Line in differences between recorders - is it noticable?

[Jema]

Once again a subject that I haven't found much on, though I read quite a lot about different recorders. Some say there's a difference, others not so much. As some of the frequent readers of the audio forum know, I recently aquired a mixpre, and now I'm looking at recorders to match it with. I currently own a FR2-LE, and that seems fine except for the unbalanced consumer line input (having to go tape out from the mixpre). I was wondering to what extent this makes a noticable difference compared to a professional +4dB balanced line input in terms of SNR, and if some recorders have better line ins than others? Is there a difference between say, a sony M10 (that is actually able to take an unbalanced professional line in, and I've been thinking of using as a backup unit) and some of the higher end recorders? Maybe the A/D makes a bigger difference on noise level?

The mid-level recorders that can take a +4dB line in through XLR or tele seems to be PMD661, R-44, and DR-680. The abovementioned FR2-LE, as well as the HD-P2 and 671 have line in, the two latter through RCA.

Following post takes up some measurements done on hand held recorders.
http://taperssection.com/index.php?...e8cccaeae5&topic=124639.msg1722123#msg1722123

Also, Wingfeldaudio also presents their sound clips with one done through line in for comparisons.

 

[Noiz2]

For that short a run it really shouldn't make any difference at all. On a long run the fact that the +4 is balanced and has a bit more power will probably make some difference. The +4 standard came about for tube equipment of the day. Other than being less susceptible to RF because of being balanced there is not any "quality" difference. You get more robust connectors so that is a plus but in the bag using right angle plugs I haven't had an issue.

That said I'm mostly doing FX so... I would be more hesitant to use RCA and mini plugs on a shoot I was getting paid serious change to do production on. Not for quality but for the robustness of the connection.

 

[Noiz2]

Looking at that site... Something you need to watch out for is that a lot of people in sound, especially those who do more playing around that actual working, have a lot of time on their hands to obsess over often trivial things. And they LOVE conducting "scientific" tests that are done completely wrong invalidating anything they came up with. They will however argue endlessly about the results and how they prove XYZ.

There were years of huge debates over the differences between 48K and 44.1K and whether 48 was "vastly superior" or "undetectable".

I didn't read it all but almost all of the "noise" in the graphs is below -100dB (none of the line levels got above -100dB). A perfect 16 bit recording has I believe 103? 106 dB of dynamic range. If someone tells you they can hear a -90 dB signal they are on drugs or in an isolation chamber and have excellent hearing. In a normal environment you loose audibility somewhere around -50 to -60 dB.

The highest levels are all down in the inaudible or almost inaudible range. You don't need to worry about any "noise" below 20Hz, you should be rolling that off anyway. In a production environment I suspect most are rolling at 80hHz or so.

Also if I read it correctly he had nothing plugged into the inputs, so that invalidates the reading right there.
And he's recording with a handheld recorder at 88.2/24 which is just silly.

So it's a bunch of pretty pictures that mean pretty much zilch.

Try it, use your ears and in the end inside general categories there is very little difference. One $200 recorder as a rule is not going to sound vastly different than another (same vintage). There are only two kinds of "tests" that you can as a rule "trust". The first you will almost never see. That would be an actual scientifically planned out test that gets the playing field pretty flat. Zacudo.com has done some of this on DSLR cameras. And the pure opinion piece. If you know that you tend to have the same taste as me and I write up something on recorder Z then you can generally "trust" that your experience will be similar. The next best is opinions from folks who do what you do, or want to do, and are working at the level you want to be working at.

Generally stay away from music boards (for this kind of thing) because they are loaded with folks who have very little experience and very strong opinions. The may be right or they may be wrong but their pinions are based on other peoples opinions and not on experience. Also music recording is very different from production sound.

 

[Jema]

One of the reasons I'm interested in something that could take a professional line output from the mixpre is just as you mentioned - the robustness of an XLR compared to an 1/8 plug. On that short a distance I agree that it's highly unlikely that it would do any difference if it's unbalanced or not - I was more interested in if the increase in the signals strength going balanced also has some other benefits.

Regarding music boards and the like - I've done quite a lot of reading on different forums, in educational purposes and for when I'm about to do a purchase. I've seen many threads doing the kind of scientific tests that you describe, arguing endlessly about it for no avail. The more I learn about and use different pieces of equipment, the more I start to value the responses of people working with them giving a more practical insight, like many people here or the experienced folks over at jwsound. Still, music boards are still valuable for finding out about some problems peaople have had with various equipment.

About the noise, it's a good thing you brought up the facts about 16 bit recording (it has 96 dB dynamic range by the way) and that most cannot hear down to about -60 in normal situations. What's important to point out though, is that while the FR2-LE for example state a dynamic range of about 95dB at 24 bit going line in, if you record something peaking at -20dbfs, this gives an effective dynamic range of 75dB. All noise is cumulative, so it's one more thing adding to the noise of the mic and preamp. In post, as you start layering sounds for atmosphere and FX, the noise from all tracks quickly adds up, so I think it's important to have a noise floor that is better than "good enough" for one channel.

 

[Rick R]

About the noise, it's a good thing you brought up the facts about 16 bit recording (it has 96 dB dynamic range

That's a best case scenario, if one is utilizing the full 16 bits. Many who record at very conservative levels, (necessary or not) are sacrificing 1-bit for every 6dB below full scale. So for instance, recording peaks at -30dB would be the equivalent of a 12bit resolution system. (all else being equal)

 

[Noiz2]

All noise is cumulative, so it's one more thing adding to the noise of the mic and preamp. In post, as you start layering sounds for atmosphere and FX, the noise from all tracks quickly adds up, so I think it's important to have a noise floor that is better than "good enough" for one channel.

While technically this is true, practically in film it's not really an issue. It's one BIG difference from music post. In sound post you are actually intentionally adding tons of "noise". So the backgrounds and FX and Foley don't really have an additive noise problem. And dialog is going to be centered on it's own with Foley so you don't really have a build up there either. Most of the electronic noise your talking about will also not make it into the mix. If your going to 35mm optical then you will be mixing to the Academy curve so basically nothing over 10K. Even films going to digital surround generally have very little high freq content. Part is out of habit but also because it's an area where there is a huge variance in your audience. If you put in some loud 15K sound most guys over 25 won't be bothered much but their wives and girlfriends will run screaming from the room.
If you look at most commercial release films you will find very little info above 13k or so.

Point is you do not want to have a noisy hissy dialog track, but you don't really need to worry about noise build up.

Self noise over -100 dB is well past "good enough", it's far in excess of anything possible till recently. Until the late 90's production tracks on big budget films were almost entirely recorded on 1/4" tape. DAT was often used, but as the "back up" copy. Very few films before 2000 or so were mixed digitally or rerecorded to digital dubbers. A major films final sound was mostly made up of recordings that had gone through at least five analog generations before duplication.

I didn't mean to imply that music sites were worthless just that the "signal to noise ratio" tends to be very low on those sites. Places like JWSOUND have a very high signal to noise ratio. And most certainly a music background is very useful in film sound BUT they are very different skills so being highly skilled in one is only a good head start in the other.

 

[Jema]

While technically this is true, practically in film it's not really an issue. It's one BIG difference from music post. In sound post you are actually intentionally adding tons of "noise". So the backgrounds and FX and Foley don't really have an additive noise problem. And dialog is going to be centered on it's own with Foley so you don't really have a build up there either. Most of the electronic noise your talking about will also not make it into the mix. If your going to 35mm optical then you will be mixing to the Academy curve so basically nothing over 10K. Even films going to digital surround generally have very little high freq content. Part is out of habit but also because it's an area where there is a huge variance in your audience. If you put in some loud 15K sound most guys over 25 won't be bothered much but their wives and girlfriends will run screaming from the room.
If you look at most commercial release films you will find very little info above 13k or so.

Thanks for this info - I didn't know the upper limit was so low in a lot of films, and I haven't thought about additive noise that way. It's very good to know.

Self noise over -100 dB is well past "good enough", it's far in excess of anything possible till recently. Until the late 90's production tracks on big budget films were almost entirely recorded on 1/4" tape. DAT was often used, but as the "back up" copy. Very few films before 2000 or so were mixed digitally or rerecorded to digital dubbers. A major films final sound was mostly made up of recordings that had gone through at least five analog generations before duplication.

I'm well aware of the amazing progress that have been done in regards to self noise when digital equipment appeared, and I feel fortunate to have started working on film in this time period. I've heard many say that the audio quality of todays hand held recorders are almost as good as the best analog equipment used 20 years ago going to tape (preamps and such excluded of course). Going through five generations of tape does not sound too fun, but it shows how there have been good sound tracks made with technology that can now be seen as inferior.

Thank you for all the replies! If anyone else have something to add, please do.