Processing clips

[g]

Legacy radio engineers are laughing at this thread. They know you judge a processor's performance over a broad range of material. Not by playing some one-off track using some amateur setup.

Goran claims that he started this thread for fun and not for product comparison. He's been the one who has offered the most processor A vs B claims. After reading back many of the posts, he has been wrong in most of his claims.

Time for the moderators to dump this thread.

Methinks a few people here are just getting cranky and annoyed because a few carefully chosen test tracks are making their favorite audio processors sound bad compared to even some antiquated analog boxes.

If you look at the bigger picture, there are some big processors with big DSP power, big names, and big price tags, which will make your station sound louder and brighter than anyone else on the dial, on the majority of source material. That's a grand accomplishment and is the reason why these processors are so popular with Top 10-market stations. But a "majority" does not mean "all."

Some source material can make some processors go from sounding very good to very bad, and in some rare cases this is simply due to the box's processing structure and can't be "adjusted out" to the user's satisfaction. In that case, there are other choices which can be investigated, even if it means using something you won't find in the racks on the 85th floor of the Empire State Building.

-g

 

[cgould]

Methinks a few people here are just getting cranky and annoyed because a few carefully chosen test tracks are making their favorite audio processors sound bad compared to even some antiquated analog boxes.

If you look at the bigger picture, there are some big processors with big DSP power, big names, and big price tags, which will make your station sound louder and brighter than anyone else on the dial, on the majority of source material. That's a grand accomplishment and is the reason why these processors are so popular with Top 10-market stations. But a "majority" does not mean "all."

Some source material can make some processors go from sounding very good to very bad, and in some rare cases this is simply due to the box's processing structure and can't be "adjusted out" to the user's satisfaction. In that case, there are other choices which can be investigated, even if it means using something you won't find in the racks on the 85th floor of the Empire State Building.

-g

Hi, g

From my point of view, its more a case of a post of a song with processors sounding really bad -- and when *properly* adjusted, it doesn't sound as "exaggerated" as the posted clip.

What is shown is the best that a particular user was able to do with their available time and energy, but it doesn't mean *you* won't be able to get a processor to do better.



Yes, there is some obvious (audible) HF control with one structure vs. another, and as "reality checker" (And Goran, to be fair) pointed out, its a matter of taste. Do you want some mild HF ducking but "open" highs, or more of a "HF smear" that doesn't duck. There is no right or wrong answer there...just a personal preference.

Yes, every processor has its weak spots, and the ideal processing unit for you is the where the weak spots do not cause issues for you.

I have no issue there...the real problem I have is stated in the opening paragraph I wrote above.

In reality, as I said before, I think it just comes down to a comfort level with different processing algorithms. If you are uncomfortable with a particular system, you are not going to be able to make it play at its best.

This is no different than cars. The more "at one" you feel with your car, the more amazing things you can do vs. someone who just can't get the feel for the thing.

Not a criticism, just a fact of life...and audio processors are no different.

-C

 

[Fonorban]

Radiotoday did not open any “Pandora’s box” I think here is an open forum else will be “locked” so you, I, everybody can write his thoughts, experiences, opinions etc about processing, Goran if there is a different point of view for the same theme don’t take it personally in other words “cgould” say it better “I think everyone's comments are directed at others who start making sweeping statements about ANY processor based on these clips” correct Mr Cornelius and thanks. That I want from this thread is to see a more professional behaviour as professional is the processing techniques and shows some “respect” on our “audio teacher’s” which spend his valuable time to write here, actually I am not with any of two major brands (if you think so) but as I grow up I learn to show “respect” to a “wiser teacher” even more if he is older from me, I am sure “Legacy radio engineers are laughing at this thread” as notice it by “PaulyBoy”.

Thanks cgould for correction about rms detector but that was not mine report it was Goran’s as he responses instead of you. Oh cgould by the way I prefer “a more a "HF smear" that doesn't duck” than of “mild HF ducking but "open" highs”

Yes I totally agree “just a fact of life...and audio processors are no different”
For some people takes Looooonger time to grown up


Regards to all & Peace.
XFon

 

[Tom Wells]

I've enjoyed this whole thread, and have done my own critique of radio audio quality for 35 years.

I remember many changes in that time. Always the problem as cgould has neatly summed, is how to get a time constant (and it must be continuously changing) that will make our audio on receive mimic the logarithmic curve of perceived loudness, even at low levels.

Most of the older hardware designs did use voltage control bias adjustment, where the multi-time constant is vital.

I bought a very inexpensive compressor/limiter based on ( haven't opened it yet) an opto-isolator, presumably leds or a lamp,
directed onto a cadmium sulfide cell, this varying resistance being used in biasing a 12AX7.
The little searching for flaws on the web showed claims that this method may produce some "ringing". I don't hear any.
I am delighted with the results by using the 2 channels separately in a two pass method I described in another thread.
It doesn't make every problem recording fine, so I'm leaving it where it makes most of them as good as can be.
I make notes and re-eq specific songs.
The only good sounding volume expander circuits used in very expensive radios of the 1930's were based on lamps,
where the problem of time constant is largely negated. And it was easy to find a voltage/current/brightness ratio that was effective.
I bought the compressor based on working on such circuits.

I woud like to ask the professionals here what their thoughts on these designs are, as they have surely encountered them.

I am using it for AM processing at home. aircheck links on pocast board, latest one is since adding the compressor.

 

[cgould]

I bought a very inexpensive compressor/limiter based on ( haven't opened it yet) an opto-isolator, presumably leds or a lamp,
directed onto a cadmium sulfide cell, this varying resistance being used in biasing a 12AX7.
The little searching for flaws on the web showed claims that this method may produce some "ringing". I don't hear any.
I am delighted with the results by using the 2 channels separately in a two pass method I described in another thread.
It doesn't make every problem recording fine, so I'm leaving it where it makes most of them as good as can be.
I make notes and re-eq specific songs.

<edit>

I woud like to ask the professionals here what their thoughts on these designs are, as they have surely encountered them.

I am using it for AM processing at home. aircheck links on pocast board, latest one is since adding the compressor.

The compressor / limiter designs that use light for gain control are fun to work with. The earliest version of my multiband processor (that still fits into the family of designs that I still develop to this day) used LED / Photocell for gain control. This was mostly due to me not finding a VCA I liked for my new design. So, I gave LED Photocells a go around.

The interesting thing about Light Dependent Resistors (LDR's) is that each photocell of different types has certain response characteristics, and they naturally exhibit a sort of multiple time constant type behavior, and you can add additional "outboard" time constants to complement the characteristics of the photocell.

The ones that sound really neat are the ones with a "spongy" type response (like the Urei compressors). Nice 'warm' tone.

The problem comes into play when the LED's and / or photocells start to age. As we all know, LED's change in brightness over time, and this changes the dynamics of the Opto controlled compressor. Usually, it isn't too much of a problem with a single band compressor. Multiband compressors...especially stereo ones, on the other hand!!!

-C

 

[Tom Wells]

Well, I'll see if I can get spares, if data is available. In mono two pass operation I suspect I could get by quite a while by twiddling
ratios and thresholds.

This circuit point seems ripe for experimentation. I have quite a number of cds cells and lamp/led combinations I could try.
I'm quite happy to enjoy for now.

 

[gtomas]

It isn't an attack on you, Goran...just some of us putting the breaks on some folks who are taking these clips in a direction they were not intended to be in. Being the creator of the thread, I'm sure there's a part of you taking this personally, and feel we are all talking directly to YOU...well we do talk to you when you take it personally... but I think everyone's comments are directed at others who start making sweeping statements about ANY processor based on these clips.

The Internet communication can often be deceiving as there are no nonverbal cues for us to determine the tone, context, etc in which way something was said. So let me assure you in writing that I don't feel attacked in any way!

I've started this thread true, but I don't take "ownership" over it, in fact my intention was for everyone to discuss and share opinions, comments and clips of processing and different sound stations have around the world. Some people choose to deliberately take things out of the context and make this thread and sharing of clips look "bad" and for some reason want to shut it down. I hope that this won't happen because in my opinion that shrinks the processing world smaller... And as a processing enthusiast I'd like to see it grow bigger!

To get back to your comment, the thing is there are different preferences in processing, different tastes, different processing styles and different brand favorites... Why would I feel attacked if someone else doesn't find some aspect of certain processor, that for example bothers me, problematic?

I stand behind what I think is a problem with Omnia.6, I've spend so much time with this for anyone, even you or Frank (and I have great respect for both of you and Frank!) to tell me it can be fixed with simple parameter adjustment, in a satisfactory way. I don't mean this to sound pretentious or disrespectful, only firmly proven as far as I am concerned. I hope you can understand this. For example, would Bob be able to convince Frank a few decades ago that he can get that sound he was looking for with parameter adjustment on 8100/XT2? I'm exaggerating a bit, but I hope you get the picture!

I'm sorry if you feel I'm "copping out" but I've sent e-mails to Frank going back to 2004 about this. The thing is my focus is moving elsewhere now and I don't have time to further prove my point more than I already have. Frank will decide if there's something there to be addressed or it's just a few people's preference. It's his processor and he knows his customers best.


Regards,
Goran Tomas

 

[littlejohn]

The lights and photocells - or electroluminescent panels and photocells - have a smooth sound which a lot of vocalists like. Take a look at the prices paid for old Teletronix/UREI boixes using this technique. The LA-2 and -3s bring premiums. Studio guys tell me their talent loves to sing through them. For on-air use, the BL-40 Modulimiter used the same technique. It was a single band, but could make your AM sound pretty good for its day.

 

[rrsounds]

The lights and photocells - or electroluminescent panels and photocells - have a smooth sound which a lot of vocalists like. Take a look at the prices paid for old Teletronix/UREI boixes using this technique. The LA-2 and -3s bring premiums. Studio guys tell me their talent loves to sing through them. For on-air use, the BL-40 Modulimiter used the same technique. It was a single band, but could make your AM sound pretty good for its day.

Speaking of photocells and limiters... I really don't think you get the full effect of "photocell processing" unless you use a lightbulb to drive the cell. LEDs don't count, Corny.

I've uploaded a clip of a song ("Machine Gun" by the Commodores) processed by a lamp/photocell compressor I built (and recorded) when I was a teenager. I made the recording as part of a movie soundtrack for a project during my first year in Community College. I got an "A+"

Sorry about the quality; the recording is way more than 30 years old. The original was a good 1/4 stereo two-track, but I could not align the tape deck before ripping it to mp3, so there is some serious phase error. You may note the source material was a well-played 45.

The compressor was based around a cheap CDS photocell with a Radio Shack 1.5V grain-of-wheat lamp in the control loop. I found the smaller the lamp, the faster the time constant...but still very 'RMS-y.' This was the smallest lamp I could find. I think you'll hear what I mean.

At any rate, I guess you might say that this compressor was, in essence, Ariane's great-great-great grandmother.

Hope you enjoy the recording. No attempt was made to prevent listener fatigue.


Kind Regards,
David

PS: BTW, the original analog Ariane used the previously-mentioned THAT 4301 RMS detector and VCAs.

 

[cgould]

I stand behind what I think is a problem with Omnia.6, I've spend so much time with this for anyone, even you or Frank (and I have great respect for both of you and Frank!) to tell me it can be fixed with simple parameter adjustment, in a satisfactory way. I don't mean this to sound pretentious or disrespectful, only firmly proven as far as I am concerned. I hope you can understand this. For example, would Bob be able to convince Frank a few decades ago that he can get that sound he was looking for with parameter adjustment on 8100/XT2? I'm exaggerating a bit, but I hope you get the picture!
omas

Yes, I do!

Now, if you prefer the HF sound of the Orban, there is no way to really get an Omnia to emulate it. Same for the Orban box in regards to the Omnia. They are two different beasts.

The Omnia can get through the "Sweet Jane" cut if you let the Omnia ride it through. What I mean is that it'll let it fly as-is, and the silibant part of the song will be loud and overbearing (much as it is in the recording). If you want it to CONTROL that kind of stuff with the multiband sections, and adjust it to do so, it will, you'll just hear it ducking on that material. You can use setup A, and have the sibilant energy blast through, and the the "silk" control, which will knock down the HF energy in a manner that causes minimal ducking, but the overall sound will be a bit duller than before.

A lot of this is due to the Omnia doing what it does...try to keep the highs "open" sounding as possible. The Orban DSP does it with really fast responding limiters that squash out this difference...The Orban analog boxes did it with HF clipping (8100 - a combination of really fast HF limiting and clipping, the XT-2 - pretty much with distortion canceled HF clipping).

Again...two different beasts, and two different takes on the approach to processing HF energy.

This may clarify things on my stance...not necessarily trying to defend a 'favorite' processor, just pointing out what the 'problem' really is here.

;D

-C

 

[FFoti1]

I stand behind what I think is a problem with Omnia.6, I've spend so much time with this for anyone, even you or Frank (and I have great respect for both of you and Frank!) to tell me it can be fixed with simple parameter adjustment, in a satisfactory way. I don't mean this to sound pretentious or disrespectful, only firmly proven as far as I am concerned. I hope you can understand this. For example, would Bob be able to convince Frank a few decades ago that he can get that sound he was looking for with parameter adjustment on 8100/XT2? I'm exaggerating a bit, but I hope you get the picture!

