Modartt user forum » Recordings featuring Pianoteq and Organteq » Actual Steinway B vs. Pianoteq Steinway B: you be the judge
Five short excerpts in quick A/B comparison. See if you can tell which is which. Enjoy.
https://soundcloud.com/george-tingley-8...tual-piano
For me, in all the fragments, the "a" version is live. (the last passage is just too short to judge)
I second that. The A fragment was the acoustic one. The resonances felt slightly more natural. However, I didn't expect them to feel so similar. I expected there would be a somewhat larger difference - like there was for the Model D comparison on youtube with the Chopin Nocturne.
Listened through Focal Alpha 65 which should have good enough quality to present the nuances in the recordings.
Now I wonder if I could get them to more closely match by simply tweaking
the resonance and by applying some EQ or trying another reverb.
I think you could increase the detuning in ptq a little, or use the 'condition' slider to do that and a some other things. Also, 'A' sounded a little more compressed to me. Maybe trying different microphone types in ptq could reproduce some of that sound? In short, 'B' sounded a bit more perfect and also more perfectly recorded to me.
Nice job! How did you mic the acoustic piano?
Fascinating! They sound very similar indeed, only the attack sound gives away the real Steinway B for me. I think the real one is always B although the last one is too short to be sure.
Have your B’s gtingley, any missing pedal noise?
"A" sounds real to me. I also think there's much more happening in the sound of "A", like there's more timbral variation, both from note to note, and within in each note over time, whereas the "B" notes sound both more "samey". In Pianoteq, soft bass notes tend to sound like the notes just turn on without much of an attack, and I can hear that here.
For me though, the playing experience can be magical, so while the other aspects of Pianoteq might improve in the future, I'd happily stick with 6.2 for the rest of my life.
I agree with the majority of posters here - 'A' sounds like the real Steinway, I think. But they are remarkably close! A very interesting comparison.
I don’t agree with the majority of posters. At first I thought that A is real Steinway. But, after listening 10 times, I changed my mind. Because, I hear some resonance or something like i have in my modified Ptq when I change the condition slider. So, A is Ptq. (I think). There is a difference, but they are quite close.
I vote that the real thing is "A" in the first set, "B" in the second, third, and fourth sets, and "A" in the fifth set.
- David
I maintain that all the first fragments (A) are the acoustic, although I can't be certain for the last segment.
But the differences are truly subtle, and the fact that one can't say for sure and that there are people who think that the B fragments are acoustic, or that they're mixed between the 5 segments, is absolutely fantastic. It just shows how good the Pianoteq sound is now. And who knows, until we get a confirmation from the OP, I may just as well be wrong in my guess, although I don't think so
I maintain that all the first fragments (A) are the acoustic, although I can't be certain for the last segment.
But the differences are truly subtle, and the fact that one can't say for sure and that there are people who think that the B fragments are acoustic, or that they're mixed between the 5 segments, is absolutely fantastic. It just shows how good the Pianoteq sound is now.
Me too. I can hear something in the 'A' segments which strongly suggests 'acoustic piano'. If I'm correct, I'll reveal what it is!
mcoll wrote:I maintain that all the first fragments (A) are the acoustic, although I can't be certain for the last segment.
But the differences are truly subtle, and the fact that one can't say for sure and that there are people who think that the B fragments are acoustic, or that they're mixed between the 5 segments, is absolutely fantastic. It just shows how good the Pianoteq sound is now.
Me too. I can hear something in the 'A' segments which strongly suggests 'acoustic piano'. If I'm correct, I'll reveal what it is!
Exactly, I also caught some strong hints in the sound. If it weren't for those, I honestly think I would've been completely uncertain which is which.
While I can clearly hear the differences between the samples, I don't think I could have guessed whether one was an acoustic or modeled piano, except for the fact that you told us. Even if I had thought one was modeled, I wouldn't have been sure. Kudos to Pianoteq, that's quite a good showing.
My vote:
A is the Acoustic
B is Pianoteq
Both, A & B sounds very good. Nice work.
Indeed A sounds more natural, or we should perhaps just say full, with more variances.
Pianoteq always had a much better performance for more elaborated music rather than key by key slow melodies. Anyway the progress they did in the last 5 years it's great. Now the entire range sounds good, and even slow key by key sounds very decent for middle range. It's so good you need to comapre to the real thing in slow playing to try to spot and point things.
It's possible that changing mic position and other setting you can get even better ressonances from pianoteq, I presume.
I finally tracked down (pun intended) my microphone guy who did the setup for the actual Steinway B recording which was held at 25th Street Studios in Oakland. His name is Steve Jarvis - you may want to google him after you read this to know something of his background. Anyway, these are the mics we have used at my three sessions with him.
1. A pair of Neumann small diaphragm KM88i for inside the piano
2. A pair of Sennheiser MK800 condenser mics for close to the piano
3. A pair of Sanken condenser omnis for the room
4. Schoeps CNC6/MK4 condensers in the curve of the piano
5. Vintage AKG C414EB for up near the hammers (rock style)
6. Sennheiser MKH800 in ORTF stereo outside the piano just beyond the curve
Just sort of took this for granted at the time. Lol!
Wow! And that's more than we can do with Pianoteq by many microphones.
Of course, Pianoteq is B.
This can be heard from the first second and there is no doubt for me, even for the last excerpt.
That said, your recording could be perfect, but there is a typical Pianoteq sound aggressivity in the high and mid frequencies, among other multiple details that betray the recording.
The only remedy is to properly configure Pianoteq’s tuning and voicing parameters, then to make a slight equalization (although eq is not necessary to realism if you correctly place the microphones).
If you are not used to it, it is also easy to be fooled by a recording with a loud volume.
Normalize your recordings 20 dB above those of Decca, and you will have the illusion that your recordings sound better ...
In fact, the raw sound of Pianoteq is already extremely faithful to that of the original instrument, but the problems begin when you switch to recording mode.
Here is one of our last recordings, if that can give you an idea of what I'm talking about: https://www.forum-pianoteq.com/viewtopic.php?id=5777
Thanks for your feedback. This is all quite new to me and so I am taking notes. I will have to google "normalize 20dB above Decca." Perhaps you could fill me in on how this is done. I do my recording in LogicExpress. Thanks again.
Of course, Pianoteq is B.
This can be heard from the first second and there is no doubt for me, even for the last excerpt.
That said, your recording could be perfect, but there is a typical Pianoteq sound aggressivity in the high and mid frequencies, among other multiple details that betray the recording.
The only remedy is to properly configure Pianoteq’s tuning and voicing parameters, then to make a slight equalization (although eq is not necessary to realism if you correctly place the microphones).
If you are not used to it, it is also easy to be fooled by a recording with a loud volume.
Normalize your recordings 20 dB above those of Decca, and you will have the illusion that your recordings sound better ...In fact, the raw sound of Pianoteq is already extremely faithful to that of the original instrument, but the problems begin when you switch to recording mode.
Here is one of our last recordings, if that can give you an idea of what I'm talking about: https://www.forum-pianoteq.com/viewtopic.php?id=5777
I will have to google "normalize 20dB above Decca." Perhaps you could fill me in on how this is done. I do my recording in LogicExpress. Thanks again.
I just wanted to say that the Pianoteq sound always looks better and realistic when the volume is loud.
For example, some Decca piano recordings are normalized to a rather low volume, and these sometimes seem ridiculous when you compare them with louder recordings. But of course, it's just an illusion.
This is a paradox, but if you put your head in the piano, you're too close to hear the flaws.
P.S.: Normalization is just the '' standard '' of sound volume commonly adopted for a musical genre ...
M.A.P. - I'm confused as to the reason for your comment - doesn't everything sound better louder, including the real thing? It brings out all the nuances. Certainly works when A-B'ing audio components, as it's easy to influence people that way.
Hi David
Thanks for your comment. This allows me to specify that Gtingley’s recording is very well normalized for this genre, and his comparison with the authentic Steinway is perfectly calibrated.
I was only surprised to read that many experimented users hesitated between A and B, while we can hear from the first note that Pianoteq is B, especially because of the excessive aggressivity in the high frequencies, attack, reverb, depth and stereo diffusion ... among many other acoustic details that make all the difference between a '' real '' and a '' virtual '' recording.
This is the reason why I notice that a loud volume can have a considerable influence on the perception of realism.
In reality, Modartt's rough model is already extremely realistic and close to the original instrument, and personally, I hear absolutely no difference.
... of course there is a difference, but it is very insignificant.
The real difference is how you configure Pianoteq's settings.
Cheers
Mickael
I had never recorded with Steinway B, but here's a first comparison test I've done with a Christiane Karajeva’s recording:
https://soundcloud.com/modelling-audio-...rajeva-ptq
I did the mix on a small Denon hi-fi, because I'm not at my studio these days, but that should give you a preview.
My idea here is to make a similar record as realistic as possible, rather than making an exact copy of the original recording.
If you can listen to it on your hi-fi system at a reasonable volume, forgetting that it is Pianoteq, it is that the objective is reached;)
I will make a new mix later if necessary ...
Here, the high-quality audio files, including Christiane Karajeva excerpt: https://drive.google.com/open?id=1lCFDD...xRC3oduXWM
Here, the Karajeva’s full album: https://www.prestoclassical.co.uk/class...d-scriabin
Here, an other version:
https://soundcloud.com/modelling-audio-...eva-ptq-v2
I only pushed back the main mics a bit to get a little more sound from the room in the high frequencies ...
Here, the high-quality audio file: https://drive.google.com/open?id=1eMr0X...3KpnrAS9Cq
Not easy to choose!
I may prefer the second version, a little less intimate and more airy, but it's very subjective!
Here, two new raw recordings with the Steinway B:
https://drive.google.com/open?id=1e8It5...VLZhEO5JIx
I only made a small eq correction, but there is still a lot of work to make it sound like the CD ... I'll come back to it later.
These pieces come from the superb Michael Dulin’s solo piano album: