|
Selecting correct sound card for software editing...
|
View this Thread in Original format
| bachatu |
Alright guys,
Here is my question my issue/question:
I understand that when you do most editing with software such as cool edit or soundforge, that they use processor and memory to perform functions, such as normalizing, etc.
I just purchased a M-Audio audiophile 2496 (24bit) sound card. I will be recording in 24bit (finally), and I will be obviously using the 2496 sound card to do so. However, Ive noticed that this sound card has RCA outs (Left & RIGHT) and RCA ins (Left & Right) and midi in, digital in & out. Now, I want to continue using my monitor desktop speakers (PC speakers) and earphone jack out to my headphones to do my editing, because of convinience.... and i dont want to always have to turn on my dj mixer on to have to listen or playback tunes on my PC.
So I plan on utilizing my other sound card for playback (santa cruz... similar to sound blaster live and also have a sound blaster live oem card) and setting it for playback, while the other for recording (24bit for recording). Im wondering if that will affect performance on editing functions through the editing program, by using an inferior sound card. Does the process any of the editing functions through the sound card driver, or is it all done through memory and processor?
Im gonna check Radio Shack to see if they have a mini jack female to RCA male connector, to connect my PC speakers, so i can use my Maudio sound card for everything... recording and playback. |
|
|
| Dj Thy |
Sadly, the soundcard will affect editing. The problem doesn't really lie in the editing part, but in the way the program handles the soundcard.
Let's assume you made a recording in 24 bit 96 kHz (an example). The audiophile will handle this without a prob. Now assume you have configured your program's so it uses the output of one of your consumer cards. The program will see that that card isn't 24 bit compatible (and also can't handle the high samplerate). Because at playback the file really stays in 24 bit, is passed through your DAC and so gets converted to the analog signal you hear. The DAC's of the consumer soundcards can't handle the 24 bit stream so give an error.
There are programs that have kind of a protection against that. It converts the bitdepth to a format compatible with your output (so 16 bit with your cards). But then you lose the advantage of your 24 bits. And on good monitors this can really make a difference (an effect will sound different in 24 bits than in 16 bits for example, so you can make different judgements based on different situations).
So my advice is : use the consumer card for whatever else you want (games, multimedia...), but stay on the Audiophile while you are editing audio or making music... |
|
|
| bachatu |
Thanks for DJ Thy for the quick reply. Very helpful information and makes sense too. I will see if i can purchase that adapter so i can use my headset through my pc speakers.
I didn’t get to install it yesterday and test it... can't wait to get home and install it, do some 24bit recording! Definitely looking forward to it.... |
|
|
| DjJade |
| also... converting from a higher bitrate to a lower bitrate sampling causes noise that wasnt there before. soundforge has some extra tools to move the noise to frequencies that are out of the range of our hearing but... i donno the technicalities behind that...i just started recording myself |
|
|
| Dj Thy |
Hmm, yes and no.
Reducing bitdepth without any other precaution will indeed add nasty noise, but mostly on silent parts (fades and such). That is because you have too few bits to store the information contained in the extra part of the higher bitdepth.
For example 24 bit to 16 bit. Basically with bitdepth reduction, you'll only use the 16 first bits of the 24 bit signal to obtain your 16 bit sample. The problem is you still had information in those 8 extra bits. Without any further precaution, the 16 bit will be rounded to cope with the value of the 8 extra bits (the signal flow of that is represented by those 8 bits is too little to be detected properly by the 16th bit, so it detects erratically). And that produces bad results.
To remedy (well in fact it's not really a remedy, it just keeps the problem smaller) is dithering. Basically dithering is just add random noise, but in a controlled manner. Why in the name of god can adding noise improve the signal???
Well remember I told you the remaining signal was too small to detect properly? Adding noise will improve that. It just gives the signal just that little more to be detected right, so it can be reconstructed nicely, without showing the artifacts you'd normally have without dither. The trade off is you'll hear a little bit more noise on silent parts, but it's FAR less annoying than the erratic stuff you hear when you just reduced bitdepth.
So remember, if you are reducing bitdepth (you will eventually, for example when burning on cd, you gotta go to 16 bit, 44.1 kHz) always dither... Most programs have adequate dithering algorithms (remember to use 16 bit dither when dithering down to 16 bit. There is also 24 bit and 8 bit dither. Always use the bitdepth you are going to). And it should always be the last step.
When the dither is performed correctly, it's safe to say a well recorded 24 bit sample dithered down to 16 bit will sound better than the same sample originally recorded in 16 bit... |
|
|
| bachatu |
So dithering is the way to go, if you plan on burning the recorded wave on cd...
Is there any difference in quality of dithering between programs? For example, lets say Cool Edit Pro, Wavelab, Soundforge. Would the human ear be capable of detecting a difference in noise or sound quality by choosing a different program? |
|
|
| Dj Thy |
Only dither when you are reducing bitdepth. If you stay at the same bitdepth it has no point, you'll only add noise.
And yes there are several algorithms for dithering. When you have good ears (and a good monitoring system) you'll be able to tell the differences between dither. Better dithering algorithms provide noise shaping. With that, the dither spectrum is shaped so it still provides the advantages, but is less audible.
A very good guide about dithering can be found on izotope's site.
http://www.izotope.com/products/aud...ozoneguide.html |
|
|
| bachatu |
Nice link there, tons of information.. .I didn’t get to check everything out, only first pages. I will look into it with more detail later.
I have to mention though… I extremely impressed with the performance of the M-Audio audiophile 2496. All I can say is there was a great improvement. I tested it and just messed around a bit behind the deck and recorded a 60 min set at 24 bit. I went converted it to 24 bit to 16bit, selected to dither 1 bit to peak (rectangle). From there, I made a mistake in saving the wave as 128kbp mp3 fie, because it was set default in sound forge from me compressing one of my other sets a few days ago. I then recorded the mp3 onto a cd (uncompressed). Even though the file was compressed and uncompressed, I still noticed a huge difference in quality over my other consumer sound cards (specifically Sound Blaster Live and turtle beach Santa Cruz).
What is different and noticeable?
Well, the main thing is that the sound is more dimensional than before. It has so much depth to it and it overall sounds much more natural. It sounds like every instrument and sound is layered separate, playing live. Sounds like live music, very nice!
Anyhow, ill stop myself from talking, but just wanted to give a mini review on the card… excellent card and great investment, starting at $150, I think it’s more than worth it. |
|
|
| DjJade |
| yeah the audiophile is so smooth... thats the best way i can think to discribe it. bachatu you should use lame encoding. i think i heard better results after using it over the one built into soundforge |
|
|
| bachatu |
Actually, now that you mention it.. I realize that i dont use soundforge as my main encoder. I normally use cdex. But I cant remember now what i was doing with the program that I had saved something as mp3...i think i used it as a mp3 splitter, for like a long set... to put in two different cds. It works fine as a splitter....though this last time, it did not intend on saving my file as mp3, its just that when i saved it, default was set to mp3 and i wasnt really paying attention.
I also have lame... but i use it as a plugin for winamp, when I stream live sets... it encodes before it streams the info to the relay servers. |
|
|
|
|
