What's new

Hollywood Strings play Wagner + Hermode Test

parnasso

Member
Hi all,

this actually started out as a template test, I wanted to see how well the Hollywood cinematic style libraries would fare in a classical late-romanticism environment. Then Cubase 7 with Hermode tuning came along and so I decided to make a comparison between a version with Hermode tuning applied and one without.

I'm interested to hear what you think, for me the difference is subtle but the Hermode version sounds more brilliant as it makes the tuning generally a bit higher and it corresponds more to how a real orchestra would tune (a have a recording of this piece with Karajan and the Berlin Philharmonics where the tuning is still a bit higher though). However the sound tends to get a bit thinner in my impression, maybe it's a consequence of the slight pitch-shifting that gets applied in the sampler in order to retune the samples.
The version without Hermode tuning is flatter and thus a little darker and less brilliant but maybe also warmer. What do you think? The difference at the beginning with the Cello section is maybe not that big but listen to when the other string sections come in one by one and finally when the brass hits in, there I find a noticeable difference.
Of course I'm also interested in any other comment on this mockup, this is my first project with these libraries and I'm still learning how to use these samples and mix them… many thanks!

The piece is the prelude to the 3rd act of Meistersinger by Richard Wagner (a beautiful, slow and highly contrapunctual composition) and the libraries used are Hollywood Strings, Hollywood Brass, Cinebrass Pro and Berlin Woodwinds.

So here it is:

Version with Hermode tuning:

mp3: https://www.box.com/s/eiqpa2l1lk9a72t21wsg
wav: https://www.box.com/s/4vs4sldq3zn4ec3a7885

Version without Hermode tuning:

mp3: https://www.box.com/s/refwwcifaefklr7s34fh
wav: https://www.box.com/s/t19mvirp2ng69rwvjv09

Thank you very much for listening and I'd be grateful for any comments!
 

jamwerks

Senior Member
Good example. I'm impressed by the sound of EW, your programming(!), and also the Hermode thing.

To my ears, the Hermode example sounds a tad more "real". Apparently, for example, when it detects a 5th, it raises the upper not by 2 cents, a major third by 2 or three (I think), and a minor 7th by 3 or 4.

Makes me think that all daw's will one day have that.
 

XcesSound

Member
Thank you for these 2 great demos. The difference is very subtle but I do prefer the one with Hermode.

The mock-ups are great! I especially was drawn to the brass sound, I did notice some of the bumps in dynamic but overall brilliant. Thanks
 

Gerd Kaeding

Active Member
Hi all,

this actually started out as a template test, I wanted to see how well the Hollywood cinematic style libraries would fare in a classical late-romanticism environment. Then Cubase 7 with Hermode tuning came along and so I decided to make a comparison between a version with Hermode tuning applied and one without.

I'm interested to hear what you think, for me the difference is subtle but the Hermode version sounds more brilliant as it makes the tuning generally a bit higher and it corresponds more to how a real orchestra would tune (a have a recording of this piece with Karajan and the Berlin Philharmonics where the tuning is still a bit higher though). However the sound tends to get a bit thinner in my impression, maybe it's a consequence of the slight pitch-shifting that gets applied in the sampler in order to retune the samples.
The version without Hermode tuning is flater and thus a little darker and less brilliant but maybe also warmer. What do you think? The difference at the beginning with the Cello section is maybe not that big but listen to when the other string sections come in one by one and finally when the brass hits in, there I find a noticeable difference.
Of course I'm also interested in any other comment on this mockup, this is my first project with these libraries and I'm still learning how to use these samples and mix them… many thanks!

The piece is the prelude to the 3. act of Meistersinger by Richard Wagner (a beautiful, slow and highly contrapunctual composition) and the libraries used are Hollywood Strings, Hollywood Brass, Cinebrass Pro and Berlin Woodwinds.

So here it is:

Version with Hermode tuning:

mp3: https://www.box.com/s/eiqpa2l1lk9a72t21wsg
wav: https://www.box.com/s/4vs4sldq3zn4ec3a7885

Version without Hermode tuning:

mp3: https://www.box.com/s/refwwcifaefklr7s34fh
wav: https://www.box.com/s/t19mvirp2ng69rwvjv09

Thank you very much for listening and I'd be grateful for any comments!

Hi Leo ,

first of : I really like your programming !

Concerning the tuning : I actually prefer the Hermode tuning .

Thanks for sharing !!!


Makes me think that all daw's will one day have that.

Logic has those tuning options/possibilities for quite a while now ( http://help.apple.com/logicpro/mac/9.1. ... tasks=true ) , including the Hermode tuning . However this only works with Logic's internal EXS sampleplayer and its Synths , not with thirdparty AudioUnits.
I'm not sure if this is due to the AudioUnit standard ( opposed to Steinbergs VST 2/3 ) , or to Logic's specific limitations.
After hearing the example I'd love to use the Hermode tuning on my own HS/HB/... arrangements now. ... ( ... Why oh why is Cubase always one step further ...??? ... :| )



Best
Gerd
 

Sid Francis

Senior Member
Funny that there are differences: In the german Cubase forum, someone mentioned, that he was sad since the anticipated Hermode tuning would only be applicable to the instruments that are delivered with Cubase 7 and not to all VSTIs. But I did not yet delve into this new feature....
 
I don't hear any difference...sorry.
And as far as I know (and tested) Hermode tuning doesn't work with Kontakt or Play...
I would be very happy to be proved wrong though...
 

TGV

Senior Member
I thought I could hear a very subtle difference, indeed when the brass joins the party at 1:22, but ... omg, did I wet my pants. What an beautiful rendering, apart from the brass that's a bit wavering, as XcesSound points out.
 
OP
P

parnasso

Member
Thread starter
  • Thread Starter
  • Thread Starter
  • #9
Hi guys, thank you very much for your nice comments, I'm very glad that you liked my rendering!

I admit that also for me the difference is very subtle but for example at 0:51 when the 2nd violins kick in their F seems sharper in the Hermode version and very flat (nearly an E...) in the non Hermode version (by the way, I find the tuning in HS to be generally a bit on the flat side in comparison to real orchestras, does anyone have this impression as well?). The same goes for the G minor chord at 1:15 and most evidently to me in the brass parts in general and in the G major chord at 1:22 in particular.
But maybe I'm only imagining differences that are not there and I get tricked by my expectations, who knows? :?

In the Cubase manual Hermode tuning (HMT) is described as follows (p.454):

HMT: Follow
Activating this button for a track applies Hermode Tuning to the notes played on this track. Hermode Tuning retunes the notes you play and creates clear frequencies for every fifth and third interval, for example. Retuning only affects individual notes and maintains the pitch relationship between keys and notes. The retuning is a continuous process and takes the musical context into account.
When you apply Hermode Tuning to tracks that use VST 2 instruments, the played notes are retuned with every keystroke. Dynamic retuning while notes are playing is only possible with VST 3 instruments that support Micro Tuning and Note Expression. For VST instruments that support Note Expression, Hermode Tuning also works in MIDI Thru mode.

There's written nothing that Hermode tuning would work only with Cubase's own instruments or that in general it would have special requirements - it's supposed to work even with VST2 instruments, although not with dynamic retuning which requires "VST 3 instruments that support Micro Tuning and Note Expression".

That being said, I just tested HMT with Kontakt hosted directly in Cubase and it doesn't emit any sound when HMT is activated. However, in my Meistersinger project I had Play and Kontakt hosted in VE Pro and connected it to Cubase via the VE Pro VST3 plugin and that worked.

XcesSound said:
The mock-ups are great! I especially was drawn to the brass sound, I did notice some of the bumps in dynamic but overall brilliant. Thanks
TGV said:
I thought I could hear a very subtle difference, indeed when the brass joins the party at 1:22, but ... omg, did I wet my pants. What an beautiful rendering, apart from the brass that's a bit wavering, as XcesSound points out.

XcesSound and TGV, thank you very much for your kind words, but may I ask, what do you mean by the brass having "bumps in dynamics" and "wavering"? Do you mean the vibrato that sometimes creates a bit of a pulsating and wavering sound?
I think that a big brass section, especially at high dynamic levels, does naturally waver a bit as those are instruments that are hard to control and pretty exhausting to play, especially with long notes. Could you show me some places in the piece where you find that the brass could be better? I would be very interested in this as I would like to improve my mock-up skills, thank you very much!
 

Markus-Music

New Member
I've done an interesting and very in-depth Interview with the Inventor of Hermode Tuning Werner Mohrlok. Feel free to read and comment! His statement is, that NI doesn't support HMT at all... So here's the Interview (in German Language) that has just been updated with a statement from VSL!

http://www.cremerseele.de/interview-werner-mohrlok-der-erfinder-des-hermode-tuning/
 
OP
P

parnasso

Member
Thread starter
  • Thread Starter
  • Thread Starter
  • #11
Markus, thank you for posting the interview, that's very interesting!

I can now confirm the bug with Kontakt as I have written in my previous post, even though initially I didn't notice it because I had all of my Kontakt instances hosted in VE Pro and there the notes played as expected ...

So apparently Kontakt doesn't support the "Detune" parameter, but maybe Play does? Mr. Mohrlok points out in the interview that Detune is a well-known parameter and many plug-ins already support it, so maybe Play is one of them. In my project most instruments are Play-based so that would explain the difference I'm hearing. But it's entirely possible that I'm just imagining differences that don't exist :lol:
 

Fagotto

New Member
From the inventor of Hermode Tuning:
Firstly a great compliment for "parnasso" and his impressive musical example. Many thanks also for the encouraging commendation and appreciating words.

A little hint for all, especially for "dfhagai": The human ear is a very sensible tool but it requires for an optimal working about 10 or 15 minutes. You will realize this by visting a symphony concert - and oftenly the orchestra also needs the same time for being "in tune". Frankly, when listening the first time to this musical example I also feeled only little sound differences. But when comparing three times the horns from 1:18 on or the flutes and high strings from 3:10 on, the sound difference seemed me to be enormous.

To "parnasso": The Cello unison in the beginning is not corrected by Hermode Tuning, as for an unison the equal temperament is an adequate tuning method.
Hermode tuning doesn't tune in principal "higher", it corrects the notes of the detected chords and note combinations partial higher, partial deeper than with equal temperament. A typical tuning correction for a major chord (C-major) is C +4, G +6 and E-10 Cents, for a minor chord (a-minor) it is A -6, E -4 an C +10 Cents. In this way the sum of tuning deviation is always near to 0. At least in default. Due to compensatory measure at complex harmonic sequences the average level could be corrected until +/- 15 Cents, but these are exceptions. Against this little increasing and lowering of the avarage tuning height happen continuousl. A specific note could be retuned to a maximum of +/- 25 Cents.
These maximal retuning values concern the "Classic" and the "Baroque" mode. The mode "Jazz-Pop" inducts the "natural seventh" in "little major chords". For example the G in A-C#-E-G. This natural seventh is very deep. The tuning values in default are A +6, C# -8, E + 8, G -25 (!!!). In this way there could come audible retuning steps. On the other hand it sounds great, especially when such chords come up in Tonika function on the end of a phrase.

Regarding the actual situation:
In LOGIC Hermode Tuning is a real time function and it works perfectly with its internal plugins (EXS24 and so on). For external plugins LOGICS sends its tuning messages by the MIDI message "Single Note Tuning Real Time". Therefore the referring plugins have to support this message. AFAIK this is actually only done by the Vienna Symphonic Library". But there exists a new problem: Hermode Tuning works correctly in the arranger window. But when bouncing the result "offline", the tuning jumps back to equal temperament. This is as Logic has forgotten to add time stamps to the variable tuning messages. The bug is known by Apple and they promised to fix it, but until now it hasn't been done and I am tired for reminding it.
So, if working with an external plugin supporting Hermode Tuning you have to record the result online.
IN CUBASE the plugins have to support VST DETUNE which sends the start tuning values of each note. This DETUNE parameter is calculated in advance. In this way Hermode Tuning is not a real time function. But as sometimes notes have to be retuned when changing its harmonic function on the flight, these new retuning values are sent by VST Note Expression. Therefore for a perfect tuing the plugin has to support both messages.
The Vienna Symphonic Library is preparing the support of both messages. The Kontakt and Kontakt player of NI have problems by supporting these messages. Steinberg told me that DETUNE messages let fall in silence all applications of NI, but I cannot affirm it by own experience as I don't possess NI applications.
 

FriFlo

Senior Member
While I find the idea of Hermode tuning brilliant, I see a major problems:

Only VSL samples are tuned very exact, its patches only being a few cents off the the "perfect pitch". Other libraries tend to be a little off pitch more or less. A good example, where this is not necessarily bad (a matter of your personal POW, of course), is LASS. Here, by combining multiple layers of divisi, tuned slightly differently, you get the effect of real players: They never hit it 100% accurately.
Other libraries not following the divisi concept are also more or less out of tune.
This leads me to the conclusion, that hermode tuning will be only useful to a very narrow range of libraries. And even then, you would want to tweak it with pitch bend or tuning facilities to make things more natural. Hence, I believe the best quality is only achieved, if you truely have the ears to do those modifications by hand (or rather pitch bend wheel, ribbon controller, etc.). But this takes REALLY good ears. Usually, only string players have that ability. Otherwise, you will sit there for hours and hours and the result might even turn out worse than equal temperament ...
That is why I find the possibility of hermode tuning really interesting! But the lack of support and interest by most people will probably mean its failure. Sorry to say this ...
 

Fagotto

New Member
FriFlo: you are right: Not all instruments of all sample libraries are tuned to the precise values of equal temperament. And not all instruments of one and the same library are tuned to the same pitch. In case, the difference is little, this could be advantageous for a living, realistic sound. In case, the pitch differences would be higher, it would be more problematic due to unwanted beats.

But in all cases, Hermode Tuning will bring a more lively sound by simulating the tuning behaviour of human string and wind instrument players. The intonation of such musicians is not always perfect but much better than with equal temperament.

A practial example: I possess some recordings of Haydn's Symphony 103. This begins with a timpani roll, followed by a bass unison. Than the wind instruments start with a dominant seven chord. The intonation of such a chord by the first entrance is extremly difficult and at many of these recordings not perfect in tune. The best result is recorded by a live concert of Gerard Schwarz with his Scottish Chamber Orchestra in 1991.
The mentioned start chord shows the notes F-A-C-Eb. A is the highest tone, played by the flute.
The correct values of a tuning system with perfect tuned fifths and thirds are:
F 0, A -14, C 2, Eb 0.
(The numbers mean the deviation from equal temperament in Cents).
The values of the Scottish chamber orchestra show the values
F 0, A -12, C 2, Eb 0
( I have set the root onto the value 0 for a better understanding. In comparision to an equal temperament with A = 440 HZ the pitch of this orchestra is a little higher).

I feel, you haven't worked with Hermode Tuning until now. So you don't know that you can control the tuning by selecting the correction degree with the help of 11 steps between 100 % (full correction), 90 %, 80 % and so on, until 0 % (equal temperament).
In this way you could reduce the degree of correction for instance to 90 % so that you would get the same result as the Scottish Chamber Orchestra. But it could be that 100 % will make you still more happy.

Besides, Hermode Tuning reflects by each correction how to change the pitch below audibility. In case of conflicts it reduces automatically the correction, so that the result will not be in every case a theoretical perfect intonation. Besides, there exist a lot of note combinations which cannot be tuned to perfection. One example: An augmented triad (C-E-G#/Ab) consists in three major thirds. In case we would tune them all to 5/4 we would destroy the octave. And this is not possible. In such cases Hermode Tuning tunes to an inequal temperament with priority to inaudible retuning steps.

