I think the difference is obvious, comparing the synchronization of the current Sony Alpha cams with the obsolete Sony Nex-5, half of them with some mis-sync.
Now, a posteriori, after the message of Turbguy en Phereo, it’s evident that the path that “br_foxy” suggested to me (and that I followed withaut any doubt given my inexperience in that type of measurements:) in order to measure the synchronization in milliseconds is not enough exact.
As “gl” wrote in 3D-Photo: “With the 60Hz LCD, the best you can see is 1000/60 = 16.67ms temporal resolution. So if both cameras are showing the same image (eg. a counter), they could still be up to 16.67ms out of sync with each other (unlikely but worst case, and you have no way of knowing).”
Now, after reading also the lucid message of Wayne Karberg (Turbguy:) it is evident that we need a monitor with a refresh rate of 1000 Hz (or alternativy a analogue stopwatch with display in miliseconds) to get that exact quantification in milliseconds, It seems clear that this type of monitor and this type of analogue stopwatch, if they exists, they will not be easily accessible.
Randy showed in Phereo another path, linking the website with the programm made by Peter Wimmer... but I'm thinking about a problem: How we can verify that the program works fine if now that program is the only way to measure the synchronization with the accuracy of one millisecond or less?
I believe that only a solution based on LED screens, until milllisecons, as already mentioned there, could be fully scientific and facilitate easy measurements but... producing words is very easy and it is very doubtful that anyone will be able to build it, using leds:
Dan, that's the bottom line, if the sync looks good for fast objects, water drops , birds etc then it's good enough. But to test it and come up with actual numbers you need a source that is moving at a known speed.
OK, had to think about it. This method cannot measure sync any smaller than 16msec, as a direct result of the refresh rate of the monitor (assuming a 60Hz refresh rate.) That makes my earlier comment completely null and void.
LCD monitors and rolling shutters complicate trying to take measurements. The old method used a CRT scanning electron beam as the sync reference: https://www.3dtv.at/Knowhow/Synctest_en.aspx
I'm glad you posted this, Jose. It prompted me to test my Samsung NX1000 rig using the same method. After many different shots at different times of the day, I'm happy to say the NX1000 rig has also logged 100% accuracy, if that's the proper term. I also tested it in burst mode - the first 7 shots were also "perfect". but the last burst shot, when the buffers gets full, is delayed from the rest, and is not perfect. I don't see any reason to post the images - just be aware that the Sony cameras that Jose and Randy use, and the NX1000's appear to be really good performers.
These four photos are made with the Samsung Galaxy S7 (cha-cha method) with the intention of quantifying in milliseconds the perfect synchronization of the current Sony Alpha cameras.
This morning, (last day NSA Convention 2018 in Ohio, USA) , simultaneously (that is, with perfect synchronization :) happened
On the one hand I saw that a colleague ("br_foxy") of the Yahoo 3D Photo group had provided me with a good website to measure the synchronization in mileseconds of the cameras:
https://www.online-stopwatch.com/
On the other hand, my wife was waiting for me impatiently (on Saturdays more people come) to go to the beach.
For this reasons, due to the excessive rush, the four photos have a poor composition and the second one has excessive deviation, because I have not had time to obtain photos with better framing, composition and depth.
The photos have shown (and have also quantified) what was previously evident with a detailed observation:
There is perfect synchronization is all the shots (the milliseconds always coincides in both screens), but for that is necessary to make the photo in two steps (as in the Samsung NX cameras):
- First step: focus (half press)
- Second step: shoot (full press)
If, hastily, the full pressure is made directly (focus and shot immediately followed) the result in the synchronization is very incoherent and can oscillate a lot: From 541 vs 568: (27 ms difference) until the best syn; 342 vs 336: (6 ms difference).
Directly full pressure (focusing and shoting simultaneously ) implies bad sync: from 6 ms (best) to 27 ( worst sync).
I hope this helps all the colleagues for whom it is important to quantify the synchronization in milliseconds and visual verification is not enough.
OK. You are right, Randy.
The problem is only in my fingers... not in the cams because it’s necessary focusing before shooting ... :) in order to get a perfect sync.
This is the only shot from the series with a sync issue. I've found this happens with my a5100s if the focus has not been achieved by both cameras before the shutter is tripped.
I think the difference is obvious, comparing the synchronization of the current Sony Alpha cams with the obsolete Sony Nex-5, half of them with some mis-sync.
http://phereo.com/album/5b4487abe7e5644f1400000e
Already commented in the forum:
http://forum.phereo.com/viewtopic.php?f=3&t=296&start=10