[grimer]

Any experts on voice recording out there? I want to record some interviews at best possible quality.

I want to have these sounding good for possible use as podcasts.

Does anyone know about acoustics and how best to record? In reality these are going to be done at home, usual room with furniture etc. What are best distances from speaker etc?

I researched recording devices a bit and have bought a Zoom H4N to do them on. Is there a particular format / compression etc that I would be best recording at?

Thanks in advance.

[Serpico]

Record in mono at 44100khz pcm, 16 bit, set your levels between -22dB and -12dB max. This is way too big a file for internet uploading, but it's your best uncompressed format for editing. Use a basic editor (like Goldwave) to top and tail your recordings and compress to a lower bitrate. The bitrate you compress to will depend on the acoustics of your room and the quality of your encoder, I recommend the LAME encoder as I've found this deals with room reverb better without artefacts, at lower bitrates, than others like the Faunhoffer. As a starting point try compressing to 64kBps mono mp3, if this sounds ok you could try as low as 32kBps mono, because you recorded in WAV (uncompressed PCM) you'll always have the originals if it sounds shit.
As for the room, go for a quiet room (obviously), lots of soft furnishings, and get the mics as close as practical to the speaker(s), no more than a meter if possible. If the mic is on a table, put it on a foam sheet (or thick felt) if possible to cut down on reflections from the table. If the recorder has a high pass filter (hpf) at around 100Hz, switch it on.  Always do a test recording first and listen to it on a good quality set of closed headphones.

[Obi-Wan is lost...]

Not an expert like Serp, but done a bit in the past and if your on a budget try and get hold of some foam packing sheets, maybe try computer shops etc.

Really useful for covering hard surfaces if your setting up a mobile studio in someones home etc.

This kind of stuff


Also Audacity is a good free editor if you haven't got one.
http://audacity.sourceforge.net/

[Probes]

When ive recorded singing, pining duvets or blankets on the walls made a massive difference in cutting out reflections, also carpet floor is good not lamenate. 

[Probes]

...or depending who it is, just throw the duvet over the top of them  this has been done withh a highly amused bunch of mates.

[Serpico]

...or depending who it is, just throw the duvet over the top of them  this has been done withh a highly amused bunch of mates.

'In bed with Grimer' - a series of intimate interviews. Straight to the late night slot on Channel Five.

[grimer]

I could be the Paula Yates of UKB.

Thanks for the advice.

[Probes]

...or depending who it is, just throw the duvet over the top of them  this has been done withh a highly amused bunch of mates.

'In bed with Grimer' - a series of intimate interviews. Straight to the late night slot on Channel Five.

and coupled with its sister program 'the joys of old age sex'   

.. nothinglike insulting someone you dont know 

[Serpico]

Record in mono at 44100khz pcm

Error... should be 44100Hz (44.1kHz).
And 'Faunhoffer' should have been Fraunhoffer.

[grimer]

.. nothinglike insulting someone you dont know 

Have you insulted me?

[Serpico]

.. nothinglike insulting someone you dont know 

Have you insulted me?

He implied you're a geriatric in the bedroom.
An insult, or just factual?

[Probes]

.. nothinglike insulting someone you dont know 

Have you insulted me?

Maybe i inadvertantly complimented you then?

[sjw]

Record in mono at 44100khz pcm, 16 bit, set your levels between -22dB and -12dB max. This is way too big a file for internet uploading, but it's your best uncompressed format for editing. Use a basic editor (like Goldwave) to top and tail your recordings and compress to a lower bitrate. The bitrate you compress to will depend on the acoustics of your room and the quality of your encoder, I recommend the LAME encoder as I've found this deals with room reverb better without artefacts, at lower bitrates, than others like the Faunhoffer. As a starting point try compressing to 64kBps mono mp3, if this sounds ok you could try as low as 32kBps mono, because you recorded in WAV (uncompressed PCM) you'll always have the originals if it sounds shit.
As for the room, go for a quiet room (obviously), lots of soft furnishings, and get the mics as close as practical to the speaker(s), no more than a meter if possible. If the mic is on a table, put it on a foam sheet (or thick felt) if possible to cut down on reflections from the table. If the recorder has a high pass filter (hpf) at around 100Hz, switch it on.  Always do a test recording first and listen to it on a good quality set of closed headphones.

What he said. All I would add is, to achieve a decent loudness you could consider dynamic compression or at least 'normalising' when editing/pre mp3ing.

