H9 Sample Rate question

[kevin_jazz]

According to the Eventide H9 FAQ:

"What is the sample rate?

Because there is no digital I/O the sample rate is optimized for maximum audio performance for each effect and is program dependent.

"

Are they saying that the H9 is an analog device?  Surely not. If it is a digital device, how can there be no A/D D/A converter??

Kevin

[nickrose]

They are neither saying that it is an analog device, nor that it has no A/D/A converter.

The statement says "Because there is no digital I/O" – this means that it has no AES, S/PDIF or other digital I/O. It is entirely possible to have a digital device with only analog I/O, as is the case here. You go in analog, through an A/D to digital, perform your digital processing, then send to a D/A, which produces an analog output.

[jsaliga]

If the H9 has a DSP then there is both ADC and DAC functions.  You can’t do digital signal processing on an analog signal without them.  So whether or not there is digital I/O is irrelevant, and the question about the bit depth and sampling rates used intenally by the Eventide H9, or other pedals in the Eventide lineup that use DSP is completely valid and on point.  I can’t find this information anywhere, and I won’t be buying any Eventide pedals unless and until these basic disclosures are made.

Boss, Strymon, and others have no problem supplying customers with this information.  Not sure why it is such an issue for Eventide unless it is using something like 16bit/44.1KHz.

[noodle1]

jsaliga wrote:

So whether or not there is digital I/O is irrelevant.

Wrong! Its entirely relevant. And has to do with the decision of how sampling is done. There are also things like asynchronous sampling, i.e. avoiding sequential sampling requirements. In this case, what sampling and methods they use are part of their own proprietary information which can be released or held back at their own discretion. Trying to pressure them or using rude hyperbole about them hiding only shows what you are not considering in your thinking.

jsaliga wrote:

Not sure why it is such an issue for Eventide unless it is using something like 16bit/44.1KHz.

So scrap basically every great effects box that was used with synthesizers and other instruments in the past to make all those great tones? Since they didn’t have what you call “fidelity”. Perhaps I’m biased but I think its interesting, the rise of sterile, lifeless effects and effects boxes, which dominate today, all in the name of the new criteria of “fidelity”. 

 

[nickrose]

We use 10 bit 22.05kHz. It's a wonder they sound so good.

 

[Given To Fly]

nickrose wrote:

We use 10 bit 22.05kHz. It’s a wonder they sound so good.

Aww, come on! Now my H9 sounds bad and I’ll have to throw it away in the trash! I guess I’ll buy Strymon products now because I can hear sound quality when I read specs. It is kind of like chromesthesia. 

 

[VFuente]

nickrose wrote:

We use 10 bit 22.05kHz. It’s a wonder they sound so good.

I’ve trying to find information on the H9 DAC bit depth and sample rate. Only answer I find is this one and this quote from the H9 FAQ:
“What is the sample rate? Because there is no digital I/O the sample rate is optimized for maximum audio performance for each effect and is program dependent.”

Is it really a max of 10bit/22kHz?? Can someone please verify this or point me to the actual information or official source (unless I’ve missed it manual and specs don’t say anything). Looking at my H9 board all I’m able to find is the reference to the DSP chip (DSPB56367AG150-K41R) and as far as I know, I’m very ignorant in this matter, does not perform the actual A/D, D/A conversion.

Thanks in advance

[Perfumado]

haha wtf eventide, yes the guitar is a shit instrument, so the samplerate is irrellevent, it will sound better pre-effects. but if she/he own a 20k synth, and want to add Fx the samplerate is not irrelevent at all, delay and reverb can be discussed, but with things like distortion, higher samplerates are in 9/10 cases “better”. 

why not just give info on the samplerates and explain to people why your algorithms sound better at low samplerates? with your digital delay unit and rack effects unit, you are very proud of the samplerate and bitdepth hmm…

[Fender17]

Well, as stated in the manual, there is not just one samplerate used. 

It may be interesting, but the resulting sound is what makes the effect, not the knowledge of sample rate used. 

There already is loudness war, megapixel war, …. Should we join sample rate and bit-depth war? Probably more important than sample rate is quality od A/D and D/A conversion and quality of DSP algorithms. And many other technical things.

Why not to look at it as a blackbox. It eiher does what one needs/likes or does not. No need to bother by innerworkings.

 

[jsaliga]

Fender17 wrote:

There already is loudness war, megapixel war, …. Should we join sample rate and bit-depth war? Probably more important than sample rate is quality od A/D and D/A conversion and quality of DSP algorithms. And many other technical things.

That’s a false equivalence.  And I find it rather striking that you seem to be arguing that ignorance is bliss, or knowledge is irrelevant.  Knowledge is power.  It always has been and always will be.  I’m not saying the quality of the algorithms aren’t important, but I disagree that data rates aren’t important.  A sampling rate of 22.05KHz has a Nyquist frequency of  11.025KHz.  So anything above that will be cut off.  Now if you want to make the case that these higher frequencies are irrelevant and that 10bits of sampling depth provides sufficient dynamic range and low enough noise then fair enough.  That might be acceptable for an electric guitar, but I would not use your effects pedals with a synthesizer.

[Fender17]

Have you tried H9 with the synth? Does it sound good or not? That is what is important. IMHO

But, if you want to find out what the sampling rate is and if the bit depth is satisfactory and what is the quality of AD and DA conversion, feed it with a sine wave of higher frequencies and compare input and output.

[Perfumado]

there is no war, samplerate and bitdepth is never irrelevant. Ensoniq asr samplers “14 bit 44100” sounds “great”, but if you are after a clean/faithfull sound a emu ore yamaha sampler would at the time be the better choice. Yes as a example, using numerical integration/continuous model approach, nice sounding distortion can be achived at a lower samplerate, but even then samplerate does not become irrelevant. so instead of talking about conspiracy and war, why not just give people the info, so they can decide. What sounds great is a matter of personal opinion.

And yes A/D D/A converters is never irelevant either.

i just felt bad for kevin and wanted to support him, since his fair question was being rediculed.

now i am thinking since the samplerate is dynamic how cool would it be if it was user definable! 🙂

[Fender17]

Sample rate and bitdepth is indeed irrelevant when it exeeds certain level. Sample rate and bitdepth should match the purpose intended by designer and desired by user. The purpose of pedal is to change the sound not to maintain the fidelity of the sound. One may like the sound tweaks one may not. That is what is important. Knowledge of innerworkings may be interresting, but it says almost nothing about resulting sound….

I do use h9 with synths and I like the sound. That is enought for me to know about H9.

[jsaliga]

Fender17 wrote:

Sample rate and bitdepth is indeed irrelevant when it exeeds certain level. Sample rate and bitdepth should match the purpose intended by designer and desired by user. The purpose of pedal is to change the sound not to maintain the fidelity of the sound. One may like the sound tweaks one may not. That is what is important. Knowledge of innerworkings may be interresting, but it says almost nothing about resulting sound….

I do use h9 with synths and I like the sound. That is enought for me to know about H9.

You get to speak for yourself, but you don’t get to speak for everyone else.  Your remarks about bit depth and sampling rates are nebulous at best.  I’m glad the pedal works for you.  But I have already ruled it out due to poor fidelity, which may not matter to you, but it does matter to me.  We clearly have different needs.

To each his own.

[Perfumado]

 

how to desire unknown samplerates and achieve true self inside through the blackbox principle

[Fender17]

Well, if one desires certain samplerate, than may make decison based on specsheet . If one desires certain sound than he has to try the unit in flesh. I both cases you have blackbox. 

[jsaliga]

Fender17 wrote:

Well, if one desires certain samplerate, than may make decison based on specsheet . If one desires certain sound than he has to try the unit in flesh. I both cases you have blackbox.

That comment doesn’t even pass the laugh test, and it’s getting increasingly difficult by the post to take you seriously because you’re starting to sound like an Eventide fanboy.  Sampling rate and sound quality are not mutually exclusive, and you seem to be arguing that one must chose one over the other.  I have both, fancy that!     Sample rate goes directly to fidelity.  Moreover, no sensible person would buy a pedal without trying it out.

 

[Fender17]

jsaliga wrote:

 Sampling rate and sound quality are not mutually exclusive, and you seem to be arguing that one must chose one over the other.  I have both, fancy that!     Sample rate goes directly to fidelity.  Moreover, no sensible person would buy a pedal without trying it out.

My agument is that no sensible person would sort out certain pedal (made by renoved manufacturer) only based on samplerate it uses, without trying it out.:-)

Higher samplerate may bring better sound, but it may also bring worse sound if not engineered well, if algorithms used are poor, and so on. I may patch 192 kHz/32 bit pedal for you, but I give you garranty that it will sound terrible at best. 😉

Have you tried H9 in person?

I a not Eventide fan, but I have technical educaton and I know that implementation is quite often more important than datasheet numbers. And that things often surprisingly work very well even thought one may thing it may not work well. Engineering is mosly about compromises and the result is, what counts, not the means used to gain the result…

[jsaliga]

Fender17 wrote:

My agument is that no sensible person would sort out certain pedal (made by renoved manufacturer) only based on samplerate it uses, without trying it out.:-)

You’re preaching to the choir.

Quote:

Higher samplerate may bring better sound, but it may also bring worse sound if not engineered well, if algorithms used are poor, and so on. I may patch 192 kHz/32 bit pedal for you, but I give you garranty that it will sound terrible at best. 😉

That’s a non-sequitur.

Quote:

Have you tried H9 in person?

I don’t need to.  It is a very crowded and competitive market.  I don’t need to try every single effects pedal on the market.  I narrow the contenders to a list of products I want to try based on features, reviews, and yes, specifications (such a dirty word around here).  What I do not do is go through a buying process that meets with the approval of someone on the internet who I don’t know.  The only person I am interested in pleasing with my equipment purchases is me.  And for the record I never said the Eventide sounded bad.  What I did say is that it has poor fidelity and I stand by that assertion.  You might care to take a look at this…

https://www.cco.caltech.edu/~boyk/spectra/spectra.htm

From a fidelity perspective I consider 24bit/96KHz the absolute bare minimum.  There are plenty of excellent effects pedals with DSPs that use a 96KHz sampling rate.

Quote:

I a not Eventide fan, but I have technical educaton and I know that implementation is quite often more important than datasheet numbers.

I have two degrees: one in electrical engineering and another in computer science.

Quote:

And that things often surprisingly work very well even thought one may thing it may not work well. Engineering is mosly about compromises and the result is, what counts, not the means used to gain the result…

Meaningless doublespeak.

Let’s just agree to disagree.  I have lost all interest in taking this exchange any further.  Feel free to have the last word if you must.

[Fender17]

jsaliga wrote:

And for the record I never said the Eventide sounded bad.  What I did say is that it has poor fidelity and I stand by that assertion.  You might care to take a look at this…

https://www.cco.caltech.edu/~boyk/spectra/spectra.htm

From a fidelity perspective I consider 24bit/96KHz the absolute bare minimum.  There are plenty of excellent effects pedals with DSPs that use a 96KHz sampling rate.

OK, I understand and accept that for you techspec is important.

That Caltech article is interesting. I interpret it in the way that 20kHz+ sounds switch the brain to alpha-wave state which goes on for some time after 20 kHz+ sound diminishes . But one does not hear the sound above 20 kHz. It just alters the state of brain and possibly perception of sounds below 20 kHz.

But at least I know, why I went into this conversation, now. The info that 20 kHz+ sound induces alpha state may be important for my “research” on influence of sound and movement to mind state. So, thanks for that info. It will be helpfull for me in future.

[jsaliga]

Feel better now??

[noodle1]

jsaliga wrote:

Feel better now??

Just pointing out the output sampling of FX on certain designs isn’t necessarily the same or a fixed sampling rate

[Fender17]

Have you managed to try H9, yet?

[jsaliga]

Fender17 wrote:

Have you managed to try H9, yet?

No.  Why would I?  I am perfectly happy with my current lineup of effects.  There is no need to fix something that I don’t think is broke or requires change.  And just to be completely open, some of my favorite effects are not even guitar pedals: they are Eurorack modules.  For example, the most interesting reverb I have ever heard is the Erbe-Verb module from Make Noise.  I use it mostly with my modular system but I have also run guitars through it from time to time.  I absolutely love it but will readily concede that it’s not for everyone.  Since it has no presets it can get pretty wild and there is something of a learning curve to getting it dialed in.  But that said it is massively flexible in ways that other reverbs are not.  For example, you can modulate many parameters with a control voltage or LFO for some pretty mind blowing effects.

I suggest that you guys not get too wound up because I am not that impressed with the Eventide technical design.  I pretty much said what I had to say here four years ago and my position hasn’t changed nor is it likely to.  Eventide seems to be doing just fine without selling me any pedals and I am fairly certain my misgivings about the brand is not going to cause any loyal Eventide users to trash their pedals and abandon the brand.

My number one rule with gear is not to care what someone else thinks of my choices and preferences.  I don’t buy gear to please them…I buy it please me and to fill a musical need or add a capability that I lacked previously.  What someone else makes of my choices isn’t relevant and so I try to avoid situations where I feel as though I have to defend my preferences and choices.  That is wholly unnecessary.

[noodle1]

Maybe he asked if you tried the H9 yet because you were here asking about the device. Well, by asking I mean demanding eventide tell you the sample rates at the cost of your business. And now you’re saying “Why would I?” Well, nobody would care to ask you, if you weren’t here asking about the gear in the first place so don’t act so surprised.

jsaliga wrote:

But that said it is massively flexible in ways that other reverbs are not.  For example, you can modulate many parameters with a control voltage or LFO for some pretty mind blowing effects.

Eventide FX processors can do this, with ANY parameter(s), at 24bit/96K too!

jsaliga wrote:

I pretty much said what I had to say here four years ago and my position hasn’t changed nor is it likely to.

