• 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

OMNIA 6 EXi Analog Distorsion

stha said:
RealityCheckr is right. Other tests with different audio processors are giving the same kind of results (not only Omnias...). This is an harmonic distorsion introduced by the clipper and ANY clipper (as far as I know). But I'm still surprised to see how this effect is perceptible in some conditions, even if the clipper drive is at the minimum. Again, it was not against the Omnia processor, only because it's the one I have.

STHA,

I'm not sure about your findings, as you have not posted a detailed test setup and methodolgy. FM processors 'embed' the preemphasis network into the final limiting section, and unless you know and understand, how to test the emphasized signal, it is possible to get incorrect readings.
 
RealityCheckr:
Yes, Cornelius Gould have made a very interesting website that I visit very often.
I certainely still have a lot to learn, that's why I'm asking your advices with humility and passion.

Mr Foti:
It's the same if there's no preemphasis. With a normal use of the processor, I can hear sometimes a little distorion on the midrange frequencies, it's more with female voices or some instruments. To make that effect more evident, I've made a simple test with an audio input to the Omnia but with a 4khz bandwidth (using for example my 6200DAB for it's high pass output filter, but the processing is by passed). When monitoring all the Omnia6 FM outputs the distorsion is now evident to the ears. If I measure with a spectrum analyzer, the HD output is crystal clear and there's nothing higher than 4khz. When switching again to FM, I can clearly "see" the distorsion higher than 4 khz. It's the same with an OmniaFm hot, but also a 8200, 8100XT2...

Thank you all for your time and your patience.
 
stha said:
RealityCheckr:
Yes, Cornelius Gould have made a very interesting website that I visit very often.
I certainely still have a lot to learn, that's why I'm asking your advices with humility and passion.

Mr Foti:
It's the same if there's no preemphasis. With a normal use of the processor, I can hear sometimes a little distorion on the midrange frequencies, it's more with female voices or some instruments. To make that effect more evident, I've made a simple test with an audio input to the Omnia but with a 4khz bandwidth (using for example my 6200DAB for it's high pass output filter, but the processing is by passed). When monitoring all the Omnia6 FM outputs the distorsion is now evident to the ears. If I measure with a spectrum analyzer, the HD output is crystal clear and there's nothing higher than 4khz. When switching again to FM, I can clearly "see" the distorsion higher than 4 khz. It's the same with an OmniaFm hot, but also a 8200, 8100XT2...

Thank you all for your time and your patience.

STHA,

Well, there seems to be two viewpoints going on here. Glad you wish to learn, as there are many who can help. Not sure it was a wise idea to claim our product suffered a design issue.

A few thoughts...

1. You still have not offered any test criteria or methodology, and because of this it is difficult to understand your concern
2. It appears you are hearing the differences between a clipper and look-ahead limiter. A clipper will generate harmonics, even with distortion control, whereas a look-ahead limiter produces less harmonic distortion.

Based upon your last post, it is not clear how you are doing this test. I suggest you contact our Support Engineers and they can help you. Much easier than using a public forum to figure this out.
 
stha said:
RealityCheckr:
Yes, Cornelius Gould have made a very interesting website that I visit very often.
I certainely still have a lot to learn, that's why I'm asking your advices with humility and passion.

Mr Foti:
It's the same if there's no preemphasis. With a normal use of the processor, I can hear sometimes a little distorion on the midrange frequencies, it's more with female voices or some instruments. To make that effect more evident, I've made a simple test with an audio input to the Omnia but with a 4khz bandwidth (using for example my 6200DAB for it's high pass output filter, but the processing is by passed). When monitoring all the Omnia6 FM outputs the distorsion is now evident to the ears. If I measure with a spectrum analyzer, the HD output is crystal clear and there's nothing higher than 4khz. When switching again to FM, I can clearly "see" the distorsion higher than 4 khz. It's the same with an OmniaFm hot, but also a 8200, 8100XT2...

