HD Radio Sample

[chriscoxradio]

Alright... for all of those people who think HD's a Farce and have never actually heard HD Radio, listen to this. It's a recording from Q101 WKQX in Chicago. I recorded it on my Accurian Radio, and did not edit a thing. It shows the change in quality from analogue, to HD. ALSO: I am 40-ish miles from Chicago, and I'm using the Stock Antenna that came out of the box.

Enough of my blabbering, here's the recording.

http://www.box.net/shared/g4qkx079hj

Just click Untitled.mp3 to get download the MP3.

 

[tomservo]

40-ish miles from Chicago, I believe that would still be in the 60dBu contour of the station, possibly. Not bad, though.

The music sample wasn't very good for this purpose, it sounded like someone just tweaked the treble where the switchover occurred. And gosh, there didn't seem to be any stereo seperation in that song at all. What jumped out in your sample though was the DJ's banter, it was completely noise-free in the background, perfectly quiet. Very crisp.

Unfortunately, I'd have to say this sample doesn't really speak highly of HD.

 

[raydofan]

Unfortunately, I'd have to say this sample doesn't really speak highly of HD.

My Tannoys agree. The top end of the music is watery. Perhaps a longer sample-the bitrate is 320kbps, correct?

 

[PhilJSmith67]

Looking at the FCC 54dBu Service Contour map for a Chicago station atop the Sears Tower (94.7 WZZN), it appears that 40 miles puts you right at the 54dBu line.

http://www.fcc.gov/ftp/Bureaus/MB/Databases/fm_tv_service_areas/maps/FM537174.gif


Re-encoding previously-encoded audio doesn't work well. All artifacts become heavily exaggerated, even if you're re-encoding at 320k. Try re-encoding audio taken from a Sirius or XM receiver -- It sounds phenomenally worse than what came out of the satellite receiver. I've re-encoded audio from a 384kbps digital TV streams and heard hollow, distant treble that I didn't perceive in the original broadcast.

Aside from that, I appreciate that you took the time to transcode the audio and post it for others to hear.

 

[raydofan]

Re-encoding previously-encoded audio doesn't work well. All artifacts become heavily exaggerated, even if you're re-encoding at 320k. Try re-encoding audio taken from a Sirius or XM receiver -- It sounds phenomenally worse than what came out of the satellite receiver. I've re-encoded audio from a 384kbps digital TV streams and heard hollow, distant treble that I didn't perceive in the original broadcast.

What program are you using to re-encode? What was the input method?

Using a Fraunhofer MP3 encoder, I can make pretty accurate recordings from many sources-including SR. Also, Jetaudio makes a decent recorder with Ogg Vorbis encoding, which, to my trained ear, sounds really accurate.

I have hours off of Sirius recorded in Ogg Vorbis, and it sounds great.
Use a USB input UW10 Yamaha, another USB box for output.

You?

 

[Tom Wells]

Too much treble after the switch-over. Do ya think I'm hard of hearing , or are ya tryin' to burn out my hearing?
A live band doesn't even have this much treble unless you're on stage next to the drum set.

And the DJ's got the zizzlies on her voice.

The re-encoding problems are usually worse sounding than this.
This is much like the effect seen in print when a halftone-screen gets reproduced again in another halftone screen.
The effect is called a Moire pattern, a repeating inteference patttern produced by the inexact synch between the reproduction dots.
Re-encoding is the same effect in audio.
I don't think thats an issue with this recording. What I hear is purely the 96k rate zizzly.
If the sample rate of the later dub is sufficiently high, we will only hear artifacts from the original encoding.

 

[chriscoxradio]

I'll make a new one tomorrow, and encode it at a better bitrate.

 

[raydofan]

I'll make a new one tomorrow, and encode it at a better bitrate.

320kbps is very high-I'm not sure how much higher you can go without going to 44.1 wave.

If you do record, try getting more music.

 

[PhilJSmith67]

Fraunhofer or LAME @ 320k work very well on uncompressed audio (analog or PCM). But re-encoding audio that has already gone round-trip through compression and decompression yields audible artifacts. At high bitrates it's very much tolerable, but 44.1 PCM won't exaggerate artifacts or distortion from the initial compression/decompression cycle like adding another cycle would.

I listened to that 320k audio clip. If my Sangean sounded that bad across the board on FM I would have returned it. I'm not sure if that's the fault of the station, the two compression/decompression cycles, the receiver itself, or what, but the sound quality is poor. Fortunately, I've not heard any station on my receiver with such a pronounced level of high-end distortion or compression artifacts. I can tell you one thing, that the codec for HD Radio does not produce that signature whooshy, wishy-washy midrange+treble that is totally a characteristic of MP3, which occurs even at 320k.

I am kind of curious why the audio clip starts out muddy/muffled, and then becomes distorted with treble artifacts. Was this recording acoustically coupled to the encoder, or hard-wired?

 

[raydofan]

Fraunhofer or LAME @ 320k work very well on uncompressed audio (analog or PCM). But re-encoding audio that has already gone round-trip through compression and decompression yields audible artifacts. At high bitrates it's very much tolerable, but 44.1 PCM won't exaggerate artifacts or distortion from the initial compression/decompression cycle like adding another cycle would.

I listened to that 320k audio clip. If my Sangean sounded that bad across the board on FM I would have returned it. I'm not sure if that's the fault of the station, the two compression/decompression cycles, the receiver itself, or what, but the sound quality is poor. Fortunately, I've not heard any station on my receiver with such a pronounced level of high-end distortion or compression artifacts. I can tell you one thing, that the codec for HD Radio does not produce that signature whooshy, wishy-washy midrange+treble that is totally a characteristic of MP3, which occurs even at 320k.

