attenuation of high frequencies when mixdown

  • Thread starter Thread starter SuperBlade
  • Start date Start date
S

SuperBlade

New member
Hello

I'm a beginner in CEP, can somebody help me, please ?
When mixdown to a file (WAV), I can notice certain attenuation of high frequencies of the original track.
This happens though I use only one block in one track.

If you edit the block in the Edit view and Save as (WAV), you don't have this attenuation, the result is like the original.
I think I can have something wrong in the configuration of the mixdown.
I don't have configured any filter

Is this normal ?
Does anybody know how to avoid it ?

thank you
 
You're mixing down to 16-bit, right?

What other settings do you use? What sample rate do you use? What resolution? What kind of dither?
 
yes, I mixdown with 16 bits

I've just used the default settings for the mixdown. But when I realised the attenuation of high frequencies, I tested with :
* Multitrack sheet :
- Dithering options : all of them are disabled
I can't choose any sample rate, because it is disabled and I can't access to it
- Premixing : 32-bit
- Panning mode : L/R Cut (log)
* Data sheet :
- Dither transform Results: enabled
- Use symmetric dithering : enabled
- Smooth Delete and Cut boundaries : enabled
- Smooth all edit boundaries by crossfading : enabled (3 ms)
- Dither amount for saving 32-bit data to 16-bit files 1 bits
* Display sheet :
- Windowing function : Blackmann-Harris
- Resolution 256 bands.
- Window width : 100%

I don't know if you need any think more ...
 
Okay, you're using Cool Edit 2.0.

Now, describe what you do when you do a mixdown.

You do it from multitrack? You click on Edit and Mixdown? And then when it takes you to Edit view, what do you do then?
 
Yes, I use Cool Edit Pro 2.0
Yes, I mixdown from multitrack view.
then I've tried
- Edit -> Mixdown to file -> all files

then, in the Edit view, I save as a wav file, and then I listen to it with Winamp.

I've tried too to Mixdown in the multitrack view doing :
- File -> Save Mixdown as
and selecting Windows PCM (.WAV)

with the same result.


If, for instance, I mixdown one only song (.mp3 file) that I put in the track1, the result is the same song with the high frequencies attenued.
If I load the same song in the Edit view (without doing anything), and I save it as a WAV file, I don't have this attenuation.
It is for this that I think it is made some kind of equalization when mixingdown, and I was wondering if it is possible to avoid it
 
Now you mention MP3 files. Does this happen with .wav and MP3 or just the MP3? What bitrate are the MP3's encoded at? If they are under 128k you will definately notice a loss in fidelity.
 
It happens not only with mp3, because I converted this mp3 to wav, and it occurred the same high frequency attenuation
 
Converting the MP3 to .wav is still an issue. ANYTIME you convert a file to MP3 it loses some if it's high end definition. How much depends on the bitrate you encoded it at. Converting it back to .wav wont restore that information because it has been lost forever.

Do not use MP3 while's recording or mixing. The only time you should need to use MP3 is at the absolute last step in your recording process when you convert your final mixdown .wav file to MP3 for sharing on the net. Keep that original .wav file because that is what you will use to burn to CD and anything else.
 
TexRoadkill said:
Do not use MP3 while's recording or mixing. The only time you should need to use MP3 is at the absolute last step in your recording process when you convert your final mixdown .wav file to MP3 for sharing on the net. Keep that original .wav file because that is what you will use to burn to CD and anything else.
Click to expand...
Truer words were never spoken. By the way, I noticed that you said you really noticed it when you played the file in Winamp. Winamp has it's own EQ settings, so you might want to make sure that you don't have some EQ preset (like "Rock") which hypes the highs a lot.

But if you mixdown from the multitrack session straight into edit, that .wav file should sound basically the same as the tracks sounded in the mixing session...unless you've mixed down too hot and clipped everything.

Let us know how it turns out.
 
hello

I have Winamp with the Equalizer disconnected.
Sorry, my english isn't as good to understand :

"unless you've mixed down too hot and clipped everything"

what does it mean mixing down too hot ?


another thing : can someone test in your system if the high frequencies are attenued by the mixing down ? it takes only a minute !
What I did was create a new multitrack, take a song with rich high frequencies, put it in the first track and mixdown.

The result should be the original song, shouldn't ?

I listened with Winamp (wich has a graphical equalizer) the result of mixing down in my system, and the result had the high frequencies attenued ... (?)
 
Don't apologize for the English; I wasn't really using mine very well. :D

What I meant is that if the overall level of all the tracks playing together in the multitrack is too loud ("hot"), then your mixdown will lose fidelity because of "clipping," which is really just distortion attained from mixing down too loud.

You can tell by looking at the little VU meter at the bottom of the screen... (red [or green] line). If it's always causing the red lights on the far right to go off, it's "clipping," and you need to turn the overall volume down, or turn down the volume of each track by the same amount until the red light is not going off anymore.

NOW HERE'S THE CONFUSING PART... :confused:
If you took a song and just put it into track 1, how did you "mixdown?" Was there some other track you mixed with it? By simply clicking on the .wav file so that you were looking at it in edit view, nothing should have changed at all.

I'll be happy to help you check your system if I can, but keep in mind that when you convert the file to MP3 for me or somebody else to check, it'll change it, so I'm not sure that a test of this kind will help you.

Oh, and answer this too, because I couldn't tell. DO YOU HEAR THE ATTENUATED HIGHS IN COOLEDIT after mixdown, or only when you play it in WINAMP? Even if the EQ is disabled in Winamp, there are about 50 different output plugins in Winamp that can drastically change the sound, and you won't even know if you have them unless you check in winamp under the "output" options. Obviously, if you're only hearing the problem when you play the file in Winamp, it's not the file that's the problem...it's the Winamp.

Good Luck - I'll wait to hear from you.

Chris
 
Hello

thank you for your help Chris.
Now I understand what hot means :) It's not the problem, because I have the volume unchanged (0 dB), so the result should be the same that in the original song

You have reason, it has no sense to mixdown one only song, but it can serve as a test of the mixing down process.
I did it to verify that mixing down in my system attenues high frequencies.

The test consists on :
- get a song in WAV format (44,100 Hz, 16bit stereo) rich in high frequencies.
- Put it in the track1 of an empty multitrack session (44,100 Hz).
- Mixdown to a file, and save it in WAV format (44,100 Hz, 16bit stereo)

Then I have two WAV files : the original, and the result of the mixing.

I listen to them with Winamp, with the EQ flat to 0dB.
With the original WAV, I can apreciate good levels of high frequencies.
With the result WAV, there's a frequency above wich the level is almost null (as if it was filtered, but I didn't configure any filter)

I don't know if it can be due to windowing, or dithering, or something like this ... (?)
There are no sampling frequency changes, so there shouldn't be any antialiasing filter or something like this ... (?)

If you can do the same test, do you notice the same attenuation of high frequencies in your system ?
It's important to me to know it to know if it's problem of the configuration options in my system

thank you very much
Franky.
 
We don't even have to test it to tell you CEP does normally not cut off the highs. There is definately something weird going on if all you are doing is mixing down from one wav to another. With only one track in the original file the mixdown should be identical.

Hopefully somebody familliar with 2.0 can tell you if there are any settings that would cause that.
 
Well I'll be damned...

I have to admit, I did the test FULLY believing that it would make no difference whatsoever.

First, I used an MP3 file of a high hat track and put it in the mixer. I hit "mixdown" (I had no idea it would do anything at all b/c I've never "mixed down" a single file), but it did convert it into a wav file, which makes sense. The result was what you would expect. There were MORE highs in the .wav file, b/c an MP3 file cuts off everything over 16K. So far, the world was working the way that I thought it should.

Then, I too the wav file, and put it into the mixer by itself. I hit mixdown again (again, I was still surprised that it did anything...I mean, what was it mixing???), but it did.

I then listened to the original wav file and the "TEST" wav file in Winamp, and the test file had less highs than the original. I have no idea why. I mean, on tape, it makes sense...you introduce noise, which uses up ions that could otherwise be storing high end "music." But digitally, it doesn't make sense to me at all.

I'm really glad you posted this, because I've got a bad habit of mixing down drums to a single file (just to make them easier to work with) before I mix them with everything else. This little exercise is telling me that I'm losing some fidelity, even in a digital environment where the bit rate doesn't change.


WEIRD.

I can't explain it, but it does appear to happen.
 
thank you all for your patience :)

it's a pity that Cool Edit makes this low pass filter ...
Perfection is always disturbed by any smallness

I hope we'll be able to configure CEP to avoid this filter
If I can do it, I'll post here the solution


goof luck with your music !! ;)
Franky
 
Wow, Chris. That's pretty scary. You guys should email Syntrillium and see what they say about it.

In the meantime, to avoid using the mixdown feature just record the final mixdown on a new track in realtime. If you set your windows mixer to record "What you hear" or "Wave" you should be able to record a new mixdown. Then export only that new stereo file.
 
I'm gonna try the test later today when I get some free time. If I get the same result, I'm gonna post a link to this thread on the Syntrillium board.

Interesting stuff. And scary, like Tex says. But there's gotta be something we're missing here. I mean, nobody over on the Syntrillium board has noticed or mentioned this - and it *would* have got noticed by now by that crew...
 
Guys, I think the problem here lies in the mixdown settings. If you go to Options/Settings/Multitrack and then go to "Dithering Options" for 16-bit mixdowns you'll notice a "low-high quality" slider.

If you check the help file in CE you'll see that a lower setting will speed up the process, but will result in some high-end rolloff.

Now, according to the manual, this should only be a factor when you're adjusting the *sample rate*. So, Superblade & chris: when you did these tests were you going from 48khz to 44.1, perhaps?

No need to panic.:)

Chris
 
groucho said:


No need to panic.:)

Chris
Click to expand...
*wringing hands* ... BUT, dithering??? I mean... AAAARRRGGGHHHHHHHHHHH. THE WORLD IS ENDING!!!!!!!!

(okay, so I set the "quality" to 999, that should fix it, right?). All in all, this is an excellent thread that has taught me some stuff.

Cool.

bye,
chris
 
But try it. I've done it 2X since I originally did it, and I've screwed around with every parameter I can find that would possibly matter. (this is considered "fun" for me...seriously).

Anyway, everything's been 44.1, so the dithering shouldn't matter, right? I would be interested in the outcome of DOBRO's test.

And yes, I'm probably missing 12 things...I'm definitely no expert. Kinda sucks though...I'm doing a tune now that has 8 drum tracks, and I'd like to mix them down...but I don't dare. LOL. Seriously...the loss is slight, but it was there on mine.

ANYBODY ELSE???
 
You must log in or register to reply here.
Back
Top