cool edit 32 bit question

I got a cd-r of some material to remaster using cool edit pro. i was told the file was recorded from vinyl to cool edit in 32 bit. however, i seem to be having some trouble using the file. when i play it, all i get is static. I tried copying it to my hard drive using EAC, but that didn't work. Then I copied it to my hard drive using windows explorer. finally, I tried just opening the cd track in cool edit, but that didn't work, either. Am I doing something wrong here? Are there any things I should do to work on a 32 bit file. I have never gotten 32 bit files before - only 16 that I upsampled to 32/96k. Help!!

 

Is the CDR sent to you as data or CD audio? You realize EAC will only work under the redbook standard and stream in 16 bit. Also, what sound card are you using and what version of Cool Edit?

If someone was going to send you a disc of CD audio, it's 16bit. If somoene was going to send you anything in high bit, it should be data. Data Cds make the best backups.

I've not seen ONE instance where sampling an LP helped in a high bitrate. If you intend on listening to the LP rip as CD audio, try to work as plainly in 16 bit. Trust me on this.

What I think he sent you was a ISO9660 disc with data on it. Again, I'm guessing, but if you work in a high bitrate, the logical thing to do is to throw you the raw PCM data files on a CDR Rom.

 

You know, I have had a similar problem in the past with Cool Edit. I can't remember how I resolved this, but it hasn't happened since I installed A LOT more ram on my computer. One thing I do remember: Try opening this file in Cool Edit from the disc your friend sent. Once this file is opened, zoom way in and look at the wave-form; if it or one channel is a solid block, then that file will not work no matter what you do to it, it's corrupted on the disc he sent. I had to scrap a few long sessions when this occurred during regular use and recording, after saving a file and then reopening at another time. Again this is before I had added more ram. Be aware though, that I'm not certain that more ram would help here, but it might be worth a shot.

In addition, you should really check out the Syntrillium user's forum, and their tech support guys will answer emails quickly (within one day) and you will have your answer, they are both very helpful.

Let us know what you come up with.

Michael

 

Sckott:
I use a M-audio Delta 66 card...

I've always done LP transfers that were destined for cd as 32 bit, 44.1k recordings- (I've been under the impression that 32 bit will capture more detail), and then convert to 16-bit. (This is essentially all I do for transfers, no EQ, little or no noise reduction, I even accept a pop or click now and then just to keep things pure as possible). I am aware that problems multiply when converting from, say, 96k to 44.1, assuming the final product was going to be a red book cd, so I stay in 44.1. But 16-bit over 32-bit... this troubles me! Please help me understand -Thanks!

Michael

 

There are two 32-bit floating point file types used by Cool Edit. 1.2 and older use a propriety 16.8 version. The new 2.0 version uses the standard IEEE 0.24 Type 3 version that all the other audio editors use.

If this is the case, all you need to do is open up the files in 'Open As", choosing the appropriate file type, and "Save As" to the other, depending on which software you use.

All versions of Cool Edit can read both types of files, but they have their respective defaults.

 

I JUST got a Delta 44 24/96 card. Now you guys are saying to NOT use 96K and still use 44.1 instead? Is this a mathematical problem with dithering or something?

BOB

 

From what I understand, it is indeed a mathematical dithering issue. If you are going to work at a higher sampling rate than 44.1, then it is best to use a sampling rate that is a multiple of 44.1 (e.g. 88.2). Of course, this assumes that you are going to need to dither down to 16/44.1 to burn to redbook CD.

As far as upsampling a 16/44.1 file for cleanup, it is certainly best to convert to 32 bit, but I doubt that there's any advantage to increasing the sampling rate.

 

I figured by using 96k, then I would be potentially "capturing" more information. So, you are suggesting possibly using 88 instead?

3 questions:

1) What happens if I use 96 and dither to 44.1? Does the sound get drastically altered? Will it introduce artifacts?

2) How are real CDs mastered at 24/96 without encountering the same problems?

3) If you were to record at a higher sampling rate in CEP (either 88 or 96), what dither selections would you make?

BOB

 

Part of the deal is that if you record at higher sample rates you will have a higher resolution recording that you can play on your computer, and will be ready for newer media - like when you can burn your own DVD-A disks. There are bound to be some losses when downconverting, but your mileage may vary on how noticeable it is.

BTW I don't think "dithering" is the right term here. Dither has to do with improving the performance at the quietest end of the dynamic range.

 

I'll leave it up to you to experiment, but anyone sampling in a higher rate for the sake of an LP transfer is wasting their time. If your product is 44.1/16 bit, that's what you should work in. No use in "reducing the soup" if you want LESS MANIPULATION in the way it's handled. If it's chicken soup, it should taste like chicken soup, but don't treat it like pizza. Taking an analog signal to digital should be simple, just as most good LPs are made. Less in the chain, the better.

You CAN go 24 bit for some tape transfers that warrant the high capture, but then again, I'd only do it if I was doing 1. Mastering FOR that formatted sound or 2. archiving.

If you take an LP and use high bitrate to catpure it, you're just leaving more work for your computer to do later to give you nothing very different, but digitally manipulated. There's a reason to go to higher bitrates, and LP rips ain't one of em.

Less cooks in the soup!

Have fun anyways !

 

Scott, do you mean that I should not be doing click removal, editing, and EQ in 32 bit float? It would seem to me that if I am doing any sort of work on files, be it LP transfers or anything else, that it would be beneficial to do the work in 32 bit float. Cool Edit even recommends this. Or are you saying the initial LP transfer should be only 16 bit?

This is the first time I have tried to use anything other than a 16 bit file for LP remastering. I usually just recorded the LP to my stand alone CD burner, then copied the cd-r to my hard drive, upsampled to 24/96, and worked that way. However, it would seem to me that if the file I am working on was transferred to 32 bit, it would be one less step (not having to upsample to do my remastering). Am I incorrect in thinking this way?

I must say that the 32 bit float remasters I have done sound very good to me, but if there is a better way, I'm all ears!

 

Gotta clear up something that commonly gets confused. Dithering has NOTHING to do with sampling rates. Dithering only comes into play when downconverting from a high bit-depth like 32-bit or 24-bit to 20, or 16-bit. Downsampling involves going down from, say, 48k to 44.1.

Some people feel that recording in 88k is better than 96k because 88 divides evenly into 44.1 for compact disc.

Sampling rates have to do with frequency response.
Bit depths have to do with resolution.

I highly disagree with Scott in that if you record an LP in 16-bit that that's where you should stay. No. IF you process that file, it is best to convert to 32-bit first. You preserve the information in that 16-bit file by converting to 32-bit so you don't lose resolution with processing. You DO lose resolution if you process at 16-bit. I can plainly hear it, especially if I do A/B tests with the LP and resulting CD-R.

If you want to get the best, and your card can handle it, record your LP at 24-bit, and save at 32-bit (float).

If you do NOT process anything, and do no more than trimming, stay with 16-bit.

 

Chris, keep doing what you're doing. But, there is no real need to go through the trouble of upsampling to 96k for a CD-R project. Not only are you wasting time, but also damaging the sound of your files more than if you just dither down to 16-bit.

I record and process at 32-bit (float)/44.1k. If my final resting place is CD, why bother with anything higher? See, I can get away with recording at 44.1 because my card does excellent 48k (it's internal sampling rate) to 44.1. In MOST cases, most cards record at 48k. I do recommend staying there until it's time to go to CD-R. THEN you downsample to 44.1 before the conversion to 16-bit.

In the end, you should do your own experimentation to find what methods sound best to you for your needs. There is NO right or wrong answer. You see that everyone of us who deal with digital have their own beliefs and methods. I guess Steve would say, "whatever floats your boat".

 

Not to keep asking questions, but would it be best to say record at 24 or 32-bit to get a higher resolution, but keep it at 44.1 because there is no need for the headroom (I know that vinyl is way under 22k anyway)? Then dither it down to 16 and burn from there if there is no need to do any declicking or noise reduction? It would seem to make more sense to capture a bit more detail.

I am assuming that reason commercial CDs are done at 24/96 is because the master tapes probably have much higher frequency response to begin with.

BOB

 

Metalbob, if you just want to transfer your stuff straight, with NO processing, just do 16-bit/44.1.

 

Analog master tapes don't have much more, if any frequency response than what you can get at home on the consumer level with good equipment. Even if tape and vinyl did have a higher frequency response, the actual sound you hear will be limited by something in the playback chain, which is usually just up to 20,000kHz anyway.

 

Interesting reading on dither, and how to keep a signal pure:

http://www.digido.com/ditheressay.html

 

OK, but what would be the benefits of converting from 16 to 32 bits for processing? My way of making CD-Rs from Vinyl is 1) recording to my Sony PCM-M1 DAT @ 16/44.1. 2) Then I transfer it digitally to my hard disk, 3) open the file in Cool Edit and use the ClickFix Extension (not on the entire WAV file but only on the portions which need to be treated). I do not normalize the tracks nor do I record from LP in 16/48 since I have found the conversion to 44.1 does no good to the sound.
Would a conversion to 32 bit before the declicking process increase the accuracy of the declicker?

Arne

 

Yes. You ideally should process in 32-bit. You will also not destroy your sound. I proicess in 16-bit sometrimes and I can hear the sound degradation taking place.

Again, ANYTHING you do to a 16-bit file, aside from simple trimming, will degrade the sound.