I have tested a Capture project on four different systems:
- MacBook Pro M1 Pro, 16 GB RAM
- MacBook Pro M5 Pro, 24 GB RAM
- Gigabyte G5 MF Gaming Laptop Intel i5-12500H, RTX 4050 with 6GB VRAM and 64 GB RAM
- Desktop PC Ryzen 7 5700G, RTX 3060 with12 GBVRAM and 16 GB RAM
Test Conditions: The exact same Capture project was tested, featuring: 32 Wash and Spot Moving heads, 12 Astera Pixel bars, Haze/Fog, Beam effects, Animated Musicians and Crowd, Trusses, Curtains and Stage Elements.
I tried Adaptive Qualtiy on and tried Priority prio set to FPS or detail After that i switched Adaptive Quality Off and tried all the resolution limits (from 560p to 1440p) in both cases: Prio set to FPS or detail Additionally, I later optimized the project by: Disabling "Throws Light" and "Spill Lightinh" for the pixel bars and hiding unnecessary layers (so only fixtures visable). This had a significantly greater impact than expected.
Results: - The M1 Pro is still usable for basic visualization, but performance drops sharply during scenes with haze and beams. Initially, it sometimes reached only 5–10 FPS. After optimization, it often achieved 40+ FPS. It remained the slowest machine when handling complex beam/haze scenes.
- The M5 Pro was the biggest surprise. It offered roughly double the performance of the M1 Pro. At 1080p with practical, real-world settings, it reached around 40 FPS. At 720p, it exceeded 50 FPS. Even with haze and beams, it often performed on par with the RTX 4050. Subjectively, it provided the smoothest and most pleasant visual experience.
- Gigabyte RTX 4050 -the gaming laptop generally achieved the highest consistent frame rates—up to 60 FPS. It delivered very strong performance in Capture. However, it was always extremely loud; the fans would immediately ramp up to full speed. When running on battery power without a power adapter, performance drops significantly.
- RTX 3060 Desktop The system equipped with an RTX 3060 was significantly more competitive than expected, performing close to the M5 Pro, depending on the settings.
The most substantial performance gain resulted from optimizing the Capture project. Disabling "Throws Light" and "Spill Lighting" for the pixel bars led to an increase in FPS that was - in some cases - greater than the jump observed when moving from the M1 Pro to the M5 Pro.
An interesting observation was the difference in how Capture behaves on macOS (Apple Silicon) versus Windows. On both MacBooks (M1 Pro and M5 Pro), the FPS increased significantly whenever there was movement: M5 Pro: Static: approx. 40–50 FPS Rapid camera movement or pan/tilt movements of the moving heads: increase to 120 FPS. M1 Pro: Static: approx. 45 FPS Movement: up to approx. 90–100 FPS. On the Windows laptop (with an RTX 4050) and the desktop PC, the opposite was observed: The FPS remained largely constant—or even dropped slightly—during camera pans and moving-head movements. It was particularly striking that, subjectively, the motion appeared smoothest on the M5 Pro, even though the static FPS values were sometimes lower than those of the RTX 4050-equipped laptop. I do not have a technical explanation for this, but it appears that on the Mac, certain calculations are prioritized differently when movement occurs.
Hi,
First of all: interpunction exists ;-)
Having said that, my impression (without your thorough testing) is roughly the same. But I must say that framerate doesn't tell the full story. The quality indicator is much more important most of the time. But it is important to mention that even with those two indicators at a low number, high quality renders can still be made. Sometimes the picture perfect image is not very important when previsualising, as long as the rendered images are good. These will always come out perfect, even with a very low quality in the Live window.
@Floriaan Hello I edited the text, which was transtlated via google - you are right, it was quite confusing. During the tests it was easier to compare the fps with prio set to detail and a fixed 100% quality. To get the focus more on quality i would say it like this. You have to optimize the project in any case. Throw and Spill Lighting is not possible when using Pixel Bars. Stage Elements should be also reduced to the minimum for a smooth live view and a nice sharp picture when preprogramming. Now lets have a look what would be the highest resolution that is possible with the best compromise of FPS and detail quality: (Adaptvie Quality switched off) -M1 Pro MacBook: best setting would bei max. 720p and prio all the way to FPS, -M5 Pro MacBook: here you can have a resolution of fixed 1080p and have the prio switch in the middle between FPS and quality -Gigabyte M5 Laptop RTX 4050 resolution 1080p and can have the prio to full detail - Desktop RTX 3060 like the M5 Pro
Thank you for sharing these insights! My experience is that leavinf adaptive quality switched on, gives a very good balance and on a mac overall performance cannot be beat. Even though sometimes the FPS reads a s very low, beams still move rather smooth.
It also depends a lot if you are building a static look or if you are building effects with a lot of constant movement in position, gobo's and color. In my theatre work I do a less of the latter.
I switched from an M1 macbook Air to an M4 macbook air with a 10 core GPU and 16Gb of memory. That was a huge difference and it is a very capable machine for on-the-go work. At the office we have a windows PC with a proper graphics card for the more demanding work.
I must say that I am really impressed with however bad the live graphics might look on my macbook air when building complex dynamic looks, rendered movies or stills always come out perfectly. Iit does not make much difference with the renders from the big guy at the office.