Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 16, 2017 19:56:02 GMT
Frans, this is aimed at you because you might have an answer..... (not headphone related though)
I'm currently using three portable digital recording devices with an Audio Technica at2020 microphone which is a basic, run of the mill no frills, studio mic. Noise figures are ok considering it's price, but it's frequency response is lovely.
The problem that I have is, I think just a mismatch and I'm looking to see whether there is an electrical way around the problem.
The really cheap (but excellent) Zoom H1 recorder has a mic input which doubles as a line in and works perfectly. The AT2020 sounds superb. The Tascam DR40 can be switched between line in and mic in so that's not a great problem. It's just a bit big for portable.
The Tascam DR05 seems to have no real 'line in' and the mic input doesn't double up as a 'line in'. The result is a lot of noise from anything plugged in via my portable pre amp. Also, the gain sets it right into clipping (plus noise)
So it looks as though the Tascam DR5 can't take a balanced input or line in signal via its mic input. There are two variations of line in I think.... one for consumer stuff (-10dbu) and another (higher value +4dbu) for pro stuff, but this is purely consumer.
Is there an electrical way to get the socket to behave like a line in (consumer levels)? I'm sure that the noise is some kind of mismatch and is there whether there is zero gain or full gain from the preamp.
I'm doing some location recording, so will be taking the Zoom around with me I guess, but would prefer the Tascam if I could get a match. It'll be put onto a video soundtrack.
Two mics hanging from ceiling for 'stereo', feed. I'll use the Tascam DR40 fixed for that. The DR05 backstage in mono if I can get it to match with the preamp and AT2020 plus the Zoom H1 moving around because it's tiny. All to be mixed onto one soundtrack in the end. (Plus a shotgun on top of one camera, while the other camera is fixed with another shotgun) I'm tracking a show being made up to the final production.
Noise levels need to be fairly low if possible so that combinations of soundtracks don't compound noise issues so I want to use preamps rather than recorder preamps to make sure all signals are pretty robust.
I'll be using the AT2020 with phantom power at 48v going into a cheapo Saramonic Smartrig+ and the other feeds will also be fed into Smartrigs which allows 4 microphones into each recorder. It's just this DR05 which has the problem.
I don't want to spend another £200 on a Zoom H4 if possible because it's also a bit big.
I know that you most likely won't be familiar with the gear, but you may well know of a workaround to be able to use a mic input from a line output. It's not as simple as levels, it's a total mismatch and I'm guessing, impedance might well have something to do with it. I was thinking maybe an interface before the recorder? DI box?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Sept 17, 2017 7:18:57 GMT
The difference between mic levels and line levels are about 40dB to 60dB lower than line levels. 40dB is a factor 100 in electrical levels, 60dB is a factor 1000. This means that when a line level signal is applied to a mic level input the signal is 100x to 1000x too big resulting in severe clipping of the signal and background noise levels to be 40dB to 60dB louder. Basically noise that would be almost inaudible on a mic input would become quite loud when line level signals are applied. The way most console inputs work (mixing consoles and studio type DAC/ADC's) is by using the same input amplifier for these signals but the alter the gain of the input circuit. For this reason you will always find a small pot with either markings such as 0dB (or -4) - +60dB or has a 'line to mic' level indicator. Cheaper or smaller gear may have a switch which toggles between line and mic level. The frequency response should be about the same (flat) but is determined by the quality of the used circuit. Some cheaper inputs may roll-off slightly in the treble. Some inputs (those with input transformers) may roll-off in the subbass area. The technical solution that is used is to simply provide more 'feedback' for higher signals (line level) in the exact same circuit. More feedback in a high gain amplifier (such as a microphone input) results in: a: smaller input sensitivity so bigger signals can be applied. b: lower noise levels of the circuit itself c: higher bandwith (unless bandwidth limiting filters are used) d: lower amounts of distortion e: less sensitive to nasties on power supplies These signal levels simply are too far apart to be handled in one single circuit with a fixed gain. So what to do: To apply a mic level to a line-level input you need a mic pre-amp. To apply a line level signal to a mic level input you need an attenuator. The problem with an attenuator is that what one basically does is reduce the signal level (let's say -40dB so 100x smaller) and insert that in an amplifier that amplifies it 40dB again (the mic input) In the end ... line level output signal will be line level again in the device and not clip. But ... frequency response might have a slight rol-off in the treble. Background noise levels are higher and may be audible. Distortion levels may be slightly worse. Something like this may be needed: fentronix.wordpress.com/2015/01/23/audio-attenuation/depending on the mic circuit in question (how much headroom/gain it has got) Fleabay sells these for cheap.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 17, 2017 8:21:11 GMT
Thanks Frans. I suspected FR might go adrift which is why I was asking. I also wondered whether headroom might suffer as well. Oh well. The cheapest (clean signal) way is to just get another Zoom H1 which is cheap as far as recorders go. Just trying to avoid yet another recording device.
When I can, I try to use the Smartrig Preamp. Remarkably cheap for the facilities it offers in that it can take four mics (albeit in a slightly crude configuration) but does mean that you can close mic more than one person and get great results. Noise levels are good as we; especially once you get mics close.
These digital recorders are extremely good and allow excellent results; ideal for film soundtracks because of portability and simplicity. I often leave one running in my pocket with just a lav mic attached to me somewhere and before filming something, just clap a few times for a sync point and edit the (say) two hour soundtrack later down to just the 'clapped' sections for the film. Makes a massive difference.
Cheers - I suspected that it wasn't that simple!!
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Sept 17, 2017 9:04:08 GMT
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 17, 2017 9:54:02 GMT
Thanks Frans. I was a bit concerned about any effects on FR since the at2020's that I often use are great and nicely extended, and I don't want to curtail it up there too much. For the stereo hanging Mics, I'll be using shots but the at2020's elsewhere so Phantom power and preamp are needed.
At least an attenuator might work and it's cheaper than another Zoom. I'll give it a go. Hopefully, 25db is enough to get consumer level line? The noise and artefacts are excruciatingly loud. Including digital noises being transmitted from the player.
I am loathe to put a pot in the line as well ... I wondered whether different vol settings might also alter FR and noise depending on which setting I use, so at least a fixed value would be consistent and allow 'mass' eq later on to save time!!
At best, I could get 80db noise values but more likely with mics to be around 55 - 60 which is just about ok. 60 is the target if possible.
I'll try the less fancy option. You never know ... it might work.
Cheers
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Sept 17, 2017 14:38:24 GMT
Here's an idea. Spend 2 pound more and buy the fancy one. This will give you stepless control over the sound from those mics so you can get the balance between mics right. Something that cannot be done afterwards. And it is more versatile. The potmeter isn't going to be worse than the fixed (resistors).
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 17, 2017 15:33:29 GMT
OK, I'll give it a try. Thanks Frans.
I think you're right. I'm aiming at 60db noise floor if I'm lucky, but 55 can be accepted so I have to keep an eye on the figures. The AT can do 75db and preamp 80 or so, so I'm just factoring in two mics going at the same time or any other artefacts added.
If necessary, I'll play around with the sound envelopes in processing. (A bit dodgy, but can give good results on quieter passages) Trying to avoid noise gates if possible. Messes up transients.
It will eventually be sent out on a closed system but they won't consider anything with unacceptable noise levels so I'm trying to plan what I can do in advance without buying new gear.
Quite an exciting project with a theatre company who are bringing in pro coaches to train singers and produce a show. Nice not to have to be performing and just observing!!
I'll let you know what happens!!
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 21, 2017 13:52:42 GMT
Gave both a try.
They both take the noise away but the attenuation is a bit too much ... even 25dB!
Result is that the preamp has to be at full whack and the mics sound lacking in dynamic range.
I think I might just use the software on my computer to remove noise perhaps. I'll take a hiss profile at different volume settings in dead quiet and then apply them to my files. Did a rough noise removal this morning and did get above 60db which is good enough for video.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,882
|
Post by solderdude on Sept 21, 2017 14:48:40 GMT
When you know how much attenuation you need I can probably modify it so it does what you want.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 21, 2017 17:31:54 GMT
Probably more like 10 Frans, which is quite weird. I was expecting 40 to be the correct amount.
The digital ticks from the recorder and hiss are totally eliminated even with 25dB reduction. Just that mics have such a tiny output that it's difficult to judge levels I guess.
I had a play around on the computer this afternoon and found a mixture of things that I can do to effectively eliminate noise as far as the ear is concerned. (Not technically though) I basically applied noise reduction by sampling silence. That has to be used very carefully because it can impact on the sound quite audibly.
I compressed speech which raised levels (and noise) quite a lot. Used the noise reduction. Then EQ'd to bring clarity back in speech Normalised to -3dB. Then a crafty little bit of jiggery pokery, where I altered the shape of the sound envelope at critical points. So just after I finished speaking, I compressed the envelope right down so absolutely no noise came through and then opened the envelope again for the next word.
It meant a lot of upping and downing the sound envelope, but in the end, the perceived sound is remarkably clean and in critical points, way down below the target of 60dB because I shut the sound off, similar to using a noise gate, except I was controlling it more exactly than a gate would. (Only at critical points)
So it looks like I have a good solution to the problem of mic noise and am able to control it via software quite easily, so that in sections where noise is masked, then less manipulation is needed, but in those really quiet areas, where you can hear a pin drop, I shut the envelope down so that noise got banged on the head, giving the impression of a limitless noise floor!!
It does work quite well as long as you get the timing of the fade in and out of the envelope to coincide with the wave shape.
The stage stuff doesn't worry me with regards to noise, since it just gets masked by the music, but as part of the project, I will be recording interviews which will be done in dead silence and it's those moments that show the weaknesses of most microphones.
It does help a lot if mics are as low noise as possible though, since then, less manipulation is needed. It takes time to sort out and then add the time to sort out the video picture, let alone put the whole lot together and add any soundtracks/effects - it becomes a very long haul job.
Mics are such noisy devices and tbh, I don't think it matters what I use .... it will still need editing in the end. Can't take a direct feed from a mic and plant it onto the video - they're all too noisy!!
|
|
|
Post by drumdrym on Sept 22, 2017 13:15:28 GMT
Hi Ian, found your last post quite interesting, I process in Audacity about 3.5 hours of speech each week as audio for the blind. I have seven readers reading in their homes on a variety of recording devices so the audio is very varied. They send their audio copy to me by email, one reader sends from Manitoba Canada the rest are within a few miles of me. Noise is always a problem but I've never tried to use the envelope function. I do use the Nyquist noise gate plugin and stock Audacity noise reduction effect. Used with care I get good results on noise. My main line of attack is with a plugin from Tokyo Dawn Labs called TDR NOVA a very versatile EQ. If you've not come across this plugin it's free and well worth having, here is a link to it.......... www.tokyodawn.net/tokyo-dawn-labs/ I also use a limiter called Limiter No.6 and normalise to -2db. Video on TDR NOVA Introduction to NOVA and NOVA GE by Dan Worrall
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 22, 2017 14:36:33 GMT
Thanks Alan. I'll give that plugin a go.
Trouble with this video is that it is being watched and the company demand a certain noise floor as well as a loudest volume setting. They kind of set their standards which have to be met.
The show itself presents no problem because the noise floor is way under ambient, but as you say ..... spoken word in a quiet environment is a very difficult one to avoid hearing something; even with digital. Only as good as the mic that you use in reality.
I'm being a cheapskate in that I'm trying to avoid buying yet another mic just for the talking head stuff. In a way, what I'm doing is a workaround rather than absolutely meeting the criteria but the envelope shifting trick is a bit more subtle than an out and out gate. Sometimes, it's possible to hear a gate opening and closing, depending on what you're recording. The worst example of a gate being used is Phil Collins snare drum!!!
I know that you also used the AT2020 but a USB version. Mine is an xlr so needs phantom power etc. They're not bad, but still a tad noisy. They quote 74dB s/n but the reality isn't that if you check. You get around 65 unless you start bellowing into it!! Sound is great. I've even used two for stereo recording and they were lovely.
I found out the magic qualities of a good mic though. A couple of months back, I was using a borrowed Senn top of the range mic where noise was just not there for all intents and purposes. That leaves you much freer with mic placement and recording levels than the Audio Technica. I've found over and over with cheap mics that placement and input volume is absolutely essential to get right or else noise becomes a problem. Cheap mics are way more touchy in this respect whereas with the Senns, you could literally chuck it on a settee and pick up someone on the other side of the room with very minimal noise. Makes you lazy but there's no doubt, they work well.
So my editing technique is going to be tested here I think and playing around with envelopes has helped me disguise what is going on. In fact, if you make the envelope too small, it feels as though the sound has been sucked away, so you do need to be subtle on occasion.
I tested it by recording on a lav mic on the other side of the room with low volume. So we're talking really poor quality and using a combination of noise reduction, eq and envelope manipulation, I rescued the recording up to a point and noise stayed very low in the crucial bits. Very weird though, to hear a recording taken further away with room acoustic sounding like its right up against your ears after cleaning it up. It kind of made the sound seem very close but retained room acoustic!! Useable though. So I can rescue it but that plus the video editing is going to be an enormous amount of work.
I'm thinking about getting a Rode studio mic just for this job. Not too bad at around £150 I think but the noise levels are stunningly low. (Rode NT1a) It has a slight 'u' shape response rather than as flat as I'd like, but eq would fix that. Might actually be nice for music though.
Mics are an absolute minefield; even in video. In fact, you can get away with poor video and people accept it, but not poor sound. Maybe it's time I lashed out and got a better mic.... I'm just being a bit tight, but in terms of hours saved in audio editing, it might save me more than £150!!!
Just before I went on holiday, I did a show with just a cheap shotgun setup. Worked out fine. I had three cameras ... one fixed in the centre at the back. One fixed at the side of the stage and the other, I held and moved around. The editing was hilarious since I had thee films in effect of the same thing so I simply planted them on top of each other, synced up and literally just went through, cutting out the two camera angles that I didn't want.
Sound went from the shotguns into a four track digital portable made by Tascam and I just synced this with the video at the start before editing it. So the sound stayed consistent while I changed camera angles at will!! When you're a one man band, that's the sort of thing you have to do I guess.
I was pleased with the result, but it was hours of work.
I'm basically trying to save money by editing, but the more I think about it, it might not be worth the time I'd need to sort out noise levels. The Audio Technica is only just about good enough really.
I'll try that plugin though. Thanks Alan.
Btw, if you have seven readers and want consistency a lav mic is a great way to go. They pin them close to their mouths so at least their voices are planted much higher up than the noise floor. I've even used a lav mic right in front of my mouth with a foam cover as a handheld mic and believe it or not, they sound terrific.
A cheapish one that I use is a Rode Smartlav+ which isn't the quietest and I also have a couple of real cheapos that have long leads that sound ok... a Boya m1, I think it's called which has a nice sound sig and an Aputure Lav. They are both quite quiet and are cheap. Even cheaper ones work really well as a handheld mic for speech though tbh, as long as the speaker has them very close to their mouth. They have a sound that is designed for speech, so no deep bass and hopefully not too much sting in the treble. They're not like full frequency studio mics.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Sept 22, 2017 19:53:54 GMT
Had a play with the noise gate, Alan. Very effective as long as the noise is reduced first so that you don't hear it coming in and out. Combining that with envelope shaping basically eliminates noise virtually completely.
I'll be looking for the worst quality mic next, just to see what I can do with it.
So my workflow at the moment is...
Normalise (-3dB) Compress if needed and renormalise. Noise reduction. (I reduce to get it to -60dB) Noise gate which is useful in that it analyses noise floor and suggests where to set the gate. Envelope shaping of anything really quiet.
Voila .... no noise!!!!
The file that I played with was recorded on a lav at the bottom of the bed so about six feet away at very low volume so that the original waveform was barely there on the screen!!
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Oct 6, 2017 11:04:24 GMT
Alan, I don’t know whether you’ve ever considered them, but I have some cheapo dynamic Mics that I use sometimes for speech. While they aren’t so toppy as condensers and are less sensitive, the strange thing is that signal to noise is really quite good. I use a preamp with EQ so can brighten the top up while recording if I want and these things are built like tanks.
I think the lack of electronics inside help an awful lot with self noise.
The one I’m using is a Behringher Ultravoice XM8500. Ridiculously cheap at around £17 but results can be really good from it. (As long as you haven’t developed a ‘house sound’ for your books.
While the AT2020is a great mic with low noise, it can be quite ‘hi fi’ sounding and very prone to proximity effect. So to get a balance, you naturally get away from the mic and turn up the preamp to compensate. Then you’re fighting with preamp noise as well as mic. With these dynamics, you can go straight in with minimal proximity and get the gain as low as anything, even though they’re less sensitive than powered Mics.
Useless for distance video because pickup is limited to close proximity, but great in the rejection of household noise. I had Lucy dancing beside me at about four feet away while I spoke and the mic rejected the noise easily. It is a cardioid as well so that helps enormously. The dynamics are often overlooked these days other than on stage.
Surprising, what you find going back a step to an old technology, only to find that it isn’t really as bad as you might think. I just did an interview with the Ultravoice and literally had no cleaning up to do. I did a bit of EQ, but tend to go easy on that if I can.
|
|
|
Post by drumdrym on Oct 7, 2017 1:29:28 GMT
Hi Ian that's a good shout, I'm about to buy one for a try out. Would be good partnered with the £35 UMC22. I have one dynamic, Audio-technica ATM63HE coupled to a £23 Saramonic SmartRig, inputs on XLR with optional 48v and outputs to TRRS suitable for connection to Smartphones, bumps up the recorded phone audio making them sound very reasonable. House sound..................probably, but not easy taking eight quite different sound sources and outputting them as one homogeneous whole. I like the AT2020 USB and can get quite close to it, but you have to wrap it up to cut back on the plosives. I had some abrasive sheet a bit like pot scourer material, memory fails me as to what it's trade name is but I'll find out. The white sheets have no abrasive embedded I use them to apply a hard wax oil coating on wood. Wrapped two sections round the whole mic body with two elastic bands and bingo best plosive shield I've ever used. (pic to follow) Like the usability of USB mics plug in direct to PC and record. Of course the resultant audio quality greatly depends on the PC. The cheap Behringer UMC range of preamps are very quiet. Like Google EQ is my friend in taming the worst of my eight readers audio output. I think if you heard it you'd run a mile. Alan
|
|