hiss removal on a good quality VHS tape?

[lordsmurf]

The page is updated now. Go to it again.
http://www.digitalfaq.com/dvdguides/restoration/audio/soundforge/soundforge.htm

The December 2008 filter update has more than 100 presets in it, and will include all the ones mentioned here.

[zee944]

The thing with my tape is that its quality is very crappy... And that it has embedded subtitles... So I would prefer not to use it...

I didn't mean to use it... at least not for authoring your final DVD. The shitty image would only be for syncing purposes. If your good quality audio and good quality video are from different sources, they very likely will not match as easily as you may think. Not even after you do the NTSC->PAL speed-up thing. You'll need to do some syncing job anyway. So my advice is to transfer your VHS tape to DVD through a DVD recorder with high audio quality (uncompressed LPCM) and low video quality settings. Doesn't matter how bad the video quality is.

[lordsmurf]

I do a lot of mismatched "fix" jobs, often from many sources. It's not easy as a one-click, but it's rarely as hard as you suggest either. The big issue is to handle the video in 5-10 minute sections, watch it for sync issues. If one source has a gap, work around it so that such a flaw does not affect the output of the final fixed version. That might mean cutting something, or it might mean a spot of black video or silent audio. It needs to be dumped into a big timeline program, like Adobe Premiere.

I don't necessarily have a handle on this project yet. I just showed some examples of audio fix work, and peeked at a video sample that isn't too bad off. I'll have to read this again tomorrow, going to bed now.

[lordsmurf]

Should I normalize at 80% after all the filters? (thats what the guide says)
Something else I could do to the track? Or thats as far as it'll go in quality?

To be honest, I didn't imagined it could sound so good considering the source.
You have basically recreated the "echo" right? Added some reverb perhaps?

If this clip represents the overall audio file, then no, I'd leave it as is. It needs no more normalization. The only augmentation I'd make is to lower the STEP 1 normalize to a lower %, but ONLY IF the filters from STEP 2-4 push the audio into a "red line" type distortion. I hope that made sense for you.

I tried a few more filters, and no, I think that's about the peak performance you'll get from SoundForge. Other manual settings (creating new filters) were also experimented with quickly, but no gains were to be had. I opened it in a few other programs, and the improvements were not noticeable, but did take a lot of time to set up -- not worth it. This sort of audio correction work is where SoundForge excels.

I did not add anything to the file, no. My procedure was outlined in full in those 4 steps. All I did was remove enough muffle and noise that the natural echo returned, uncovered again. I don't know anything about this scene's video, so I'll just trust you know it's supposed to sound that way.

[lordsmurf]

I would save the audio as 48kHz stereo prior to any corrections, and then leave it that way. I actually did not change the English audio first, I left it as 44.1kHz stereo, saving it to 48kHz only when done restoring it. The Spanish one was saved prior to any work.

The Spanish one was better. It has a richness of tonal values that I'm afraid to damage too much. A little hiss can be removed with little to no detriment. Remember that most all televisions "hide" some of the hiss at very low levels, so what little is still left over here will likely only be detectable on a computer, when the volume is loud, and you're paying attention to it or trying to hear problems.

Steps used for Spanish:
1. Open audio in SF, save as stereo 48kHz prior to work, open new file to work from
2. Normalize to 60%, audio too low to work with otherwise, hiss does not become apparent until this is done.
3. Use Hiss Cut Alt 0 filter, removes some hiss without being too damaging

Again, attached as a FLAC.

spanish-sample-stereo-lsfix.flac

[sideskroll]

I would save the audio as 48kHz stereo prior to any corrections, and then leave it that way. I actually did not change the English audio first, I left it as 44.1kHz stereo, saving it to 48kHz only when done restoring it. The Spanish one was saved prior to any work.

Hello lordsmurf, so the spanish track was almost "as good as it gets" right? cool then, I'll just remove some hiss like you said. I see that you saved the file in stereo. Should I do the same when upping to 48kHz or should I just leave it in mono? The sample I uploaded (as the original track) is mono.

f this clip represents the overall audio file, then no, I'd leave it as is. It needs no more normalization. The only augmentation I'd make is to lower the STEP 1 normalize to a lower %, but ONLY IF the filters from STEP 2-4 push the audio into a "red line" type distortion. I hope that made sense for you.

Sadly, the other track(the one you worked on earlier) sounded great overall. But, in some scenes the audio was unbearable, I mean, there are some scenes in which there's a record player sounding in the background (LP, and you know how they make LPs sound in films most of the time, the scratching and whatnot) and to make it worst the music has trumpets in it so it sounds terrible. Like "cracking" or something like that, I was hearing it at mid volume and when that part came I had to lower the volume. Same whenever a woman screams or something that has a higher tone than the mans voice in the sample that you worked on. The speakers sound as they were about to burst or something.

Remember that most all televisions "hide" some of the hiss at very low levels, so what little is still left over here will likely only be detectable on a computer, when the volume is loud, and you're paying attention to it or trying to hear problems.

Does this also applies to home theaters? :S I always watch DVDs with mine.

You know what? I'll try to copy my VHS tape to DVD using a recorder. And we'll filter that track. Because the worst part of the english one is that it was saved in 28 or 32 kbps so to make things worst, it wasn't just the crappy sound, it was the crappy compression.
After that I'll see how to sync it to the video.
Thanks again for all your help. Hopefully I'll be uploading a sample of the new audio track later today.

PS: Should I upload more samples? Or is it okay to work only with such a reduced portion? I wouldn't like to get the same error as in the english track.
Or did they appeared due to the increased High notes restoration?

[sideskroll]

Hello again. I managed to copy my VHS tape to DVD and extract the audio.
Here I attach the new sample (work your magic lordsmurf )
Thanks again.

PS: I tried to attach it already converted to stereo but couldn't because the file size was too big, and I didn't wanted to upload in FLAC because I thought that some of the roiginal quality would be lost and therefore the filtering process would be affected.
Anyway, I understood what you told be about saving it in stereo 48 khz prior to filtering so thats no problem. I should do that with both audio tracks right?

PS2: this is kinda weird, like I told you before, the avi that I have is PAL (25 FPS) and the VHS (therefore the new audio) is NTSC (or so I thought) but when I listen to it watching the avi they sync perfectly, and when I tried changing the framerate it was a bit too fast... How can this be?

english.vhs.sample.wav

[zee944]

PS2: this is kinda weird, like I told you before, the avi that I have is PAL (25 FPS) and the VHS (therefore the new audio) is NTSC (or so I thought) but when I listen to it watching the avi they sync perfectly, and when I tried changing the framerate it was a bit too fast... How can this be?

Probably your NTSC tape already runs with PAL speed. A messed up conversion with blended frames, I guess.

[sideskroll]

PS2: this is kinda weird, like I told you before, the avi that I have is PAL (25 FPS) and the VHS (therefore the new audio) is NTSC (or so I thought) but when I listen to it watching the avi they sync perfectly, and when I tried changing the framerate it was a bit too fast... How can this be?

Probably your NTSC tape already runs with PAL speed. A messed up conversion with blended frames, I guess.

So, will that be a problem when I try to synch both audios and video? I've noticed also that the DVD I Made from the tape has almost twice the number of frames than the avi when I load them both into VirtualDub.
Thats giving me a hard time to sync, because I can't use delaycut properly because of the difference in frames... What could be causing this? I mean, I've seen some movies start with delay of X frames and end up with XX frames of delay, but in this case the ovie starts and the difference is lets say 133 frames, and at the end of the movie, the difference is something like 333897 frames!!!!!

[zee944]

I mean, I've seen some movies start with delay of X frames and end up with XX frames of delay, but in this case the ovie starts and the difference is lets say 133 frames, and at the end of the movie, the difference is something like 333897 frames!!!!!

This 300+ thousand frames is way off. This just can't be, there's a bug somewhere.

If it's really a messed up conversion with PAL speed, then you "only" have to deinterlace/decimate your NTSC source (dropping/blending frames) to have about the same frame number as your PAL source, leaving the audio stream totally intact. Only in case you want to use my method for syncing, of course.

[sideskroll]

Could you tell me your method zee944? I want the NTSC audio track to sync with the PAL video source...
The thing that I've noticed is that for every image, the NTSC source has more frames composing it. LEt me try to explain: for example a credit (at the beginning of the movie) is composed by 100 frames (from black to black) in the PAL avi.
In the NTSC m2v the same credit is composed by 200 frames (like: black, gray, credit appearing, credit, credit dissapearing, gray, black) while the PAL source is something like (black, credit, black)
I don't know if I'm explaining myself correctly... Hopefully I am..
Thanks.

*Edit: I don't know if this helps or not. But I've just compared almost frame by frame and the differences in frames seems to be increasing. I mean, the first frame of the movie, a black screen, has a difference of 226 frames between sources. and the last frame of the movie has 5066 frames of difference between sources. what could be causing this and how could I solve it?
To be honest, I don't care about the video because the video will be from the avi source, but I can't sync the audio this way because there's no reference point.
This thing is proving to be way harder than I expected

[zee944]

I understand everything you've said so far about the video.
My method is (as I alluded to it before) to compare the two videos streams instead of the audios. I do it on per frame basis. It doesn't mean you have to go through all the frames - you just have to find all the missing and the plus frames in the poor quality video. Then on every critical spot you have to add or delete something from the audio stream to sync it to the reference video. This is a very reliable solution if the two videos are in the same format.

In your case, something is wrong with the NTSC. It's interlaced, which is OK, but due to the messed up PAL->NTSC conversion, it has probably ghosted frames too. And if it really has over 300,000 frames, then probably it's also bobbed (fields converted to frames). We have to make it to have about as much frames as your reference video has.

Measure three things:
- A short scene which is about 100 frames long in your reference video.
- A longer scene which is about 1000 frames long in your reference video.
- Measure the largest portion of the movie (practically almost the entire movie) which you can really tell they had the "same" first and last frames (reference points) in both videos. This is what you've done but with inaccurate manner. I mean, don't use a black frame as reference point. Let the first ref. point be a frame of an easily recognizable scene in the beginning of the movie, etc.

I'm expecting something like this:

short scene: 96 frames vs. 210 frames (good video vs. bad video)
longer scene: 1085 frames vs. 2385 frames (good video vs. bad video)
largest portion: 130,512 frames vs. 285,430 frames (good video vs. bad video)

Then it'll be much easier to find out what happened to your video. Also check what fps your videos running at, I'm just assuming your PAL video is 25 and your NTSC is 29.97.

[sideskroll]

Ok, heres the deal:
(Im guessing here that by "reference video" you mean the one I copied from my VHS right?)
Under that asumption, I'd have to say that I didn't do any compression or transcoding whatsoever. I recorded a long time ago my video from an original (so that leaves transcoding aside)Now, I don't know if this has something to do with anything, but this movie was STV (straight to video) I'm not sure if thats telecine, or dropframe or I don't know... Just wanted to give as much detail as possible.
What I did was, I connected my VCR to the DVD Recorder through mono RCA cable (VCR only had one audio out jack) there was no capture involved and I'm not sure if the recorder does some kind of framedrop or deinterlacing or something (I don't think it does) the poitn of all this jibberish is that there was no capture at all, no settings no nothing. Just straight from the VCR to the Recorder and to be honest I don't know the procedures of the recorder to, well... record :P Is the resulting DVD interlaced? Deinterlaced? dropframe? telecine? or what?
Do you think that perhaps the problem is from the avi? By the way, there are som scenes in the DVD (the one I recorded) that, I don't know how to call it, "flip" or become fuzzy. And thats because the tape is bad in those parts. It doesn't stop or anything, but the sound does get a little "funky" for a second or so, although It shouldn't affect the lenght because it keeps going, no slowdown or anything. Could this be a problem? I didn't care too much though because like I said, I only wanted the audio from my tape.
Sorry for the long answer, I just wanted to try and give as many details as I can.
Now, to answer your questions:

which you can really tell they had the "same" first and last frames (reference points) in both videos. This is what you've done but with inaccurate manner. I mean, don't use a black frame as reference point. Let the first ref. point be a frame of an easily recognizable scene in the beginning of the movie, etc.

I know thats what I'm supposed to do, but its kind of hard, because like I stated before, lets say that the first scene is that of an apple ok? In the avi (good video) theres a black frame, a "grayish" transparent frame and the apple in all its fruity glory, same thing when the scene is changing.
Now in my m2v (bad video, good audio) the same scene goes like this: black frame, grayish frame, lighter grayish frame, even lighter grayish frame, transparent apple (x5), apple completed, transparent apple (x5) etc...
So I'm not sure which frame to select as the starting point.
Hopefully you understand what I'm trying to say here...

Also check what fps your videos running at, I'm just assuming your PAL video is 25 and your NTSC is 29.97.

Thats the first thing I did with gspot (should I use a different tool?) you're correct. avi is running at 25 and m2v is running at 29.97 fps.
Again, sorry for this huuuuge post :P
Thanks in advance.

[zee944]

(Im guessing here that by "reference video" you mean the one I copied from my VHS right?)

reference video = your video with good image. Your AVI file, I guess.

Do you think that perhaps the problem is from the avi?

Your AVI is probably alright.

And thats because the tape is bad in those parts. It doesn't stop or anything, but the sound does get a little "funky" for a second or so, although It shouldn't affect the lenght because it keeps going, no slowdown or anything. Could this be a problem? I didn't care too much though because like I said, I only wanted the audio from my tape.

It's a different problem than the one I was talking about, but it can be a problem, too. First, as you've noticed, it's noticable in the audio. Second, you may have dropped or added frames at these spots. More syncing job for you.

Now in my m2v (bad video, good audio) the same scene goes like this: black frame, grayish frame, lighter grayish frame, even lighter grayish frame, transparent apple (x5), apple completed, transparent apple (x5) etc...
So I'm not sure which frame to select as the starting point.

I thought you have figured this out by now Let's say your movie goes like this: we see black screen, then an apple, than a pear. You want to find out how long is the apple scene. When the apple starts to appear, let's say that's the first frame. Following this logic, when the first transparent pear appears, that'll be the last frame. That's it. It'll be accurate enough for us.

This transparent thing is must be an error in transfering. Maybe it has something to with the format compatibility, or the cable used, some sort of format settings, anything. I doubt your tape actually plays like that. By the way, did you record your tape with uncompressed audio?

Gspot is OK.

[sideskroll]

I thought you have figured this out by now Let's say your movie goes like this: we see black screen, then an apple, than a pear. You want to find out how long is the apple scene. When the apple starts to appear, let's say that's the first frame. Following this logic, when the first transparent pear appears, that'll be the last frame. That's it. It'll be accurate enough for us.

Yes, I understood that. I think I didn't explained myself... I'm measuring with the title credits ok? Because I find It a lot easier (If I shouldn't do it that way please, tell me so) what I would like to know is how to recognize the first and last frame? Lets say I'm measuring with the credit "Tom Cruise", what I've been doing is to take the first frame after fully black screen (when the letters kinda start to appear in the background) as the first frame of "Tom Cruise" ok? Then, as last frame, the last frame before the black screen appears (when the letters kinda dissapear but you can still see them) Is this correct? Is this what you mean? I've been doing it this way because it is much easier than to try and sync with the video itself.
Please tell me if I should do it the other way (with the movie instead of the credits) because I've already started writing down the times beginnning with the credits
Thanks again.

[sideskroll]

short: avi 148 > 243=95 frames
m2v 260 > 380=120 frames

longer: avi 148 > 1156=1008 frames
m2v 260 > 1521=1261 frames

longest: avi 148 > 133676=133528 frames
m2v 260 > 167099=166839 frames

Those are the times (done the way I explained in my previous post)
I think I know what the problem is,in the m2v file there are a lot of "dissolvences" (I don't know how you call it in english) It's when instead of cutting you black from one frame to another, you "smoothly" change scenes through a "disolvance"...
On the other hand, in the avi the same scenes just change from one to another with a clean cut.
I meassured the frames by going to the last frame with an image prior to a black screen. Is that correct? Is that what I should've done? Thats what I've been trying to explain in all my previous posts. There are no disolvances in the avi so I need to know what to consider the "last frame" in the m2v file.
Thanks for all your help. I already estarted to hate this movie :@

[zee944]

Hey, these figures are perfectly normal. If you take a closer look, these parts in your M2V file are 1.25 times longer in frames. This implies proper 23.976p-->29.97i telecining.

What you've written about the frame numbers before was misleading. You made it seem the credits use twice as many frames etc., but most importantly you said 333,897 frames difference, which was probably a typo.

Lets say I'm measuring with the credit "Tom Cruise", what I've been doing is to take the first frame after fully black screen (when the letters kinda start to appear in the background) as the first frame of "Tom Cruise" ok? Then, as last frame, the last frame before the black screen appears (when the letters kinda dissapear but you can still see them) Is this correct? Is this what you mean?

