• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

Best sampling rate question

Accepting that 32kbps is never going to sound like 96 or 128 kbps, but desiring to add some 32kbps slots for our local listeners who are stuck on dial up to 128 connections, thereby leaving a little band width for the 128'ers to do something else other than listen to our stream I am wondering what would be the best sampling rate or best configuration to use to stream music on 32 kbps.

I tried running our current stream settings for our 96 kbps stream sampling at 44.1 khz in stereo on a 32 kbps stream and it sounded super horrible. I've listened to other 32 kbps streams and the sounded "good enough" to listen and work to.

Any suggestions - I imagine lose the stereo right away ...

Thanks,
Rickity
gulchradio.com
 
This is an area where I have curiosity, but not much trustworthy advice.

If I am going to put content out there at 32kbps, I don't think I would encode it in stereo. Use a mono mix in hopes of getting a better sound.

I produce church recordings for podcast delivery at 40kbps in mono and they are not the best, but not shabby either. Speech only. Once in a while I may sneak in a snipped of musical sound at the end.

I experimented with speech at 16 and 20 kbps. I was not happy with the results. Even in mono.
 
I'll agree with GRC; mono is a good option. At 32 kbps a mono 22.1 kHz sample rate should sound OK. If you must have stereo, 16 kHz sample rate may be tolerable if you roll off your audio above 8 kHz. A rule of thumb I've heard is the sample rate should be 2.5 times the highest audio frequency.

I run a stream at 24kbps/22kHz mono with no complaints from listeners but it's real oldies. I call it the "glorious AM radio sound". :)
 
stuckinthe50s said:
I'll agree with GRC; mono is a good option. At 32 kbps a mono 22.1 kHz sample rate should sound OK. If you must have stereo, 16 kHz sample rate may be tolerable if you roll off your audio above 8 kHz. A rule of thumb I've heard is the sample rate should be 2.5 times the highest audio frequency.

I run a stream at 24kbps/22kHz mono with no complaints from listeners but it's real oldies. I call it the "glorious AM radio sound". :)

I stopped posting podcasts of my AM 1620, which I uploaded as mp3 at 192k or 128k in order to get acceptable reproduction of the full fidelity, but then Pod-O-Matic started transcribing everything down to 96k, which sounds plain awful. No longer worth the bother.
AM in and of itself doesn't "have" any particular sound, but the poor receivers often used makes people think it has to be that way.

The streams I've heard which give different choices of bitrate are mis-named. The one called "AM" should rightfully be called
"old fashioned telephone land line".

Glorious AM sound is better sounding than FM. 8)
 
Thanks guys for the thoughts - been fooling around with it and got it to where it sounds not too shabby. Set it to mono with a sample rate of 16 khz, for no other reasoning than figuring 32 is one third of 96 kbps and the standard for that seems to be 44.1 khz, so 16 came the closest to one third of 44.1 - not very scientific. Will probably try different rates tomorrow.

If you want to listen to see what you think hopefully you can tune in on: http://live.str3am.com:2580/listen.pls

Comments welcome, kind of booms and thumps - the cruiser kids would approve.

"This is a test, for the next 60 seconds this station is conducting ..."

It'll be up for awhile.

rickity
gulchradio.com
 
I hear a continuous "zizz" that rides on top everything when there's audio....it even follows the "tail-off" on your voice
during the reverb decay. I love the reverb, and the overall balance of tone is maybe a bit heavy on the bass, but not too much.

The zizz is gone when there's no audio, so it's not the stream per se.
Perhaps if you roll off the highs at 8 khz, with a peak AT 7-8 khz, it wouldn't have this, but I suspect there is a ringing
"brightness" artifact from the reverb creating the "zizz".

Keep us informed and best wishes.
 
scanman1 said:
You could try using the AAC+ codec as well. It actually sounds quite decent at 32K.

Can Joe Sixpack click on your link and get sound.... or will he have to find new codecs, new drivers, new software?

I haven't kept records but, best guess, one out of every three audio sites I try to listen to results in failure. I need something different on my computer that isn't there, and the failure message does not give me a clue as to what I need.

No brag, just fact went the line went years ago in the TV series. I get up a little earlier than the average bear, I read more computer books than the average bear. I track down more "smart people" in person and on the Internet than the average bear. As I sometimes with a hint of arrogance mixed with sarcasm say to salespeople in computer retailing: "I eat computers for lunch, mister. That explanation you just gave me doesn't fly."

If I have trouble figuring out what I need to do to get my computer to unpack someone stream and turn it into something pouring out of my monitor speaker, what is poor Joe Sixpack going to do when a stream won't flow?
 
Goat Rodeo Cowboy said:
scanman1 said:
You could try using the AAC+ codec as well. It actually sounds quite decent at 32K.

Can Joe Sixpack click on your link and get sound.... or will he have to find new codecs, new drivers, new software?

I haven't kept records but, best guess, one out of every three audio sites I try to listen to results in failure. I need something different on my computer that isn't there, and the failure message does not give me a clue as to what I need.

No brag, just fact went the line went years ago in the TV series. I get up a little earlier than the average bear, I read more computer books than the average bear. I track down more "smart people" in person and on the Internet than the average bear. As I sometimes with a hint of arrogance mixed with sarcasm say to salespeople in computer retailing: "I eat computers for lunch, mister. That explanation you just gave me doesn't fly."

If I have trouble figuring out what I need to do to get my computer to unpack someone stream and turn it into something pouring out of my monitor speaker, what is poor Joe Sixpack going to do when a stream won't flow?
Yes to all of the above. I think the difficulty of tuning in is one of the great limiting factors to streaming radio. Using someone like LoudCity to cover royalties introduces another tune in step that even I can't get to work sometimes. Frequently I have to give detailed instructions on how to tune in, which is time consuming and frustrating and kills listener growth. If you tune in straight to shoutcast it's easy, but no royalty payments. I have tried tuning many major radio stations and gotten nowhere with their tune in's so I give up on them.

Last night, when I first tested this stream I had it streaming our 96 KBPS stream that we use to monitor the stream - feeding the 32 by suing the playback source (itunes) as the source for the 32. Currently it is streaming direct from the main computer and I backed the 8 KHZ off a bit as suggested by Tom Wells. As for AAC+ it's not a popular format, like mp3 is so with limited time and skills I'll have to stick with this.

I note that the audio quality varies from player to player - when played out of iTunes it has a crisper and cleaner sound than with the shoutcast player, for some reason realplayer won't recognize the stream and tells me to drop dead.

A note about our music files - everything we play is recorded as an AAC (mp4a) at a minimum of 128 kbps, we use Apple computers and stream with software called Nicecast. Nicecast does have "effects" options, which we don't apply, except as in the case where Tom suggested backing the range off a bit. We don't mess with the music (we don't reprocess it), except to edit out long applause, etc.

Thanks for the conversation, more welcomed. I've got plenty of opinions on the pros and cons of streaming radio - in one short sentence it's not as easy as people who haven't done it think - but we can save that for another time. My wife is sitting here expecting me to pay attention to her ... what's with that?

rickity
www.gulchradio.com
 
To be honest I think 32k aac+ sounds just as good as 128k mp3,, although I havent had much luck getting listeners using AAC+ as alot of people never upgrade their players, and therefor they wont support AAC+... it could have alot to do with your Soundcard,, you dont want to over EQ or pump abunch of processing into a dial up stream cause it will clip and destort. Simply run a 32k 22hhz mono stream through raw, flat, no processing and it should sound pretty good.
 
rickityone said:
Comments welcome, kind of booms and thumps - the cruiser kids would approve.

I listened and it sounded good. Much better than I expected from 32/16. You've done a good job of encoding the tracks.

I'd stay away from AAC+. Most listeners can't listen to that format.
 
Goat Rodeo Cowboy said:
scanman1 said:
You could try using the AAC+ codec as well. It actually sounds quite decent at 32K.

Can Joe Sixpack click on your link and get sound.... or will he have to find new codecs, new drivers, new software?

I haven't kept records but, best guess, one out of every three audio sites I try to listen to results in failure. I need something different on my computer that isn't there, and the failure message does not give me a clue as to what I need.

No brag, just fact went the line went years ago in the TV series. I get up a little earlier than the average bear, I read more computer books than the average bear. I track down more "smart people" in person and on the Internet than the average bear. As I sometimes with a hint of arrogance mixed with sarcasm say to salespeople in computer retailing: "I eat computers for lunch, mister. That explanation you just gave me doesn't fly."

If I have trouble figuring out what I need to do to get my computer to unpack someone stream and turn it into something pouring out of my monitor speaker, what is poor Joe Sixpack going to do when a stream won't flow?

Download Winamp. It's easy to install, and handles AAC+ just fine.

I get what you're saying, however. Most folks just want to use what's already installed on their computers when they buy them. So any stream that can be listened to with Windows Media Player is going to be more desirable since it comes standard with all PCs. In that case, staying with the original idea of low bitrate MP3 is in the best interest of the average listener.
 
For our station we stream at 96K and for my music it is ripped at 192K. The only issue I have is that for some reason winamp shows 96K at 22 Khz eventhough we all have our respective sams set to stream at 96K 44.1Khz... Wish I knew the trick to get the 44Khz..... CC1
 
CrazeeCarroll1 said:
For our station we stream at 96K and for my music it is ripped at 192K. The only issue I have is that for some reason winamp shows 96K at 22 Khz eventhough we all have our respective sams set to stream at 96K 44.1Khz... Wish I knew the trick to get the 44Khz..... CC1
I take it you mean the SAM encoder. Do you have it set to "High Quality"? That will override what ever sample rate you set. Change it to "Fast Sampling".
 
Maybe I'm confused, but around here, dial-up is 56K. That being said, programming to that group seems marginally beneficial.
 
The best sound and to be balanced with the best potential listeners is transmitting 64 K stereo and sampling most of my files at 160K. Most people listen at the office or even at home with 10-30 watt speakers. It's not like there listening in on power towers tuning in to Classical. You don't need to go AAC+ (unless you want to save bandwidth) because the majority just wants something clear and clean and basic too listen to.
 
pellmell said:
Maybe I'm confused, but around here, dial-up is 56K. That being said, programming to that group seems marginally beneficial.
A "56k dial-up" never actually attains the 56k download speed. I think that number was dreamed up by some marketing department. Even with the best phone line and modem you'll rarely get over 40k.
 
stuckinthe50s said:
I'll agree with GRC; mono is a good option. At 32 kbps a mono 22.1 kHz sample rate should sound OK. If you must have stereo, 16 kHz sample rate may be tolerable if you roll off your audio above 8 kHz. A rule of thumb I've heard is the sample rate should be 2.5 times the highest audio frequency.

I run a stream at 24kbps/22kHz mono with no complaints from listeners but it's real oldies. I call it the "glorious AM radio sound". :)
I agree with GRC & Sit50s. Although if you can pull it off (And you'd need something like EdCast to do this), you might be able to go 32/22 STEREO which sounds about as good as 24/22 mono. But again, you'd have to use a fancier DSP than what SHOUTcast, JetCast or even the latest version of SAM provide to do it though.

And yes, 24/22 mono is more indicative of what a 50,000 AM blowtorch like KOA Radio here in Denver would sound like in terrestrial radio.

That said, I wouldn't podcast at that bitrate though as anything below 64K is considered sub-standard.

Hope this helps.....

Cheers :D
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom