Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico
Hi Enrico, The bottleneck is undoubtedly the expensive surface computation happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting. You could speed it even more by dropping full lighting and instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users chimerax-users@cgl.ucsf.edu wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Okay, thank you very much Eric ! May I have some tutorial for such alternative surface visualization and its coloring ( I am also interested in the transparency effect). Here is how I do it at present moment:
surface protein rainbow protein target c palette ^paired-12
# transparence options transparency #1 80 rainbow protein target s palette ^paired-12 transparency 14
And here is for lighting:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4 depthCueEnd .9 lighting gentle color floral white intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows finer
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 01:27 Eric Pettersen pett@cgl.ucsf.edu ha scritto:
Hi Enrico, The bottleneck is undoubtedly the expensive surface computation happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting. You could speed it even more by dropping full lighting and instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users <
chimerax-users@cgl.ucsf.edu> wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several
Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never
mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with
an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives:
https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Hello, The previous reply from Eric already told you the menu option or command that you could use to show spheres. These are just a space-filling atom representation, not a surface. You can make them transparent with the "transparency" command and "target a" ... however, then you will not see atoms inside of them because the spheres are the atoms. So this may not meet your needs. If you really need a surface with atoms also shown inside, you would have to tolerate the slower calculation.
Showing atoms as spheres is independent of lighting, you could use the same lighting as you already were. However, as Eric said, the simple lighting will be faster than gentle. So it's a choice... fancier appearance with slower calculations, or less fancy and faster.
Elaine ----- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Mar 26, 2024, at 12:46 AM, Enrico Martinez via ChimeraX-users chimerax-users@cgl.ucsf.edu wrote:
Okay, thank you very much Eric ! May I have some tutorial for such alternative surface visualization and its coloring ( I am also interested in the transparency effect). Here is how I do it at present moment:
surface protein rainbow protein target c palette ^paired-12
# transparence options transparency #1 80 rainbow protein target s palette ^paired-12 transparency 14
And here is for lighting:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4 depthCueEnd .9 lighting gentle color floral white intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows finer
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 01:27 Eric Pettersen pett@cgl.ucsf.edu ha scritto: Hi Enrico, The bottleneck is undoubtedly the expensive surface computation happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting. You could speed it even more by dropping full lighting and instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users chimerax-users@cgl.ucsf.edu wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
Okay, thank you very much Elaine !
Actually you are right, I use the transparent surface to visualise additional cartoon representation inside, so probably it would be a good choice for me.
I have two additional questions:
1 - From the technical side, I checked the CPU loading on a new mac and found that the Chimera-X uses only 15% from the m2 pro chip. Are there any possibility to use more resources during movie sessions with Chimera-X ? ;-)
2 - about the lighting, I've just checked my scirpt and found that two parallel options are activated:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4 depthCueEnd .9 lighting gentle color floral white intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows finer
Since I like this rossy feeling, would something like the below command work for me with the same effect but less loading ?
lighting gentle color misty rose intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows normal
Many thanks in advance !
Enric
Il giorno mar 26 mar 2024 alle ore 16:11 Elaine Meng meng@cgl.ucsf.edu ha scritto:
Hello, The previous reply from Eric already told you the menu option or command that you could use to show spheres. These are just a space-filling atom representation, not a surface. You can make them transparent with the "transparency" command and "target a" ... however, then you will not see atoms inside of them because the spheres are the atoms. So this may not meet your needs. If you really need a surface with atoms also shown inside, you would have to tolerate the slower calculation.
Showing atoms as spheres is independent of lighting, you could use the same lighting as you already were. However, as Eric said, the simple lighting will be faster than gentle. So it's a choice... fancier appearance with slower calculations, or less fancy and faster.
Elaine
Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Mar 26, 2024, at 12:46 AM, Enrico Martinez via ChimeraX-users <
chimerax-users@cgl.ucsf.edu> wrote:
Okay, thank you very much Eric ! May I have some tutorial for such alternative surface visualization and
its coloring ( I am also interested in the transparency effect). Here is how I do it at present moment:
surface protein rainbow protein target c palette ^paired-12
# transparence options transparency #1 80 rainbow protein target s palette ^paired-12 transparency 14
And here is for lighting:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4
depthCueEnd .9
lighting gentle color floral white intensity 0.9 ambientColor rosy brown
shadows true qualityOfShadows finer
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 01:27 Eric Pettersen <
pett@cgl.ucsf.edu> ha scritto:
Hi Enrico, The bottleneck is undoubtedly the expensive surface computation
happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting.
You could speed it even more by dropping full lighting and
instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users <
chimerax-users@cgl.ucsf.edu> wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several
Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never
mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with
an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives:
https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives:
https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
(1) I don't know, sorry
(2) you have to just try it and see ... because it is your opinion that determines whether it would "work for you," right?
Elaine
On Mar 26, 2024, at 8:34 AM, Enrico Martinez jmsstarlight@gmail.com wrote:
Okay, thank you very much Elaine !
Actually you are right, I use the transparent surface to visualise additional cartoon representation inside, so probably it would be a good choice for me.
I have two additional questions:
1 - From the technical side, I checked the CPU loading on a new mac and found that the Chimera-X uses only 15% from the m2 pro chip. Are there any possibility to use more resources during movie sessions with Chimera-X ? ;-)
2 - about the lighting, I've just checked my scirpt and found that two parallel options are activated:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4 depthCueEnd .9 lighting gentle color floral white intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows finer
Since I like this rossy feeling, would something like the below command work for me with the same effect but less loading ?
lighting gentle color misty rose intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows normal
Many thanks in advance !
Enric
Il giorno mar 26 mar 2024 alle ore 16:11 Elaine Meng meng@cgl.ucsf.edu ha scritto: Hello, The previous reply from Eric already told you the menu option or command that you could use to show spheres. These are just a space-filling atom representation, not a surface. You can make them transparent with the "transparency" command and "target a" ... however, then you will not see atoms inside of them because the spheres are the atoms. So this may not meet your needs. If you really need a surface with atoms also shown inside, you would have to tolerate the slower calculation.
Showing atoms as spheres is independent of lighting, you could use the same lighting as you already were. However, as Eric said, the simple lighting will be faster than gentle. So it's a choice... fancier appearance with slower calculations, or less fancy and faster.
Elaine
Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Mar 26, 2024, at 12:46 AM, Enrico Martinez via ChimeraX-users chimerax-users@cgl.ucsf.edu wrote:
Okay, thank you very much Eric ! May I have some tutorial for such alternative surface visualization and its coloring ( I am also interested in the transparency effect). Here is how I do it at present moment:
surface protein rainbow protein target c palette ^paired-12
# transparence options transparency #1 80 rainbow protein target s palette ^paired-12 transparency 14
And here is for lighting:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4 depthCueEnd .9 lighting gentle color floral white intensity 0.9 ambientColor rosy brown shadows true qualityOfShadows finer
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 01:27 Eric Pettersen pett@cgl.ucsf.edu ha scritto: Hi Enrico, The bottleneck is undoubtedly the expensive surface computation happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting. You could speed it even more by dropping full lighting and instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users chimerax-users@cgl.ucsf.edu wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using several Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane with an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
That's right, no problem ! Probably the lag was due to depthCue true and all the rest options in the first command.. Anyway, I hope one day (with the m5 chip of Apple) we may do the speed movie of 1 ms trajectory !! All the best, Enrico
Il giorno mar 26 mar 2024 alle ore 16:37 Elaine Meng meng@cgl.ucsf.edu ha scritto:
(1) I don't know, sorry
(2) you have to just try it and see ... because it is your opinion that determines whether it would "work for you," right?
Elaine
On Mar 26, 2024, at 8:34 AM, Enrico Martinez jmsstarlight@gmail.com
wrote:
Okay, thank you very much Elaine !
Actually you are right, I use the transparent surface to visualise
additional cartoon representation inside, so probably it would be a good choice for me.
I have two additional questions:
1 - From the technical side, I checked the CPU loading on a new mac and
found that the Chimera-X uses only 15% from the m2 pro chip. Are there any possibility to use more resources during movie sessions with Chimera-X ? ;-)
2 - about the lighting, I've just checked my scirpt and found that two
parallel options are activated:
lighting gentle depthCue true depthCueColor misty rose depthCueStart .4
depthCueEnd .9
lighting gentle color floral white intensity 0.9 ambientColor rosy brown
shadows true qualityOfShadows finer
Since I like this rossy feeling, would something like the below command
work for me with the same effect but less loading ?
lighting gentle color misty rose intensity 0.9 ambientColor rosy brown
shadows true qualityOfShadows normal
Many thanks in advance !
Enric
Il giorno mar 26 mar 2024 alle ore 16:11 Elaine Meng meng@cgl.ucsf.edu
ha scritto:
Hello, The previous reply from Eric already told you the menu option or command
that you could use to show spheres. These are just a space-filling atom representation, not a surface. You can make them transparent with the "transparency" command and "target a" ... however, then you will not see atoms inside of them because the spheres are the atoms. So this may not meet your needs. If you really need a surface with atoms also shown inside, you would have to tolerate the slower calculation.
Showing atoms as spheres is independent of lighting, you could use the
same lighting as you already were. However, as Eric said, the simple lighting will be faster than gentle. So it's a choice... fancier appearance with slower calculations, or less fancy and faster.
Elaine
Elaine C. Meng, Ph.D. UCSF Chimera(X) team Resource for Biocomputing, Visualization, and Informatics Department of Pharmaceutical Chemistry University of California, San Francisco
On Mar 26, 2024, at 12:46 AM, Enrico Martinez via ChimeraX-users <
chimerax-users@cgl.ucsf.edu> wrote:
Okay, thank you very much Eric ! May I have some tutorial for such alternative surface visualization
and its coloring ( I am also interested in the transparency effect). Here is how I do it at present moment:
surface protein rainbow protein target c palette ^paired-12
# transparence options transparency #1 80 rainbow protein target s palette ^paired-12 transparency 14
And here is for lighting:
lighting gentle depthCue true depthCueColor misty rose depthCueStart
.4 depthCueEnd .9
lighting gentle color floral white intensity 0.9 ambientColor rosy
brown shadows true qualityOfShadows finer
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 01:27 Eric Pettersen <
pett@cgl.ucsf.edu> ha scritto:
Hi Enrico, The bottleneck is undoubtedly the expensive surface
computation happening every frame. My advice would be to drop the surface computation from your script and “simulate” a surface by using space-filling spheres for the atoms, either with one of the “Space-Filling” presets in the Presets menu, or the Sphere style in the toolbar (command: “style sphere”). The will give some of the same feel for the space the molecules occupy without a surface computation. The time to record a movie should then be vastly sped up. It will still look very nice if you use full lighting.
You could speed it even more by dropping full lighting and
instead using a single light source (“lighting simple shadows true”).
--Eric
Eric Pettersen UCSF Computer Graphics Lab
On Mar 21, 2024, at 1:34 AM, Enrico Martinez via ChimeraX-users <
chimerax-users@cgl.ucsf.edu> wrote:
Dear ChimeraX users !
I wonder about the performance on the modern Apple CPUs: using
several Apple's models as a benchmark, I tested Chimera-X on the 2015 and 2017 iMacs equipped with the i7 8 cores and then on the 2023 Macbook Pro equipped with the Apple m2pro chip.
In the end, although the modern hardware was much more stable (never
mind when loading large netcdf trajectories, as I reported in the case of the 2015 iMAC), I only noticed a slight increase in performance during film editing.
As a benchmark, I took a very large model containing the membrane
with an embedded oligomer and loaded a trajectory of 5500 frames in Chimera-X. I used surface rendering with 0.4 grid spacing for everything. While everything worked fine, I was able to produce a short 30-second film of 750 frames in about 1h30 on the m2pro, which was almost the same as on my imac 2015. Is there any way to improve the performance of film-making on modern Apple CPUs?
Many thanks in advance
Enrico _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives:
https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Archives:
https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/
participants (3)
-
Elaine Meng
-
Enrico Martinez
-
Eric Pettersen