Topic: Zoom H4Essential

[wildlife]

Can someone tell me their experience with this recorder?

I am considering this recorder because it is capable of having XLR, TRS and LINE IN inputs connected to it and also because it seems convenient to not have to worry about dicking with the levels due to the 32bit float recording capability.

My main concern is that, when I was looking at the manual I noticed the following warning:

NOTE
On the H4essential, only the INPUT 1 and INPUT 2 jacks have dual A/D converter circuits.

From what I understand, in order to record without requiring level changes, the recorder makes two recordings: one at a high level and one at a low level. These two recordings are then processed together into one 32 bit file that covers the whole range of sound via the A/D converter. So if the only the Input 1 and 2 jacks use the A/D convertor, would that mean that when recording with the LINE IN jack I will still need to adjust the level when recording? I often use Church mics with a battery box to record shows so it's an important factor for me.

I'm sure my question can easily be answered by someone who has used the recorder so, if you have any other anecdotes/warnings/praises about the recorder I'd appreciate that too.

I did do a search on the forums before posting but I couldn't find anything specific to this model. The warning I mentioned above does not appear on the manual for the other HXessential recorders either.

[Dan33185]

Not sure if this is exactly what you mean, but when I record, it creates two files. One records at a lower level than the other, I'm assuming as a "just in case" file, but I've never needed to use it.

[Ozpeter]

Hmmm.  Naughty.  It's advertised as dual a/d conversion - with no mention that it is only on the XLR inputs.  Another reason why reading the manual online before purchase of any device is not a bad idea - not that I always do that myself...  I just checked the H1 XLR manual and that has no such restriction but it only has two channels, so only two preamps (well, one stereo preamp) so the issue would not arise.  The H2 essential doesn't mention dual a/d converters in the specs or the manual.  Ooops, I never noticed that.

[TheJez]

Not sure if this is exactly what you mean, but when I record, it creates two files. One records at a lower level than the other, I'm assuming as a "just in case" file, but I've never needed to use it.

This seems to be the moment to advertise the SafetyTrackMerger software that takes the main recording and the backup/quieter/safety recording and turns these into a 32bfp recording during post-processing, just like multi-adc recorders would do in realtime!
See https://taperssection.com/index.php?topic=206443.0

[wildlife]

Thanks for the answers.

Not sure if this is exactly what you mean, but when I record, it creates two files. One records at a lower level than the other, I'm assuming as a "just in case" file, but I've never needed to use it.

To put it simply I want to know what impact the lack of an A/D convertor on the LINE IN input (and built-in mics, though I don't plan on using them) will have on the recording process and the recording in general. If I understand correctly the difference is that when recording via TRS/XLR the files get merged together in realtime and with the other inputs you just get two files at the end?

This seems to be the moment to advertise the SafetyTrackMerger software that takes the main recording and the backup/quieter/safety recording and turns these into a 32bfp recording during post-processing, just like multi-adc recorders would do in realtime!
See https://taperssection.com/index.php?topic=206443.0

Thanks for this suggestion.

[Ozpeter]

I have just spent about an hour online trying to find any worthwhile explanation of the practical outcome of using 32 bit float dual converters vs single converters, and I am finding nothing.  I would expect that a dual converter would make better use of the 32 bit float format dynamic range, as it uses one converter to deal with lower level signals and the other to deal with higher level signals (instantly - this isn't some kind of record-first-process-afterwards thing - it doesn't produce two recorded tracks).  But the existence of single a/d 32 bit float converters doesn't seem to get a mention. 

Yet given that single converters for 32 bit float seem to be installed on lower cost 32 bit float devices, but not mentioned by the manufacturer, makes me wonder whether in fact devices using single converters are actually more common out there than dual converters (in the sense that for instance Zoom probably sell more H series devices than F series, for instance).  Wireless mics using 32 bit float are now pretty common, but reading the blurb for a Rode device of that kind there is nothing mentioned about dual converters, which makes me suspicious.

Does anyone have a link to a good explanation of the difference between dual and single a/d converters?  I'm worn out looking!

[TheJez]

Not sure if this is exactly what you mean, but when I record, it creates two files. One records at a lower level than the other, I'm assuming as a "just in case" file, but I've never needed to use it.

This seems to be the moment to advertise the SafetyTrackMerger software that takes the main recording and the backup/quieter/safety recording and turns these into a 32bfp recording during post-processing, just like multi-adc recorders would do in realtime!
See https://taperssection.com/index.php?topic=206443.0

After reading the manual of the H4E, I think I need to retract my remark about the SafetyTrackMerger for being irrelevant in this scenario. It seems the device stores each recorded channel, as well as a 'stereo mix' according to the mixer settings. It seems that this mix is the "just in case" file as mentioned by Dan33185. As this whole 'mixing thing' is done completely in the digital realm, it is not a classical 'safety track' as seen on recorders like the Roland R07. It means there is no benefit in merging the normal track and this mix track. If the signal was too hot and overloading the analog amplifier and/or the ADC's, then the resulting distortion will be present on all tracks.

[TheJez]

But the existence of single a/d 32 bit float converters doesn't seem to get a mention. 

...because there are no single 32bfp ADC's, AFAIK! Although this is a matter of persective: If you'd put two 24-bit ADC's and a Digital Signal Processor that combines the outputs of the ADC's to 32bfp onto a single chip or PCB, could we call that a single 32bfp ADC?
Anyway, this Zoom H4Essential feels very weird. As I understand, the input gain is fixed, both for XLR and mini-jack inputs. I guess the analog input stage for the mini-jack and the first ADC are tuned/matched in such a way that they can manage the expected maximum input signal levels without distortion (and without too much noise for quiet signals). However, the XLR inputs must be able to handle a wider range of input levels, which is too much for a single ADC to handle without compromising the noise level too much. Hence the 2nd ADC. (Of course, if there would be gain control, also XLR could be handled well with a single ADC. But apparently they wanted to create a 'set-and-forget' device without gain control.)

[Ozpeter]

But the existence of single a/d 32 bit float converters doesn't seem to get a mention. 

...because there are no single 32bfp ADC's, AFAIK! Although this is a matter of persective: If you'd put two 24-bit ADC's and a Digital Signal Processor that combines the outputs of the ADC's to 32bfp onto a single chip or PCB, could we call that a single 32bfp ADC?
Anyway, this Zoom H4Essential feels very weird. As I understand, the input gain is fixed, both for XLR and mini-jack inputs. I guess the analog input stage for the mini-jack and the first ADC are tuned/matched in such a way that they can manage the expected maximum input signal levels without distortion (and without too much noise for quiet signals). However, the XLR inputs must be able to handle a wider range of input levels, which is too much for a single ADC to handle without compromising the noise level too much. Hence the 2nd ADC. (Of course, if there would be gain control, also XLR could be handled well with a single ADC. But apparently they wanted to create a 'set-and-forget' device without gain control.)

But if Zoom are careful to tell you after purchase that the H4e does not have dual A/D converters on the 3.5mm inputs, but the output remains 32 bit float (it seems), then single converters must exist?  And why do they not claim to have used duals in the other He series devices that I mentioned if singles don't exist?  It's confusing...

[wildlife]

Thanks Ozpeter and TheJez. I think I am understanding it a bit better now.

What I gather is that the dual A/D converter on each XLR input records at a fixed low gain level and a fixed high gain level, which then comes out as a single 32 bit file per channel (plus a stereo mix of that). However with the LINE IN input 32 bit file/s(?) are produced from one fixed level (i.e. a "single" A/D converter) as opposed to two. Is the assumption here that the fixed gain level is sufficient to cover the range of most LINE IN inputs when converting to a 32 bit format? At least according to the manufacturers...

[Ozpeter]

As those not interested in the H4e may not be reading this thread, I have raised the matter of how 32 bit float can be obtained from a single converter, in the long running 32 bit float dedicated thread.

[Ozpeter]

Wildlife, you have no idea what you started with your innocent question.  See this link for more...

https://www.reddit.com/r/audioengineering/comments/1ipo4wb/revealed_devices_are_being_marketed_as_32_bit/?sort=confidence

I didn't acknowledge you by 'name' but I did at least acknowledge the forum where you had posted!

[wildlife]

Okay so I got my hands on this recorder and did a few tests.

First, I hooked the recorder up to my computer and overloaded the input. For LINE IN the input clipped and a message flashed on the screen indicating it was overloaded. For XLR and TRS no such warning appeared despite cranking the volume loud enough to distort the signal. If I put the output a reasonable volume, the resulting waveform for XLR and LINE IN are similar at an equal output level. TRS gives a quieter waveform probably because it is intended for consoles while the other inputs are for mics. But anyway this test is not realistic so there is not much else I could learn from it.

Second, I took the recorder to a loud rock show. It was a small venue and I positioned myself less than 2 meters from the stack. One act was recorded with the LINE IN input and the other with the XLR input via an adapter. The mics were CA-14 with the ugly battery box. Neither recording had any particular issue with respect to capturing the volume, and they both produced a similar-sized wave form so I decided to try the LINE IN input in a more stressful situation.

Third, I recorded a performer who is notorious for playing at hearing damage levels in a small venue, less than a meter from the stack. Perhaps this test is also unrealistic but for me it is not an uncommon situation. This time I used the LINE IN input only with the same mic mentioned above. The waveform display on the recorder was basically a brickwall the entire time however I only noticed the overload warning appear twice for the duration. I'm not the type to constantly check the level though. After the performance I checked the waveform peaks and only found one portion with audible clipping.

So if you are going to be using this recorder for recordings where clipping is a concern I would not recommend to use the LINE IN input or to use an external preamp so that clipping can be mitigated. To me it seems kind of silly to have to use another piece of equipment to set the level because this is a basic feature on pretty much all recorders up until today.

Wildlife, you have no idea what you started with your innocent question.  See this link for more...

https://www.reddit.com/r/audioengineering/comments/1ipo4wb/revealed_devices_are_being_marketed_as_32_bit/?sort=confidence

I didn't acknowledge you by 'name' but I did at least acknowledge the forum where you had posted!

Glad I could be of some help.

[wildlife]

Wildlife, you have no idea what you started with your innocent question.  See this link for more...

https://www.reddit.com/r/audioengineering/comments/1ipo4wb/revealed_devices_are_being_marketed_as_32_bit/?sort=confidence

I didn't acknowledge you by 'name' but I did at least acknowledge the forum where you had posted!

Apologies for the double post. I read this discussion and thread about the possibility of using some kind of compressing/decompressing technique to get a 32bit float recording got me thinking: one of the points at which I noticed the overlevel warning during the recording was not even the loudest part of the resulting file according to the waveform. Plus, the point where I found an actual clip was not the loudest part of the recording either. I attached a screenshot of the waveform peaks. Mind you that even with the LINE IN input the resulting file does exceed 0dB so in order to see the entire waveform I ran "amplify" in Audacity to reduce the volume by about 17dB, making the peak -0.1dB. Since I wasn't staring at the recorder often there were probably more instances of the overlevel warning appearing than what is noted. But anyway to me the anomaly is that despite three indications of clipping, the spike on the waveform is different for all three of them. The reason being that if the audio clips, it always clips at the same level and therefore the height of the spike should be identical for all three instances. Can someone comment on this?

[cpferris]

Question about this recorder.  New to taping and have soundboard access tonight to a regional band.  I have a zoom h4e.  I will bring a pair of XLR and TRS cables (as I am not sure what the board has available to me).  The Zoom h4e has 2 XLR/TRS inputs on the bottom.  My goal is a single stereo track recording. Is my best option to plug each XLR into inputs 1 and 2?  I wasn't sure if you recorded 2 inputs at the same time if they record 2 different tracks (that would need to be combined)?  Or do I get a Y adapter and run both XLR into input 1?

Just a little nervous and want to be prepared given I am getting SBD access.