Sounds good, i will wait for next release befor i change anything. Iwill give NAJDA one more chance...
With a 90 mph limit I wouldn't buy it, but as you probably well know, a lot of German cars are electronically limited to 155 mph, despite the vehicle being capable of higher speeds - the point being that the higher speeds are not useful.
Please tell me what difference frequencies above 25 kHz make to the sound you hear...
Ok how i told befor i dont want to start a pro and cons hires discusion specialy english isn,t my motherlanguage.
But i found a link with the right description of the most problem with lo Res equipment.
High-res audio: the science behind the numbers | What Hi-Fi?
"The case for increased sampling rates is stronger. 44.1kHz was chosen for CD because it allowed an upper frequency limit of just over 20kHz – the upper limit of what humans can hear. You’ve got to be pretty young and have pristine ears to do it though.
The way digital works means that there are an awful lot of unwanted signals generated above that upper frequency limit. These have to be filtered aggressively; otherwise they’ll result in more distortion in the audible range.
That filtering introduces its own distortion which folds back into the audible range. Raising the sampling rates ever higher means that the filters can be set to work at far higher frequencies taking them and their unwanted effects further away from the audible frequencies.
The raised upper-frequency limit also means that the upper harmonics of instruments can be represented better, even if science strongly suggests we can’t actually hear them."
I did much blindtests with diferent Equipment and files and for me hires sounds better richer in Details more room between the instruments and so on....
But im not that person listen music on mp3 player or in the background.
When i listen a new record i take time for it and do nothing other in Background! If i go Running with mp3 player then i guess i wont be able to hear the differences but thats not the point!
The Other Point is i Bet the more details in the source is the better works the filter calkulation in the DSP ...but please dont start a debate about the Math behind it!!
Are we talking "hires" (whatever that means), or low-pass filtering?
English isn't my first language either - it is my third.
That applies to the analog-to-digital conversion, not digital-to-analog. You need to filter *before* the ADC - or oversample and then filter digitally. Both are valid methods. Neither imply that you can hear anything above c. 20 kHz.
So basically "we are going to disagree with science"?
Would you be able to share your test setup and your blind test result logs?
Definitely not the point - so trying to paint anyone questioning your claims as "someone who goes running with a mp3 player" is a pretty cheap shot that actually undermines your credibility.
Why not? Can you please present the maths? I would love to see it?
I've been following these discussions about the ASRC and have already attempted to answer Frunse's questions via email.
Summarizing what happens with the ASRC:
- The ASRC is included in the DSP and is documented in the DSP reference manual (chap 19). There's nothing such as 'hardware' or 'software' mode.
- Every digital input goes through the ASRC. The input sample rate is converted to the user selected sample rate, which is one of 48, 96 or 192 kHz.
- Frunse has observed that digital signals going though the ASRC are band-limited to 0->24 kHz. This is correct, and that's how the ASRC works. However, the signal rate is still one of 48, 96 or 192 kHz - this doesn't seem to be well understood. So to make it clear: the ASRC converts the input signal to the selected rate, but band-limits it to 24 kHz.
Hope this clarifies how the ASRC works.
Cheers
Nick
Summarizing what happens with the ASRC:
- The ASRC is included in the DSP and is documented in the DSP reference manual (chap 19). There's nothing such as 'hardware' or 'software' mode.
- Every digital input goes through the ASRC. The input sample rate is converted to the user selected sample rate, which is one of 48, 96 or 192 kHz.
- Frunse has observed that digital signals going though the ASRC are band-limited to 0->24 kHz. This is correct, and that's how the ASRC works. However, the signal rate is still one of 48, 96 or 192 kHz - this doesn't seem to be well understood. So to make it clear: the ASRC converts the input signal to the selected rate, but band-limits it to 24 kHz.
Hope this clarifies how the ASRC works.
Cheers
Nick
ASRC-question
Hi Nick,
the bandwith of the output of the ASRC is limited to 24 Khz as you told. My question: how is this achieved? What kind of filtering is applied? Do this filter introduce any disturbance into the audio band of 0-20 Khz?
If I use the 3 I2S outputs for another DAC, depending on the SR selected ( 48,96,192) I will get outputs at 48,96,192 SR and band limited to 24 Khz?
Cheers
Sigi
Hi Nick,
the bandwith of the output of the ASRC is limited to 24 Khz as you told. My question: how is this achieved? What kind of filtering is applied? Do this filter introduce any disturbance into the audio band of 0-20 Khz?
If I use the 3 I2S outputs for another DAC, depending on the SR selected ( 48,96,192) I will get outputs at 48,96,192 SR and band limited to 24 Khz?
Cheers
Sigi
Disturbance? Well, if there were, wouldn't they had it implemented in a different way? The limit is hardcoded in the ASRC, can't tell you why, but would guess because limited internal DSP calculations resolution (24bit) and internal audio core frequency. Same can be seen with the i.e. 28/56bit ADAU DSPs from Analog Devices. You can run them with non standard sampling rates (96/192kHz). This rises internal resulution in trade for processing words but the internal AD/DA-stage is still running at 48kHz. So incoming/outgoing audio is 48kHz, internal blown up.
The Freescale DSP actually isn't recommend for new design anymore.
The Freescale DSP actually isn't recommend for new design anymore.
Last edited:
Hi, what you commented might be true, but it has to be kept in mind, that up to now, at least for us, Najda was best with regard to sound quality when compared to DSPs costing more than 8000 $ !!
Thx's
now i understand what you tried me to tell, the data below 24kHz cutof isnt downsamplet to 48kHz, right?
The Resolution of the source is still there (at fs:96kHz still 96Khz but highs above 24kHz are "cleaned" or filtered out, why ever ...no free Openair concert for Bats at all)
PS: sorry for my long wire in my Had, but sometimes......not all is green or red
resolution
Hi,
may be an example: if you have two tweeters. One is going up to 21 KHz, the other up to 28KHz. But you can hear the first one is much more finer and
with very high resolution?!
I also think that the Najda ist a phantasic solution for multiway horns.
And we all are satisfied with the Najda! But it is our hobby too. And DIYers always try to optimize......
regards Ulf
Hi,
may be an example: if you have two tweeters. One is going up to 21 KHz, the other up to 28KHz. But you can hear the first one is much more finer and
with very high resolution?!
I also think that the Najda ist a phantasic solution for multiway horns.
And we all are satisfied with the Najda! But it is our hobby too. And DIYers always try to optimize......
regards Ulf
I guess the problem is that we all have a different definition of "optimizing"
To an engineer, "optimizing" means making something as good as possible, but no better.
Going for bigger numbers when it has no audible relevance is waste, not optimization.
I try to descripe what i mean...
What Means Samplingfrequenc?
at 44,1 kHz you has 44100 measuremenst per second at 24Bit
so Bits are the steps at X-Axis to descripe the Amplitude for each measurement
and the Samplingfrequencie are steps at Y-Axis
so if you has 96kHz Samplingfrequencie you has 96000 Measurements per second and even at 1kHz Sinewave you has twice smother steps to descripe the form of the sinewave at 1kHz digitaly = more Reselution...right?
i guess specially for digital calculations more data = less loss from Rounding error and other calculations problems...? or not?
At the end of the digital chain is the outputfilter , it makes the stepds of the sinwave smotth as well to Analog domain so may be it isnt, much important for normal dac but the DSP makes digital calculations and that why i is important to know i guess.
Sorry for my English dudes
No. The actual wave coming out of the DAC is just as smooth in both cases, but the 96kHz oversampling gives you (theoretically) 6 dB better SNR - but 24 bits already gives you about 48 dB more signal-to-noise ratio than any real-life recording, so it really doesn't matter.
Yes, intermediate results in a calculation require the 24 or 32 bits to preserve the original precision - but sample rate doesn't make a difference.
Right. The calculations are already made at a sufficient precision and sample rate - the issue we are discussing is the low-pass filtering above audible frequencies.
No. The actual wave coming out of the DAC is just as smooth in both cases, but the 96kHz oversampling gives you (theoretically) 6 dB better SNR - but 24 bits already gives you about 48 dB more signal-to-noise ratio than any real-life recording, so it really doesn't matter.
ok , i found an good description abut this but it is writen in German language only ...sorry
https://goldohr.wordpress.com/2012/...oher-samplingfrequenzen-bei-hd-digital-audio/
so we could say an 32Bit DSP DAC combination at 48Khz makes more sense than a 24Bit/ 96Khz if all mades fine around the Digtal elektronics , right ?
at the studios the Equipment for 24/96 recordings was used already long time befor it comes to the Mainstream so thats what iwant to have at home i dont think we will get much benefit from 24/192, Reason for this is Marketing i guess to sell new Equipment
I was happy with the sound of the DVD-A Recordings (24/96) long time ago specialy you compared it to a normal CD
so you have 2 benefitts: Outputfilter at 48Khz and 24Bit for resolution and SN/R
No prob - I can read German pretty OK (I had a couple of years of German in school 40 years ago, but lost most of it when I learned Dutch).
32 bit 48 kHz makes more sense than 24 bit 96 kHz in DSP and processing, yes. 96 kHz *when recording* makes sense to be sure to avoid the requirement for steep analog filters, but once recorded/digitized, the signal can safely be bandwith-limited.
Indeeed. A lot of studios still record at 44.1 or 48 kHz. Anything above 96 kHz is overkill and doesn't get you any benefits (and can actually be harmful, see the reference to modulation products at the end of the German page you pointed to).
How much of that was better recordings in general, and how much was the actual format?