Actually, you came and sh*t on eventide insinuating they were “hiding” their sample rates, demanded the information rudely, and claimed that 24bit/96k was the bare minimum for any effects, tried to prove it with an ultrasonics whitepaper, and now hilariously are still here, except now you’re talking about your favorite (48Kbps btw) reverb pedal.

jsaliga wrote:

What someone else makes of my choices isn’t relevant…

True, except it is relevant when you make specific claims about your choices on a public forum. (And even mislead people when it comes to effects)

 

[Fender17]

jsaliga wrote:

 

For example, the most interesting reverb I have ever heard is the Erbe-Verb module from Make Noise.

Time to hear H9 Max… Or other Eventide’s processors, if you require even higher sound fidelity. 😉 H9000 is a good starting point.

[noodle1]

Sorry about that, I needed to chill, I was upset about the fanboy and eventide comments. But just to answer the question, all the big brother Eventides can do 24/96, but that also it’s not always as simple as a simple answer. Sorry for the conflict, I just feel like people nowadays are too “spec” driven, when a lot of the time the magic is how these things are utilized.  That being said, I did a lot of A/B comparisons of 24/96 to 24/48 on an eventide rack unit and I have to say the 96 sounded a bit more open and better but it seemed like there was a volume drop. I will concur with your statement earlier than some effects are better off with the higher bitrate, I just don’t want people to think the H9 is inferior. A lot of it has to do with how the unit is hooked up too! parallel wet is usually the best way to mix an eventide, especially when they are crunching bits/samples. Honestly I feel like nickrose’s 10bit/22khz comment was actually a troll/joke comment, but I’m not a huge expert on the H9s so I may be wrong. I do know that the H9 is not 24/96K tho!
Wish you all the best.

[jsaliga]

Please do me the basic courtesy of not putting words in my mouth or spinning what I have said in a way that alters my meaning.  If you want to refer to something I said then QUOTE it.  

You have posted twice here and done it both times…like you are eager to get into a confrontation.  Either it stops or I stop commenting here.

[jsaliga]

Funny how you quote me and then completely distort and mangle what I actually said.  We are done and I won’t be back to comment further.  The fanboys have won.

[jsaliga]

Oh, and I would like to thank the Eventide folks who responded and  answered my question about the sampling rate.

For everyone else: Relax, and take a chill pill.  At least one of you needs it.

Bye now.

[Funkstar De Luxe]

It operates at 44.1kHz. I can’t measure the bit depth accurately but the noise floor isn’t low enough to go beyond 16bit anyway (<96dB)

High frequencies are properly attenuated at ~20kHz, so no need to worry about aliasing.

H9 sounds lovely and it’s sampling rate isn’t really relevant. However it does show that Eventide lack confidence in their engineering (and perhaps their customers) that they won’t publish this information.

[Fender17]

I guess, that the thing is that different algorithms run on different bitrate or bitdepth. So, if specifying bitrate/bitdepth Eventide would have to list which algorithms use this and which use that. Not speaking that things may be even more complicated….

Even order, two orders more expensive rackmounts like H9000, H8000, Eclipse use different bitrates for different algorithms…. And some users complain, they want the same bitrate for all.

[brock]

VFuente wrote:

nickrose wrote:

We use 10 bit 22.05kHz. It’s a wonder they sound so good.

,,, Is it really a max of 10bit/22kHz?? Can someone please verify this or point me to the actual information or official source …

You can’t get any more official of a response than from Nick Rose.

[VFuente]

@brock, my english ain’t so good, I read booth this comments and I’m still confused:

Funkstar De Luxe wrote:

It operates at 44.1kHz. I can’t measure the bit depth accurately but the noise floor isn’t low enough to go beyond 16bit anyway (<96dB) High frequencies are properly attenuated at ~20kHz, so no need to worry about aliasing. H9 sounds lovely and it’s sampling rate isn’t really relevant. However it does show that Eventide lack confidence in their engineering (and perhaps their customers) that they won’t publish this information.

noodle1 wrote:

Honestly I feel like nickrose’s 10bit/22khz comment was actually a troll/joke comment, but I’m not a huge expert on the H9s so I may be wrong.

I undestood that part, I’m asking for the A/D converter maximum capabilities don’t need each individiual algorithm values:

Fender17 wrote:

I guess, that the thing is that different algorithms run on different bitrate or bitdepth. So, if specifying bitrate/bitdepth Eventide would have to list which algorithms use this and which use that. Not speaking that things may be even more complicated….

@nickrose Once again sorry for my language limits, 10bit/22khz is the actual real response and no joke or irony? (⌒_⌒;)

[Funkstar De Luxe]

Again, the sample rate is 44.1kHz. When Nick said 22kHz he meant that was the highest frequency, not the sample rate.

Lastly, the dynamic range of the device does not exceeds 16bit.

The H9 is 44.1kHz with an actual dynamic range less than 16bit. It shocks me how much misinformation is in this forum.

[VFuente]

Thanks and sorry, it’s so much easier when is just listed on the technical specs.

[Author]

Posts