Portable analog compressor?

[Rick Burnett]

Some opinions.
Vinyl does have a better signal, for about 5 plays. After that you have worn off the the high feq. parts of the groove. You get more plays with really high end gear but every play is destructive no mater what you do. Well there were those laser readers but oops now your back to digital...
I could be called wrong but analog (recordings) never had the dynamic range digital has. Analog sit's on top of a sea of noise and at the top you have mechanical (or magnetic) compression. You could never get away with the low level sounds that digital can capture. Tape sounds better when saturated than digital when overloaded but that doesn't constitute added dynamic range. The effective dynamic range is between the noise and the distortion and digital wins. The real issue is that a lot of people like the "inferior" signal of tape and vinyl. That is a mater of taste. So own it say you like the sound better. It doesn't really matter what the "best" is 99.9% of the time. If it did MP3 would not exist.

And I think I posted a long thread on this somewhere before but with Tape VS "digital", it is inherently a bogus argument. Why? Because magnetism is digital. When you get down to the lowest level it either is or it isn't, there isn't any "partly magnetized" particles. Tape is just a unclocked random distribution digital signal that is read by an averaging decoder. The "problem" , if there is one, with digital systems today is not that they are digital but the way that they encode/decode the signal. And electrons come in a finite size also so all electrical signals are also inherently digital.

Compressors: They are used differently in the music world than they are in production sound. In a lot of cases compressors are used to alter the sound in music production. That is exactly what you should not be doing in production. The poster who talked about listening to movies on DVD and realizing that they recorded a lot of noise and it was important, got it about 1/2 correct. Most of that ambient noise you heard (if they were big films) was added in post. But yes you don't want to go crazy trying to get a studio sound, it's not going to be possible. As a post person if I had my choice I would get production tracks that sounded totally clean and close miced. I don't want any of the "ambience" of the set on the tracks if possible. They will just get in the way of post. I spend a lot of time trying to get rid of all that stuff in post. Most sets sound nothing like what the director wants the location to sound like. Robins in a day for night scene are murder.

Limiters are often a good thing IF they stay out of the way except when that unexpected bang/ shout/ etc happens. AGC is not a limiter it is a very poor compressor like function. The problem with production sound and limiters is that you have a LOT of noise and a limiter is going to make that more of a problem. An expander could be useful, until the leading lady decides to whisper an important line. So really your best bet is to capture as clean a signal ass possible. Hot is good but clipping is not. It's a balancing act because too far down and you slide under the mud of the ambient sound (that post will try to make go away). One of the good reasons to hire someone who has some experience.

And lastly, having worked with a slew of boards big and small and an awful lot of mic preamps, I don't remember any having compression built in. THere was one but it was a pre / compressor as two units in one box. I have never seen compression on the channel strip. A lot of field mixers have limiters. And I have seen some boards with per channel limiters but not a lot. Now if your saying it's built into the A/D converter you might be right, but that's the converter not the mic pre.

The SD limiters, at least my read on the docks, look at the signal after the initial gain stage (MixPre also looks at the output) and acts on the signal before the initial gain stage. So in this case both of you are correct. The limiter is looking at a post mic pre signal but it's triggering a limiting function on the mic level signal.

I think that is highly misleading to try and tell people analog is digital because it's all the mechanisms for how the system works that is the actual differences in why they behave differently. The magnitude of difference between the sampling capability and the sampling distribution or SO radically different, they are not comparable in these terms. Analog refers to a particular type of electronic circuit. Just like the fact that Digital circuits are still analog circuits at the end of the day, but there is enough precedence in engineering to understand the difference. The argument shouldn't be Tape vs Digital, it should be Analog Tape vs Digital Tape, just like the differences between AM radio modulation and FM radio modulation. Sure, they are just modulated differently, but the results are VERY different.

To me digital means synchronous circuitry operated in a sampling mechanism. Even computers work this way.

Pertaining to listening to movies to get a starting point, this helps people see how much noise is actually going on in a scene. It's quite a bit, and goes against what recording music is typically like. I agree that I very often change my ambience in certain scenes, based on director needs, but that is NOT the rule of everything I have worked on or seen, or even talked about with other professionals as well. Most people record enough background tone to add in and slice out between the dialog if necessary. Controlling the ambience of a room is very hard, and cleaning out ambience not only takes a lot of time in dialog, it's very hard not to cause flanging even with the best noise reducers. I try to get as usable stuff on set as possible, ADR if conditions just did not warrant that (like recording in an echoey warehouse) and then supplement the scene with additional ambience.

Totally agree on the limiters. I would avoid an expander like the plague as well.

Most the the boards I am referring to are high end recording boards, like a Neve Genesis console. A friend just got one and every channel has a compressor on it. Many of the mic-pre rack mounts are just pulled from boards, like the SSL version, and it's just a tool there if you need it. On actual smaller or concern variety mixing boards, I have seen them on some of them, but many times they have a limiter on the input and run compression in the box digital it appears (like the D-consoles).

When I was shopping for a mic-pre, EVERY mic-pre I looked at that was $1500 to $5999 had a compressor built in and an EQ. Almost all of these were modern design. But, they are used when I am micing musical instruments and NEVER for voice EVER.

I'm actually surprised they are doing ANYTHING to the signal at the mic-pre signal. With those low levels, each level of additional circuitry adds noise. Of course, depends on the circuitry, maybe what they are using is LESS than the ADC step and by adjusting it there, the pre-amp noise is probably consistent.

I will say this, dynamic range in digital depends on audio or video. And, it depends on not the bit depth, but the SNR. Remember all the old 16-bit recording devices? Many of those were VERY noisy. Today, I absolutely agree that I think audio that is digital can get more dynamic range. That said, it still has ADC noise, and digital algorithms still introduce quantization errors, especially with algorithms like low pass filters. To me though, I'll take digital anyday!

If we consider film an analog process (although I am sure you are going to say it is digital since each of the molecules that reacts to light is finite! ) digital sensors are just not at that same level of dynamic range just yet. We are getting closer (with the Alexa and Epic hitting 13.5 stops).

But honestly, at the end of the day, I know NO ONE that records audio on set to analog tape. Digital tape, yes, like a DAT, but not analog.

 

[Rick Burnett]

Not exactly true. I mean it is, but it is NOT the production recorders job do get the ambience, that is done in post. They should be concerned if they are getting a lot of noise. But the cure is probably to get the boom in more aggressively, you want to be as close as possible.




Two points. If it's just bringing down the top, it's a limiter not a compressor. Second "digital" compressors are modeled after analog ones and they act the same. A compressor limits the top and raises the bottom, if it's just playing with the top it's a limiter not a compressor.

I agree it's not the job, but the point is, the gain control is not going to change the difference between the ambience and the talent. So if you record that at the top or the middle of your 24-bit dynamic range, that difference is still there. The only think you are fighting is the ADC noise but with todays recorders, it is WAY lower than the ambience typically (unless you are recording just ambience, and you gain to that ambience, you can sometimes get into situations where you hear it. But, once you bring that ambience back in and set it down to the levels that make sense in scenes, that ADC noise is pushed back down again.)

I agree with getting that boom mic in there, mostly because the more you get the talent, and the less you capture the verb in the room too high, the better it sounds AND the better you can tweak it in post.

Compressors, limiters and expanders are all the same essential circuitry used in three different ways, and many times, with certain parameters fixed. If I bring the top down on a signal, I am sorry, but it is compressing the signal. That is EXACTLY what you are doing to the waveform even if you just push the highs down, because you brought up the lows so you could compress the dynamic range into a smaller one. The difference typically between limiting and compression in terms of a technical level is your ratio. Most limiters I have used are infinite ratio, whereas when you discuss compression, most people use a ratio depending on the source material and what sound right, so that the compression is just a pulling down of those peaks by a ratio, instead of a hard edge. Same thing with an expander, you are expanding your signal into a larger range.

Just because you change the gain on your preamp and not changing the gain ON the compressor gain setting doesn't all of a sudden make what you are doing different, it's the same thing. You are increasing the lows in both locations and pushing down the highs. SAME thing. Again, the difference is how aggressive you are with pushing those highs down. Setting a compressor to the max ratio means you are functioning like a limiter, you want to guarantee that no output EVER goes above a certain amount.

And digital or analog makes no difference in terms of what you are accomplishing. People emulate analog compression to get the characteristics of how that compression sounds, including saturation inside the circuits. That said, there can be digital compression as well, it's just most people don't like the sound of it as much because it's too clean and doesn't color the signal). I use digital compression when I don't want color on a sound. Just like digital EQs. I use them when I need a surgical fix that is ESSENTIAL not to change the character of a sound.

 

[Rick Burnett]

I wanted to add, great discussion. It is interesting to read about others approaches to working on audio. It's an often overlooked aspect of film making and most independent films I see are SEVERELY lacking in this area. It is amazing how much audio can make a not as well shot movie better, but that bad audio ALWAYS ruins a beautifully shot movie.

Like a sound recorder once told me. "The best audio in a movie is the audio you don't notice. That means everything feels natural, sounds good, and doesn't detract from the experience." I've always thought that was good advice. The last thing you want to do is distract the audience from the story.

 

[timbook2]

as I said before: ride the volume fader during record! forget compressors!
and yes: vinyl sounds better than any 44.1 KHz 16 bit CD will ever be able to ....but its so easy to use

 

[Noiz2]

Compressors, limiters and expanders are all the same essential circuitry used in three different ways, and many times, with certain parameters fixed. If I bring the top down on a signal, I am sorry, but it is compressing the signal. That is EXACTLY what you are doing to the waveform even if you just push the highs down, because you brought up the lows so you could compress the dynamic range into a smaller one. The difference typically between limiting and compression in terms of a technical level is your ratio. Most limiters I have used are infinite ratio, whereas when you discuss compression, most people use a ratio depending on the source material and what sound right, so that the compression is just a pulling down of those peaks by a ratio, instead of a hard edge. Same thing with an expander, you are expanding your signal into a larger range.

Actually having built them this isn't really true. A compressor, old school analog anyway, rides gain to an extent a bit like a AGC and you are affecting the low level signals also. A limiter doesn't do this. It is true that all use similar circuitry but they perform very different functions. A screwdriver and a hammer are both tools made of steel but they, generally, are not interchangeable. In a general kind of way you are right in that the ratio tends to be higher in limiters than compressors but that is not hard and fast. I have some compressors with very high ratios and I have seen limiters than can go to pretty low ratios. If you count anything that changes the dynamic range a compressor then that would also include EQ and overloading your inputs and well just about everything in the chain has some affect on dynamic range, or at least the usable dynamic range.

Just because you change the gain on your preamp and not changing the gain ON the compressor gain setting doesn't all of a sudden make what you are doing different, it's the same thing. You are increasing the lows in both locations and pushing down the highs. SAME thing. Again, the difference is how aggressive you are with pushing those highs down. Setting a compressor to the max ratio means you are functioning like a limiter, you want to guarantee that no output EVER goes above a certain amount.

This kind of contradicts your "the difference is the ratio". The circuits are similar they are not the same and the functions are different.

And digital or analog makes no difference in terms of what you are accomplishing.

That is what I said. Someone else said that "digital" compressors worked on the signal differently.

That said, there can be digital compression as well, it's just most people don't like the sound of it as much because it's too clean and doesn't color the signal). I use digital compression when I don't want color on a sound. Just like digital EQs. I use them when I need a surgical fix that is ESSENTIAL not to change the character of a sound.

And this thread is on production sound and the job of the production mixer is not to change the sound but to get as clean and accurate a dialog recording as possible.

 

[Noiz2]

I think that is highly misleading to try and tell people analog is digital because it's all the mechanisms for how the system works that is the actual differences in why they behave differently. The magnitude of difference between the sampling capability and the sampling distribution or SO radically different, they are not comparable in these terms. Analog refers to a particular type of electronic circuit. Just like the fact that Digital circuits are still analog circuits at the end of the day, but there is enough precedence in engineering to understand the difference. The argument shouldn't be Tape vs Digital, it should be Analog Tape vs Digital Tape, just like the differences between AM radio modulation and FM radio modulation. Sure, they are just modulated differently, but the results are VERY different.

To me digital means synchronous circuitry operated in a sampling mechanism. Even computers work this way.

Well, hmmm. AM and FM are far more different than you think but that is a tangent. Bit stream digital is close to analog in the way it deals with the bits of information that comes in. Also all the compression algorithms are in one way or another modeling analog signal flow. And they are all digital. It's fine to say they don't match what YOU thing of as digital but it's not OK to expand that to a more general definition. It is I think a better way of looking at the continuum to realize what is really going on rather than marketing names. A lock step fixed sample size digital stream is what is most common for "uncompressed" digital signals. But it is not what is most common for compressed formats which are a lot smarter about how they package the information. And it IS important to break down the false distinctions between "digital" and "analog". Once you realize the problem isn't digital VS Analog but one form of encoding VS another THEN you can have useful discussions about the state of things and how to make them better. As long as your locked into a 1980's definition of what "digital" is and have some fixed idea of what "digital" sounds like VS what "analog" sounds like there is no chance for a real dialog.

The future (as if I knew) I think is in bit stream type formats. They work much like analog in that it's looking at concentrations of information rather than stupidly looking at a slice of time and giving the same importance to dead air as it does to a complex signal. MP3's and such have a bad rep because they have been used to shrink the information flow. But if you reverse that and keep the flow about the same but instead make your encoding smarter... An MP3 at the same data rate holds a LOT more detail than a WAV file. Even computers are not really working in the way you describe anymore. With multithreading it is not operating in a "synchronous circuitry operated in a sampling mechanism" way. It's arguable that each thread is on it's own doing that but then your back on my side in saying magnetic tape is digital, that the true nature of the system is defined by it's basic components not the general impression of it's functioning.

Pertaining to listening to movies to get a starting point, this helps people see how much noise is actually going on in a scene. It's quite a bit, and goes against what recording music is typically like. I agree that I very often change my ambience in certain scenes, based on director needs, but that is NOT the rule of everything I have worked on or seen, or even talked about with other professionals as well. Most people record enough background tone to add in and slice out between the dialog if necessary. Controlling the ambience of a room is very hard, and cleaning out ambience not only takes a lot of time in dialog, it's very hard not to cause flanging even with the best noise reducers. I try to get as usable stuff on set as possible, ADR if conditions just did not warrant that (like recording in an echoey warehouse) and then supplement the scene with additional ambience.

Well since you mentioned watching DVD's of films and the implication was that you were talking about major features not small indi no budget films I would have to say you were talking to the wrong professionals. The rule of thumb on a Hollywood film is that EVERYTHING except dialog is done in post. Yes it does take a fair amount of work, less if your working on shots from a big budget film since they have actually taken steps to eliminate as much of that as possible. Shoes on a big film, for instance, have a padding added to the soles so they don't make much noise. Cups and plates etc also have a pad on the bottom to cut down on the noise they make.

So if you were listening to films with budgets under say $100,000 then maybe you were hearing a lot of production sound. If it was say $5mill. plus then it's very unlikely than you heard ambiences from production and if it was $30 mill. and up you weren't hearing any production Amb at all.

When I was shopping for a mic-pre, EVERY mic-pre I looked at that was $1500 to $5999 had a compressor built in and an EQ. Almost all of these were modern design. But, they are used when I am micing musical instruments and NEVER for voice EVER.

We have an Avalon it's in that range and doesn't (or at least didn't when we bought it). But I haven't been shopping in awhile so maybe that is the new thing? I have a variety of pre's here that are just under that range (~~$600-$1300) and none of them have compression built in, but again they are all at least ten years old so?

If we consider film an analog process (although I am sure you are going to say it is digital since each of the molecules that reacts to light is finite! ) digital sensors are just not at that same level of dynamic range just yet. We are getting closer (with the Alexa and Epic hitting 13.5 stops).

I'm not going to go that small, and color film uses dye's. Ironically I did once work on a process for doing "digital film". Essentially a halftone but with the small dots at around three microns. This was a purely digital film but the resolution was finer than the silver particles in film and printing paper so resulting prints looked like conventional film, except that it was archival and you could blow it up quite a bit and not loose detain since what you were printing on was a lower resolution than the film. Visual needs a lot bigger bandwidth and yes indeed the dynamic range of video is not up to film. I wouldn't agree that the same is true with sound. If I were going to be picky I would say that film and audio have the opposite issues at the moment. That sound needs improvement in it's resolution more than dynamic range and visa versa. They are all pretty damn good though.

But honestly, at the end of the day, I know NO ONE that records audio on set to analog tape. Digital tape, yes, like a DAT, but not analog.

Wow DAT, I didn't know you could still get tapes. Another touch of irony is that Nagra's are still used a bit in FX recording. Not a lot but that tape compression helps on a few things. I can't imagine a feature using tape though for production. The cost would be a factor (mostly in time).

 

[Noiz2]

vinyl sounds better than any 44.1 KHz 16 bit CD will ever be able to ....but its so easy to use

That is just such a silly statement. Far more depends on the quality of the recording, mastering, playback system, media, etc. I can pull any number of LPs out of boxes in the basement and unless your a big fan of snap crackle and pop they don't sound better than an MP3 of the same thing off a not great CD.

CD's got a bad rap because of what was being put on them and the early players sucked. People go off on this tangent either because it's the "hip" stance or because they are comparing a high end turntable connected to a great system do a low rent CD on a crappy system.

But if your serious I'll give you my paypal address and you drop some change in there and I'll ship you a couple of crates of LP's. Cover art and nostalgia is about all they have going for them.

 

[bang]

here some good infos about LPs vs CDs
http://www.audioholics.com/educatio...ology/dynamic-comparison-of-lps-vs-cds-part-4

and here a tape vs digital recording test
http://www.youtube.com/watch?v=AgH99wyRoVc&feature=related

 

[Rick Burnett]

My statements were not contradictory in terms of compression and limiting, they are doing the exact same function with different ratios. In both scenarios the gain is driven up before the compression circuitry. Both circuits have a threshold to which once that threshold is passed, a gain reduction circuit is used to lower the level of the incoming signal. The difference between the two is in how much that level is reduced, how fast the reduction is done, and how fast it recovers. In most cases, a limiter is a compressor with fixed values that you cannot change, but the circuitry is essentially the same (I only say it is different in that, if you aren't making a variable control, like knee, attack, etc, then that circuitry will be different).

And I hate to break it to you, but multiprocessing systems still work is a sampled syncronous system. The hand-off of data is still happening off a common clock, it's just there are multiple CPU's that can perform simultanously, but synchronously as well. It makes NO sense to make an asynchronous system because you degrade performance for no gain. I should know, I've been a chip designer for 15 years and have quite a bit of IC experience.

Understanding the difference between analog and digital is not different today. People who lack knowledge of digital are constantly getting into problems, especially in large digital system requiring common stable clocks to keep from introducing artifacts and glitching into a digital system. Even further though, it's the sampling nature of a synchronous system that defines the difference between a digital and analog system. Once you sample data at a specific periodic rate, you allow the system to operate under a totally different set of parameters. For instance, once you've gone past your ADC, noise in the system is not as critical because the threshold for bit detection is quite large, so you've eliminated a source of distortion in the system. With analog, this isn't so. The very nature of analog means EVERYTHING you signal goes near, or through, is a potential problem. EVERY component in the system, whether resistors or capacitors, op-amps, etc have the possibility of not only coloring the sound, but causing imbalance. tolerances are HORRIBLE in analog equipment, which rightly so, is why they are so expensive. The whole reason people turn their amps on a half hour or more before they use them in high end circuitry is so the system can warm up, and equalize in operation due to the analog components.

You won't have to argue with me about digital versus analog, I'm digital as I have found the speeds which we run digital now, if you are careful to work WITHIN the exact dynamic range of your digital bit-depth, are AWESOME today. It's just certain things, like amplifiers and speaker cabinets, that NEVER work exactly the same (yet) because the physics modeling hasn't happened. Most of these systems are sampled with impulse responses, but not all analog systems are non-causal. Speaker cabinets are DEFINITELY causal as are some compressors, (which essentially means that the history of what they are doing affects what they are doing now).

It's also nice that digital offers other options like look-ahead compression, which I've personally not witnesses in an analog compressor.

But back to the topic (sorry for the tangent), as said, there are multiple reasons for a compressor. For recording purposes, it has always been that if the dynamic range of the input source is too high, like drums for instance, you can use the compressor to bring down the high parts and allow the low parts to get recorded higher. However, since the SNR of digital recording is so good these days, typically the room noise is louder than amplifying the ADC noise. Therefore, the determination of using the compressor really depends on how far up you need to bring the low parts of a recording compared to how quiet your room is. Having recorded in an anechoic chamber before, (NO echos AT ALL), if you had a room somewhere between this extreme and just a regular room, I could see using a compressor. I use mine with a VERY small ratio just to pull off some really high parts from clipping. As clipping in the digital world is MUCH worse than the analog world for the most part. (The analog world doesn't typically hard clip like the digital world which just maxes out with no saturation).

And, the Avalon I was looking to purchase DEFINITELY has a compressor:

http://www.sweetwater.com/store/detail/VT737SP/

But the truth is, you can but their products individually, or in one unit.

This is what I use:

http://www.sweetwater.com/store/detail/ISA220/

(which contains both a compressor AND a limiter at the end for obviously different purposes)

It is true I have no worked on multi-million dollar films myself. However, I have talked with many an engineer who has, as they took my work after I finished to do the final mixdown in 5.1 of the films I worked on. For smaller budget films, it's true, we try to get as much GOOD sound as we can with the onset recordings. In fact, I discussed these same topics with the DP of Phantasm just last year, and he told me how they preferred to record sound (as I definitely needed help), and they too tried to get as much as they could in the microphone and then only fix what they needed to fix in post.

That said, we NEVER wanted sounds going off under the actors voices so we always avoided that. Second, while I said we always wanted ambience in the recording (and I still stand by that) the separation of the ambience to the vocals we want as high as possible. We want the boom as close to the actors as possible. In addition, if the location has air conditioning vents, we try to make sure that when we move the boom mic around, we aren't capturing the wind say of the vents in the mic.

But, truth is, everyone has their own method of working. Test test test, it's the best way. Always think about the end result, and before investing a whole bunch of time into your process, go through it with something short, and make sure that the methods you are using are acceptable to wherever your work is going

 

[Noiz2]

My statements were not contradictory in terms of compression and limiting, they are doing the exact same function with different ratios. In both scenarios the gain is driven up before the compression circuitry.

This may be true for some analog compressors but certainly not all. Optosensors used in at least some compressors alter the gain of the compressor and turn it UP when the signal gets below a certain point. I have built those so I can tell you that it is not a fixed gain system. Assuming that "digital" compressors have the same capability as software compressors, they also can vary the gain. The compressor I use the most has four bands that can compress/ expand in various ways. If all you saw was the out put you would see compression but in actual fact there is band limited gain change going on also. In fact you could set it up to compress up only so it didn't lower any levels but just pushed up the low level ones.

I think the useful info here is that the similarity is that both limiters and compressors "compress" the dynamic range. As a general rule limiters are preferred for production because in theory they won't alter the sound until you would have clipped and they thus offer a safety net. You may be able to do the same thing with some compressors but it's not the function they were designed for.

And I hate to break it to you, but multiprocessing systems still work is a sampled syncronous system. The hand-off of data is still happening off a common clock, it's just there are multiple CPU's that can perform simultanously, but synchronously as well.

As I said it all depends on how fine you want to break it down. You want to go down to the "molecular" level of computing but balk if I go to the "molecular" level of analog. I think you need to look at both on the same scale or it's a pointless argument.

It makes NO sense to make an asynchronous system because you degrade performance for no gain. I should know, I've been a chip designer for 15 years and have quite a bit of IC experience.

Two points. First there can be great advantage in asynchronous systems in terms of speed and robustness. Synchronous systems are inherently fragile. A simplistic example is the storage of sound. Files on an old RLL formated HD from the 80's are completely useless. Why? because of the fragility of the technology. You couldn't fin a machine today to even mount that drive on. If you did the format is probably not readable of either the drive or the file. CDR's just fade out, digital optical tape needs to be rewritten every 6 mo. to a year because of data drift, Floppies are quickly fading from the playable category, 5 1/4 and 8" floppies have already disappeared. A lot of sound was stored on Pinacle MO drives, I have a few for archive material but you can't get them now. JAZZ drives and ORB drives were used a lot or sound, neither is usable anymore. DAT tapes are going away. This is all technology "fragility" and it applies because it is those technologies that are required to support synchronous data storage. If you look at the asynchronous data storage you have film, Lp's, Dictaphones, paper, magnetic tape, magnetic film, optical film, wire recorders, cylinder recorders. Of all of those only wire recorders and dicktaphones would give you much of a hunt to find a playback device. Some of those are over one hundred years old and still functional (if you include music written on paper you go back a lot longer than that). Some have a level of age fragility but most have fared better than digital storage in that area also. If you go closer to the core of the systems the asynchronous is more robust because the difference inherent in the systems. Synchronous falls apart if small amounts of data are missing. There are ways of building in some redundancy but inherent in the system is the need for things to be where they are expected. Asynchronous is very error tolerant. You can loose a tremendous amount of your original data and still be able to extract the core information.

Oh the second point is that experience gives one knowledge of the past, but not nec. a good clue about the future. That doesn't mean I have any better clue about the future, you might be dead on? But it is kind of a pointless argument when talking about the future. I have a fair amount of big budget film experience. I can tell you with some authority how things were done and with slightly less authority how they are done now (I haven't worked on a big budget in the last couple of years years) but I can't tell you with any credibility of how things will be done. I'm fairly good at seeing trends but in the end it's just my opinion.

Related to that part of my opinion on the advantages of asynchronous storage is based on my experience with archiving shows and the problems that our new digital age have brought us in the film world. There are shows less than ten years old that will never be able to be remastered because there is no viable copy of the post sound work. You can in most cases get the elements of any film back into the thirties or so, but a lot of films in the late 80's and 90's don't have anything unless they printed stems to mag. It's been a BIG issue for Hollywood because that represents a lot of lost future revenue.

Understanding the difference between analog and digital is not different today. People who lack knowledge of digital are constantly getting into problems, especially in large digital system requiring common stable clocks to keep from introducing artifacts and glitching into a digital system. Even further though, it's the sampling nature of a synchronous system that defines the difference between a digital and analog system. Once you sample data at a specific periodic rate, you allow the system to operate under a totally different set of parameters. For instance, once you've gone past your ADC, noise in the system is not as critical because the threshold for bit detection is quite large, so you've eliminated a source of distortion in the system.

I would agree to a certain extent. You have also added a whole new layer of possible distortion though. Your swapping analog noise and distortions for digital distortions. The key being that you in theory have a better chance of eliminating digital distortions than you do eliminating the analog ones.

With analog, this isn't so. The very nature of analog means EVERYTHING you signal goes near, or through, is a potential problem. EVERY component in the system, whether resistors or capacitors, op-amps, etc have the possibility of not only coloring the sound, but causing imbalance.

This is also true of digital, it's just more software oriented. Some bad math in an algorithm can trash your sound as fast as a blown capacitor. The big difference is that one is repairable. And chips go bad as often as components.

tolerances are HORRIBLE in analog equipment, which rightly so, is why they are so expensive. The whole reason people turn their amps on a half hour or more before they use them in high end circuitry is so the system can warm up, and equalize in operation due to the analog components.

By what standards? Tolerances are horrible by digital standards but asynchronous systems are much more forgiving and don't need the high tolerance of a synchronous system. Now we are talking about sound here and it's important to remember that. Your dealing with a very subjective field where Joe calls it distortion and Bob calls it pleasantly warm. If this were brain surgery the "loose" accuracy of analog would be a problem but were talking about a system where no mater what you do it is going to sound different to every individual listening and in every different playback space.
So I agree that compared to a digital system analog is loosey goosey, but it actually in most cases doesn't matter much or at all.

You won't have to argue with me about digital versus analog, I'm digital as I have found the speeds which we run digital now, if you are careful to work WITHIN the exact dynamic range of your digital bit-depth, are AWESOME today. It's just certain things, like amplifiers and speaker cabinets, that NEVER work exactly the same (yet) because the physics modeling hasn't happened. Most of these systems are sampled with impulse responses, but not all analog systems are non-causal. Speaker cabinets are DEFINITELY causal as are some compressors, (which essentially means that the history of what they are doing affects what they are doing now).

I'm with you. I love analog gear, I grew up with it and it just works. But I almost never use it because it's just so much slower in a post workflow where real time seems slow.
Side note. I did some picture editing the other day. used some filters and FCP wanted 6 hours to render a 3 min piece!!!!

And flexibility. There is no way to do the kind of surgery I do every day if I were on an analog system.

It's also nice that digital offers other options like look-ahead compression, which I've personally not witnesses in an analog compressor.

I don't think it would be possible. You would have to delay the signal and doing that in the analog realm is not usually a tasty choice.

But back to the topic (sorry for the tangent), as said, there are multiple reasons for a compressor. For recording purposes, it has always been that if the dynamic range of the input source is too high, like drums for instance, you can use the compressor to bring down the high parts and allow the low parts to get recorded higher. However, since the SNR of digital recording is so good these days, typically the room noise is louder than amplifying the ADC noise. Therefore, the determination of using the compressor really depends on how far up you need to bring the low parts of a recording compared to how quiet your room is. Having recorded in an anechoic chamber before, (NO echos AT ALL), if you had a room somewhere between this extreme and just a regular room, I could see using a compressor. I use mine with a VERY small ratio just to pull off some really high parts from clipping. As clipping in the digital world is MUCH worse than the analog world for the most part. (The analog world doesn't typically hard clip like the digital world which just maxes out with no saturation).

If I didn't say it before, there aren't really any rules except what gives you the best results. Compressors are not typical on location recording but I'm sure there are a few experienced pros who do it because they have made it work for them.

It is true I have no worked on multi-million dollar films myself. However, I have talked with many an engineer who has, as they took my work after I finished to do the final mixdown in 5.1 of the films I worked on. For smaller budget films, it's true, we try to get as much GOOD sound as we can with the onset recordings.

I often end up using a fair amount of production sound on small budget films also, you don't have the time or crew usually to a Hollywood style post.

In fact, I discussed these same topics with the DP of Phantasm just last year, and he told me how they preferred to record sound (as I definitely needed help), and they too tried to get as much as they could in the microphone and then only fix what they needed to fix in post.

This was the DP. I don't know him so he might be a closed sound guy, but DP's often don't know squat about sound except that they don't want that guy with the furry stick anywhere near "his" shot. And even the production mixer know hardly anything about what goes on in post on a big film. They are all long gone by the time post sound happens. So with no intention to slight the guy in anyway I would take anything he said about the sound work with a healthy heap of salt. There is also very little to get on a big budget film. As I mentioned everything has been dampened down and they have all this noisy equipment going. A low budget film by it's nature has a lot more that can be gotten on set. Do what works for you, but I wouldn't modle it by what a DP told me was done, they probably don't really know (just think they do).

That said, we NEVER wanted sounds going off under the actors voices so we always avoided that. Second, while I said we always wanted ambience in the recording (and I still stand by that) the separation of the ambience to the vocals we want as high as possible. We want the boom as close to the actors as possible. In addition, if the location has air conditioning vents, we try to make sure that when we move the boom mic around, we aren't capturing the wind say of the vents in the mic.

But, truth is, everyone has their own method of working. Test test test, it's the best way. Always think about the end result, and before investing a whole bunch of time into your process, go through it with something short, and make sure that the methods you are using are acceptable to wherever your work is going.

That IS the bottom line.
Cheers
SK

 

[Rick Burnett]

This may be true for some analog compressors but certainly not all. Optosensors used in at least some compressors alter the gain of the compressor and turn it UP when the signal gets below a certain point. I have built those so I can tell you that it is not a fixed gain system. Assuming that "digital" compressors have the same capability as software compressors, they also can vary the gain. The compressor I use the most has four bands that can compress/ expand in various ways. If all you saw was the out put you would see compression but in actual fact there is band limited gain change going on also. In fact you could set it up to compress up only so it didn't lower any levels but just pushed up the low level ones.

I think the useful info here is that the similarity is that both limiters and compressors "compress" the dynamic range. As a general rule limiters are preferred for production because in theory they won't alter the sound until you would have clipped and they thus offer a safety net. You may be able to do the same thing with some compressors but it's not the function they were designed for.

I will definitely give you that there are exceptions to the rule all over the place. Especially with multi-band compressors because they are essentially band pass filters with compressors so that you can apply different compression ratios to different bands of audio. I use two of them actually, the iZotope Ozone 3 for mastering and the Multiband precision from UAD. While I am sure there are hardware equivalents to some of them, many of the digital ones that are not trying to recreate a certain piece of gear definitely have added in A LOT of bells and whistles (which has been nice).

And yes, initially compressors and limiters are designed for different purposes, however, understanding how they work is how you'd know that in most cases, you can use a compressor as a limiter (or vice versus) if there are characteristics of those pieces of equipment or software that you want to use. I've used both for both in certain circumstances with great results. But then again, I like to experiment with gear and find out just how useful it can be!

As I said it all depends on how fine you want to break it down. You want to go down to the "molecular" level of computing but balk if I go to the "molecular" level of analog. I think you need to look at both on the same scale or it's a pointless argument.

I am not going to the molecular level of computing. I'm comparing from the same level of each as much as possible. This is why I brought up AM vs FM. At the molecular level (if you will since they are both RF modulations) they are the same thing. Both are RF energy. The difference is in how they are modulated, and as I am sure you are aware, FM has a superior retention for amplitude and frequency information because of the way it is encoded. This is the comparison I am making between analog and digital. Sure, each has their strengths and weaknesses, but to say they are the same is misleading, because at a practical level of looking at the pros/cons, they ARE different. And that's what people care about, and what makes sense. Talking about them at the molecular level really adds no value to discussing them in this context. It's like saying, hey, humans are made of molecules and so are DVD players, so they are essentially the same. Trying to compare the least common denominator doesn't help prepare you for understanding what they to expect when using them.

Two points. First there can be great advantage in asynchronous systems in terms of speed and robustness. Synchronous systems are inherently fragile. A simplistic example is the storage of sound. Files on an old RLL formated HD from the 80's are completely useless. Why? because of the fragility of the technology. You couldn't fin a machine today to even mount that drive on. If you did the format is probably not readable of either the drive or the file. CDR's just fade out, digital optical tape needs to be rewritten every 6 mo. to a year because of data drift, Floppies are quickly fading from the playable category, 5 1/4 and 8" floppies have already disappeared. A lot of sound was stored on Pinacle MO drives, I have a few for archive material but you can't get them now. JAZZ drives and ORB drives were used a lot or sound, neither is usable anymore.

First, RLL and even MFM drives are using a technology that is not inherently digital, in fact, the technology they are using, just like modern hard drives, is magnetism. Just like tape, which analog has been recorded to for a LONG time. This is not a digital issue, this is a storage issue, and analog suffers from these same issues. There are TONS of analog and digital data that get ruined from cheap storage means. Just like paper, and if you don't use archival strength paper OR CDRs, they do have shelf lives.

The first CDs from the 80s still work just fine, and sound just as good as the day I got them, after millions of listens. Unlike an analog record, which I also have a few, that degrade rapidly with every listen. In this case, for retention of data, the digital is superior.

And I have floppy disks from probably 10 years or more that still work. Your issue is with magnetism and I agree it's not very good. But analog recordings can lose data as well, and both have to be rewritten. Which one is easier? Depends on the damage. I've seen both cases and each have been difficult to repair the missing information.

DAT tapes are going away.

In audio recording yes, but in digital data retention absolutely not. There aren't any solutions that can backup huge amounts of data. We use this technology where I work ALL the time because it works. I've restored project data from YEARS ago with no issue. Since we backup thousands of gigabytes of data, this is our only option and we've not had a problem yet. Our systems are robotic for changing out the tapes. DAT's are going away from audio because they are linear recording devices and the other digital mechanisms, that are just file bases, are SO MUCH easier to work with. It all depends on the freedoms you need and the size you need.

This is all technology "fragility" and it applies because it is those technologies that are required to support synchronous data storage. If you look at the asynchronous data storage you have film, Lp's, Dictaphones, paper, magnetic tape, magnetic film, optical film, wire recorders, cylinder recorders. Of all of those only wire recorders and dicktaphones would give you much of a hunt to find a playback device. Some of those are over one hundred years old and still functional (if you include music written on paper you go back a lot longer than that). Some have a level of age fragility but most have fared better than digital storage in that area also.

Again, you are confusing the media with the data from a few specific instances. Lp's are not cheap to make, I know, we had a lathe at a studio I've worked with. Just like if you master a CD and not use a CDR, it will last for A LONG time. Just like if you use a tape player (analog style) and tapes get stuck, break, the head wears out, all these problems plagued analog for YEARS. They could record digital onto every one of those mediums for all the advantage and disadvantage. The problem here isn't with digital, it's with recording technologies in general. They are all pretty bad.

And the funny thing is, I can get a JAZZ drive, I can get a ZIP drive, because I know people who were smart enough to get enough of the hardware in case the company went out of business. When that happened, they transistioned off the hardware. Heck, if someone wanted to they could print out digital on paper and then read it in with OCR technology in the future if needed. Again, the issue isn't analog vs digital, which this was about, it is about modern recording techniques of ANY data type and how they are not archival safe. (which many people don't know).

I personally use RAID arrays.

If you go closer to the core of the systems the asynchronous is more robust because the difference inherent in the systems. Synchronous falls apart if small amounts of data are missing. There are ways of building in some redundancy but inherent in the system is the need for things to be where they are expected. Asynchronous is very error tolerant. You can loose a tremendous amount of your original data and still be able to extract the core information.

Again, that depends on a lot of things. Data critical systems use checksums and parity information to deal with these issues. It is true an asynchronous system handles small errors better, but those are NEVER the problems I typically see. I've NEVER had a digital audio file go bad, ever. It all comes down to understanding where faults can happen. Digital gives you more protection of contamination of your data because it adds another level of encoding to it. Granted, that encoding also introduces distortion, but you are trading a smaller distortion (in most cases) for a larger protection in your data.

Related to that part of my opinion on the advantages of asynchronous storage is based on my experience with archiving shows and the problems that our new digital age have brought us in the film world. There are shows less than ten years old that will never be able to be remastered because there is no viable copy of the post sound work. You can in most cases get the elements of any film back into the thirties or so, but a lot of films in the late 80's and 90's don't have anything unless they printed stems to mag. It's been a BIG issue for Hollywood because that represents a lot of lost future revenue.

Again, these aren't digital problems, these are archival problems. RLL and MFM drives are not impossible to extract from, but it will cost money. I agree that a LOT of bad decisions were made in many of the recorded arts. Not only with audio either, but how many films have been lost, erased, that studios have looked for.

One thing I always hated about analog film, if you don't go see the movie around opening night, the film looks HORRIBLE after about a week of viewing. At least with digital projection, the film doesn't degrade.

I would agree to a certain extent. You have also added a whole new layer of possible distortion though. Your swapping analog noise and distortions for digital distortions. The key being that you in theory have a better chance of eliminating digital distortions than you do eliminating the analog ones.

Absolutely agreed (as with my statements above), you are making trade-offs

This is also true of digital, it's just more software oriented. Some bad math in an algorithm can trash your sound as fast as a blown capacitor. The big difference is that one is repairable. And chips go bad as often as components.
[/QUOTE=Noiz2;2065082]

Bad math yes, absolutely, but pertaining to chips going bad, when they go bad they don't go undetected. I have NEVER EVER lost audio from a chip going bad, nor had distortions caused by such a thing. I can say however, that analog synths and analog power supply issues have BOTH given me problems. This goes back to my discussion on the data protection.

By what standards? Tolerances are horrible by digital standards but asynchronous systems are much more forgiving and don't need the high tolerance of a synchronous system. Now we are talking about sound here and it's important to remember that. Your dealing with a very subjective field where Joe calls it distortion and Bob calls it pleasantly warm. If this were brain surgery the "loose" accuracy of analog would be a problem but were talking about a system where no mater what you do it is going to sound different to every individual listening and in every different playback space.
So I agree that compared to a digital system analog is loosey goosey, but it actually in most cases doesn't matter much or at all.

I'm referring to the ability of the audio to retain it's data without distortion. Digital works on thresholds where they can be designed with a wide enough voltage differential that even really harsh noise won't flip bits. That same level of noise in an analog system with the same voltage differential capability has no protection to the noise. The digital can remain absolutely clean with no extra distortion from passing through whatever transmission medium. Understanding what the limitations on each encoding mechanism is important to understanding not only what compromises you are giving up, but what limitations you need to be worried about. Which again is the whole reason I use a $2k 2 channel ADC to get the most pristine digital input I can get and then work with it in digital. Since I know that my material will be listened to on many different playback systems, I want it as perfect as possible so poor systems don't sound worse because of my source material.

But then again, this is a huge tangent, which illustrates that, between even me and you who are very knowledgable, we both have a lot of differences in how we feel and treat analog and digital. Just imagine how it feels to hit these questions when you are new to the subject? EVERYTHING we've discussed is important, ESPECIALLY archiving data. Yet, they aren't discussed that much.

They each have their strengths and weaknesses, understanding that is important.

I'm with you. I love analog gear, I grew up with it and it just works. But I almost never use it because it's just so much slower in a post workflow where real time seems slow.
Side note. I did some picture editing the other day. used some filters and FCP wanted 6 hours to render a 3 min piece!!!!

Yeah, I agree. I used to patch in a bunch of gear into my system (as I use Logic Pro, ProTools and Digital Performer) but when I can't bounce non-real time anymore, I was like whoa! This is taking WAY TOO LONG!

The editor I work with was having that discussion with me the other day about FCP and filters. I do 3D rendering as well. I had a 30 second helicopter I animated take 28 hours!!!

And flexibility. There is no way to do the kind of surgery I do every day if I were on an analog system.

ABSOLUTELY. I worked on a film where the sound tech let the recording peak out on some yelling segments. I had to go in by hand and draw in the waveforms That was SO painful, but this yelling scene was not able to get recut and the guy was yelling in italian (we couldn't get him back) and none of us spoke italian to ADR the scene! Now I meet with the sound guys first to discuss

I don't think it would be possible. You would have to delay the signal and doing that in the analog realm is not usually a tasty choice.

True, like a tape or optical delay, which I've seen, but they add another level of circuitry and honestly, cause time delay problems. Nice thing about current DAW technology is that if you use a lookahead, the DAW (except protools native) will alter the playback of ALL tracks so they are all in time. Absolutely fantastic!!!

If I didn't say it before, there aren't really any rules except what gives you the best results. Compressors are not typical on location recording but I'm sure there are a few experienced pros who do it because they have made it work for them.

True. I would never use one on recording because I don't want to doctor up the spoken audio. I tell people to record voice with and without a compressor and listen to them and they will see just why they don't want to do that for film.

This was the DP. I don't know him so he might be a closed sound guy, but DP's often don't know squat about sound except that they don't want that guy with the furry stick anywhere near "his" shot. And even the production mixer know hardly anything about what goes on in post on a big film. They are all long gone by the time post sound happens. So with no intention to slight the guy in anyway I would take anything he said about the sound work with a healthy heap of salt. There is also very little to get on a big budget film. As I mentioned everything has been dampened down and they have all this noisy equipment going. A low budget film by it's nature has a lot more that can be gotten on set. Do what works for you, but I wouldn't modle it by what a DP told me was done, they probably don't really know (just think they do).

He's gone on to do producer work as well now, so he's involved with the entire process and not just shooting anymore. I've spoken with him quite a have extracted some good information. The problem though is he still shoots like the 1980s. But yeah, I agree. He had come to my studio to give us a hand with critiquing my audio for a film. We talked about it for quite a bit. I came away from that scenario understanding just how much work goes into sound, which most low budget productions don't realize sadly.

That IS the bottom line.
Cheers
SK

Yes. Very true. And thanks for the discussion It's not often I can wax technical audio wise around here

What was that about compressors in the field? Haha!

 

[Chadfish]

That has got to be the longest post I've ever seen! I think it's safe to say that this thread has melted into deep digression.

 

[gpforet]

That has got to be the longest post I've ever seen! I think it's safe to say that this thread has melted into deep digression.

Analog or digital digression?

 

[Chadfish]

Compression digression is the worst kind of session. That's my confession.

 

[timbook2]

Omg

 

[Rick Burnett]

Hahahah! But I digress

 

[Noiz2]

Hey Grimepoch,
Wow just noticed the name ;~)

You know we don't actually disagree all that much. When I lump the storage with the data your right it is two separate issues. But to the end user it is really pretty much one experience. I'm sure you can find a place that will restore an old MFM drive but for all practical purposes it's not possible. So to the end user it doesn't make much difference if the storage or the system has failed the botom line is that it's not recoverable "by reasonable means". Floppies are a good point. Mac's have not supported floppies of any flavor for quite a few years. I have a USB drive downstairs but the drivers ended with OS9, I think. Now it is true that the floppy hasn't "failed" it's still probably fine. But the means of getting information off it doesn't exist anymore so it doesn't really mater on the macro level that the floppy is good. That was my point anyway. I can right now for the cost of pulling it out of the box play back 1/4" tape from the 1960's, I can not get a file off a Floppy from 2002. So for all practical purposes the digital file from eight years ago is a "dead language" and the analog "file" from 50 years ago is still fine.

It's not the analog nature of the storage that failed it's the digital nature of the retrieval that has "failed".

And your not storing "data" on DAT tapes. You might be using DDS tapes, or more likely if you really need a lot of storage DLT tapes. But DAT is Digital Audio Tape. DDs looks much the same and probably at the core isn't that much different but they are not the same thing. DAT is dead. DDS is still probably around because of the large corporate user base.

Damn, I was going to say "shake hands and call it a day" but got side tracked. It's that ADD kicking in.

I think we need a new dead horse. This one's looking kind of gross.

I find if you digress long enough you start to regress and then the whole thing starts to depress.

Cheers
SK

 

[Noiz2]

What was that about compressors in the field? Haha!

Make me thing we should have a thread on decompressors after being in the field.

I need to go to the homebrew forum........

 

[Rick Burnett]

Haha, yes, I think we mostly agree as well I embrace digital as I like freedom and simplicity and at the end of the day, I am more efficient.

Side story, I backed up a drive once onto a zip drive with Iomega's backup utility. I had moved that file off the drive onto another drive. Well, four years latter I needed something off that archive. There was no easy way to get the data out of the file and at that point Iomega was gone. I definitely know what you mean. Using proprietary technology can be dangerous. The KISS principle applies here (keep it simple stupid)

Oh, and floppies do work with OSX if you get one that attached with USB. I had to do this because one of my synths (Korg Triton) only updates with floppies. I was able to do it from OSX.

Homebrew? I used to do that. I need to do that again!!!

 

[Noiz2]

Now here's a topic!

I haven't brewed in about a year. I had taken over the (unused) dark room. Then my wife decided to use it for storage!
You can go to Basic Brewing (.com)?, it's a site and they do podcasts. Check out a few of the video podcasts and it will get you out brewing again.
MMMMM.

If we move to Detroit we will probably have enough room that I can set aside some for a brewery. One up side.
Cheers
SK