I dont know which monitors you guys using, but maybe thats the difference... but i can hear a difference i dont know if omnisphere sound better or sonar x1 but with 96 khz there is a clearer sound then with 44,1 khz.

i will do a bounce and upload it for you.

cheers



Edit: tried to upload the files but i can only upload a filesize of 2mb and the 96khz file is 15 mb for 10 seconds.

Guys try it yourself the difference is in the mids or high mids with 96khz there are more dominant then with 44,1khz atleast for me with my setup.

 

You can upload only 2MB? Where the hell did you try this?

Use fileserve, hotfile, wupload, or whatever...

 

i tried to upload it here directly... but i used fileserve now here are the two files its exactly the same patch the same notes everything is the same. only one time with 96khz and one time with 44,1 khz.

http://www.fileserve.com/file/y2YKn9x
http://www.fileserve.com/file/yj36VBY

 

Uploading samples doesn’t prove anything; Advantage of higher sampling rate is in recording/mixing process. Each of us should make conclusion for them separately and act according this. In this story is most important is it author (creator) of the song happy, if he is happy sound engineer should be happy to.

 

@Gulliver Omnisphere samples are 96 or 192 I believe (or at least that's what they proud themselves to be)... I don't see application for 96 unless ur doing film... 48 also... I use 88.2 now and I just asked my remixers for 88.2khz/24bit unmastered or 44.1khz/24bit mastered so I can send their versions of my track to the label or master it myself and bring it down to 44.1khz... It seems it's workin out that way! Also you can definitely hear the diff between 320kbps and WAV - it's subtle but you can hear it (at least I can)

 

I downloaded the files, thank you. They have a bitdepth of 64 bit, how is that possible?
I couldn't import them in Cubase. It's difficult to compare them properly anyway, because every time I should switch the samplerate on my audio-interface.
But I could play them with my media player, and they sound different indeed. But I don't know if the samplerate is the reason, something else could be going on there, how can you be so sure?

Actually the version with 44,1 kHz sounds better to my ears. It has more highs, and sounds much clearer.
The 96 kHz version sounds duller and a bit muddy in comparison.

But I don't know the reason for that, how did you compare the two versions?
If you don't change the samplerate on your audiocard, it has to do some resampling for playback, I suppose (or the software).
Btw, could you tell us the name of the patch?

Would be nice to read other opinions on this one.

I just checked the Spectrasonics homepage, and couldn't find any information about that. Where do you have that info from?
It would be very unusual for a company to make a virtual instrument with 96 kHz or let alone 192 kHz samples, because that would increase the size of the library dramatically. So I higly doubt that.

I would be curious, if you would survive a blind test between 320kbps and WAV, and also other people here who claim to "definitely" hear the difference.

 

Hello Gullivere,

im working with sonar x1 and i edited a ini file from sonar x1 where you can change the bith depth i think this is the reason for the 64 bit, else there is a option in sonar x1 where you can check and uncheck the 64 bit option....

the 44,1khz file is smoother, the 96 khz file has more dominant mid highs at least for my ears... in think the 44,1 khz file is missing some of the frequencies because of the lower sampling rate... cause the 96khz file has more information, this is my opinion, maybe someone has another opinion on this one.

The name of the patch is adagio transparent strings warm

i compared them in sonar x1 i closed the project and changed the settings for the sampling rate and opend the it with other sampling rate, but you can also compare them in vlc player it does open both files.

 

It's interesting how perceptions can differ.
For me, the 44,1 version is like if you remove a blanket from the speakers, after listening to the 96 version.

Try also listening again, with good headphones, and without prejudice.

Btw, speaking about Omnisphere: if I browse trough the patches with my mouse wheel, no matter if I scroll down or up, it always scrolls down in the list. Is it like this also for you?

 

So ist very person i think everyone hears different.

I have no problems using the mousewheel in omnisphere do you have the latest updates installed?

 

Yes I have, expect the very latest one (1.5.5 I think), but I always had this issue.
I can use the mouse wheel, but it always scrolls down, not matter which direction I turn.
Maybe the mouse is the reason, I will try a different one.

No other comments about the examples from suchenderxxx?

 

Hi,

Regardin 16bit vs 24bit: each bit will give you 6db range for regording:
24bit = audio range of 144dB
16bit = audio range of 96dB

So you will have more headroom and you will not need to run your preamps super hot as in the early 16bit digital recording era (with the danger of distorting or use compression/limiting while tracking).

Regarding 44.1khz vs. 48khz the choice is mostly relevant considering the end product. If you mix for dvd you may choose the latter. if you mix to CD/mp3 44.1 is enough. Why? humans have an effective range of hearing from 20hz to some theoretical 20khz. To capture this you would need twice the resolution in khz. 44.1 can capture more than you need. If you want to be "future proof" or swear you get more air with 48 khz you might chose to record this way. Personally, 24bit/44.1khz is more than enough.

opty

 

Actually 16 bits gives you Dynamic Range of 92 dB, not audio range. Dynamic Range is also known as "headroom".


