Syntrillium Backup Forums :: View topic

Jerry Meloon · Posts: 4

We recently upgraded to v1.1 of CE2000 and now find that we are extremely limited in what we can do with MP3 recordings. Anything we try to save as MP3 "MUST" be done as 44.1 stereo. Anything else and we get an error message saying that we have to change the sampling rate and convert to stereo. What has changed? We used to revel in 22,050 mono for commercial distribution and now it's taking 4x as long to upload an ad.

Jerry Meloon, WDBA/WREQ
jerry@wdba.com

Syntrillium M.D. · Location: USA Posts: 5124

This is certainly not normal behavior. In fact, you should have more options. When you go to save as MP3, click the OPTIONS button in the Save As dialog. Once the MP3 options window is open, you should click on the pull-down menu (below VBR/CBR) and you should have a host of encoding options. If you for some reason had it on a 'custom' setting or one of the presets, perhaps that is why it was forcing 44.1...

Let us know.

---Syntrillium Support

Craig Jackman · Location: Canada Posts: 909

Good Lord! Why would you want to distribute mono 22k commercials? That would sound like crap over my airwaves! Even my wife would notice the lack of frequency response (11k). Don't you care about your product? Let me guess, you also use a 96kbs bit rate too ...

Syntrillium M.D. · Location: USA Posts: 5124

Whoa, whoa! Getting a little heated here.

Be nice guys...

---Syntrillium Support

Edited by - syntrillium support on 05/04/2001 10:26:19 AM

Craig Jackman · Location: Canada Posts: 909

You think _that's_ heated?? Oh boy ...

Pardon me for my opinion, but I'm passionate about radio. If Syntrillium is hosting this forum about radio - and earning $$ selling their fine products to a)radio stations, or b)people who work at radio stations - they must realise that the passion for radio exists out there too.

I've been working with MP3 files for 3 or 4 years now, using it to transfer commercials between stations, between advertisers, and for transfering voice tracks between station and "big voice guy". During this time I've heard some spectacular MP3 files, and I've also heard some that sound like ... well lets just say less than spectacular. Maybe it's the people I've worked with, or maybe it's my ears, or maybe it's just that I care way too much, but mono MP3 files sampled at 22,050 SOUND LIKE CRAP!!!!!!!!!!!!!!!!!! There is enough on the radio now to make listeners want to change the station or put a tape or CD on, or go read a book or WHATEVER. Every element of your radio station from music to commercials to promos to stings/sweepers to the announcers that string them together has to sound great ... you have to give the listener a reason NOT to change the station! I mean is there anything worse than sitting through 4 min/240 sec of really horrid cliche ridden commercials? Compound that with a mono commercial with a frequency response that comes to a dead end at 10,804 Hz. I just shake my head and have one simple word. WHY? What is the listener to think when they hear that coming out of their favorite song that was in stereo and took the entire available frequency response of the FM signal.

Even if you are in a small town in Pennsylvania, there is no good excuse to use such a small sample rate. Both of the listed stations are FM, with a theoretical frequency response of 15k. In reality that frequency response may be only greater than 12k, but that's still more than Mr. Meloon is using. If it's an issue of storage, geez spend the money and buy a bigger hard drive, get one in trade for airtime, they're cheap now! Better yet, make a stereo dub of your commercial to CD or DAT or reel or cassette or VHS, while it's still a WAV file. That way your work is stored at some reasonable quality level should you need it again and isn't taking up space on your hard drive, THEN make your MP3 copy to send to your client.

Jay Marble · Posts: 145

Craig: I guess the way I feel about MP3 and radio is: It is the best thing radio producers who share files, have at this particular time. Here is Alaska, public radio reporters are using mp3 and ftp to share stories. Those reporters with good sound cards(darla and gadget labs mostly), and with decent microphones, and who use a reasonable sample rate, have pretty good sounding stories. When we first started using our file sharing system we used a sample rate of 22050. It certainly sounded better than a phone feed, but still pretty thin. We currently post our stories on the ftp at a sample rate of 32K(mono of course). But there is a huge difference between 22 and 32K. We all have dial up connections. Under the best of conditions it takes us 2 to 3 minutes to transfer 1 minute of sound. Eventually we will get broad band connections. When we do, we will say "goodbye" to mp3, and switch to .wav. MP3 is one giant compromise. While I love what it lets us do, I will not mourn its passing. As I see it, the only reason to use mp3, is if you are transfering the file over the net. If you are doing in-house production, use .wav.

Syntrillium M.D. · Location: USA Posts: 5124

hi all. Please don't misunderstand me...Passion is good. In fact, I venture to say that there isn't enough of it in most areas of the biz... Just keeping an eye (and an ear out) that's all!