Bringing the things on the point: Hermode Tuning is no abstract architecture following the idea of a theoretical perfect just intonation. It is in contrast to this an elastic system, emending the tuning and as a result of this ameliorating the sound. Independend of the quality of the sound material.
 

jamwerks

Senior Member
Very interesting read Fagotto.

As I understand it, Hermode is built into Cubase? Do you have any upcoming deals with other daws (DP?). Also, can you say a bit about cpu usage for the realtime pitch correction?

Thanks !!
 
A little hint for all, especially for "dfhagai": The human ear is a very sensible tool but it requires for an optimal working about 10 or 15 minutes. You will realize this by visting a symphony concert - and oftenly the orchestra also needs the same time for being "in tune". Frankly, when listening the first time to this musical example I also feeled only little sound differences. But when comparing three times the horns from 1:18 on or the flutes and high strings from 3:10 on, the sound difference seemed me to be enormous.

How could there be any difference if hermode tuning doesn't work with Kontakt nor Play?
 

germancomponist

Senior Member
An interesting thread and it is nice to see you here, Werner Mohrlok. o-[][]-o

I was and am always at this theme. Thanks for all the informations.

Interesting: Some month ago I opened a thread here about this theme, and many people had only a smile at my tuning statements. :mrgreen:

Gunther

@Leo: To get more attention to this thread, I would change the title.

For example like this: "Hermode Test with HS strings, playing Wagner"
 

XcesSound

Member
XcesSound said:
The mock-ups are great! I especially was drawn to the brass sound, I did notice some of the bumps in dynamic but overall brilliant. Thanks
TGV said:
I thought I could hear a very subtle difference, indeed when the brass joins the party at 1:22, but ... omg, did I wet my pants. What an beautiful rendering, apart from the brass that's a bit wavering, as XcesSound points out.

XcesSound and TGV, thank you very much for your kind words, but may I ask, what do you mean by the brass having "bumps in dynamics" and "wavering"? Do you mean the vibrato that sometimes creates a bit of a pulsating and wavering sound?
I think that a big brass section, especially at high dynamic levels, does naturally waver a bit as those are instruments that are hard to control and pretty exhausting to play, especially with long notes. Could you show me some places in the piece where you find that the brass could be better? I would be very interested in this as I would like to improve my mock-up skills, thank you very much![/quote]

Hello. What I meant by the "bumps in dynamics" is more to do with the timbre which the french horn (or most brass instrument) changes as dynamic increases the tone becomes more metallic and brighter. In your track for the Wagner, at 2'00'' and the next 10 seconds, I could hear the F Sharp sticking out quite a bit in the melodic line, due to the brighter tone associated with a higher dynamic. A live player would have made it sound smoother due to the burst of the air being more gradual. But other than that it's extremely convincing mock-up!
 

Fagotto

New Member
... As I understand it, Hermode is built into Cubase? Do you have any upcoming deals with other daws (DP?). Also, can you say a bit about cpu usage for the realtime pitch correction?...

Upcoming deals: Actually no. Cakewalk temporarily was interested in it as they possess already Hermode Tuning in the z3ta synth. But not in Sonar. Actually they wrap themselves in silence.

It may be that for such applications and also for note edition programs still in 2013 there will come out a new orchestral sample library of high quality, embedded in a sampler which will content Hermode Tuning also as an internal feature.

The load of the CPU is minimal as Hermode Tuning function bases almost in tables, followed by less calculation steps on the end. When looking to the CPU loading display at my "Logic", I cannot aware a difference when working with Hermode Tuning or working with equal temperament.
 

muk

Senior Member
Just wanted to say that these mockups are fantastic, really impressive! To me the differences were subtle, but I liked the one with hermode tuning slightly better
 
Top Bottom