Topic: Preliminary Zoom H4 bench test results

[guysonic]

While I intend to do fuller review of Zoom H4 deck, a customer purchased one and drop shipped it here for knowing compatibility with DSM mic, and if needing PA-3SX preamp.  Below is a copy of email I just sent her that should be of interest to this group

-------partial email about H4 results---------

Spent about 6 hours testing of the H4 deck with/without preamp, and as external USB recorder to computer recording software (as powered by only USB connection).
 
In short, Yes this deck will function in at least OK manner as portable audio deck ONLY with external preamp.  Mic input VERY noisy, at least with power supply noises when running on internal batteries. 
 
Also, H4 is quite awkward to operate as a deck, being very much like Sony MD that forgets user set record quality settings with every ON/OFF cycle.  Also awkward is no stereo ganged-together REC level adjust, so each channel needs menu steps for each that takes total of 20-30 seconds time, or so it seems.   
 
Most are having severe overload clipping distortions using the deck, and found the cause to be REC level adjust has 127 steps for setting record level for good VU level.  BUT out of the 127 steps, ONLY #100 - #127 range works to control level.  Adjusting REC level below #100 will lower VU levels, but 1st INPUT STAGE IS THEN CLIP DISTORTING!!!  On/Off cycle defaults REC level to #100 adjustment setting on both channels in STEREO mode.
 
So far no one I know has mentioned this deck's limited REC adjustment range while recording from analog inputs. 
 
In other words, using full 127 range for level is ONLY NOT a problem when using the deck's mixer function that works on ALREADY RECORDED mix-track FILE levels, and this (4 track mode) does have full 127 step range without overload distortion liability. 
 
TIP: When using external preamplifier with H4, ALWAYS upon turning on for recording manually set each channel recording level to #115. 
This will allow for +/- 7 dB (14 dB total) #100 - #127 adjustment range, and when set to #115, my PA-3SX preamp's clip light is EXACTLY the same for deck input near-clip condition.
 
This is why it is important to test these decks for functioning.
 
Advanced 'Oxy nickel' Alkaline Batteries seem to last no more than 3-4 hours before flashing low battery shows, which is OK, but likely less time on regular lower costing Alkaline.
 
BAD NEWS in this deck DOES NOT WORK because of horrendous noise (see photo) as external computer soundcard recording function, at least while running on USB adapter without AC adapter attached.  I should do test using AC adapter, if this does any better as the USB connection turns on and powers deck automatically.  Research indicates Zoom engineers know this, and maybe connecting AC adapter reduces most noise?
 
WORST NEWS is this deck DOES NOT HAVE smooth frequency response.  Two reductions of 6000 cycles and another roll-off at 15000 cycles (see below).  Highest quality 96K sample rate shows another serious reduction knee at 35000 cycles.  There is little excuse for the declining high frequency that looks like the mediocre audio performance of a prosumer camcorder. 

Less troublesome is the LINE and Headphones output frequency rolls off even more severe, sounds like AM radio with 5000 cycles bandwidth!
 
In summary, suggest forgoing all the interesting H4 musician features.  Instead get R-09 with built-in piece of mind refinements that's far easier to use with lower costing PA-3SX preamp (no dual 1/4" molded plug cost) with stereo mini plug output.  Also get the Viewcase for R-09 for outdoor OLED viewing
 
I will pay for return shipping on the H4 as thank you for opportunity to know this 'piece of work' by Samson. 

------clipped from email-------------

There's more awful truth about this deck, but I think Samson needs go back to engineering with this one.  Not likely at this point as H4 model seems to have been around for going on 2 years with Samson saying everything is fine in their opinion.
 

[Roving Sign]

Are any of these issues fixable by firmware updates?

[guysonic]

Are any of these issues fixable by firmware updates?

Good question. 

Firmware might correct for lost user settings with ON/OFF cycle by either storing in system memory, or on the SD flash card in a systems folder.

Less likely firmware has much to do with the mid-high frequency roll-off, but hard to say if hardware circuitry or digital post process from firmware is at cause.

No chance firmware has anything to do with high frequency noise issues leaking into the mic levels.  This is typically circuit and board design shortcomings, but as shown, using an external preamp corrects for noise, but preamp benefit becomes moot solution for making this a good performing deck because of that terrible high frequency reduction

[bdasilva]

I'm just going to move this so its in the same thread. The power supply will not be fixed by firmware,

Here's a note, reportedly written by a Zoom engineer, describing the
H4's "noise problem."

"We have figured out that the noise problem is caused by the analog
circuit board layout. So, it's not possible to solve by updating the
system software...Noise was available because battery does not apply
enough to load change. So noise is disappeared when you use a supplied
AC adaptor...Also, noise level depends on a SD memory brand, because
power consumption is different."

So far I have not seen reports listing which SD cards might be more
likely to cause noise issues due to higher power requirements.

Another web site has links to recordings that demonstrate the "noise
problem".

"There is an audible (and visible as a wave form in a DAW) periodic
tone (corresponding to the blinking of the H4 access light) on all
recordings made with battery power. (This tone may also be there at
24/96, but I cannot hear it.) There is no tone present when using the
AC adapter.

Here are links to sample recordings:
1. access light tone 44.1.wav http://www.box.net/public/xe7yvzpezb
2. access light tone 48.wav http://www.box.net/public/sobave36ct

Some people have added three 1000 mfd 4v tantal condensors over the
voltage rail.

[Church-Audio]

I was not going to say anything about ZOOM but now that the cat is out of the bag. They make the worst guitar processors in the world... That's what this company does they make guitar effects units... They thought they could make a recording device but...... Well if you cant make a guitar effects processor sound good how the hell are you going to make a digital recorder??? Zoom as the worst service reputation in the business. My friend who manages a music store part of a very popular chain of stores in Canada says... They all dropped the line because the service was so bad and the quality of the product was so bad they did not want the hassle.. So I say lets have a zoom bond fire!!! I think Guy is right I cant see a firmware upgrade fixing Frequency response issues unless they crippled it on purpose to save space??@!

So stick to Edirol sure they are part of Roland and Roland a part of Boss, A real company that makes real guitar effects that sound really good... And has a long tradition of making excellent products for musicians. Or Zoom a company that makes things out of cheap plastic and have the R&D and development skills of a 5 year old.

[Church-Audio]

That's what this company does they make guitar effects units... They thought they could make a recording device but......

Just a guess but I don't think they make the unit.  I think someone in china/taiwan/etc does most everything regarding development and production and it is just being sold under the Zoom brand.. Like any outsourcing effort, it is difficult to fully specify everything necessary to ensure quality.  It is fairly easy and common for the other party to deliver something that meets the letter of the contract but which is not an entirely viable product. We have heard similar rumors about the microtrack's product history..

I hope the Korg units do not suffer in that way.. Given that the MR-1000 is apparently too sensitive on the mic inputs and has no phantom on the TRS line-in, I already have some concerns.

I am pretty sure it comes from the same place.. As for Korg no fears they have always made a great product. They have very good development and they know what sounds good. I have several Korg products and I have always found the build quality to be very good and the sound quality to be even better.

[flintstone]

Since others here (e.g., JoeShambro, Spyder9, Will_S) have reported some success recording with the H4, we have to consider the possibility that the unit tested by Guysonic was defective.  Wouldn't be the first time a bum recorder found its way out the factory doors.

Flintstone

[Church-Audio]

Since others here (e.g., JoeShambro, Spyder9, Will_S) have reported some success recording with the H4, we have to consider the possibility that the unit tested by Guysonic was defective.  Wouldn't be the first time a bum recorder found its way out the factory doors.

Flintstone

You could be right. But I doubt it, I would love to get my hands on one to test. I have an idea if I can send a file to someone that has one transfer the file via the line out on a good sound card into the line input. I could at least measure frequency response of the returned file. It would have to be in a WAV format. Anyone interested let me know. This would not be a perfect test but it would give us an idea of 20hz to 20khz I could do much better if I had a recorder in my hands.

[Will_S]

I was thinking of doing some really simple testing of the frequency response myself.  I don't have the tools that Guysonic or Chris has, but is there something fundamentally wrong with this:

In sound studio, I will create a wav file that starts out with a 1 kHz sine wav at -6 dB, and then transitions to a 10 kHz sine wave at -6 dB.  I will play this back through my Presonus Firebox line in to the Zoom.  Then I'll open up the resulting wav file and see if the peaks are indeed ~ 4 dB lower for the second half of the file as you would expect from Guy's tests.  I was surprised how bad the FR reported by Guy was and wonder if he did indeed get a bum unit.  I've never tested mine but have been satisfied with the sound, maybe it's just a good match to my bright AT mics.

I could also try the file Chris suggested if it's likely to be more informative.

And there may be a lull in my taping between March 17 and April 21 that would allow me to loan my Zooom out to Chris.  Any chance you (Chris) could do the 4.7K mod to my AT853s while you have the Zoom and we're shipping things back and forth anyway?

As to the noise figures, the Zoom preamp is not super-quiet for sure but I've never heard the access light noise some people report on any of my actual music recordings (see link in signature, and look at the more recent stuff eg Waybacks 12/1/06).  I've heard it depends on the SD card you run, the one the H4 comes with may be a current hog where as my Trasncend 4 GB may not.  (And/or it may help that I use good quality NiMH batteries that may deal better with the current draw???)

As to using it via the USB connection, I have no interest in doing so and never tried it, although I believe noise through the USB connection IS an issue the first firmware update was supposed to address.  I never cared enough to check.

[Church-Audio]

I was thinking of doing some really simple testing of the frequency response myself.  I don't have the tools that Guysonic or Chris has, but is there something fundamentally wrong with this:

In sound studio, I will create a wav file that starts out with a 1 kHz sine wav at -6 dB, and then transitions to a 10 kHz sine wave at -6 dB.  I will play this back through my Presonus Firebox line in to the Zoom.  Then I'll open up the resulting wav file and see if the peaks are indeed ~ 4 dB lower for the second half of the file as you would expect from Guy's tests.  I was surprised how bad the FR reported by Guy was and wonder if he did indeed get a bum unit.  I've never tested mine but have been satisfied with the sound, maybe it's just a good match to my bright AT mics.

I could also try the file Chris suggested if it's likely to be more informative.

And there may be a lull in my taping between March 17 and April 21 that would allow me to loan my Zooom out to Chris.  Any chance you (Chris) could do the 4.7K mod to my AT853s while you have the Zoom and we're shipping things back and forth anyway?

As to the noise figures, the Zoom preamp is not super-quiet for sure but I've never heard the access light noise some people report on any of my actual music recordings (see link in signature, and look at the more recent stuff eg Waybacks 12/1/06).  I've heard it depends on the SD card you run, the one the H4 comes with may be a current hog where as my Trasncend 4 GB may not.  (And/or it may help that I use good quality NiMH batteries that may deal better with the current draw???)

As to using it via the USB connection, I have no interest in doing so and never tried it, although I believe noise through the USB connection IS an issue the first firmware update was supposed to address.  I never cared enough to check.

If you want me to do that mod not a problem...

[guysonic]

I admit to having a bad feeling about the H4 when a few customers asked about it last year.  And one kind taper on this list offered a loan a few months back that I passed on because of this feeling, and having ability to do a quick evaluation/return was questionable.

But one musician customer using a Sharp MD was motivated to buy the H4 for having 4-track mix features felt useful for her purposes, so I agreed to test the unit.   

Initially found the H4's preamp full of high frequency switching noise (see spectrum below), and the Line/phones output was severely bandwidth limited as noted in the post so tested noise/bandwidth ONLY with external preamp. 

I reasoned if H4 is going to be usable at all for mic recording, it had to be in the lowest sensitivity 'line level' setting using the unbalanced dual TS input.   As the graph previously posted shows, with external PA-3SX preamp the H4 is acceptably quiet enough for serious consideration. 

However, the limited high frequency bandwidth (which was measured several times because I couldn't believe it to be that bad) was the showstopper removing any hope of serious use as a recorder.

Doubtful the "M/H" input settings have more available bandwidth.  This is moot because noise is way excessive on all input settings on this deck. 

Using this deck with AC adapter to reduce noise to acceptable level is a possibility, but makes this deck useless for the purposes of versatile portable recording, so not a "fix' option unless only interest is AC powered computer-connected recording where lower noise is claimed 'not-a-problem.'   

Also consider all noise, including switching noise, at output of audio circuits is ALWAYS reduced by rolling off the high frequency audio bandwidth response like was found.  So I feel the limited bandwidth is a 'noise fix' done with firmware same as the 'mic modeling' feature operating with the internal mics?

If the bandwidth is being limited (at the three 'knee' frequencies noted) to reduce the noise with 'modeling' process firmware, then possibility to remove this when in 'LINE' level mode so external preamp gives H4 a chance of decent quality recording not yet possible in the (defective?) unit tested. 

Of course, if full bandwidth flat to 45k cycles is established like it should have, high frequency noise in the graph is going to rise, and if not already noticing, the low frequency noise in the H4 is really the worst I remember seeing in a deck. 

Bottom line, unless a complete redesign,  in my opinion only hope for H4 is restoring flat bandwidth to >40K cycles AND using external mic pre.

Sans the bandwidth restoration, to me external preamp seems a necessary upgrade for virtually ALL low cost flash decks used for at least acoustic recording purposes with moderately low output mics.   

In perspective, it took Sony 2 1/2 generations of DAT deck refinement for mic preamp to start to get really usable, and the best was 4.5 generation M1 model.  More realistic to expect low costing flash deck refinement in 2 more generations where mic preamp is not so needed.

[hummat]

I'm kinda new to whole recording thing.  Thought the H4 would be a great 'all-in-one' starter box (if it seems to good to be true, it probably is).  Took it out a few times and had terrible results, did some digging on the internet and found all of the issues/problems are in the unit.  I've since moved on.  So. . .

Chris, if you want a unit to test for an open-ended length of time, PM me. 

[Church-Audio]

I'm kinda new to whole recording thing.  Thought the H4 would be a great 'all-in-one' starter box (if it seems to good to be true, it probably is).  Took it out a few times and had terrible results, did some digging on the internet and found all of the issues/problems are in the unit.  I've since moved on.  So. . .

Chris, if you want a unit to test for an open-ended length of time, PM me. 

Yes I would like to take a look at this thing. I need a small memory card for testing as well if you have a 64 meg card that would be fine.

I will pm you with my address

[Will_S]

Ouch.  By my own primitive measurements, my Zoom seems to measure a bit better than the unit Guy tested, but not by much.  Mine "only" seems to be down ~2.5 dB at 10 kHz but looks to be 5-6 dB down by 20 kHz.  Attached as flac files are the source ("Input") file with sine wavs at a couple different frequencies all at - 3dB, and the result ("Output") when I recorded it from the 1/4" outs of my Presonus Firebox into the 1/4" ins of the Zoom.

That said, for a $300 unit that provides phantom power, has XLR inputs, and runs for over 3 hours off a pair of AA batteries, I have to say that I'm farily satisfied with the sound under real world conditions.  I'm quite happy with the sound of my Waybacks 12/1/06, JGB Band 10/27/06, and Dan Bern 11/9/06 pulls (all on the archive) using mics > Zoom.  If I have room to set up Presonus Firebox > laptop I'll go ahead and do that, but when space is tight I still find the Zoom gets the job done.

http://homepage.mac.com/satterwill/Input.flac
http://homepage.mac.com/satterwill/Output.flac

[Will_S]

In fairness to the Zoom, I repeated this test with my iRiver, going line in.

The iRiver was flat to 10 kHz and down 1 dB at 20 kHz.  Hmmm.  Not good, Zoom.