768 Khz

Hello,
just saw this from RME
https://www.rme-usa.com/adi-2-pro-fs.html
Do you guys know any software that can record at 768 Khz.  The highest i found was Audacity that can do 384Khz
Many Thanks

ubxf said:

Hello,
just saw this from RME
https://www.rme-usa.com/adi-2-pro-fs.html
Do you guys know any software that can record at 768 Khz.  The highest i found was Audacity that can do 384Khz
Many Thanks

Click to expand...

That is approaching the lower end of AM radio.....  Do not expect to find audio software to do that, of course commercial or scientific solutions surely exist.

JR

in my limited experience when i used scientific software in the past it was unusable sound wise.
384 is fine but since they are making the hardware i figured software must not be far behind.
i'm doing extreme pitch and time  processing and looking at the best sound quality possible

ubxf said:

in my limited experience when i used scientific software in the past it was unusable sound wise.
384 is fine but since they are making the hardware i figured software must not be far behind.
i'm doing extreme pitch and time  processing and looking at the best sound quality possible

Click to expand...

I can not hear up to 768kHz...

JR

ubxf said:

in my limited experience when i used scientific software in the past it was unusable sound wise.

Click to expand...

Why? there is no reason for that. SR is just one parameter out of many, the next one is bit depth, but don't forget the rest, like the analog path.

384 is fine but since they are making the hardware i figured software must not be far behind.

Click to expand...

What does 384k gives you that DS or QS don't?

i'm doing extreme pitch and time  processing

Click to expand...

These algorithms rely on oversampling and downsampling; that's where the quality is (or not).

and looking at the best sound quality possible

Click to expand...

I'm not sure that 384 ot 768 guarantee best quality possible.
In the last century, dbx made converters running at 700k; they were not bad, in fact much superior to anything else at the time, but they are largely outclassed today by any decent DS converter.

And remember that most current AD converter sample at 6.144MHz.

JohnRoberts said:

I can not hear up to 768kHz...

JR

Click to expand...

Me neither but for example  i use hydrophones to record ultrasound and need to pitch down a few octaves to listen to the results. I used scientific gear both software and converters 384khz and while it worked  it was not very good sounding. I did similar recording with my sound devices gear at 192khz and it sounded ok. RME gear in general sounds good so i'm thinking of giving it a try.

ubxf said:

Me neither but for example  i use hydrophones to record ultrasound and need to pitch down a few octaves to listen to the results. I used scientific gear both software and converters 384khz and while it worked  it was not very good sounding. I did similar recording with my sound devices gear at 192khz and it sounded ok. RME gear in general sounds good so i'm thinking of giving it a try.

Click to expand...

Heterodyning was invented in 1901... 

down shift the signal before you record it.

JR 

I need to do some studying to figure out what that is.

I may not understand it well enough to explain it succinctly but I think it is a variation on intentionally creating beats or alias frequencies folded down into usable frequencies.

My old SW receiver had a BFO (beat frequency oscillator) to shift morse code to a more intelligible pitch.

Back in the 70s I had a old used spectrum analyzer (only $700 used) that I am pretty sure used hetrodyning to extract pitch info..(but I can't swear to that).


My larger point is that if you apply your post processing to make the sounds more audible "before" you record them it might solve your problem.

JR

Thanks it makes sense but it's a little bit like the chicken and the egg thing. The only way i know how to pitch shift since i stopped using tape is using some kind of digital pitch shifter. i don't know how it would work with ultrasounds

The general idea, and likely problem, in a sketch. A local oscillator beats the HF signal and one intermodulation product falls at a lower frequency. However "harmonics" are not preserved, and the bandwidth is not reduced.

ubxf said:

Thanks it makes sense but it's a little bit like the chicken and the egg thing. The only way i know how to pitch shift since i stopped using tape is using some kind of digital pitch shifter. i don't know how it would work with ultrasounds

Click to expand...

Heterodyning uses a local oscillator which signal Flo is "mixed" with the incoming baseband signal Fx. It results in a upper product at F2=Flo+Fx and a lower product at F1=Fx-Flo. If the incoming signal spreads over the 200-300kHz range, a Flo of 200k Hz would result in F1 spreading the 0Hz-100kHz, that can possibly be converted at Fs=192k.
Now there are at least two caveats:
"Mixing" implies a non-linear element that results in undesired artefacts. These are usually dealt with by filtering.
Since the transposition is algebraic, the intervals become non-harmonically related, e.g. two baseband signals at 210k and 280k (a perfect 4th interval) transpose as 10k and 80k (a triple octave). I guess that's what PRR means by " "harmonics" are not preserved".
Does it matter for your investigations? I don't think so, but I may be wrong.
Now there are instrumentation data acquisition converters that operate at several MHz sampling rate, but they often lack bit depth.
Check the National instruments catalogue.
A 16 bit 1.2Msample/s card will set you back about $3k. And you'll need several things, like software and adapters...

Reaper handles 768 kHz just fine. I suspect some other DAWs do too.

Reaper isn't an analysis tool, but you can do almost anything through the various supported scripting languages (python, jesusonic, lua...).

This is interesting

I work with film audio postproduction and a trick used here is to record sound FX at high sample and then pitch down a couple of octaves. Transformers as an example has a lot of these kind of sounds in it. For someone who worked with tape this is like... DUH...

MKH8040 from sennheiser has a range up to 40k

Sanken even makes a microphone popular with the crowd that can record up to 100k.

Of course with overtones not being correct the Heterodyning technique has its musical limitations but that might also be interesting from a FX point of view.

I have to investigate this a bit


Thanks JR!

Edit: Would this then be something like a ringmod?

S

synthiaks said:

Edit: Would this then be something like a ringmod?

Click to expand...

Ring modulator is an extreme form since it slices the signal instead of multiplying it, which results in more products than the basic upper sideband and lower sideband. It may or may not make a significant difference. It does when the heterodyning frequency is much lower than the signal, which is usually the case with synths. Not so much in the specific case of baseband shifting.

Thank you abbey

I now have something to read up on! I only understand 20% of your answer 

> something like a ringmod?

Ring Modulator is just technology. Like all technology, it can be used for good, or it can be used for evil.

A good RM can be driven just into its multiplying range to give simple intermodulations. But bad RMs or greedy users can go into the chop&slice range mentioned, make a real mess of the signal.

There are other multipliers not in rings. But the 4-diode RM is simple and was common 1940s-1990.

Those sample rates are for the various flavors of DSD.

If you are getting into radio frequency range, perhaps looking into SDR (software defined radio) would be an option?  I am not thoroughly familiar with the technology but folks are doing all kinds of signal analysis with it and there is a very active open source user base.