Not really. You should've taken the first frame of "Tom Cruise" fading in; and the first frame of "Tom Cruise" fading out. Or you could've taken the last frame of both. Instead of this, you've taken the first frame of fading in, and the last frame of fading out. That's no good. Anyway, it doesn't matter anymore. Take a DGIndex. Open your .M2V file, then Save Project (results in a .D2V) file.

Then create the following AVISynth script with your Notepad:

LoadPlugin("DGDecode.dll")
LoadPlugin("Decomb.dll")

MPEG2Source("your_source.d2v")
AssumeTFF()
Telecide(guide=1)
Decimate(cycle=5)

#Lanczos4Resize(720,576)
ChangeFPS(25)	# ChangeFPS(25) for improper source (blended PAL-->NTSC conversion), AssumeFPS(25) for proper source

Get a Decomb.dll (v5.2.2 or later), and put DGIndex's DGDecode.dll into the same directory as your .AVS and .M2V file. Pay attention that the .D2V file (you can open it simply with Notepad) points to your .M2V file. Open the .AVS file in VirtualDub, and we'll see what happens.

[sideskroll]

Thanks, I'll try it overnight, see what happens.
What are we trying to accomplish here? render the number of frames to the same/near the same as in the avi?
Would that help me to sync the freaking audio??? (I've already tried changing from 29.97-23.976 to 25 and even from XX.XX to 23.976-29.97) the weird part is that I accomplished the best sync when I changed the framerate treating the wav as if it would've been encoded in 23.976 to 25.00... It started perfectly in sync and at about 20 minutes into the movie the audio began to lag behind...
Another thing that will become a headache is that the avi has the beginning cutted short as well as the ending. Although that should be no problem if we accomplish the frame conversion (I could use delaycut after all the editing to sync the delay)
Thanks again zee.

EDIT* Should I delete the previous version of decomb? (I think its 5.2.1) or is it okas if I keep both? (I'm not sure if rebuilder is working with 5.2.3, the one I got. Nevermind, I think I'll just rename the other one for the moment )

[sideskroll]

Ok, although the number of frames is lower now, the difference is still there, for example from point a to point b (measuring by the first frame of every credit like you told me) there are 113 frames in the m2v file and 108 in the avi file. Now, from the first frame of the movie to the end there are 133695 frames (m2v) 128275 (avi).
Whats next?

[zee944]

What are we trying to accomplish here? render the number of frames to the same/near the same as in the avi?
Would that help me to sync the freaking audio???

Yes. Why else we would do all this? Of course you can go another way, and sync the audio without looking at the bad video. Then you can skip all this crap.

Another thing that will become a headache is that the avi has the beginning cutted short as well as the ending. Although that should be no problem if we accomplish the frame conversion (I could use delaycut after all the editing to sync the delay)

Delaycut won't be useful here. Don't expect the two versions to match perfectly (not even after you adjusted delay). Besides, I've assumed you've transfered your tape with uncompressed audio. Delaycut handles only AC3 and DTS, as far as I remember.

[zee944]

Ok, although the number of frames is lower now, the difference is still there, for example from point a to point b (measuring by the first frame of every credit like you told me) there are 113 frames in the m2v file and 108 in the avi file. Now, from the first frame of the movie to the end there are 133695 frames (m2v) 128275 (avi).
Whats next?

Then delete the "ChangeFPS(25)" line. It'll probably be 128,219 frames long, which is fine.

[sideskroll]

Delaycut won't be useful here. Don't expect the two versions to match perfectly (not even after you adjusted delay). Besides, I've assumed you've transfered your tape with uncompressed audio. Delaycut handles only AC3 and DTS, as far as I remember.

There was no option to record with uncompressed audio I just had to settle for HQ recording
Yes, delaycut only handles those file types. But that will be mygoal right? I'll end up transcoding the audio to AC3 for the final muxing.
Is there a way to force the audio to sync? Maybe I could insert a couple of seconds here are there, depending on the final number of frames. Whats the duration of a frame?
Will the audio match the video now? I mean, I've decreased the number of frames (therefore, the duration also) of the video, so wouldn't it be logical to assume that it won't match its own audio track now?

LoadPlugin("DGDecode.dll") 
LoadPlugin("Decomb.dll") 

MPEG2Source("your_source.d2v") 
AssumeTFF() 
Telecide(guide=1) 
Decimate(cycle=5) 

#Lanczos4Resize(720,576) 
AssumeFPS(25) for proper source

Should I delete the "AssumeFPS" line ? VirtualDub tells me it doesn't know what "for" means in the 10th line.

[sideskroll]

Ok, now the m2v file is shorter than the avi...

Short
M2V: 209 > 317= 108
AVI: 148 > 256= 108

Larger
M2V: 209 > 2631= 2422
AVI: 148 > 2569= 2421

Largest
M2V: 209 > 133680= 133471
AVI: 148 > 133676= 133528

[zee944]

It seems fine now. Theoretically, you could start working on the audio part.
I still don't understand what do you want to use Delaycut for.
Duration of a frame is 1/23.976 or 1/25 seconds.

Something you said about the videos earlier wasn't true. Can't both videos have the same running time (you stated this when you told they're almost in sync) while both has 133,500 frames and different fps (23.976 vs. 25). It's up to you find out where you've made your mistake.

[sideskroll]

It seems fine now. Theoretically, you could start working on the audio part.
I still don't understand what do you want to use Delaycut for.
Duration of a frame is 1/23.976 or 1/25 seconds.

Because the video from my VHS coopy has all the FBI and distributor crap before the actual start of the movie and the AVI doesn't. It just starts from the beginning of the credits.

Something you said about the videos earlier wasn't true. Can't both videos have the same running time (you stated this when you told they're almost in sync) while both has 133,500 frames and different fps (23.976 vs. 25). It's up to you find out where you've made your mistake.

To be honest, I don't understand what you're trying to say here. I stated what when? :S The only time bot videos where in sync for 20 minutes more or less was when I converted the freamerate of the audio to 23.976. I'm not sure why, but they were.
Ok, so let's say that I want to sync the audio now, what should I do? I mean, do I have to change the framerate of the audio track? From 23.976 to 25 right? Is this the original framrate of the movie? 23.976? (audios are completely out of sync right now, the avi one is going much faster then the one from my vhs, even if the beginning is in sync it starts to lag behind almost inmediately) I haven't saved the m2v file with the avisynth scripts, is this ok? I think they were only meant to allow me to close the gap between videos correct?
Thank you for your help.

[sideskroll]

Maybe I'm doing something wrong here... But it is impossible to sync.
What should I do? What was the point of fixing the number of frames? It only modified the video... Or maybe I'm overlooking something here? How should I go about fixing this sync issue? Thanks again.

[sideskroll]

At last ! I managed to perfectly sync the movie with my VHS audio. Sadly, the AviSynth didn't work zee I had to do it "by hand" comparing sound waves and cutting and adding time (as well as changing the tempo). But, the good news (for me that is ) Is that it worked and now I am the proud owner of an avi with 2 audio tracks YAY! (I know it doesn't mean sh*t to you but c´mon! give a guy a break here, I've been working on this pet project of mine for almost 3 weeks to the day...)
Now, I would like to request the invaluable help or the master restorer "lordsmurf" for the final steps prior to authoring.
Could you take a look (or in this case, a "hearing" ) at the sample attached? It is much, much better than the previous english track you heard. There's some hiss though... But it shouldn't be a problem. There are no artifacts of compression (unlike the previous one that sounded metallic)

The last step would be for you guys to help me out with my avisynth script to encode to mpeg2. I will be using CCE SP 2.70 for that so the quality should be quite good on the output.

Thank you all for your time and help.

sample-english.wav

[zee944]

Sorry, I had no time to write. I'm surprised you didn't find out what to do. You'd only had to speed up the audio by 25/23.976, then sync by looking at the videos, which is much easier.

CCE is quite a bad encoder.

[sideskroll]

Sorry, I had no time to write. I'm surprised you didn't find out what to do. You'd only had to speed up the audio by 25/23.976, then sync by looking at the videos, which is much easier.

CCE is quite a bad encoder.

No, I knew what to do. It just didn' sync... I had to fix it by "ear" or by comparing one stream to the other. The video had nothing to do as a matter of fact. I had to speed some parts up, while leaving a couple as they were and cutting and expanding other parts as well. The good part i that now it is COMPLETELY in sync. From the first frame to the last one, not one ms of delay or nothing.
Thanks for your help.
How come you say that CCE is a bad encoder? I've always read that its the best. or one of the best at least... Which encoder would you recommend?

[brassplyer]

Here's an example clip of a hissy source:

I played with Soopafresh's sample a little.

http://www.fileden.com/files/2008/3/11/1809238/dehissed3%20mono.wav