|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
213.7.81.168
In Reply to: RE: What is incredible posted by theob on November 22, 2014 at 04:12:03
Double Precision Single Point!
All the selections sound different and so there is really no reference for what is good or bad.
Follow Ups:
> > All the selections sound different and so there is really no reference for what is good or bad. < <
My guess is that fixed, floating and double precision refers to how internal processing is done. Fixed and floating point are 32 bit while double precision is 64 bit. There might be a rounding error on the 32th least significant bit, but there should be absolutely no difference if the pcm is saved as 24 bit (this can be verified by doing binary comparison).
Since the best d/a converter is able to do 21 bit it is either amazing that you can hear a difference or my guess is wrong or there is a bug in the software.
The issue isn't when what happens at the 32nd bit. The issue is what happens at the 24th bit. It is possible to convert 24 bits to 32 bit floating point and get back the original 24 bits, but if there has been any digital signal processing other than the straight format conversion then this may not be what sounds best, because there will be distortion introduced when the bit depth is reduced unless dithering is used. On the other hand, if dithering is used, then a round trip from 24 bits to 32fp and back to 24 bits will be degraded by the addition of dither noise affecting the low order bit of the 24 bit file. Good audio editing software allows for various options in these format conversions and it is possible to use this capability to train oneself in how these different approaches sound.
In some cases changing the low order bit will affect the sound quality. Theory may say that this is "impossible" at 24 bits, but critical listeners have reported otherwise. While there is certainly the possibility of a bug in the software, it is more likely to be a "feature" of converters (DAC circuitry) or analog components which are complex non-linear devices which can and do behave in non-intuitive fashion.
I do not find it at all incredible that feeding different bits to a DAC will result in hearing different sounds. That is, after all, the whole point of a DAC. :-)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
because of the amount of cash invested in making and buying 'perfect' usb dsd and high res equipment reviewed to be 'better' sounding when that sound depends on the selection of may be 30 factorial ie 30x29x28....x1 BIOS, OS, player and hardware settings.
It is well known that dithering at 16 bit, 24 bit, or even below 24 bit levels changes SQ and dynamic range perceptions.
I wish that bits are bits guys can open their minds to the broader issues of audio reproduction.
The foobar for dummies link indicates that there are several additional FIR samples included in the foobar download with cutoff frequencies as high as 60Khz.. I could not find any in my Foobar folder. Were you able to?
Go to Foo_SACD_0.7.3 text file, the filters are there and can be loaded. I tried the 60kHz one on a 352.8k 2L DXD files. Good clarity and has impactmin the bass and treble. No brightness on that file. A 176.4k file converted to DSD128 also sounds clear but sound stage is not like the orginal with less definition of the locations of instruments.
Just shows how good (or bad) web 'how tos' are.
Why the writer of the software would not include a proper Readme is beyond me.
Could not agree more. Just found the filters folders, loaded the 60Khz file, played it and another step up in clarity. Thanks for the help. With this setup in my opinion Foobar far exceeds jRiver in dsd playback. I always thought wav playback in Foo was superior.
Wrt the brightness...agree again. But isn't it ironic that when you extend the cutoff frequency from 30 to 60 kHz it gets less bright. I found this to be true in every application I have tried it.
because you are listening to the effect of the steep digital filter.
Caution: the 0 kHz filter is intensive on cpu power. My power consumption went from 25W tp 36W by using it and the cpu temperature went up from 50 to 60 C.
A good compromise for me is: DSD 128 type D and pcm to dsd none. There is a bug in the software when I engage certain settings whereby 192k pcm files are downsampled to 96k! I haven't gotten to the bottom of this yet but I shall.
Sorry to hear that. Everything seems to be ok here. Cpu usage is running 4-6%, no problem on heat.
What board and cpu are you using?
I downgraded the BIOS of my spare GA-G31MS2L From F8 to F6X today, disabled spread spectrum and this made quite a difference to playback rhythm for the better. everything else being the same!
This gets ridiculous.
I'm using the GA 31 also with an E7200 cpu. I run the mobo at default settings. When I tried slowing the cpu speed down everything got fat and wooly. So I left it at default settings.
Haven't tried; will look into it. I am having a PC issue at present, with Server 2012 crashing.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: