Curious

[John E. Cox]

I've been editing a lot of narration lately and have noticed that some voices (when looking at the waveform view) tend to be centered beneath the zero crossing point, while others might be centered slightly above the zero point. I'm sure there's nothing wrong with the equipment since the sinewave of a 1k tone is perfectly centered. Just curious if anyone knows how to explain this? Thanks.

John
CDC, Atlanta

[admin]

This is what it sounds like. The only thing I can think of is that it is coming in from your input signal while recording. It you are doing a digital transfer, check and see if your master has an offset. Maybe someone else can give you an idea how to do that digitally. :?

[geezer]

---I've never seen Microeditor induce a DC offset, but have noticed a small one coming from the product of many different converters via DAT machines, especially....

You will see the DC offset when amplifying the waveform view to the max, then performing a fadeout on noise: You'll see the noise "arcing" up or down to the center line at the fade out.

Small DC offsets are tolerable, large are not. I don't know of anyway to fix an offset within MicroEditor, but I have seen it in other software packages and hardware boxes....I think some of the RME boxes have a software interface that allows you to do a custom DC offset fix.

Correct me if I'm wrong here, but, if I'm not mistaken, an extreme low frequency digital filter will also remove a DC offset (a la the 2hz or whatever filter available in the Finalizer).

[Michael A. Phillips]

Reguarding the waveform, what you are seeing is the effects of absolute polarity. When the waveform is above the the center, there are more positive peaks. A "p" sound into a microphone produces a positive pulse with a positive (above the center line)wave. The loudspeaker moves air in the forward direction. Many pieces of audio equipment invert the polarity of the wave. That same"p" sound will make the loudspeaker move backwards first. That is why some of your tracks look different from each other. I am a believer that there is a definite sound quality difference between a voice tarck with its absolute polarity reversed. Hear for yourself. Take any track with more peaks below the line & invert the polarity.

[geezer]

----There is definitely a major difference between absolute polarity and DC offset....Although you will sporadically see a difference between the size of the waveform above and below the line, in my experience it is rare to see this as a constant factor and still have the wave correctly crossing the zero point at the line, which is what DC offset is all about.

If there is no DC offset, you will see a very quiet or silent signal intersect exactly with the zero line when you greatly magnify the waveform resolution. If there is a DC offset, then this signal will sit above or below the line....This is the sure-fire test.

[Michael A. Phillips]

In some instances the wave will be asymetrical. There will be no difference at the zero line. What I see very often is a track with many more positive peaks than negative. There was a Gates limiter that was used at many AM stations years ago that would cheek the waveform & invert it to make it positvie. On an XZ scope. this positve wave will be higher on the top. Using the micro-edit waveform it is very easy to see an inverted signal. I can send some examples if you like...

[Rich LePage]

I think the comment about using a filter such as in the Finalizer is correct. You do have to watch with TC Elec stuff though, as using the analog ins/outs, they are DC coupled, and so is the MTU i/o module for analog. You can get garbage in there real easy if not careful using an analog connection between say a TC Finalizer and the MTU i/o module. Some other gear also does this. You'll see it on the MTU meters, too.

I first ran into it when I was using the Finalizer to help condition some material I wanted to load into Medit at 22K for m/media.
The TC box's multi-band compression was very desirable for that purpose, used heavier than I might normally do for other work.
TC can provide a diagram of how to add some blocking caps to their unit.

A quick and dirty way to fix this (which I used though against my
philosophy of "less is more"!) is to take the TC digital output to a DAT deck (in my case it was a Tascam DA30 I think), put the deck in input monitor mode, come out of the thing analog (balanced) to the MTU analog inputs. That eliminated the DC the TC was introducing from its analog outs.
Sort of a quick way of setting up a buffer amp.... Yeah, more components, more electronics stages-- but solved the problem quicker than I could grab a pile of caps and a soldering iron to get the job done. And since the target was 22K anyway, I figured the penalty wasn't so terrible given the circumstances.

On the DC offset, if you are working with a WAV file, you can take it into Cool Edit Pro and remove DC offset. Unfortunately, Cool Edit won't directly recognize the SF format. I have asked them several times about making SF a recognized file type, they say they are considering it but nothing yet. (prob would not hurt if some others ask 'em too!) Cool Edit will also accept "raw" data. I have had limited success in bringing in SFs to it that way, but it is usually too much trouble, so my workaround if I run across DC offset with a SF is to open Medit, bring in the full SF,
save the whole thing as a WAV file, then take to Cool Edit Pro for
DC removal, saving it, then opening the result in Medit for working on it. Need mucho disk space of course, but that's not as much of an issue as it once was.

The downside of that is if you use seg or file flags and had them in the original file, you won't have them anymore once WAV'd etc. I use the file flags (still) all the time when recording stuff as markers (up to 99 of course). That way, they stay with the file
(except as noted above).
Often I just set up a temp project to work with when originating stuff at outside studios, then yank the drive and bring it back to my post studios for editing, often setting up a new project file
when I start editing-- so the markers are real impt to me a lot of the time.

I had quite a bit of trouble at one outside studio I use with a Ramsa digital console. I was coming out of it analog and then going to an Apogee a/d converter as a front end to an MTU system. I was getting files with DC offset in them. I also tried using the dig. outs of the Ramsa board directly to MTU i/o module dig. in (DAIO board) and got same thing. Ditto on using the Krystal digital input (either flavor). It was the Ramsa that was the prob. Might have just been that unit, might be all the Ramsa dig. consoles-- have no way of knowing.


I use that same system at a variety of outside studios, and other than with the Ramsa board, it has been pretty bulletproof. It has an MTU 2 chan i/o module with DAIO board, the unbal Krystal analog i/o, and the Apogee a/d all available. Have used with Sony, Neve, Soundcraft, Mackie, Trident, SSL and other consoles
with no problem-- excepting that Ramsa. I usually wind up using the Apogee though can easily swap to the MTU i/o module.

Reason for Apogee is that for cheap insurance, I tend to like having its soft limit function on during recording. For return to the studio's console though I nearly always use the balanced outputs from the i/o module. It seemed to me that setup was best way to go given using a variety of outside systems to originate voice track and other material, also to lay multitrack mixes or stems down into Medit. Gives a lot of options and flexibility, not to mention some redundancy, which is always nice to have when expensive talent's time clocks are ticking away on a session (not to mention the studio time clock, too)!

For a m/track mix often I will forego the Apogee and just use the MTU i/o module. (again my less-is-more philosophy I guess...)
The i/o mod. converters are very clean, the Apogee is as well.
And for my needs, I usually need balanced i/o, especially at most of the NYC outside rooms I use, though this approach still gives you the unbalanced Krystal i/o too if you need or can use that.
But balanced (with some pin 3 to 2 swap cables sometimes)
lets you just plug the MTU system in wherever the studio has a DAT deck or other 2 channel harness hanging around. I keep some XLR to bantam and XLR to 1/4" mil-type TRS plugs with it too for plugging into any of the patch bays I might run across.

Hope this helps-- Rich

[geezer]

Thanks, Rich-

I use Cool Edit almost exclusively for file conversions, which I only do 4 or 5 times a year, so had never noticed the DC offset function....Good tip....Maybe I'll bug them about SF files, as well. I hadn't even thought of it as a possibility until now.

[Rich LePage]

Cool Edit is about the best MP3 conversion program I've
run across. I had to convert about 6000 WAV files
a while back and it filled the bill quite nicely- and
ran real well as a batch processing tool. I use
Waves' Waveconvert for that too, but not MP3.


Like you I don't use it for much other stuff,
though I've had some success de-noising with it and
also doing some EQ in unusual situations. But
compared to Medit, I just find other editing programs
a real bear to use-- most seem very non-intuitive to
me. Guess MTU has spoiled us!

Rich

[clawson]

Again, folks...

CoolEdit 2000 WILL import all MTU files. You DO have to specify the data format but it DOES work right now without any re-coding from the Syntrillium folks.

Rich is correct about the MP3 coding, too. CoolEdit uses the Fraunhofer algorithm and it is the best on the market. I use it all the time with my MTU files.

Chas.

--
Charles Lawson
Senior Recording/Production Engineer
WETA Radio & Television

[Rich LePage]

Charles:
Will Cool Edit 2000 preserve the flags??
Especially file flags which go in the header?
If so I'd run right out and buy it, since I'm using Cool
Edit Pro which is (I guess) a little different.

I've worked on SFs in Cool Pro, but did have
some problems after processing one once
using Cool Pro when I tried to bring it back
into Medit. (besides of course the flags
are good bye).
I went back to the original SF and saved
it as WAV in Medit, then did same thing
to the WAV using Cool Pro, and that
went back into Medit just fine.

So maybe there are some diffs between
Cool2000 and Cool Pro I don't know about.
What's been your experience with the 2000
in terms of what you use it to do, etc?
Do you know if it has all the processing
stuff that Cool Edit Pro has?

Regards--- Rich

[clawson]

Rich asks:

> Will Cool Edit 2000 preserve the flags??
> Especially file flags which go in the header?

Yes. CoolEdit 2000 *will* preserve the file flags. Just to confirm this, I recorded a 24-bit SF while dropping file flags in on-the-fly. I then opened the 24-bit SF in CoolEdit 2000, renormalized the file (so that I could say that I had done some processing) and saved the effected file as a NEW 24-bit raw-pcm file. Microsound was able to make a segment from the new file without trouble and, when I clicked the "Read File Flags" command, they popped right up where they should be.

> If so I'd run right out and buy it, since I'm using Cool
> Edit Pro which is (I guess) a little different.

CoolEdit Pro is the fancier workstation-like version which can't hold a candle to MTU's speed and flexibility. CoolEdit 2000 is the basic two-channel sound file editor/processor which has some terrific algorithms for sample-rate conversion, broadband noise removal, comp/lim and the like. I use it extensively to process things that MTU software doesn't do or doesn't do as well (sorry, Dave). CE2K makes a wonderful companion to Microeditor and can freely exchange even the 24-bit file format. Download it right now. It's worth every cent and more.

Chas.

--
Charles Lawson
Senior Recording/Production Engineer
WETA Radio & Television

[geezer]

---I second the high opinion of CE2K....The MP3 conversion algorithms alone are worth the modest price....But, Chas., you seem to be saying that, unlike the info from Rich LePage in the other thread, that you CAN open a 24-bit MTU .SF file in Cool Edit!

Boy, you've made my day!....Are you reading this, Rich?

[Rich LePage]

Wow- who would have thought less $ and
better than the Cool Pro. I'm sold
based on what you said here.

And here I'd been using the Pro for the
same kinds of things you mention,
when I could have saved some $ though
I did buy Cool Pro a long time ago.

BTW, I never use it to edit-- it's way too
non-intuitive for me.
Off to Syntrillium for me-- thx for the input.

Rich

[Hessel]

Yes!
The CE2K works fine with 24 bit SF files. This saved my day. After import in the CE2K i removed
DC Offset and return to ME as a 24 bit file. The new file sounds perfect.

Hessel

[clawson]

I'm glad to hear that others are enjoying success with the CE2K method. If I'd realized that this workaround wasn't common knowledge, I would have posted it here a long time ago. Sorry about the delay!

Best to all,

Chas.

--
Charles Lawson
Senior Recording/Production Engineer
WETA Radio & Television

[Bill Odom]

Any one found a way to get cool edit to save a sf1 ( the raw method seems to delete the header of the sf)that can be read by microeditor besides save as wav and import as wav ?

thanks
bill

[Hessel]

To processing a 24SF in CE2K, i do in this way:

Open file as... command > i choose a SF file >
next appears two windows > choose 32bit float resolution and sample rate.
Next appers Raw data window > choose "32bit IEEE float (0.24)" and "Offset input data by 0", the "Create .DAT header file on save" is NOT checked.
I never select all waveform to processing, i exclude the file header from the selection (you will need this header intact). After do all processing use the "Save" command, not "Save as...".
ME will open the prcessed file. If this file was in a project session the file will to play with the new processed sound.

Sorry for my poor english

Hessel

[Bill Odom]

great
it was a a 16 sf not a 24 but the ideas the same
"don't select the whole file so me will still see the header and do a save rather than a save as"
are the key elements to make sf's compatable processed with CE compatable with ME.

i knew i had done it before just forgot what i had done !



BILL