I'm sorry if you feel I'm "copping out" but I've sent e-mails to Frank going back to 2004 about this. The thing is my focus is moving elsewhere now and I don't have time to further prove my point more than I already have. Frank will decide if there's something there to be addressed or it's just a few people's preference. It's his processor and he knows his customers best.

Yes, you are copping out.

Your efforts since 2004, have been very limited, and you have privately admitted this to to me in eMail. Recently, you admitted in this forum that you quit on it, when I was helping you make strides on your concern. Once you were able to illustrate the situation, I sent you one modified preset to try, and then you never got back to me. You have not been fair with regard to this, yet you are willing to make strong claims against our product.

You feel that this is an issue with the product, but you quit on it before anything could be resolved. Thousands and thousands of other Omnia uses do not have this concern. You should admit, publically, that part of the issue was your lack of expressing what you were looking for in the audio. It wasn't until you eMailed me old audio samples that it became aparent of what your desire was/is. I accept that this could have been due to a difference in 'understanding' based upon our USA/Euro cultural differences.

At some point, when many other people are able to make something work successfully, and you can't, the problem may not be with the product...

-Frank Foti

 

[cgould]

Speaking of photocells and limiters... I really don't think you get the full effect of "photocell processing" unless you use a lightbulb to drive the cell. LEDs don't count, Corny.

I experimented with Light Bulb and Photocell gain reduction too! But it was really early on, and not part of anything I have done in fairly recent times, so I didn't mention it. I was *really* new to this stuff back then.

My most vivid memory was not taking into account the light bulb rating VS th maximum voltage the control sidechain could deliver.

When a REALLY loud passage came along, and blew the light bulb out, the blast my poor ears got through the headphones blew the hat off of my head -- much like the recording engineer in one of the episodes of the Little Rascals series of short films! ;D

When not blowing the hat off of my head, do recall noticing a really cool sound coming from it, though...though I don't think I could have appreciated it at the time.

-C

 

[cgould]

The Omnia can get through the "Sweet Jane" cut if you let the Omnia ride it through. What I mean is that it'll let it fly as-is, and the silibant part of the song will be loud and overbearing (much as it is in the recording). If you want it to CONTROL that kind of stuff with the multiband sections, and adjust it to do so, it will, you'll just hear it ducking on that material.

I wanted to say "If you want to SQUASH out that kind of stuff with the multiband sections, it won't do that without ducking.

It isn't a HF squash box...to do that, you need a different processor.

-C

 

[Tom Wells]

Speaking of photocells and limiters... I really don't think you get the full effect of "photocell processing" unless you use a lightbulb to drive the cell. LEDs don't count, Corny.

That was exactly what I was thinking about, and why I now have to find out what this compressor uses.

I don't mind changing lamps or cds cells and/or calibrating to maintain a device if I'll get a better result than light emitting diodes give.

 

[cgould]

Speaking of photocells and limiters... I really don't think you get the full effect of "photocell processing" unless you use a lightbulb to drive the cell. LEDs don't count, Corny.

That was exactly what I was thinking about, and why I now have to find out what this compressor uses.

I don't mind changing lamps or cds cells and/or calibrating to maintain a device if I'll get a better result than light emitting diodes give.

Light Bulbs will be way more consistent over time than an LED will be....about the only adjustment I can think of is a new bulb when the current one goes out! Plus, you'll get way less "audio hash" through the cds cell from a light bulb VS an LED.

Nice smooth audio audio out...

-C

 

[Info-warrior]

The LA-2A uses an EL (electro-luminescent) panel. it's the same type of panel used in those flat night lights that glow green or what was used as the back light on the old Radio shack Pro 2004-2006 scanner series. The problems I hear regarding them, though no personal experience with EL compressors, is that the panels don't last long because of the high frequencies.

Someone who has an LA-2A or similiar processor please comment on how often or if you've had to replace your panel.

Incidentally, LA-2As are being manufactured again, though even the reproductions are not cheap.

 

[rrsounds]

Speaking of photocells and limiters... I really don't think you get the full effect of "photocell processing" unless you use a lightbulb to drive the cell. LEDs don't count, Corny.

That was exactly what I was thinking about, and why I now have to find out what this compressor uses.

I don't mind changing lamps or cds cells and/or calibrating to maintain a device if I'll get a better result than light emitting diodes give.

Light Bulbs will be way more consistent over time than an LED will be....about the only adjustment I can think of is a new bulb when the current one goes out! Plus, you'll get way less "audio hash" through the cds cell from a light bulb VS an LED.