Carry on, gentlemen.

---Syntrillium Support

Craig Jackman · Location: Canada Posts: 909

I would agree with Jay that MP3 is a compromise. However you can alter the degree of compromise using MP3.

When I first started using MP3 compression I took the time to do a listening test just to see the difference. Everyone has probably noticed the same thing, but anything less than 128kbs is "internet quality". 192kbs and up is transparent on the air, and it took me 5 minutes of searching to find 1 artifact at 320kbs. You can get great sounding MP3 files if you are willing to take the time.

32k sampling is what the default in our automation system is. In studio you can hear a difference between it and 44.1, but over the air, the casual listener won't pick up on it. I had a previous system that would record 30" spots at 32k and 60" spots at 22k and that was noticable on the air. Again, it'll all sound good if you are willing to take the time.

For news/voice use mono 32k is great and as long as a reasonable bit rate is used (anything from 160 and up) it should sound really good on the air. I understand the compromise involved using dialup ... I had pages of errors when I was on dial up, and broadband has made an immensse difference in my life, both for email and ftp transfers. That being said, the only way I transfer WAV files is over the internal network to the studio down the hall. Even with high speed access I still use MP3 compression to send material out of the building. Oh, I use a higher bit rate so it doesn't affect the sound quality as much, but it's still an MP3.

Jay Marble · Posts: 145

Craig: I agree...The higher the sample rate rate the better off you are. If we could get better internet speeds we would certainly up the sample rate.

I confess I don't completely understand your bit rate approach. I'm just gonna ask...How do you save at such high bit rates? When I look at my options/settings, I find stuff about 8 bit, 16 bit, and 32 bit. What am I missing here?

On a different much more rambling note, if you have broad band, why are you using .mp3? Aren't other formats more forgiving? Let me explain my situation. Our various stations have various automation systems(Audio Vault and Dillete mostly).. The one thing thing they have in common is that every imported .mp3 file has to be converted to .wav or .mp2 before it can be used. One reason we want to make the jump to broadband is so that we eliminate the coversion process from .mp3 to .wav. While NPR will accept .mp2 and .wave,even a phone feed, they won't accept an .mp3 file. My take on this is that they don't want to put an mp3 file through another conversion. Not exactly a question I guess..Jay

Jerry Meloon · Posts: 4

Sorry to be late getting back. Thanks to Support for guiding me into "Options". YES, it is showing a "CUSTOM" setting when I try to do 22050 mono recordings. The definition shows 128k at 22050 mono. But what SHOULD it default to if not the CUSTOM at the bottom of the list?

Regarding the choice of sampling, our stations use APTx compression on the air with Scott Studios AXS systems. But for dry read ads or voice tracking even the high-limit imposed by 22050 is overkill. Nobody's vocal frequency launches up to 11khz... at least nobody I know who works in radio. The other consideration is that in small markets we do not have T1, DSL, etc. connections. So we're stuck with 31kbs modem connections. Big stereo files take several minutes to upload to a holding area on our site and it takes just as long for the neighboring stations to download so we do what is practical. I apologize for being so pragmatic but once IBOC digital arrives, we'll probably ALL be scrambling to match bit rates (sorry, different subject).

Jay Marble · Posts: 145

Jerry: Please don't appologise for being pragmatic. I think most of us in the small market radio world are dealing with some variation of these problems/dilemas. Sure we would all like to be using big fat .wav files. Instead we use what we can...Jay

Craig Jackman · Location: Canada Posts: 909

Jay, what your are missing is the bit rate setting in MP3 conversion. All of my WAV work is done at 44.1/16 bit, but when I change that WAV to MP3, it's still a 44.1 sampling rate, but the bit rate is usually 192 kb/s ... not a word length of 192 bits. The bit rate refers to the amount of data compression you choose. So 192 kb/s is a data compression rate of 7.4:1 ... meaning for every part of the original file you keep, you are throwing away 7.4 parts using MP3 "lossy" data compression.

Why do I use MP3? For the same reason everyone else does, speed. Even using broad band it's faster for everyone involved to use a 800k MP3 file (for a 30" commercial) rather than 5 Mb for the same file as a WAV. If I'm sending to clients, reps, or other stations ... or recieving from agencies or other stations a total of 30+ MP3 files a day, I need all the speed I can get. Like I said before, I only send WAV files over the internal network to the the studio down the hall. Anything that goes into our automation system has to be converted (MP3, MP2, WAV), as we do not have file import capability and it too uses APT-X compression. Even if I wanted to send at the highest quality WAV file, other stations in the market on dial up would be pretty pissed at having to download such a large file.