Practice still won't lengthen the aural memory span, unless you know of actual case which a person was able to alter the DNA by practicing.
What you found and what he found have nothing to do with the aural memory span I brought up in response to your "adjustment of it, or else make your own, then give it a listen." 🙄 Perhaps you don't know how to compensate for our short aural memory span when doing such comparison. Given that, all your listening comparisons were done without such compensation, which explains why you would post things like "Dac was down that day for some modding, but a few days later I had it working again and decided to try Jam's cables and get it over with. I connected them between the XLR outputs on AK4499 eval board and the Neurochrome HP-1 headphone amp input.
Wow! Everything sounded better, less distorted, and the difference was easy to hear! I was basically stunned, never expected it."
Once again you are intentionally playing confused. Cables that I compared, given the resolution of the test reproduction system used, were easy to classify in relative terms long enough for ABX.
What required memorization and was more difficult were some of PMA's listening tests. Particularly the opamp recordings test, IMHO. Another one of PMA's that was rather challenging for me on my system at the the time involved some string instrument classical music (which has a lot of natural odd-harmonic structure). Despite the choice of music, turned out there was a place in the performance where the files sounded different but the difference was subtle and very odd to memorize. Memorized it long enough for a few trials, but it was very fatiguing to keep up for more trials in one sitting. Again, this kind of thing is what PMA had to deal with when he figured out how to pass ABX, described it as requiring sustained concentration, and as difficult compared to sighted listening of the same effects.
IME A/B is generally easier in terms of memory persistence required, and thus more sensitive. It requires less user training time too.
With that I will stop here. The questions are increasingly becoming political-rhetorical (i.e. troll bait), rather than honest queries in the interests of finding truth.
What required memorization and was more difficult were some of PMA's listening tests. Particularly the opamp recordings test, IMHO. Another one of PMA's that was rather challenging for me on my system at the the time involved some string instrument classical music (which has a lot of natural odd-harmonic structure). Despite the choice of music, turned out there was a place in the performance where the files sounded different but the difference was subtle and very odd to memorize. Memorized it long enough for a few trials, but it was very fatiguing to keep up for more trials in one sitting. Again, this kind of thing is what PMA had to deal with when he figured out how to pass ABX, described it as requiring sustained concentration, and as difficult compared to sighted listening of the same effects.
IME A/B is generally easier in terms of memory persistence required, and thus more sensitive. It requires less user training time too.
With that I will stop here. The questions are increasingly becoming political-rhetorical (i.e. troll bait), rather than honest queries in the interests of finding truth.
Truth is that you haven't laid out a single explanation on how to compensate for human's short aural memory span (seconds) when doing your "adjustment of it, or else make your own, then give it a listen." Maybe you know the aural memory problem with long break between components / parts swap but you deliberately omit it because it goes against the boutique audio business agenda that you've been try to push on this forum.
Hi ppp000, I would like to use Chinese cheap DSP card, like you do. Do you still use this project? Have you encountered any problems using it? I saw some photos in a recent thread and it seems to me that you are no longer using the chinese DSP card
Hey, I still use but if had choice I would probably go for adau1467 or similar so I can set up asrc. That sharc board is master i2s so everything u connect to it need to be slave and that cause problems.
Thanks for the reply,
so using the AMANERO board in slave mode and setting JRIVER with 96kHz output sampling I shouldn't have any problems, right?
My four DAC boards with ak4490 need MCLK so they already work slave mode
so using the AMANERO board in slave mode and setting JRIVER with 96kHz output sampling I shouldn't have any problems, right?
My four DAC boards with ak4490 need MCLK so they already work slave mode
I haven't bought the card yet. I'm currently using the Amanero board and the four AK4490 DACs with the Minisharc. As I intend to use the Minisharc with other better quality DACs, I wanted to recycle the four ak4490s into a cheap system. That's why I would like to try the Chinese DSP board
There are 2 versions of this board, one I have is master and works with sigmastudio, newer version is slave, doesn't work with sigma studio, only with provided software. Both look the same, difference is missing stm32 MCU in newer version.
So its a total mess. Also if you want to use core board only you might need some kind of adapter - goldpin connectors are 2.0mm raster instead of standard 2.54..
Honestly I would go with adau1466/67 board instead, unless you really need sharc.
So its a total mess. Also if you want to use core board only you might need some kind of adapter - goldpin connectors are 2.0mm raster instead of standard 2.54..
Honestly I would go with adau1466/67 board instead, unless you really need sharc.
thank you for the information
At AliExpress I saw an even more recent card than the one you're talking about (revber?) it's very different from the old one, it doesn't have a micro USB connector and it's light blue:
https://it.aliexpress.com/item/1005004812705223.html
https://aliexpress.com/item/1005005240777880.html
The card like yours is easily identifiable, the PC software is the older one
For DSP board connections I could use these two options:
https://it.aliexpress.com/item/32868906943.html
https://it.aliexpress.com/item/4000597517515.html
I would like to use the SHARC DSP to benefit from floating point processing
At AliExpress I saw an even more recent card than the one you're talking about (revber?) it's very different from the old one, it doesn't have a micro USB connector and it's light blue:
https://it.aliexpress.com/item/1005004812705223.html
https://aliexpress.com/item/1005005240777880.html
The card like yours is easily identifiable, the PC software is the older one
For DSP board connections I could use these two options:
https://it.aliexpress.com/item/32868906943.html
https://it.aliexpress.com/item/4000597517515.html
I would like to use the SHARC DSP to benefit from floating point processing
Last edited:
revber is developer of this board, so you can ask him directly. Yeah that blue board is 'proper' v2 of board. First v2 boards been the sama as v1, just MCU and crystals been removed. USB was used to flash it, so no sigmastudio support.
Im leaning more towards adau.
flexibility.. you can set up almost all(?) in sigmastudio + easy to add MCU and control dsp over i2c.
On v2 sharc board you use provided software. adsp21489 is like 4000taps max? With all added functionality you might never use.. how many taps left for crossover? In sigmastudio you decide what gets more or less power.
Adau1467 can do 24000taps at 48khz
Thats from https://www.audiosciencereview.com/...surements-and-rising-noise-floor.42383/page-8
Nanodigi is ADAU1445 based, IIR filters in this case and double precision for sure, looking at cdsp 32bit performance🙂
What holds me is that I cant find info if FIR on sigmadsp can be double precision or 32bit only..
Im leaning more towards adau.
flexibility.. you can set up almost all(?) in sigmastudio + easy to add MCU and control dsp over i2c.
On v2 sharc board you use provided software. adsp21489 is like 4000taps max? With all added functionality you might never use.. how many taps left for crossover? In sigmastudio you decide what gets more or less power.
Adau1467 can do 24000taps at 48khz
Thats from https://www.audiosciencereview.com/...surements-and-rising-noise-floor.42383/page-8
Nanodigi is ADAU1445 based, IIR filters in this case and double precision for sure, looking at cdsp 32bit performance🙂
What holds me is that I cant find info if FIR on sigmadsp can be double precision or 32bit only..
Unfortunately the noise floor value is only one of the parameters that define the performance of the DSP, but it doesn't say anything about the real sonic performance
For product with high end performance I think is best to use SHARC chip
If ADAU chips are the budget series of DSPs there must be a reason, right? 🤔
For product with high end performance I think is best to use SHARC chip
If ADAU chips are the budget series of DSPs there must be a reason, right? 🤔