Normalising = If the peak of your recording is, say -12dB, normalising the audio file will increase the whole lot so that it peaks at 0dB (the maximum pre distortion level). If you record at 16bit then it shouldn't introduce any noticeable background noise. You probably won't need to dynamically compress the audio if the recording is good, but it would help 'fatten' up the voices ie. like radio. The most basic audio editors normally have normalise and dynamic compression capabilities, so it's probably worth reading the help file/having a mess about before you convert to mp3 even if you decide it's not worth the hassle.

[grimer]

cheers for that Serps and sjw. The recorder came with cubase. Is this the editor (sorry for being so thick about all this).

[Serpico]

cheers for that Serps and sjw. The recorder came with cubase. Is this the editor (sorry for being so thick about all this).

Yes, but depending on which version it is you might find it a bit complicated to work with. If it's a multitrack version (which I suspect it is) then you'll have to create a 'project' to edit your audio - deciding track numbers, sampling rates, and project folders, etc, when all you really need to do is chop the silence off the beginning and end of your piece and maximise the volume (I intended to mention this in my OP, but forgot. I think compression is probably a bit too advanced and probably unnecessary for you at the moment, check if your recorder has a limiter and switch it on), this is far easier with a simple wav editor like Goldwave (the demo version is fully functional, and may be all you actually need).

[sjw]

I agree about Cubase, if you do find it totally baffling and time consuming then I'd just download a simple free editor, which will be more than enough. I've just had a look on the Zoom website and your recorder does have built in low-cut filters/compressors/limiters so as Serps suggested, just get involved with them instead.

[Serpico]

This is the recorder in question. There's a few specific things you need to be aware of re mic pattern/positioning, etc, but I've got to go out now, so I'll post later (or hopefully sjw will take the hint and do it for me  ).

[grimer]

Thank you, thank you 

[sjw]

It's getting rather geeky in here.

Simply put, use the 'HN4 True X/Y Stereo Mic' setting and not the other two in the diagram. Geekily put, since you will be converting the audio from stereo to mono, you don't want any phase difference between the left and right channel. If there is a phase difference between the two channels then it could potentially make your recording sound whack when you 'bounce' into mono. And that is bad. 

In the fear of over complicating what would otherwise have been quite straightforward for you, I'll shut up now and let Serp add anything that he thinks is vital information. Good luck with the interviews!

[slackline]

This is the recorder in question.

Is it common to refer to "stereo images" when recording sound?

[sjw]

Is it common to refer to "stereo images" when recording sound?

Yes it is, although until you pointed it out I've never noticed how odd it sounds. Wikipedia sums it up better than I could:

"Stereo imaging is an audio jargon term used for the aspect of sound recording and reproduction concerning spatial locations of the sound source(s), both laterally and in depth".

[Obi-Wan is lost...]

Regarding mono recording, one thing to watch for (about to show my punterness in this arena) is if you either record in mono or mix down to mono, when you export it beware of only having one channel in your MP3 export. If your listening on headphones and its only coming out one side you have this problem.  Not an expert as I said, but I found that MP3 didn't have mono option when exporting. Easily solved by copying and pasting the mono track onto the second channel. It may not be an issue depending on what editor your using, or asking someone who knows what they are doing rather than working it out yourself. 

[grimer]

You're an expert in my eyes Obi. I wouldn't even know how to make that mistake

[Serpico]

Because the mics on your recorder are set up as a coincidental pair you need to make sure that you and your interviewee are within the recorder's angle of acceptance (switch-able between 60 and 120deg), which means that ideally you and you victim should be positioned such that each of you has a mic pointed at you. This is likely to be impractical, so estimate the angle between each of you and the recorder, if it's greater than 60deg set the mics to the 120deg setting. The mono mix in Stereo Mode suggests that it will mix the mics down to a mono source. Do a test recording and judge the levels between you and President Nixon and adjust the mic levels if needed, or simpler - just move the recorder nearer to the quieter speaker. For more control, once you're familiar with the editing software, record in stereo - this will give you some degree of control over the two voices as one will be predominantly on the left track and the other on the right, still mix down to mono mp3 as this will keep the file size down. A stereo wav saved as a mono mp3 will definitely output both L+R channels.
Do some test recordings and post them up somewhere, or email them, and we'll give you some feedback.

[Falling Down]

This is the recorder in question.

Is it common to refer to "stereo images" when recording sound?

Yes, very common. I don't know the SCIENCE but I imagine there's a significant relationship and overlap between sound and vision. We know some senses overlap because of the way that Synesthesia has been observed and repeatedy measured.   I feel sound physically and visualise it structured in a three or four dimensional way in relation to my body and time - hard to describe but I could draw it if asked