As to the comparison of those 2 samples, I shall not get into that discussion because my hearing range has fallen below 17 KHz so I can no longer discern the difference between them.

 

jupp. dynamic range it is then. sorry for being imprecise. Anyways, the relative difference between 16bit and 24bit is quiet huge in terms of range you can cover. (if you use the 6db rule or not). In my book everything about 130db is not fun.

Regarding downsampling I think type of dithering plays a huge role. I usually have my go-to settings which are not very fancy and pretty presetish. Has anyone ever done a shootout of different dithering algorithms, settings or same algos between DAWs?

opty

PS: what is the true maximum dynamic range for 24bit. Just so I know next time.

PPS: I think we do not need to put the word headroom in quotes in this forum.

 

after having read all this I'm still not quite sure I have all the info.... are the two files 16 bits...24 bits (or even something else) or one is 16 the other 24?
cuz that would make a difference in sound as 24bits has definitely a better dynamics than 16...and when dithering from 24 to 16 (if they were originally different), you would still retain some of it...
just for the sake of numbers 16 bits offers 65,536 (2^16) levels of dynamic range (starting at 0) with a maximum of 92 to 96dB (it depends on what you use, opty was not wrong!) and 24 bits 16,777,216 (2^24) levels with a max of 144dB.

http://www.analog.com/en/processors-dsp/sharc/processors/technical-articles/relationship_data_word_size_dynamic_range/resources/fca.html

actually, on high end equipment, if one file is at 96 and the other at 44.1 then it's quite possible that you hear a difference (and 96 should sound better, not 44.1!) but after you down-sample to 44.1 all these are lost anyway and there will be no distinguishable difference at all...that's just about math and biology and someone already mentioned it...if you do have a difference then you did something wrong or you got wrong info in the first place


also is it possible to upload the two files to something else, this site is banned at the moment where i live (how about using zippyshare.com or wupload.com or ifolder.ru) thanks in advance

 

it might also be interesting when considering "when" to downsample. I would guess that the best way to do it is after mastering or (more generally maybe) after the last compression touched the material. I would not suggest to downsample before mastering. With 96 I would expect things to be tighter, especially in the bass. My HS80s don't go that low and I lack the money for proper subs and (more) room treatment anyways. Is there a way to selectively downsample (like multiband compression, but just for shaving things of? Such algorithms could be modelled using psychoacoustics or is it already (implicitly) included because of Nyquist–Shannon? Most encoders have an entropy stage anyways, so this might also be applicable to downsampling... or so I guess!

opty

 

@ lysergyk

No, the 16 vs. 24 bit discussion has nothing to do with the audio examples, it's a different topic.
The files are actually 64 bit , suchenderxxx made them with Sonar.

Stop talking about things in theory, listen first.
I have read in another thread that you use Omnisphere, too.

If you can't download the files, load the patch "Adagio Transparent Strings Warm" in Omnisphere, and play some chords.
Do this one time with your audio-interface set to 44,1 kHz, and the other time to 96 kHz (you can also make mixdowns).
Tell us, if you hear a difrerence, and if so, what.

 

yes it's definitely better to downsample only once and ofc that means at the very end, because first if you have files of a better quality you can more more precisely and then if one day you want to use your work with a better support (better than CDs), you'll also be ready. All filters or SRC create artifacts and errors so it's better to use them only when necessary especially if you work at 44.1khz, as it's more difficult to get a good filter
if you mean "selective" as in the ability to choose different kinds of filters, yes i've alreay seen that...but if you mean that you want to choose "where" to apply it on the signal, I've never heard of it.. so I dont know!

 

ok i'll do that as i have omnisphere...but as I told you already (see the other thread...like ping pong posting!) i guess you are on two different wavelength here

 

so i tried and listened again and again...but in all honesty I couldnt find one
sounding better than the other...there is a very slight difference..but i wouldnt
use that to say one sound "better", and if I dont earnestly focus on certain points
and listen to the file (that is just a few seconds long)as a whole, I cant even hear
it...but that's just me! as to what the difference is, overall the 44.1 sounds a little
"clearer"/ brighter/ with more high mids or high freq, and 96 warmer/more low mids/ fatter

spectral analysis clearly shows that there is a difference...but it wont make any
difference at all to an end user (=someone listening to your track mixed and mastered).

 

That was exactly my impression, too.
More clearness sounds always "better" to me, but that is subjective, of course.

The question is now, where does this difference come from?

Why the clear loss of highs in the 96 Hz version?
Actually it has a little less of everything, as your spectral analysis shows. I think we just hear the 96 version as warmer because of the lack of the high-end, in comparison to the 44,1 version.

Let me make one thing clear...
EVERYTHING we discuss here makes no differece to the end user (bitdepht, sample rate, preamps, different bus compressor plug-ins...).
Most of them are happy with a 128 kbps mp3, played back on crappy speakers.
So all this discussion we are helding just among us, because for US this things could be important.