Syntrillium Backup Forums :: View topic

JohnK · Posted - Thu Jun 28, 2001 3:59 am

I am using ce2k to record from voice recordings being streamed from a RealAudio server. To capture the sound I am using the "What you hear", on the record sheet of the Windows volume control. The recorded sound has great volume when recording the .wav, however after saving it in RP G2 format and posting it on the web, it seems to lose some volume. Is there a way of controlling this loss of strength, a setting in ce2k. I am recording with a sampling rate of 44,1 and 16bit stereo, and then saving as voice only/mono. Thanks for any help you can provide. John

Syntrillium M.D. · Location: USA

Hi John. Well, the obvious problem here is that you're re-compressing an already compressed audio file. This, in terms of quality reproduction, is not going to yield great results.

However, this should not necessarily effect the output volume. Try Normalizing your file *before* you encode to RMG2. If you've done that and it still doesn't do enough, I would use some Hard Limiting (for an overall loudness boost) or perhaps a little compression with some output gain boost.

----Syntrillium, M.D.

motorhead_6 · Posted - Fri Jun 29, 2001 11:09 pm

Did you try hard limiting?

JohnK · Posted - Tue Jul 03, 2001 7:09 am

Sorry, I am brand new to all of this. What is hard limiting? John

Syntrillium M.D. · Location: USA

Hi John. An easy explanation is to think of the hard limiter as a digital 'ceiling' with a boost! Essentially, you feed your signal input to the limiter which will essentially bring up the overall loudness of your file, while at the same time, preventing you from clipping (set by the 'Limit Max Amplitude' field). For most program material, a limit of -.1dB (default) and boost of +6/+8 is usually very good. You might want to adjust the release time, as I often think 100ms is too long for certain things, particularly 'some' finished music masters.

---Syntrillium, M.D.

JohnK · Posted - Tue Jul 03, 2001 8:41 pm

Thanks for the explanation on hard limiting. I am learning so much here. Actually, I tried the normalize and it did the trick, incredible boost in sound, especially during the times when the speaker dropped their volume. It is very consistent now. I have to tell you, this ce software is great, but I think this forum is more than icing on the cake. Thanks everyone for your help. My training site is really cooking now with all the new sounds. John

Syntrillium M.D. · Location: USA

Hi John. Well, that's great that you found Normalize. Keep in mind that normalize is not limiting or compressing, it is simply raising the highest peak in the file to just below 0dB (with 99% normalize...That would mean that if your entire file was registering around -12dB, but you had one sharp snare hit that registered around -2dB, if you normalized roughly 99% your entire file's amplitude would have only been raised 'about' 2dB...

(incidentally, I typically recommend that people normalize to around 96-98%, just to be safe)

So, you still (for future reference) might consider using 'some' compression to even out peaks, and THEN normalize, or experiment with the Hard Limiter - both of which can actually control and increase the apparent loudness of your mix.

---Syntrillium, M.D.

Aqualung · Posted - Thu Jul 05, 2001 5:09 pm

I can't find this "hard limit" feature, did find "normalize". I am using ce2k....
<ah, I found it's a plug-in>

BTW, I'm assuming that it's an either/or use, i.e. you _don't_ first use hard limit, and then normalize?

TIA

Edited by - Aqualung on 07/05/2001 7:25:07 PM

RobertM · Posted - Fri Jul 06, 2001 6:05 am

Aqualung,

You can certainly run any number of normalize and hard-limit transforms on a file, but if you set the parameters of the hard-limit correctly, there would be absolutely no point in normalizing afterwards. All normalizing does is amplify the entire file by a set amount, such that the LOUDEST sample point in the file is at a specific value below the maximum allowable value. Hard limiting does the very same thing to the loudest sample point, so normalizing to that same value will do nothing. You MIGHT do a normalize if it turned out that you had set the limit too low, and wanted to boost eveything up to correct. But a better way would be to 'undo' the 'hard-limit' and do it again with the correct parameters selected. This is because each transform adds low-level artifacts (noise) to your wav file, so fewer transforms=better quality. Artifacts may not be an issue for you if you are doing the processing at a high enough bit rate. You MIGHT notice them at 16bit, but probably not at 32bit.

Aqualung · Posted - Fri Jul 06, 2001 9:12 am

FreakboY · Posted - Sun May 25, 2003 9:14 pm

hard limiting?

is there any tutorial out there...
i really love that noise reduction
tutorial on this site! is really great!!

is there anything like it for hard limiting!??

thanx!

ozpeter · Location: Australia

Not that I'm aware of - have you read the help file for that process? There's not a lot one could add to it here. Do you have a specific problem with it?

- Ozpeter