Re: Maybe
LA_Guy said:
A 196 kbps MP3 VBR encoded file ia so darn close to CD quality that 95 out of 100 users would not be able to tell the difference (196 kbps VBR coding is about the same as 256 kbps CBR). The other 5 listeners would not think it sounded 'bad'. It is FAR BETTER then any FM station! can transmit!
I believe the fidelity of an unprocessed high bit rate MP3 file will blow away the average audio that comes out of ANY digital audio processor, including any that you have designed.
Please don't feel attacked. I'm not picking on you specifically, it's just that the quality of radio broadcasts has thumbled down so much in the last years and it has a lot to do with people taking perceptual coding easily and making assumptions and wrong conclusions. As someone who loves radio I'd like to see this trend reversed, hence my reply.
Btw, it would never occur to me to claim that FM processing is high fidelity. Processing is used to conform to some technical requirements of FM transmission and achieve some artistic goals, so by design it's not meant to preserve fidelity. We do however offer a high fidelity option in our processors - it's called the bypass preset!
Likewise, I feel compelled to react when people call other things high fidelity that are not. Now I'm not against perceptual coding. Often times it lets us achieve higher quality audio that would be possible by not using it. But it should be treated with caution (maybe we should have a label for codecs like on the cigarettes
) In my opinion, people should think twice before putting compression anywhere in their broadcast chain! They should consider the whole picture and carefully choose parameters such as bitrate and type of coding.
I agree with you that a 192kbps MP3 file encoded with something like lame will be very hard to distinguish from the original. What I don't agree with you are the conclusions that you draw from there. All codecs are not created equal, even if they use the same format. There are listening test out there which clearly show the differences between different off-line codec implementations for the same format. And then there are differences between off-line and real-time implementations. An MP3 file created on your computer is coded with an off-line codec, perhaps the one that will do two-pass coding can sound VERY different at the same target bitrate compared to a codec that does real-time coding. You should not generalize that if an MP3 file on your computer sounds good, a certain hardware codec will sound the same.
In particular, the Barix units in my opinion sound fairly bad. I heard a lot of positive comments about the Barix units and people saying it sounds good before actually having and opportunity to listen to the units. And when I did I was quite disappointed. At the highest quality settings which should do around 160 kbps (if not a little more than that) I wouldn't have expected to hear lot of artifacts, but in fact there are quite obvious artifacts going on. Like I said, in my opinion this is not high fidelity.
And you didn't give any consideration of the big picture. Is your music library compressed or not? Will you have cascading codecs? Where is your FM processing in the chain. Is it before or after the codec? Just this would require a whole article to explain the caveats...
What I'm afraid with comments like this is that people tend to get comfortable using coding and then the quality goes down rapidly. If you say it's great for STL, someone else will say "if it's good for STL, it's good for my music library". Even worse some will conclude if it's good for STL and sounds great, than it's good for my music library as well. Or - if I have coding in STL and it sounds great, I don't need to have higher quality in my music library. And they would all be wrong conclusions!
There's a use and benefit for perceptual coding in broadcasting, but you should not employ one easily and without taking everything into consideration. You should minimize the number of perceptual coding in your broadcast chain (from that field recorder all the way to the transmitter!) to a minimum - and that would mean 1. You should use the highest bitrate possible - very important! And finally, use the highest quality codec you can (MP3 is definitely not the greatest codec out there).
Regards,
Goran Tomas
