Topic: Zoom H5studio: New 32bit float recorder with F series preamps announced.

[Ozpeter]

Hopefully before long we will get some independent reviews of the H5studio - until then, it's a matter of depending on the publicity...

[morst]

Hello,
I am hesitating between waiting for the new H5 Studio or go with the purchase of an F3 + a matched couple of Se Electronics Se8.
This would be used for recording my progressing metal band rehearsals and concerts (when it is not possible to record the board mix).
Do you think the Se8 will offer a significant improvement over the new H5 mics? I know it is difficult to say as the H5 Studio is not available yet...
Thanks

In terms of positioning, compared to mics attached to the recorder? YES.
I've run into some Se mics on the job. They seem to work fine, and the productions which use them could probably afford any similar mic, but for whatever reason, chose those...

[Ozpeter]

I've just noticed that an independent review (it seems) came out a couple of weeks ago, in two parts -

Title -
My first impressions after trying out the ZOOM H5studio: "Wow, this is studio-class sound!"

Link -
https://youtu.be/FEs7L9L2CRU?si=PWYLKd_LeynfWAXg

Title -
ZOOM H5studio Sound Test & Feedback

Link -
https://youtu.be/cm5Wysm9tmw?si=eA0kKt4VjbxcaQcC

Here's my predication - this device will sell like hot cakes.  But not to pocket recorder people, 'cos it isn't small.  I've only listened to the samples on a laptop with no headphones but I can tell that the stereo imaging is very good.  The reviewer seems happy with self noise but unhappy with handling noise.  However, my take on handling noise is that every hand held recorder suffers from it to some degree, and the answer is to learn how to grip the device in such a way as you won't need to change your grip while recording.  Any mic suspension within the capsule housing is going to be pretty minute and there is no way it could be seriously effective. 

[Ozpeter]

Full manual -

https://zoomcorp.com/media/documents/E_H5studio.pdf

[Niels]

It sounds like the mics are extremely sensitive to air movements - even when he speaks at a distance in the studio, and especially when he walks around outside with the dead cat mounted. I don't think it can all be contributed to handling noise.
Could that be a consequence of the large diameter capsules?

I would imagine it is necessary to have very good wind protection if used outside.

From these videos I personally think the included mics would be a challenge to use for field recording.

[Ozpeter]

As I recall it he said that the wind muff was the wrong size, from the original H5 recorder.  Maybe that's got something to do with it.  But I suspect that the better the mic, the better the wind protection required, as a general principle.

I note from the manual that the device is fully MS stereo capable - can record and MS pair as L/R or as 'raw' MS.

GAIN knobs will affect the monitored sound and the recording data.  So they are real gain knobs.

There seems to be a limiter for input and for output too (eg when feeding a camera).  And a third one for the mix of the inputs.

Files are automatically saved at regular intervals during recording. If the power is interrupted or another
problem occurs during recording, an affected file can be restored to normal by playing it with the
H5studio.

Playback can be normalized on the fly.

The alternative modules that fit on the top support 32 bit float.

In the circuit diagrams, there are four inputs each with dual ADCs.  Good!

The supplied mics can handle 140dB SPL.  That's... loud.  XLR connectors are measured at -127 dBu EIN.  That's... quiet.

The manual is slightly enigmatic when it comes to input settings for 32 bit float.  For 24/16 bit, it says to ensure that levels peak at about -12dB  but it makes no recommendation for setting levels for 32 bit float.  Can it really handle 140dB SPL with the gain set to maximum?  I rather doubt it...

[Ozpeter]

It has been now pointed out to me what I should have spotted a mile off - in the circuit diagram the inputs are shown feeding straight into blocks marked "Dual ADC", and everything beyond that is happening in the digital realm, 'gain' included.  So when recording in 24 bit, it's possible to digitally raise the levels above what 24 bit can store, leading to clipping.  In 32 bit float, the digital gain will not raise the level above the 1500dB (or whatever the figure is) that 32 bit float can capture.  All in all, it makes the 16 and 24 bit options a bit pointless, unless immediate output in that format is required.  Unless I'm misunderstanding.

[grawk]

How were you able to determine the quality of the stereo imaging using laptop speakers?  That's like determining a porsche 911 gt3's performance in stop and go rush hour traffic.

[Ozpeter]

Put it this way - if the stereo image is very obvious on a little laptop (chromebook actually!) in terms of hearing very clearly where the passing sound sources were placed, then on a proper replay setup I would be confident that the image would be even better. When I get back home I will test this belief with the quite reasonable system I have there.  And hopefully I will not be disappointed.

[aaronji]

The manual is slightly enigmatic when it comes to input settings for 32 bit float.  For 24/16 bit, it says to ensure that levels peak at about -12dB  but it makes no recommendation for setting levels for 32 bit float.  Can it really handle 140dB SPL with the gain set to maximum?  I rather doubt it...

I think that, in most 32-bit float implementations, the user can't really set gain. Gain (or no gain) is applied at each ADC. If they designed the recorder correctly, the no gain ADC would have the voltage from 140 dBSPL be it's maximum input level. From the dBSPL and sensitivity specs, it looks like that would be right around the 4 dBU maximum input level they specify for the XLR inputs (by eyeball, as I didn't actually do the math). 

[Ozpeter]

Indeed, for 32 bit float, analog gain setting is rare.  What we've got here is simply digital gain, but not only applied to the 32 bit float output, it's applied to the 24 bit output which might not be able to handle it.  Hence the need to control levels to say -12 dB when using 24 bit, but no need to worry about that for 32 bit float.  Crank it up as much as you like and normalise later.

[unidentified]

Concur with the above two responses. Setting gain levels for 32-bit float seems to be a contradiction in terms. As odd as it may seem, in effect you set the gain levels in post.

[Ozpeter]

Concur with the above two responses. Setting gain levels for 32-bit float seems to be a contradiction in terms. As odd as it may seem, in effect you set the gain levels in post.

But - the device has three different limiters for input output and mix and presumably they are going to be affected by the gains set.  So there is a kind of point to the gain knobs.  Though maybe you could simply apply VST limiters in post...

[Ozpeter]

https://youtu.be/aLAlbjFzmCU?si=3WiF47wRR5AL5NDY

New review by "Dark Corner Studios", with noise comparisons to the original H5.  (Oh dear...)

[aaronji]

I think the limiter(s) are disabled in 32-bit mode. They are only useful in 24-bit.