Nice smooth audio audio out...

-C

Not to mention that you don't need any sort of rectification or time constant manipulation with a light bulb. It does it all by itself. Just drive it with raw audio.

You will note on the clip I uploaded that even with an amazingly fast time constant, there is little bass-into-mid intermodulation. Smooth is right!

You can avoid the 'popping the filament with too much audio' problem by:
1) carefully adjusting the position of lightbulb to cell (as close as possible, isolated from outside light with lots of black tape and/or shrinkwrap)
2) using an amp whose peak output is not capable of 'popping' the lightbulb when 'open loop,'
3) adding loop gain until you get the sound you want.

Set up as a voltage divider, the resistor-to-cell-ratio is pretty important, and cell-dependent. And some cells are faster than others.
Caveat: If you add too much loop gain, you may get unwanted (capacitively-coupled) audio feedback. Which you probably don't want.
So keep the lamp/cell wires away from each other.

BTW, lamp/photocell compression is VERY 'undigital.' Results are extremely variable, and dependent upon the components.
Though once set up, performance seems pretty consistent over time, you may have a hard time getting the whole thing to gel.

While I'm sure it's possible to model in digital, you can't beat the price. Probably less than $10 is you scrounge a bit.

Kind Regards,
David

 

[cgould]

You can avoid the 'popping the filament with too much audio' problem by:
1) carefully adjusting the position of lightbulb to cell (as close as possible, isolated from outside light with lots of black tape and/or shrinkwrap)
2) using an amp whose peak output is not capable of 'popping' the lightbulb when 'open loop,'
3) adding loop gain until you get the sound you want.

Set up as a voltage divider, the resistor-to-cell-ratio is pretty important, and cell-dependent. And some cells are faster than others.
Caveat: If you add too much loop gain, you may get unwanted (capacitively-coupled) audio feedback. Which you probably don't want.
So keep the lamp/cell wires away from each other.

BTW, lamp/photocell compression is VERY 'undigital.' Results are extremely variable, and dependent upon the components.
Though once set up, performance seems pretty consistent over time, you may have a hard time getting the whole thing to gel.

While I'm sure it's possible to model in digital, you can't beat the price. Probably less than $10 is you scrounge a bit.

Kind Regards,
David

Yes...All of that is very obvious to me NOW...but back in the day when I was a very young pup still learning, it wasn't necessarily the case ;-)

I've been wanting to build a nice warm vocal compressor for my home studio, and all of this reminded me of things I've long since forgotten about...think I may have to dig around what's left in the parts bin and whip something up!

Lightbulb / photocell vocal compressor should sound sahweet!

-C

 

[rrsounds]

You can avoid the 'popping the filament with too much audio' problem by:
1) carefully adjusting the position of lightbulb to cell (as close as possible, isolated from outside light with lots of black tape and/or shrinkwrap)
2) using an amp whose peak output is not capable of 'popping' the lightbulb when 'open loop,'
3) adding loop gain until you get the sound you want.

Set up as a voltage divider, the resistor-to-cell-ratio is pretty important, and cell-dependent. And some cells are faster than others.
Caveat: If you add too much loop gain, you may get unwanted (capacitively-coupled) audio feedback. Which you probably don't want.
So keep the lamp/cell wires away from each other.

BTW, lamp/photocell compression is VERY 'undigital.' Results are extremely variable, and dependent upon the components.
Though once set up, performance seems pretty consistent over time, you may have a hard time getting the whole thing to gel.

While I'm sure it's possible to model in digital, you can't beat the price. Probably less than $10 is you scrounge a bit.

Kind Regards,
David

Yes...All of that is very obvious to me NOW...but back in the day when I was a very young pup still learning, it wasn't necessarily the case ;-)

I've been wanting to build a nice warm vocal compressor for my home studio, and all of this reminded me of things I've long since forgotten about...think I may have to dig around what's left in the parts bin and whip something up!

Lightbulb / photocell vocal compressor should sound sahweet!

-C

Well, if anyone can do it, it's gotta be you, CG!!

Kind Regards,
David

 

[gtomas]

I promised an Orban 8500 clip... Couldn't use Sweet Jane I'm afraid :-\ This is not a particularly representative clip beeing recorded off-air of a station that uses two STLs (one hop) to get to the transmitter, but it's as real-life as it gets The clip is already up, 50 µs preemphasis, Loud-Hot factory preset and all the disclaimers in regards to what goes in the processor and what happens after it... And, of course, don't make purchasing decisions based on this!


Regards,
Goran Tomas