I am kind of curious why the audio clip starts out muddy/muffled, and then becomes distorted with treble artifacts. Was this recording acoustically coupled to the encoder, or hard-wired?

I think some receiver-end chain detail would be helpful.

The 'swooshy' sound does not occur @320kbps with the proper encoding.

With a program like MusicMatch? Yes.

I have A/B'd waves (USB output, Adcom Pre- and Power amp into Tannoy Reveal speakers) with high-bitrate MP3s. There is a slight difference in the audio, but not major artifacts.

As with any system, the HD signal on the receiver end benfits from proper handling on the broadcast end.

I think you heard the switch from Analog to IBOC.


And I'm gussing that you listen to different stations than the one in this recording.

More info about the recording is appreciated.

 

[radiopilot]

Alright... for all of those people who think HD's a Farce and have never actually heard HD Radio, listen to this. It's a recording from Q101 WKQX in Chicago. I recorded it on my Accurian Radio, and did not edit a thing. It shows the change in quality from analogue, to HD. ALSO: I am 40-ish miles from Chicago, and I'm using the Stock Antenna that came out of the box.

Enough of my blabbering, here's the recording.

http://www.box.net/shared/g4qkx079hj

Just click Untitled.mp3 to get download the MP3.

This really sounds bad and even the dj has artifacts clearly anyone can here.. the high end is awlful.

If this is what I'd have on my tuner I would have changed the station.

Radiopilot

 

[R.F. Burns]

Alright... for all of those people who think HD's a Farce and have never actually heard HD Radio, listen to this. It's a recording from Q101 WKQX in Chicago. I recorded it on my Accurian Radio, and did not edit a thing. It shows the change in quality from analogue, to HD. ALSO: I am 40-ish miles from Chicago, and I'm using the Stock Antenna that came out of the box.

Enough of my blabbering, here's the recording.

http://www.box.net/shared/g4qkx079hj

Just click Untitled.mp3 to get download the MP3.

This really sounds bad and even the dj has artifacts clearly anyone can here.. the high end is awlful.

If this is what I'd have on my tuner I would have changed the station.

Radiopilot

Unless the file is a WAV file, further compression due to cascading codecs will cause degradation to audio which otherwise might be fine. That's why we at the network level have moved towards linear systems as opposed to compressed audio systems.

 

[Mike Walker]

I hate to pick on ya', but the recording (the one the station's playing) is AWFUL. Not a bit of stereo separation. If pressed, I'd swear it's mono. I agree that HD can sound great. But I've gotta' say, this sure doesn't show it to very good effect! Loud and distorted in analog...loud,, BRIGHT, and distorted in HD (the recording itself...notthing you did, I don't suppose).

 

[raydofan]

Unless the file is a WAV file, further compression due to cascading codecs will cause degradation to audio which otherwise might be fine. That's why we at the network level have moved towards linear systems as opposed to compressed audio systems.

From what to what exactly?

I still hear evidence that NPR uses some sort of lossy system for the remote hookups of show guest. A bit of swish on the high end...

Y'all (in the remote broadcast end of it anyway? do pay for Sat access based on channel bandwidth/quality, or am I wrong?

 

[R.F. Burns]

Unless the file is a WAV file, further compression due to cascading codecs will cause degradation to audio which otherwise might be fine. That's why we at the network level have moved towards linear systems as opposed to compressed audio systems.

From what to what exactly?

I still hear evidence that NPR uses some sort of lossy system for the remote hookups of show guest. A bit of swish on the high end...

Y'all (in the remote broadcast end of it anyway? do pay for Sat access based on channel bandwidth/quality, or am I wrong?

Most likely the remote was coming in via ISDN. Now the problem is which codec are they using. G.722 is very robust but prone to artifacts and grizzly audio. Knowing NPR they are probably using AAC-LD, if there are X-Streams on both ends. If not they most likely are doing G.722 to the guest and either L3 or L2 back to them. Both are better then G.722 but they each have their own issues, such as delay. For more permanent installations a linear T (One that isn't compressed) is preferable to an ISDN connection. The latest is the use of the internet and low delay units manufactured by Comrex among others to deliver quality audio. Of course they use compression and it becomes apparent when more compression algorithms between the point of origination and the listener are introduced.

 

[chriscoxradio]

Alright, here's some info about the recording.

I took a 1/8" to 1/8" cable and wired it into the Line In on my PC. I recorded it using Cool Edit Pro 2.0. I used the built in MP3 Encoder in CEP to save it.

 

[R.F. Burns]

Alright, here's some info about the recording.

I took a 1/8" to 1/8" cable and wired it into the Line In on my PC. I recorded it using Cool Edit Pro 2.0. I used the built in MP3 Encoder in CEP to save it.

What bit rate did you encode it at? Also, once you convert a already compressed file into an MP3 that adds another level of compression and that causes artifacts.

 

[chriscoxradio]

I encoded it at 320k in Cool Edit.

 

[chriscoxradio]

Okay, here's one from 97.1 WDRV (Classic Hits) here in Chicago. It's 10 minutes long, recorded on the Accurian HD Radio, and encoded at 320kbps using Cool Edit Pro 2.0. You can also hear the switchover from Analogue to Digital in the beginning of the clip.

http://www.sendspace.com/file/uo1xl1

 

[sbe1]

Only 3 db of dynamic range. That really sucks. Why must record producers smash the audio so hard these days? Only thing I noticed was that the HD was louder not better.