What's new

Would this be useful? Head to head comparison for libraries

How useful would this idea be?

  • Extremely useful

  • Somewhat useful

  • Slightly useful

  • Not at all useful


Results are only viewable after voting.

QuiteAlright

Gastronomer
Hello all! Whenever somebody asks me for DAW recommendations, part of my advice is to use the free trials, and see which they like best personally. I've always thought that it's a shame people can't do that with sample libraries, and while the threads on the forum provide great advice to buyers, I wonder if we need a better way to compare sample libraries head-to-head.

I'm thinking of making a website where you can hear the same composition rendered across different libraries. Or to phrase it differently, you would hear the same "demo" but played with different products. My concept is that you would pick an instrument, and select which libraries you want to compare. Then it would present a demo for each articulation which would be the same across libraries. For complete orchestras, or libraries with a lot of complementary instruments, it could also have complete pieces written for various instruments.

Would this be helpful? I think it would make it easier to shop for libraries. But it would be a lot of work volume matching, changing velocity+mod curves to match different dynamics, and composing pieces that would highlight each articulation properly. I'd love to hear other people's thought about this!
 
@digimortal Seems like you had the exact same idea here? Would love to connect with you about this.

(side note: does putting the 'at' symbol notify people if they haven't posted in a thread yet?)
 
Are you going to play to the libraries strengths? I mean, if you use the same midi data for every library that you wrote using one library, it probably will make the other libraries sound bad in comparison.

Of course, a basic midi will make pretty much all sound bad.
 
Are you going to play to the libraries strengths? I mean, if you use the same midi data for every library that you wrote using one library, it probably will make the other libraries sound bad in comparison.

Partially asked and answered at the beginning of the thread:

it would be a lot of work volume matching, changing velocity+mod curves to match different dynamics, and composing pieces that would highlight each articulation properly.
If done right, this project would be invaluable. But doing it right is a real challenge. In addition to all the work involved in customizing the MIDI for each library -- alluded to in previous posts -- there's the issue of who is going to pay for all those libraries.

(side note: does putting the 'at' symbol notify people if they haven't posted in a thread yet?)
Yes, but individual members can turn off those notifications in their preferences.
 
Last edited:
Are you going to play to the libraries strengths? I mean, if you use the same midi data for every library that you wrote using one library, it probably will make the other libraries sound bad in comparison.

Of course, a basic midi will make pretty much all sound bad.
My concept (originally at least) is they would share the same composition but not the exact same MIDI data. So things like mod wheel or velocity would be changed between libraries for consistency. But the point here really isn't to play to each library's strengths, so much as to do a direct comparison with everything being equal. Personally, I just use manufacturer's demos to hear the "strengths" of a library, although I know that might not be good for others.
 
If done right, this project would be invaluable. But doing it right is a real challenge. In addition to all the work involved in customizing the MIDI for each library -- alluded to in previous posts -- there's the issue of who in going to pay for all those libraries.
Yes, I completely agree that the execution here is a huge challenge. And these things are very finicky to get right - I think it would be a huge timesink for me 😄

I'm not too worried about the cost of the libraries. It's possible I could find other people who are willing to help, but it's also possible that I could just start small (meaning by using ones I already own) and slowly expand until there's a good selection.
 
Could always just crowd source the demos. After all if it's the same composition I'm sure people would be willing to spend a few minutes just tweaking a few minor things and sending them over.

I doubt that a single person would be able to do this all by themselves. I think the overall idea is fantastic.
 
I think part of the key to avoiding libraries' strengths, and the answer to the question of fairness all around, the demo would need to be cleverly crafted to hit most of the dynamic range in each instruments' registers and show a wide range of styles without preferring one over any others
 
Last edited:
Not exactly the same thing since you’re talking about individual instruments and this was a “full track” (albeit very short), but still along the same lines.


My idea was more for myself, to learn the pros and cons of the libraries I owned at the time, but others started chiming in and before long we had a massive list.

While it was great fun while the thread was running, I’m a bit skeptical towards any functional value since you have no idea about how well programmed a particular entry is. You can only really evaluate the raw sound of the recording and room, but not if the performance makes the library justice or slaughters it. :)
 
I did it once, playing the same song with two different libraries. The result is indeed quite interesting, but do not understimate the time it takes to adopt the Midi Data to the strength of each library. You need to prepare two completely different Midi files to have a fair comparison...
 
No question this is an interesting thing: listening to the same music performed by different libraries.

But it's a bit like comparing real instruments played by different performers. I'd probably like a beginners cello performed by Yo-Yo Ma better than a Goffriller played by a beginner. Just using the same MIDI file doesn't work at all.

And it's not just the library itself. Every library has an own workflow (like Performancesamples vs VSL). One works without key switches while the other doesn't work without key switches at all. Different delays, CC curves, articulation switch via velocity and more differences come together.

Then there is blending. Some are harder to blend and get harsh or muddy in tuttis but on the other hand sound more convincing played solo. Some need mixing work. Especially dry libraries and modeled instruments which tend to perform much better than wet libraries.

And then there is the "feel". How does it perform under your fingers? Fun to work with or a fight against unevenness? Different for everyone. Some work for the samples. Others want to perform what they have in mind and what works in a real orchestra.

There were a comparison thread a few years ago just for legato violins and that was a big challenge already.




Don't want to be the party killer but I think it's a big task with not really comparable results in the end. At least not if performed by different people or using the same MIDI.


Mattia Chiappa made a good video using different libraries for the same mockup.

 
Last edited:
Hey QuiteAlright, it was indeed me who made the suggestion for something like this in the Choir comparison thread! Still think having something like this would be amazing.

