Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

First alexa shoot tomorrow

There has been an Alexa in Brisbane for a number of weeks now and out on many shoots...mainly TVC's. This is me having a 'play' the day it arrived. I havn't shot with it yet but i know people that have and they loved it...but i love my MX and so use that instead:)
 

Attachments

  • IMG_1035.jpg
    IMG_1035.jpg
    94.2 KB · Views: 0
First impression of Alex is good. Cinoptix supplied the body and lenses – a kit of 8 Master primes. The body is of a sturdy build well balanced and easy to handle, pretty much as expected.

The menu system is very easy to understand and navigate through.
Apparently there is a safety mechanism that ensures the integrity of a clip if the power is cut mid recording short of the last 2 seconds of the clip. And thank God there is such a thing as we had a power cut during the shoot and didn’t loose the clip.
The boot-up time is very quick under 30 sec each time. The media format was very fast as well under 15 sec each time even when we passed 50% capacity on the card (32GB).

As far as design goes the camera at this stage appears to be ARRI worthy except one thing. The SXS DTE module accommodates the protrusion at the bottom of the SXS cards when inserted the wrong way around.
Not a big deal but if inserted with a small amount of pressure the card gets stuck and requires a pair of long nose pliers to pull out. Not a pretty sight 2 ACs hunched over the rear of the camera with pliers in their hands scratching their heads. On all SXS card cameras I have worked with there is a block inside the slot preventing the card form being inserted the whole way. At the end of a long day it is very easy to insert the card facing the wrong way not very impressive but maybe they will come out with a new module that prevents this from happening.

Reel labeling is fairly simple, by adjusting the camera prefix and reel number you can assign a letter to the camera and a number to the reel about to be recorded. The reel number will go up incrementally with each card format.
This is unless a pre-existing prefix is on the card.
When I inserted a card with a recording on it this means the folder had a different camera prefix and reel number than what was suppose to be and I didn’t format the card. The camera created a new folder on the card with the correct camera prefix but wrong reel number.
The clip count was also effected as it did not continue from the last recorded clip number in the camera, rather from the clip numbers that were on the card under the different camera prefix folder.

I also think that a random digit or 2 should be added onto the end of the clip names. This is simply to significantly reduce the chances of duplicate clip names being recorded. If the same prefix, reel number and clip count are set for 2 separate setups in a day this can happen.
What are the chances? I know! But you never can tell and I’ve seen stranger things happen.
So best to reduce the chance of duplicate clip numbers via naming convention alla' Red.

I suppose these are reasonably simple things to fix seeing as the rest of the camera worked great.

We also experimented with recording QT 444 and 422 on the same card went without a hitch.

And to the interesting stuff...

Here are some frame grabs shot on a Master prime 21mm, T4, 800ASA, 180deg, 3600K, 25FPS, QTProres422 HQ, Log-C.

The mid gray patch measured 4 at 800ASA:
http://www.benrak.com.au/shelter/ALEXA_CHART_422.jpg

This is the same shot with pushed mids and shadows:
http://www.benrak.com.au/shelter/ALEXA_CHART_422_CC.jpg
 
I think both cameras (Red & Alexa) are great tools. It's stupid to dismiss something that you think isn't 'good enough' when it clearly is.
 
Let's all calm down. It's up to everyone to decide which camera meets his or her own needs. Because a great director or cinematographer uses a given camera doesn't mean the camera is great, just that we're more likely to see great results from it on a given project...

I'm a huge Arri fan but these shots don't wow me. The latitude and color rendition are superb (imo better than the red's) and the resolution looks adequate (if obviously worse than the red's)--but everything seems to fall apart completely when graded and even the ungraded shots show issues with tonality. Too much information crammed in too few bits--like superflat on the 7d.

But all cameras have their quirks and anyone who sees these frame grabs and shouts "not good enough!" is drawing conclusions too soon. Test what's available to you, then decide what's good enough between what you can afford to rent. (Trust me, I don't think the hvx is exactly good enough, but it's like all I use!)
 
I like Red a lot, as far as I am concerned Red 1 is the biggest game changer in digital cinematography so far. MX is amazing but I now also like Alexa seeing what it has to offer, on different jobs I would prefer different cameras.
There is no such thing as the best camera only the best camera for the job.

I had to upload these grabs to a friends website subsequently compressed them down to "not being a burden on him" size. These frame grabs are obviously highly compressed JPGs that I pushed very harshly in FCP. I'm sure that a talented grader with better tools would do a better job.
Also lets not forget they are not an exact or scientific measure in any way they are just a few shots I did on location between setups to post a few grabs for others to see.
Am sure others will post better grabs soon enough, worth while having a play around with one at your local rental company either way.

Nir.
 
Don't pay any attention to the JPEGs posted here

Don't pay any attention to the JPEGs posted here

Greetings to the reduser cosmos,

I'm the cinematographer of this shoot on which Nir was 2AC/DIT and he told me about his post here. Apparently some holy wars have broken out, which I have sadly missed! No doubt those posts would have amused, but in no way elucidated, an intelligent reader.

I shot everything in the ARRI LogC (Cineon-based) gamma space. The JPEGs Nir has posted have not been converted (or correctly converted) from LogC to sRGB for viewing on the Web. So, naturally, the images look rather washed out and downright unpalatable. If I delivered images that looked like that to clients, I'd be very soon looking for another line of work!

For example, here is one of the images that presumably triggered the war:

http://www.benrak.com.au/shelter/ALEXA_SKIN_422_CC.jpg


It really should look something like this:

http://www.sendspace.com/file/naa73p


I say "something like" because even this image has not been optimally converted to sRGB. I just did it quickly on my own computer to give a rough idea. (Also, keep in mind that Nir is not in the lit scene -- he's just in front of the set, so is out of the light.)

And ProRes 422 HQ, the format we shot this project in, is by no means the highest quality the camera is capable of. There is ProRes 4444 and ARRIRAW, which will no doubt show off the camera's abilities even more.

Hopefully the moderator-enforced ceasefire on this thread will last. ;-)

Sid.

Matt W. said:
I'm a huge Arri fan but these shots don't wow me. The latitude and color rendition are superb (imo better than the red's) and the resolution looks adequate (if obviously worse than the red's)--but everything seems to fall apart completely when graded and even the ungraded shots show issues with tonality. Too much information crammed in too few bits--like superflat on the 7d.
Click to expand...
 
Matt W. said:
The latitude and color rendition are superb (imo better than the red's) and the resolution looks adequate (if obviously worse than the red's)--but everything seems to fall apart completely when graded and even the ungraded shots show issues with tonality.
Click to expand...

Do you mind sharing which shots gave you that impression ?
 
Sidat de Silva said:
Greetings to the reduser cosmos,

I'm the cinematographer of this shoot on which Nir was 2AC/DIT and he told me about his post here. Apparently some holy wars have broken out, which I have sadly missed! No doubt those posts would have amused, but in no way elucidated, an intelligent reader.

I shot everything in the ARRI LogC (Cineon-based) gamma space. The JPEGs Nir has posted have not been converted (or correctly converted) from LogC to sRGB for viewing on the Web. So, naturally, the images look rather washed out and downright unpalatable. If I delivered images that looked like that to clients, I'd be very soon looking for another line of work!

For example, here is one of the images that presumably triggered the war:

http://www.benrak.com.au/shelter/ALEXA_SKIN_422_CC.jpg


It really should look something like this:

http://www.sendspace.com/file/naa73p


I say "something like" because even this image has not been optimally converted to sRGB. I just did it quickly on my own computer to give a rough idea. (Also, keep in mind that Nir is not in the lit scene -- he's just in front of the set, so is out of the light.)

And ProRes 422 HQ, the format we shot this project in, is by no means the highest quality the camera is capable of. There is ProRes 4444 and ARRIRAW, which will no doubt show off the camera's abilities even more.

Hopefully the moderator-enforced ceasefire on this thread will last. ;-)

Sid.
Click to expand...

Hi Sid
Nice to see you on Reduser.

I would like to add and or reiterate that in addition to the shots been taken out of the scenes lighting. They were not graded in any way to look "good" or closer to the intended look. These were simply an experiment in tolerance of the format and chip.

I realize now that this may have been a mistake and will not post stills like these again.

In addition these JPGS were extracted from the QT and then re-compressed for web upload. This means that much of the detail in the frame was perceived as noise.

These are in no way a representative of the look the shots will gain once processed and graded correctly.
 
Ouch... Haven't been here for a while (like 24 hours... -:), but what seems to have happened here, is exactly why I have chosen NOT to publish my test untill all is done...

If any of you guys have played a bit with the redlogFilm, you will see something slightly similar to the logC approach, and IMHO that is actually a very good approach.

Nope. It doesn't look "good" or "finnished" out of the box (one could record rec 709 on the Alex, but that would sort of be like shooting it in both kneecaps.) logC it is a very good way of conserving information for post. Like RAW is.

But RAW doesn't look like anything, it's (basically) just data.

logC/filmlog is a good place to start for a grade if you want to keep most of any of these cams latitude available in post.

My test will come, though...

Cheers!

Gunleik
 
the testframes look to me like RED in the starting days even the camera is big like a monster.

a lot has changed since then ( stay with RED )

btw: never had the idea to switch . . .
 
RED had some of the same approach to imagery before post when it was first released, as far as I can remember - but this may be just an emotional memory... will have to go back and look at some of my early recordings to verify...), and it was a very good approach for keeping latitude available and judging what was recorded.

Then came redspace (a way to heavy S curve for anything going through a DI), redgamma is slightly better, but still very brutal for judging what is actually recorded.

RED MX is a fantastic camera, too - though.
 
Gunleik Groven said:
Ouch... Haven't been here for a while (like 24 hours... -:), but what seems to have happened here, is exactly why I have chosen NOT to publish my test untill all is done...

If any of you guys have played a bit with the redlogFilm, you will see something slightly similar to the logC approach, and IMHO that is actually a very good approach.

Nope. It doesn't look "good" or "finnished" out of the box (one could record rec 709 on the Alex, but that would sort of be like shooting it in both kneecaps.) logC it is a very good way of conserving information for post. Like RAW is.

But RAW doesn't look like anything, it's (basically) just data.

logC/filmlog is a good place to start for a grade if you want to keep most of any of these cams latitude available in post.

My test will come, though...

Cheers!

Gunleik
Click to expand...

Hi Gunleik.

True RAW is just data, though these were not raw they were shot into QT prores.

The purpose of these stills was, and still is, to see how and how much detail got carried from the sensor through to compression using Log-C after pushing highlights and shadows. Not how good looking an image I can extract from the shots we took. As this could easily have been achieved by using well lit and graded shots or PS.

But I think I didn't explain this clearly enough in my first post.
 
You must log in or register to reply here.
Back
Top