point. What do different harmonic distortion spectra sound like?
If complex harmonic difference can be measured and seen, it might be a lot easier to know what to listen for. A visual analog might be the children's comic with the “same” image on 2 panels. Except that there are 6 differences from one panel to the other. Can you spot them? As simple as they are, they can sometimes be devilishly hard to spot. But once you do, they pop right out. (the smaller the image is printed, the harder they are to find). Comparing color proofs to a print run when I was in the printing biz was similar. Once the color difference is pointed out, it becomes obvious. Before that, not so...
But with the interest in a "better" test signal, such as multitone, comes the question: "How do we interpret the results?" This isn't going to be easy.
Most of these multitone signals seem to be seeking a result that is closer to our subjective impressions than the old, simpler tests. But how do we know what the measurement results mean?
For some help we might look back to the work of Matti Otala in the 70's and Sakai and others in the 60s for a good starting point. What do different harmonic distortion spectra sound like?
If complex harmonic difference can be measured and seen, it might be a lot easier to know what to listen for. A visual analog might be the children's comic with the “same” image on 2 panels. Except that there are 6 differences from one panel to the other. Can you spot them? As simple as they are, they can sometimes be devilishly hard to spot. But once you do, they pop right out. (the smaller the image is printed, the harder they are to find). Comparing color proofs to a print run when I was in the printing biz was similar. Once the color difference is pointed out, it becomes obvious. Before that, not so...
Re: not really
Thats what I meant, not knowing, pure DBT
To make change in only one channel is very effective
auplater said:
Thats what I meant, not knowing, pure DBT
To make change in only one channel is very effective
Re: Excuse Me?
2. No. And true. But we do know for sure that cable LRC can affect a system's sound, if the sound changes with LRC held constant (your proposed test) then you are correct there must be other factors.
3. Point taken but I am only discussing cables and would still exclude the components you mentioned especially loudspeakers.
4. Semantics? When I say "...reputation based on what they sound like" I mean the sound has particular qualities that can be described using words that others can understand, eg "this cable has a narrow soundstage". If others agree it will get a reputation for this particular quality. If it has different qualities in another system then you can't say that quality belongs to the cable.
5. This is back to point 2. Why make it harder? Using only LRC makes your case easier to prove as you only have to match three easily measured properties.
6. See 4 above. My "If everyone has different opinions...." is merely "If X is true then A or B follows". To clarify my point, If everyone or most agree that a certain cable has a particular and describable sound quality (across various systems) then you can say that quality is a property of the cable itself and you are proved correct. (this would be the body of anecdotal evidence I referred to earlier which if it exists, will make me look stupid)
7. Covered in 2
8. Your last argument about objectivists is good and it works for me, also bodes well for your eventual conversion 😉 But you bring up expensive cables and I have just seen some recently. If I had money to burn I would buy some as they are cool in my book even though I don't believe they would sound any better than welding cables.
9. Logical debate is good. I also started out "believing wires couldn't possibly sound different", then I believed they did, then I realised I was right all along. 😀 Hope you are onto something and can prove it, welding cables are quite boring. Cheers.
1. Rest easythetubeguy1954 (numbers added) said:
2. No. And true. But we do know for sure that cable LRC can affect a system's sound, if the sound changes with LRC held constant (your proposed test) then you are correct there must be other factors.
3. Point taken but I am only discussing cables and would still exclude the components you mentioned especially loudspeakers.
4. Semantics? When I say "...reputation based on what they sound like" I mean the sound has particular qualities that can be described using words that others can understand, eg "this cable has a narrow soundstage". If others agree it will get a reputation for this particular quality. If it has different qualities in another system then you can't say that quality belongs to the cable.
5. This is back to point 2. Why make it harder? Using only LRC makes your case easier to prove as you only have to match three easily measured properties.
6. See 4 above. My "If everyone has different opinions...." is merely "If X is true then A or B follows". To clarify my point, If everyone or most agree that a certain cable has a particular and describable sound quality (across various systems) then you can say that quality is a property of the cable itself and you are proved correct. (this would be the body of anecdotal evidence I referred to earlier which if it exists, will make me look stupid)
7. Covered in 2
8. Your last argument about objectivists is good and it works for me, also bodes well for your eventual conversion 😉 But you bring up expensive cables and I have just seen some recently. If I had money to burn I would buy some as they are cool in my book even though I don't believe they would sound any better than welding cables.
9. Logical debate is good. I also started out "believing wires couldn't possibly sound different", then I believed they did, then I realised I was right all along. 😀 Hope you are onto something and can prove it, welding cables are quite boring. Cheers.
Thanks Sy!
Hello SY. I'll do everything in my power to answer this question(s) ASAP with 7/22/09 being my goal for attempting to answer this question(s) of yours! I'm so sorry it's taking me this long to answer your question Sy but, in my defense I'm in terrible pain and cannot find comfort in ANY damn postion! These last few weeks I'm only sleeping when I finally conk out from sheer exhaustion after approx 72hrs or so without sleep!
I just got back from going to the doctor to have him provide my monthly quota of pain meds ---{oxycodone 30mg 6X a day}--- and muscle relaxers ---{soma 350mg 4X a day}--- and hopefully this will allow me to sit at the PC long enough to type out a reply for you.
Thanks for your patience...
Thetubeguy1954
SY said:
Hello SY. I'll do everything in my power to answer this question(s) ASAP with 7/22/09 being my goal for attempting to answer this question(s) of yours! I'm so sorry it's taking me this long to answer your question Sy but, in my defense I'm in terrible pain and cannot find comfort in ANY damn postion! These last few weeks I'm only sleeping when I finally conk out from sheer exhaustion after approx 72hrs or so without sleep!
I just got back from going to the doctor to have him provide my monthly quota of pain meds ---{oxycodone 30mg 6X a day}--- and muscle relaxers ---{soma 350mg 4X a day}--- and hopefully this will allow me to sit at the PC long enough to type out a reply for you.
Thanks for your patience...
Thetubeguy1954
Your health and well being are infinitely more important than internet blather about insignificant technical issues. Please take care.
Hi thetubeguy1954
You have made some very good posts and I’m really sorry to see that you have pain.
Take care and I wish you all the best.
Cheers
You have made some very good posts and I’m really sorry to see that you have pain.
Take care and I wish you all the best.
Cheers
Re: Thanks Sy!
Hi,
Speaking for hundreds of members too shy to express themselves:
We absolutely do adore you.
Cheers, 😉
Hi,
thetubeguy1954 said:
Speaking for hundreds of members too shy to express themselves:
We absolutely do adore you.
Cheers, 😉
- Status
- Not open for further replies.