Now if you would let my mind take full reign a (web)app like this it would go beyond just a VI compare and would look like this:

The app would be user driven, meaning that anyone could create a page/comparison with a source file and a question. The source file could be MIDI, audio, or external source like youtube link, in turn other users can then upload audio files, comments and images/screenshots.

This would then allow comparisons being posted like:

- 'For this MIDI legato line at 100bpm I would like to hear horn examples from Spitfire libraries only', users could then upload examples and comment on their settings (BBCSO, mix 1 f.e.) or post screenshot of their VI setting

- 'For this DI guitar file I would like to hear closest approximation of the tone of AC/DC', users can upload re-amped DI files and post their amp(sim) settings

- 'How do I recreate the synth sound heard at 1:34 in this youtube video, I have U-he Diva and Serum', users can upload examples and their synth settings

etc.

Mind you, I know this would be a lot of work, maybe opensource approach would help here but that also requires contributors/maintainers etc. A plain VI comparison would already be a good start ;)

tagging @Loerpert as he is a developer on here as well and liked the idea in the Choir thread
 
Interesting topic.

Yes, I think it would be helpful to hear a demo track performed by various orchestral libraries.

Especially if the producer puts a lot of TLC to make sure it is performed to each library's strengths, and the way it behaves in terms of dynamics, expression, ..etc.

By the way I just posted a Sample Library Comparison video Thread, where Dom Sigalas of Steinberg, compares four Orchestral Sample Libraries, playing most of the articulations of each library in real time. It's a different approach to what is suggested here, but I found it to be very helpful.

Check it out here :

https://vi-control.net/community/threads/great-orchestral-libraries-comparison-video.115292/


Cheers,
Muziksculp
 
No question this is an interesting thing: listening to the same music performed by different libraries.

But it's a bit like comparing real instruments played by different performers. I'd probably like a beginners cello performed by Yo-Yo Ma better than a Goffriller played by a beginner. Just using the same MIDI file doesn't work at all.

And it's not just the library itself. Every library has an own workflow (like Performancesamples vs VSL). One works without key switches while the other doesn't work without key switches at all. Different delays, CC curves, articulation switch via velocity and more differences come together.

Then there is blending. Some are harder to blend and get harsh or muddy in tuttis but on the other hand sound more convincing played solo. Some need mixing work. Especially dry libraries and modeled instruments which tend to perform much better than wet libraries.

And then there is the "feel". How does it perform under your fingers? Fun to work with or a fight against unevenness? Different for everyone. Some work for the samples. Others want to perform what they have in mind and what works in a real orchestra.

There were a comparison thread a few years ago just for legato violins and that was a big challenge already.




Don't want to be the party killer but I think it's a big task with not really comparable results in the end. At least not if performed by different people or using the same MIDI.


Mattia Chiappa made a good video using different libraries for the same mockup.


This is what I meant when I say play to each libraries strengths. If they use Keyswitches, for example, you can't use a midi that is based on velocity switching. I do think the same piece would be good though.

Then the question is style. Some libraries are made for a certain style. Kind of like trying to use Ark 1 for a quiet piece. Yes, you could do it, but it's not going to sound as good as say Tundra.

My thought is if you can get a few basic one or two minute pieces of music for different categories - film, classical, loud, soft, etc.... You can use known pieces. And have users make it with their favorite library? A regular user of the library will know all its quirks and how to work with them. As in, this is what it can sound like for this piece. Then people who are looking for a type of library can compare better? Limit effects to certain basics or DAW effects? Note what you use.

And this could get very complicated very fast.
 
Not just that, but there are some really bad demos out there. I mean bad to the point where I can’t work out how someone has made a good library sound so awful. There are people out there with some influence who really haven’t got much of a clue what they’re doing with orchestral libraries.

It might be better to ask someone with renowned skills - not just as a demo composer but also proven ability at mixing different libraries together - to do this properly and crowdfund them a professional fee for their time. Someone who has ability in pastiche, owns a lot of libraries and has skills as a musician to provide an outline as to how the libraries feel under the fingers. Maybe Re-peat? You can guarantee there will be no beating around the bush.
 
I think it would be extremely useful, but also a ton of work - but if you can pull it off , that would be awesome :)

Something I wished this website had , and I think from the amount of questions about which library should I buy questions when starting out, is a section dedicated to a summary of each library out there - And because it is so hard to argue objectively about which library is better, even if each post for a different library, could have "Votes on it " - One vote per account , per library - Thumbs up, Thumbs down or maybe a third option "Niche option" . That way, the best libraries would naturally get the most votes, and would rise to the top of the recommended Brass thread, or Strings etc.

This may not be too popular with the advertiser though, especially if there libraries did not fare very well. And of course it could be open to abuse, but I think it would give new users or people looking for a certain type of library a snapshot about how good a library is.

There is probably a better way to implement it than this, but discussion is good , and anything that helps people spend their money more wisely is to be encouraged.

Shame there was not some sort of "Standard" Midi test for Orchestral libraries , almost like an expected run through of a challenging piece that each library would be expected to put a raw, repeatable demo up , so that libraries could be compared more readily.
 
I would work, only if using the libraries as should, so NO using the same automation data (each library reacts differently to cc numbers and values), same midi (you might need to tweak it for a specific library, e.g. delay compensation, or note lengths, track separations (no ensembles present in the library), etc).
The goal should be to act like a library demo composer, trying to get the best out of each library.

Indeed one person or max 5, that have no emotional connection to the libs, no bias.. just plain doing their jobs to make the best out of every library showcased. Also a walkthrough is required, so people know HOW they did it per library.

Also NO extra libs can be used.. it should be solely the libraries in play.. nothing else. And the sound should be a is, not external FX, EQ etc...Not even the onces coming with the daw.
 
Top Bottom