Dynamic range tests posted

Vincent Sweeney · May 31, 2011

I just worked up a small feature budget for 16mm: Approximately $34 per minute, including Prod/Optar rental, mostly 7213 stock, process/scan to HD.

Marcus Joseph · May 31, 2011

I'm a little shocked with the Red's performance considering even the 16mm performed so well. But the Alexa just seems like a whole other beast, it would be great to see it at the top of its game with the raw codec. But I'd still have to credit Red for being half the price of the Alexa. It just makes me curious to see how the Red would perform in comparison to Sony's F3, being almost the same cost, it would be a crucial decision in whether or not it's worth paying a bit more for the Red in certain small level productions.

John, where abouts are you in Sydney? I've been looking for places that do 16mm and doesn't seem to be anything that affordable besides standard definition. I've seen a 16mm HD transfer from deluxe and it didn't look anything like that, but the reason for that could also just be that it's a product of a student production.

Very interesting test though, thanks for doing it.

John Brawley · May 31, 2011

John, where abouts are you in Sydney? I've been looking for places that do 16mm and doesn't seem to be anything that affordable besides standard definition. I've seen a 16mm HD transfer from deluxe and it didn't look anything like that, but the reason for that could also just be that it's a product of a student production.

Very interesting test though, thanks for doing it.

Hi Marcus.

I'm between both Melbourne and Sydney. The $ figure I quoted was for STOCK ONLY and didn't include either telecine or processing.

There are few places in Sydney that can do it. I'd try cutting edge. They often do good deals for students. Deluxe are top shelf, and you should be getting a good result from them. there's a lot of variables though for 16mm. Did you do a transfer with them already ? There's also "The lab" as well.

On thursday night at AFTRS I'm screening another set of tests that I did comparing 6 different cameras. 35mm, Super 16, Alexa, RED the sony F3 and the Canon 1Dmk4, shooting the same scenes. There will be a screening from both 35mm print and hopefully a DCP. It's at 630 and you're welcome to come along.

jb

Marcus Joseph · May 31, 2011

Hi Marcus.

I'm between both Melbourne and Sydney. The $ figure I quoted was for STOCK ONLY and didn't include either telecine or processing.

There are few places in Sydney that can do it. I'd try cutting edge. They often do good deals for students. Deluxe are top shelf, and you should be getting a good result from them. there's a lot of variables though for 16mm. Did you do a transfer with them already ? There's also "The lab" as well.

On thursday night at AFTRS I'm screening another set of tests that I did comparing 6 different cameras. 35mm, Super 16, Alexa, RED the sony F3 and the Canon 1Dmk4, shooting the same scenes. There will be a screening from both 35mm print and hopefully a DCP. It's at 630 and you're welcome to come along.

jb

Ah yes, stock is quite expensive down here. The commercial production I shot was with the 500T 7219 stock and it was really a decision based on the low lighting conditions we had to work with around this medical histogram. But we ended up at quite a cheap transfer with video eight at full height anamorphic in DVCPRO, but it still looked quite detailed and I was very impressed with it.

But the other short film production I saw in a HD transfer was one that came in at the same time, I'm quite sure it was with either efilm or deluxe, but it looked grainier than it should have been.

I'd be very interested in seeing this test if I can make it. How'd you find the 1D performed? I've edited stuff with it at 5000-6000 ISO and it looked phenomenal, I just wish it had a higher data rate to work with, but the sensitivity of it is unreal.

John Brawley · May 31, 2011

But the other short film production I saw in a HD transfer was one that came in at the same time, I'm quite sure it was with either efilm or deluxe, but it looked grainier than it should have been.

Could be for any number of reasons. Exposure in your original footage...Noise Reduction in the transfer...was it on an older RANK or on a spirit ?. Super 16 CAN look grainy in HD anyway. It gives some a fright. Other's like it. :-)

I'd be very interested in seeing this test if I can make it. How'd you find the 1D performed? I've edited stuff with it at 5000-6000 ISO and it looked phenomenal, I just wish it had a higher data rate to work with, but the sensitivity of it is unreal.

Well the whole point of this was to get the stuff onto the big screen. it's very....revealing....you should try to come along and judge for yourself.

jb

Marcus Joseph · May 31, 2011

Could be for any number of reasons. Exposure in your original footage...Noise Reduction in the transfer...was it on an older RANK or on a spirit ?. Super 16 CAN look grainy in HD anyway. It gives some a fright. Other's like it. :-)

Well the whole point of this was to get the stuff onto the big screen. it's very....revealing....you should try to come along and judge for yourself.

jb

Sound awesome, I'll try to come along, do I have to get a hold of a ticket or anything like that?

John Brawley · May 31, 2011

Sound awesome, I'll try to come along, do I have to get a hold of a ticket or anything like that?

Just send an email with how many you're bringing to

rsvp@johnbrawley.com

jb

Brian Drysdale · May 31, 2011

How'd you find the 1D performed? I've edited stuff with it at 5000-6000 ISO and it looked phenomenal, I just wish it had a higher data rate to work with, but the sensitivity of it is unreal.

The recent single sensor evaluation by Bob Primes included the 1D

http://provideocoalition.com/index.php/awilt/story/nab_2011_-_scce_charts/

Reading some people's comments after the screenings the 1D didn't do that well overall.

The F3 had favourable comments, especially given the price and the possibly of s-log.

The Sony FS100 would seem to be the camera if you want really high sensitivity. http://blog.abelcine.com/2011/05/27/sony-fs100-ratings-and-dynamic-range/

Marcus Joseph · May 31, 2011

The recent single sensor evaluation by Bob Primes included the 1D

http://provideocoalition.com/index.php/awilt/story/nab_2011_-_scce_charts/

Reading some people's comments after the screenings the 1D didn't do that well overall.

The F3 had favourable comments, especially given the price and the possibly of s-log.

The Sony FS100 would seem to be the camera if you want really high sensitivity. http://blog.abelcine.com/2011/05/27/sony-fs100-ratings-and-dynamic-range/

The FS100 looks great in sensitivity, I also had no idea about uncompressed hdmi output with timecode. These things are making me consider this camera as a possible purchase for its cheap price.

georg lamshöft · May 31, 2011

Thank you very much! Nice to see something else than just lines and brickwalls ;-)

Some scenes didn't show differences well on the Vimeo-limited output - but others did, I think the glass bricks in the first shot show that very well. RED admits a small difference in DR, but most real-world-shots I've seen and the (although not perfect but comparable) provideocoalition-tests show an improvement of two-four stops. The ALEXA is a different beast, it's expensive, it took them years to develop - not just sticking an prestigous "ARRI"-Label on it...

To be honest, Super16 still seems like a viable choice to me, especially if you want to differenciate from the look of reality-shows... How much more information is in the negative when this was just done on a telecine? I guess the lighting changed a bit on the last scene, I can't imagine S16 holding that much more sky-detail.

Makes you wonder when RAW is actually RAW... Of course, the debayering is done within the ALEXA when using ProRes, but it's rarely altered when processing it on a separate computer, on the other hand the data-rates are very high and they don't seem to "throw away" much sensor data in log-mode. The white-balance and the ISO-rating can be handled as flexible as with RedRaw - or has anybody proven otherwise? It's simply a question how much of the "sensor data" is still accessable in post, am I wrong?

John Brawley · May 31, 2011

To be honest, Super16 still seems like a viable choice to me, especially if you want to differenciate from the look of reality-shows... How much more information is in the negative when this was just done on a telecine? I guess the lighting changed a bit on the last scene, I can't imagine S16 holding that much more sky-detail.

There was probably only 15 mins in difference between the three cameras shooting that last scene. Long enough to do a take, swing a tighter lens and do another setup. It was after sunset, and film was the first setup, but I was surprised that that much was being held.

jb