Hacker News new | past | comments | ask | show | jobs | submit login

Well, but I am not sure that you would indeed require it to be the same millisecond. I mean I don't know how long the exposure was, but I guess its longer than a millisecond (just took a look at some of my own photos and they were at 4-5ms exposure in very bright sunlight, the blog image exposure is probably more like 10ms). The heading says that the pictures were taken in the same millisecond, but as far as I understand the post, they concluded it from the image, so I wonder what the precision is you can derive from a photo?



They said they compared EXIF metadata, so they should know when each of the shots were taken. I'm not a photographer and don't know what level of detail is recorded there, though.


The time precision that you can derive from the spray of the waves is a lot more precise that anything you will get out of the exif data - the wave will only have that exact captured spray pattern for a small fraction of a second, but the time synchronization on the two photographers cameras is almost certainly out by a larger amount of time than the wave looked that way.

Working with images from multiple photographers cameras who haven't gone out of their way to sync the time beforehand, I've usually had to assume about a 10s time skew. They do not have precise clocks and are not sybchronized often.


The time in the EXIF metadata is just based on the time/date stamp that's been set in the camera. I don't know about the newer Canon but I'm pretty sure the older model would just be something someone manually set.

Re-reading the article, I think saying the pics were shot at the "same millisecond" was just a colloquial expression to mean the look of the waves makes it look as if they were shot at the exact same time. They shared metadata but that seems to have been to share the other shooting parameters. I believe (and this would seem to confirm [1]) that EXIF timestamp data is only precise down to 1 second. added: and only as accurate as the photographer who set the date and time.

[1] https://www.media.mit.edu/pia/Research/deepview/exif.html


Both cameras (5DM4 and 60D) can be set via GPS using a Canon accessory, but I highly doubt that is the case (very few own that from what I have seen). Even then, the camera doesn't preserve millisecond data in the EXIF data.


Yeah, roughly the same 1/100th of a second seems more likely.


Shorter than a millisecond. One shot at 1/1000th, the other at 1/1600th.


Yes, you are right, didn't see that. So one is exactly 1ms and the other just 0.625ms.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: