John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
Adobe Addition's resampling filter, 192k-->44.1k-->192k, highest quality... a rather long FIR, suspposedly.

Do you see a "frequency shift" in the red trace as well? If so, you don't understand the peak of the 5x20 spectrum doesn't neccessarily coincide with the carrier when a bit of non-flat EQ is present (Bu2@19kHz, in this case), which is what happened with Hans' filter.
That is exactly why I mentioned moving the graph to provide more detail in that exact area of the spectra.

"Rather long FIR supposedly"? Is there any way to find out? Do they give the user the option of FIR length? That would make it rather easy to see if that is why Han's frequencies were so different.
Besides, filtering can *never* change a frequency, basic knowledge....

Yah, I've heard that before. You are telling the wrong guy that. Go back and look at Han's time display and tell me the two frequencies are the same.

Do you have a better explanation as to why he shows two frequencies? I'm all ears.

Jn
 
Last edited:
In this thread you have to prove everything that you mention to prevent that nonsense is generally being accepted.
Here are two square waves in two amplitudes fed to the same Fir filter that I used earlier, in this case having amplitudes of resp. 1.0 Volt and 0.3 Volt.
After being processed by the Fir filter, I have multiplied the 0.3 Volt filtered version by 10/3.
And guess what, results are exactly on top of each other, proving that ringing has nothing to do with limited slew rate or whatever but only with the transfer properties of the used filter.

Hans

View attachment 807459

All you have done is show that the ringing scales with the slew rate (and height). This is exactly as expected, no surprise. It would be odd if it didn't.

What you need to address is your earlier plots showing a change in frequency after you put it through your filter. That was a totally unexpected outcome.

Oh, thank you for your efforts,,also KSTR, thank you.
Jn

Oh, sorry. If you would, could you please overlay the blue and green waveforms from your earlier post simultaneous in time. Since the frequencies are different, perhaps line up the first zero crossings. Thank you.

Tomorrow I'll make two copies of the plot, use a pair of scissors and some scotch tape, a couple chicken scratches mimicking actual handwriting... scan to jpeg, voila..

Edit: KSTR, if you don't mind, would you do the exact same thing with your software? If you do not support a carrier frequency shift, we have to assume it was something in Han's software package.
 
Last edited:
Perhaps the dust surrounding the old EEs protects their 5 CDs from the invasion of molds which seem to appreciate the metal layer of CDs even more than the coating of photographic lenses?

Make that 6 :rofl:. Attached is about 1/3 of my DVD/CD collection (of 20+ years), in storage, the rest is "active". Never had a dang of a problem with mold or anything else, they are all as good as in the first day. And yes, I saw/listen them all.
 

Attachments

  • IMG_2160.jpg
    IMG_2160.jpg
    889.3 KB · Views: 175
Last edited:
Maybe Mr. T is storing the disks in a place where one stores their wine perhaps or in the kitchen where humidity is higher? I have never had a mouldy disk or LP either, I have them all in my basement too, it can get humid in the summer time but I do turn on the de-humidifier when it gets too high.
If you leave out some 20-20-20 water soluble fertilizer, exposed to air, when it turns to liquid, it is time to turn on the de-humidifier. Grow-op experience :) I have a Honeywell humidiy sensor and Atmel xmega code for you if that works better for you.
 
Last edited:
Maybe Mr. T is storing the disks in a place where one stores their wine perhaps or in the kitchen where humidity is higher? I have never had a mouldy disk or LP either, I have them all in my basement too, it can get humid in the summer time but I do turn on the de-humidifier when it gets too high.
If you leave out some 20-20-20 water soluble fertilizer, exposed to air, when it turns to liquid, it is time to turn on the de-humidifier. Grow-op experience :)

Now the dehu's generally come with a digital readout and the humidity can be set. Right now basement is running about 52-53% with the unit off. I store the hardwoods for my wooden gear clock making, so absolutely need to keep the RH at 50% plus or minus. Summertime, always on, set to 45.

Jn
 
Last edited:
"Rather long FIR supposedly"? Is there any way to find out? Do they give the user the option of FIR length? That would make it rather easy to see if that is why Han's frequencies were so different.
You can only set a "quality" parameter from 0(lowest) to 999(highest) and toggle if pre/post filter shall be applied. My example was done with "999" and "on".

Find attached the FR of the 192->44.1->192 resampling and an estimate for the lower bound of the total filter length (4500 samples), as derived by using a single sample pulse at 192kHz, result normalized to 0dB
 

Attachments

  • resample-filter-192-44-192__FR.gif
    resample-filter-192-44-192__FR.gif
    10.3 KB · Views: 208
  • resample-filter-192-44-192__length.gif
    resample-filter-192-44-192__length.gif
    21.4 KB · Views: 217
You can only set a "quality" parameter from 0(lowest) to 999(highest) and toggle if pre/post filter shall be applied. My example was done with "999" and "on".

Find attached the FR of the 192->44.1->192 resampling and an estimate for the lower bound of the total filter length (4500 samples), as derived by using a single sample pulse at 192kHz, result normalized to 0dB

Thanks, nice software.
Do you have the ability to duplicate Han's overlay of the 5 by 20 in time with about 250 taps?

Jn
 
I have a 50GHz Tektronix digital scope that I normally use. I can clearly see the AM sideband missing after sampling at 44.1KHz, this warms my heart with artistic emotions.
I have some weed or brownie for you, if you would like, I found it adds to the music & scope viewing experience. Booze does not have the same effect. Done enough to know the difference :) Abstainers just do not have as much fun = boring life, rock n roll and no weed, missed out on some good cultural experiences imo. I predict a good portion of the old rock n roll music had someone involved in the process that were stoned :) Oh yeah you guys listen to classical music right? talking to the wrong crowd. Even the late Louis Armstrong new this tid bit :) Do not take me too seriously, just having some fun with the intellectuals of the world :D
Back to your bits and bites hour :)
 
Thanks, nice software.
Do you have the ability to duplicate Han's overlay of the 5 by 20 in time with about 250 taps?

Jn
With Audition, no, but I could use Rephase to build a 250-tap brickwall kernel/IR for 44.1... haven't used Rephase for quite a while but it should be doable without major effort, I might give it try...

Then again, if @Hans could share his kernel in a format that I can use I could analyse and use it in Audition directly. Such formats are:
- WAV with 32 or 64 bit floats
- binary file (plain array of double), sample-rate
- text file with coefficients (could convert that with a C quickie)
 
Last edited:
Status
Not open for further replies.