Thank you all for your time and your patience.

Remember one important thing:

The clipper used in traditional FM processing *is* a distortion generating system. In its crudest form (as in original early to mid 1970's era processing) it was simply a set of diodes hung on the output of the (dynamic) FM limiter to "whack off" peaks that got past the dynamic limiting action. It gained you a few percent of loudness, but you really couldn't hit it too hard without causing objectionable distortion.

In this case, it isn't all that different than if you were playing a cheap car radio as loud as it will go. There is that point where it's "loud enough",a dn if you go any higher, it would distort. The simple fact is, more than likely, it is already distorting quite a bit,and that little bit of distortion is making it sound louder than it really is. if you were to match a tone to the peak level of the music, you'd see and hear quite a bit of distortion. In Fm this is a bit more complicated because of the use of Pre-Emphasis, which boosts the highs a LOT. Your radio employs de-emphasis (which does the opposite). It is a crude form of noise reduction as FM gets progressively nosier the higher you go in frequency. This system helps to keep the noise low. It is also the thorn in the side of any of us who design audio processing, and how well you manage this separates the good processors from the bad ones.

Moving on with the history lesson:

Fast forward to the beginning of the era of the 8100 (late 70's), and what you have is the same clipper idea -- except in this case, Bob Orban plays games with masking the distortion in a way that causes it to be so (audibly) low that it easily "covered over" by the various frequencies in the program audio. This explanation is simplified, but it is intended to get the idea of what is going on across.

In the DSP processor era, the game is basically still the same except we all have to account for the big technical differences between analog and digital audio.

These differences make it extremely difficult to simply copy an analog design into a DSP platform.

Anyway, a possible reason that you hear some distortion on a tone is simply due to the fact that at certain points during your sweep --depending on exactly what you are doing-- there could be moments where you are running across distortion that otherwise would be masked by other characteristics of the program audio.

The reason the HD outputs sound so clean is because instead of using clippers to "brick wall" the peak levels, we have to use look-ahead limiting. This gives similar results, except we gain some Intermod Distortion instead of the harmonic distortion of hard clipping. We do this because bit reduction codecs such as HDRadio, AAC, AAC Plus, MPEG Layer 2 (MP2), MPEG Layer 3(MP3), etc. do not perform well with clipped audio. The extra harmonic content makes it hard for these codecs to mask their own distortion, so we need to feed them clean audio for best results.

There are some processors out there that use lookahead limiting for FM broadcasting too, and the sound is quite different. There is also usually quite a bit of delay through them as well. Typically, you cannot get "brute force loudness" with them that you can have with traditional style FM clipper systems.

All this is covered quite a bit on my website www.cgould.com with more detail -- and some audible clips for examples, so feel free to browse around!

Hope this answers some questions for ya!

Regards,

-Cornelius
 
STHA

The two lead gurus have pointed out the fallacy of trying to test a psychoaccoustic processor using pure tones at fixed levels... it won't give accurate results. And, face it, very few of your listeners are listening for tones anyway. (I give it a 70, got no beat but it's danceble). A better method is probably to obtain some source material which is clean, and with which you're intamitely familiar. Then you can compare the processed result with the original material. I use the DGG recording of DeFalla's Three Cornered Hat, fanfare from the Third Movement. The fanfare is played on a set of herald trumpets, castanets, and a tynpani. It will pass a system which has little intermod distortion and little clipping distortion. The tympanist is playing a riff on his drum, overdoing the 'phat bass' will turn the notes to thump. Back off ahalf number. Too much clipping and the castanets will turn to white noise.. get out of the clip a hlf nuimber. Too much limiting and the trumpets become blare instead of clarity. Etc.
Another good cut is Kenny and Dolly 'Islands in the Stream', they both have a breathy voice quality which overdoing will quickly turnm to crud. Try setting them up on program... you'll like them
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom