i would want to hear them without normalization. why do people even want do that ?

rojarosguitar, post: 455970, member: 51093 wrote: The files were mono (left and right track separately) normalized to -1.7dB and mixed down using no processing whatsoever apart from dithering (Izotope 6).

What exactly do you mean by this? You have left and right tracks from each recorder, and then mixed them with each track in the centre (i.e. forcing mono)? If this is the case, why not leave them as stereo? At what point is the normalisation applied?

I assume that your reason for normalising is not for an absolute level but to get equal levels from the two recorders.

Mono normalization only means, that each of the channels is treated sperately and not coupled as a stereo pair. that ensures, that both tracks have similar (certainly not equal) energy. I do this to make sure that the different recorders are compared at similar volume (it is not possible to dial into the same recording level directly in the recorders, because the meters are not precise enough.) This is just to make sure that it's not the louder one that wins the comparison.

The tracks are as stereo as they can be with the inbuilt microphones (which are not any of the 'official' stereo arrays). The wide stereo position is maybe something ORTF-ish but quite a bit narrower than true ORTF.

I don't see a big problem with normalization soundwise; it's done in 96/24 realm in Protools 12 and just dithered down to 16 bit in the mixdown. Otherwise I don't know how to ensure level compatibility. Any suggestions?

rojarosguitar, post: 455976, member: 51093 wrote: Any suggestions?

use your ears.

@ Kurt Foster: Could you pleas substantiate your claim that normalizing defeats using my ears? Did I hurt some religious feelings here (like 'thou shallst not normalize' )?

There are reports of listening tests, that poeple usually rather go for the louer source. Even if normalization should degrade the sound (which I don't think, but you can comment on that), it does so for both sources, but then they become more comparable.

Don't see any reason for your aggressive remarks.

The whole point of normalizing to an objective standard is to expose the subjective differences. If the tracks are normalized to a subjective standard then the goal of separating the subjective and objective can't be achieved.

But until blind A/B/X is used the results will not be reliable. That is, if you know which one you're listening to you can't be sure you're being objective. How about you send me your tracks and I'll normalize them, make ten copies of each with random names and send them back. You try to separate them into A and B categories. Then we'll have valid data.

your asking us to say which recorder sounds best but then you add processing to the samples and that same o sameo's them out. no processing makes audio better, it only makes it different. all processing is a degradation. and yes i subscribe to the tenant, "thou shall not normalize". it's a cheap trick.

What are we chasing here? Are we simply trying to say that, on balance, we prefer one of the tracks, or are we trying to say none, one or both of the tracks is an acceptable recording of a classical guitar?

Even if there is a majority prefering one of the recorders on these guitar recordings, that does not mean to say that recorder is going to sound better in different circumstances.

Please forget the whole thing. It wouldn't have occurred even in my dreams that one can make a simple thing so complicated. One could just simply say, I like this better or that better. It's not a clinical study on life and death issues. Nobody is giving a guaranty that if you like this or that better, you will like it better in other circumstances, too. Personally I like one better than the other, but not by as much as to justify the price difference.

Nobody has been able so far to explain to me in which way normalizing the to different tracks to the same peak volume is a processing that renders the sound quality unrecognizable. Well, never mind, was just an idea to ask people with trained ears what they hear ...

it's simple. your serving us two different brands of vanilla ice cream (recorders) but you are putting chocolate syrup (processing) on both and then asking us which brand of ice cream we think tastes best. just post the recordings sans normalization and we can render an opinion.

I have asked you several times to substantiate your claim why volume normalizing is to the sound what chocolate syrup is to vanilla ice, to no avail.
Maybe I'm ignorant and the reason is obvious for everybody else, but still, I don't need this kind of arrogant response.

As I said, I'm really not interested in the continuation of this post, because we don't have any meaningful communication here. Please treat this post as closed.

http://www.hometracked.com/2008/04/20/10-myths-about-normalization/

Best mixing practices dictate that you never apply processing “just because.” But even setting that aside, there are at least 3 reasons NOT to normalize:

Normalizing raises the signal level, but also raises the noise level. Louder tracks inevitably mean louder noise. You can turn the level of a normalized track down to lower the noise, of course, but then why normalize in the first place?

Louder tracks leave less headroom before clipping occurs. Tracks that peak near 0dBfs are more likely to clip when processed with EQ and effects.
Normalizing to near 0dbfs can introduce inter sample peaks

When a track’s level is so low that you can’t use gain and volume faders to make the track loud enough for your mix. This points to an issue with the recording, and ideally you’d re-record the track at a more appropriate level. But at times when that’s not possible, normalizing can salvage an otherwise unusable take.

When you explicitly need to set a track’s peak level without regard to its perceived loudness. For example, when working with test tones, white noise, and other non-musical content. You can set the peak level manually – play through the track once, note the peak, and raise the track’s level accordingly – but the normalize feature does the work for you.

1. When a track’s level is so low that you can’t use gain and volume faders to make the track loud enough for your mix. This points to an issue with the recording, and ideally you’d re-record the track at a more appropriate level. But at times when that’s not possible, normalizing can salvage an otherwise unusable take.
2. When you explicitly need to set a track’s peak level without regard to its perceived loudness. For example, when working with test tones, white noise, and other non-musical content. You can set the peak level manually – play through the track once, note the peak, and raise the track’s level accordingly – but the normalize feature does the work for you
   

   rojarosguitar, post: 455990, member: 51093 wrote: I have asked you several times to substantiate your claim why volume normalizing is to the sound what chocolate syrup is to vanilla ice, to no avail.
   Maybe I'm ignorant and the reason is obvious for everybody else, but still, I don't need this kind of arrogant response.

   As I said, I'm really not interested in the continuation of this post, because we don't have any meaningful communication here. Please treat this post as closed.

   for every person who post a question we can have thousands of others accessing the info on the internet. we're not here just for you. so in that light, if needed RO's administration will decide when a thread is closed.

It's on Youtube...

You can get pretty decent sound from Youtube, but there's always a chance their resampling does something weird.

And, let's face it, even modest handheld recorders have such a high quality these days that hearing any difference boils down to difference in mics, placement, or Youtube weirdness.

I think you are all geting a bit hung up on the OP's use of the word "normalization". Normalisation in itself is like moving a fader up as high as it can go without casuing clipping - it does not impart any sound or processing to a track. If, for the purposes of this present thread, you read "normalization" as "equalising levels" such that the two recorder outputs have the guitar at the same acoustic level, then that's one of the things that's necessary when making a comparison in order not to impart an implied bias in the decision about which track sounds "best". It does mean that the noise floor may be at a different level between the tracks from the two recorders, but one has to ignore that when comparing the recordings of the instrument.

I have a Zoom H4N that I dig out occasionally when the recording circumstances mean that I cannot use mains-powered gear. However, I very rarely record using the built-in microphones, often because they are a fixed pattern and are attached to the recorder which I would be needing to place inconveniently above the conductor's head, for example. Instead, I mostly use external microphones via the recorder's XLR inputs, usually with an in-line battery-powered signal booster to improve the barely adequate S/N ratio that you get with these portable recorders. The booster also generates the +48V phantom power for the microphones, extending the recording time by not having the recorder's batteries powering this. The H4N can make respectable recordings, but the recording level window in which you have to operate is quite small.

Boswell, post: 455997, member: 29034 wrote: ... that's one of the things that's necessary when making a comparison in order not to impart an implied bias in the decision about which track sounds "best".

Exactly.

Kurt Foster: Thanks, I respond to your post because of the attempt to at least rationally explain your ideas, though I don't share your point of view. I don't see how peak level normalization falls under the category of normalization. For me this is an automated fader adjustment to a certain peak level. Ideally it shouldn't change the sound a file but only make it louder or softer. I will check that on PT12 to be sure. If peak level normalization is processing, then even a mixdown in a DAW is processing and I can't see how I would avoid processing using digital audio. But I got curious whether peak level normalization does anything to the spectral content of audio.

As to my formulation 'Please treat this post as closed' it may be my not being native speaker of English that came through. I perfectly know that threads are closed by moderators only. I tried to express 'case closed' but didn't wanted to sound like a lawyer.;)

As to the noise floor: having recorded at approximately same level in first place, if the noise floor of one recorder is significantly higher, that will also become apparent after normalization, which can also be a hint as to the quality of the recorder.

In any case I don't see how you can run into headroom and clipping issues by peak normalizing a recording to -1.7dB.

Anyway, thanks for a less laconic treatment of my OP.

@ Boswell: Thaks for your attempt to straighten it out, your point of view corresponds to what I thought. I never assumed that the perceived loudness will be exactly the same (because this depends on the spectral content and not only on peak level values), but at least it comes closer together than if untreated. I fear RMS normalization exactly because it changes the perceived sound just to gain the same perceived loudness, as I understand it.

Anyway, my post became obsolete because I decided to send one recorder back as it didn't fulfill my expectations. The other one I had already anyway, so the whole comparison became a moot point. But to exchanged with different people helped a lot to get more clarity (although dishearteningly this forum was the least helpful in this decision making). Maybe it was my fault how I formulated things, but I wanted to stay as neutral as possible...

Sorry for taking your time. Best
Robert

In that case, can you tell us which model you sent back and which model you already had?

I have had long since the Tascam Dr-40, which now goes by a street price of around 18o US$, and I had tried out the Sony PCM-D100, which is around 780 US$ street price. The preamps seem to be a bit quieter and a bit more transparent, but also lack life and solidity. The Sony doesn't even have XLR inputs for external mics, and for my side I found the improvement over the Tascam, if any, doesn't justify this price.